Evaluation of radio and microwave technology for motor vehicle warning system / Eugene F. Greneker

I GA
I T1DO. a.A-
I f3 \
I c..\o
I R3
I i9~q

OEPARTMENTALRESEARCH GOOT RESEARCH PROJECT NO. 9609
FINAL REPORT
GEORGIA DEPARTMENT OF TRANSPORTATION

EV, ALUATION

OF

RADIO ......

AND

MICROWAVE TECHNOLOGY FOR

\

\

MOTOR VEHICLE WARNING SYSTEM

OFFICE OF MATERIALS & RESEARCH
RESEARCH AND DEVELOPMENT BRANCH

TECHNICAL REPORT STANDARD TITLE PAGE

1. Report No. FHWA-GA-99-9609

2. Government Accession No.

4. Title Evaluation of Radio and Microwave Technology for Motor Vehicle Warning System

7. Author Eugene F. Greneker, III

3. Recipient's Catalog No.
5. Report Date September 1999
6. Performing Organization Code
8. Performing Organ. Report No.: 9609

9. Performing Organization Name and Address Georgia Tech Research Institute Georgia Institute of Technology Atlanta, Georgia 30332-0859

10. Work Unit No. II. Contract or Grant No.

12. Sponsoring Agency Name and Address Federal Highway Administration 400 7th Street, SW Washington, DC 20590

13. Type of Report and Period Covered Final, 1996-97
14. Sponsoring Agency Code

15. Supplementary Notes Prepared in cooperation with the U.S. Department of Transportation Federal Highway Administration.
16. Abstract The Safety Warning SystemTM (SWS) is an in-vehicle safety warning and signing system that uses microwave transmissions to
alert drivers in real-time to (I) hazardous road conditions, or (2) the approach of a police or emergency vehicle equipped with an SWS mobile transmitter. The performance of the SWS and its suitability as an in-vehicle signing system were evaluated in this study.
Three systems elements comprise the SWS: (I) the receiver capable of displaying the safety warning message to the SWSequipped motorist or speaking the message to the motorist using a voice synthesizer; (2) the mobile transmitter that sends the warning message to the SWS receiver and is designed to be mounted on police and emergency vehicles, slow moving vehicles, school buses, wide loads, highway construction vehicles, and other motorized potential hazards; and (3) a fixed site transmitter that can be deployed along the side of a highway or on a structure over a highway. The SWS can display anyone of 64 preprogrammed warning messages to a driver.
During the course of this study, the technical performance of the SWS was tested under almost perfect propagation conditions and also under highway conditions where propagation conditions were highly variable. Using worst case test results produced during highway testing, the resulting SWS performance data indicated that the system can provide advanced warning of a highway hazard to a driver. The results of driver performance studies taken from technical literature indicate that a forewarned driver will be able to stop his or her vehicle in a shorter distance than a driver who has not been forewarned.

17. Key Words Safety warning system, radar, microwave, road hazards, ITS

18. Distribution Statement No Restrictions

19. Security Classif. (of this report) Unclassified
Form DOT 1700.7 (8-69)

20. Security classif. (of this page) 21. No. of Pages

Unclassified

256

22. Price

Final Report Project A-5285
EXECUTIVE SUMMARY
Evaluation of Radio and Microwave Technology For Motor Vehicle Warning System
By: E. F. Greneker, Principal Research Scientist Sensors and Electromagnetic Applications Laboratory Georgia Tech Research Institute Atlanta, Georgia 30332 September 1999
The contents of this report reflect the views of the author, who is responsible for the facts and the accuracy of the data presented herein. The contents do not necessarily reflect the official views or policies of the Georgia Department of Transportation or the Federal Highway Administration. This report does not constitute a standard, specification, or regulation.

THIS PAGE INTENTIONALLY LEFT BLANK
II

EXECUTIVE SUMMARY EVALUATION OF RADIO AND MICROWAVE TECHNOLOGY FOR
MOTOR VEHICLE SAFETY WARNING SYSTEM
Under mandate of the U. S. Congress, the Federal Highway Administration (FHWA) contracted with the Georgia Tech Research Institute (GTRI), through the Georgia Department of Transportation (GDOT), to evaluate radio and microwave technology for a motor vehicle safety warning system in furtherance of safety in all types of motor vehicles. This study was performed by evaluating the technical performance of a product of the Safety Warning System, L. C. (SWSLC), marketed as the Safety Warning SystemTM (SWS). The SWS is an example of an in-vehicle safety warning and signing system that utilizes 24.1 GHz microwave technology transmissions to alert drivers in real-time to hazardous road conditions or the approach of a police or emergency vehicle that is equipped with an SWS mobile transmitter. The findings of the study of SWS performance and its suitability to perform in the role of an in-vehicle signing system is the subject of this report.
There are three systems elements that currently comprise the SWS: (1) the receiver capable of displaying the safety warning message to the SWS equipped motorist; (2) the mobile transmitter that sends the warning message to the SWS receiver and is designed to be mounted on police and emergency vehicles, slow moving vehicles, school buses, wide loads, highway construction vehicles and other motorized potential hazards; and (3) a fixed site transmitter that can be deployed along the side of a highway or on a structure over a highway.
The first step in the evaluation of the SWS by GTRI was the development of a set of warning applications that the SWS was thought capable of performing. A total of 20 applications that utilize the mobile transmitter were developed and a total of 24 applications that utilize a fixed site transmitter were developed. Next, a test plan was developed that defined the tests that would be conducted to determine the range at which the SWS signal could be received for each of the 20 mobile and 24 fixed site applications. Field testing of the SWS was conducted by following the general approach outlined in the test plan.
The first warning provided to the user of the SWS receiver is a first alert to the presence of an SWS transmitter well in advance (exact distance dependent on a number of variables) of the point where a line of sight can be established between the mobile or fixed site transmitter and the motorist's vehicle. The first alert is a tone or "beep" that is calculated to alert the driver that there is an SWS transmitter in the

vicinity. The first alert concept is supported by studies on driver response to hazard situations cited in Highway Engineering, by P. H. Wright and R. J. Paquett (fourth edition). It was determined that the distance required to "brake" a vehicle to a full stop by a pre-warned driver is much shorter than when the driver has not been pre-warned of the hazard.
Following the first alert, a warning message is presented to the driver on an alphanumeric display, the text of which defines the type of highway hazard that is being approached. The field tests showed that while the first alert was provided well in advance of the hazard in most applications, the alphanumeric warning message is usually displayed to the driver between three to five seconds before a line of sight is established between the SWS equipped motorist and the mobile or fixed site transmitter. This difference between the range at which the first alert and the warning message text is displayed occurs because the receiver can detect the SWS signal required to trigger the first alert at a lower relative signal level than the higher signal level required for reliable message decoding.
The SWS receiver is capable of displaying anyone of 64 different fixed text warnmg messages. However at the present time only 60 message texts are programmed for display. Four of the possible message slots are held in reserve for future applications. The text of the warning message is presented to the driver via an alphanumeric display mounted on the back of the receiver in a position that faces the driver. A voice synthesizer is also included on selected model receivers so that the SWS equipped driver is not required to look at the display to receive the warning. Lower costSWS compliant receivers without the alphanumeric display activate a light emitting diode (LED) display to provide the driver with information regarding the type of hazard to be encountered (moving or stationary).
TEST RESULTS
First, field tests were conducted at Dobbins Reserve Air Force Base in Marietta, Georgia on an unused runway. Testing on an unused runway allowed the performance of the SWS to be assessed without the effects of surrounding traffic biasing the results. Also, testing could be conducted at very slow speeds when necessary without traffic safety concerns. Preliminary testing of the SWS mobile transmitter located at the bottom of a hill was also conducted at Dobbins.
Data that show relative signal strength versus distance between the receiver and transmitter vehicles were collected and analyzed. These data indicated that the SWS would perform most of the warning applications that were hypothesized for this study, assuming that effects such as blockage of the SWS signal by traffic, highway curves, hills and other features did not further degrade the signal propagation
2

path. There was only one marginal case that was encountered during testing at Dobbins. SWS performance for the case where the receiver vehicle crossed the path of the transmitter vehicle perpendicularly was found to be marginal due to the lack of response of the receiver antenna pattern when operated 90 degrees off of parallel axis to the transmitter's antenna. The detailed results achieved
during testing of the SWS at Dobbins are found in Volume n, Appendix E of this report.
Field testing of the system was also conducted at GTRI's turntable facility. Turntable testing was conducted to measure the response of the test receiver and transmitter's antenna pattern as a function of angle around the host vehicle, and to further explain the observations made during testing at Dobbins. Received relative signal strength was recorded while the host vehicle containing the receiver was rotated around 360 degrees on the turntable. The stationary transmitter vehicle which served as a signal source was located 200 feet away. Next, the transmitter vehicle was put on the turntable and the receiver was located at the same fixed site 200 feet away. Relative signal strength data were recorded as a function of receiver heading during each rotation. The turntable tests indicated that the best SWS performance would be obtained when the receiver and transmitter antennas were pointed at each other. The next best performance would be obtained when the transmitter was behind the receiver. The received SWS signal strength was lowest when the antenna of the receiver was at an angle 90 degrees away from the
transmitter. The detailed turntable test plan and the results achieved are found in Volume n, Appendix C
of this report.
Highway testing was first conducted on two rural highways to test SWS performance over straight sections, curves and hills. On the straight sections, the maximum range at which the SWS could be received could not be determined because the straight portion of test roadway did not exceed approximately 3,600 feet and at this distance the SWS warning message was displayed from the start of each test except when blocked by another vehicle. Unfortunately, no rural highway test site with more line of sight distance could be found in the north Georgia area within the time available for test site selection. On average, the SWS provided a first alert equipped driver on straight highway sections at distances in excess of 3,400 feet.
During rural curve testing the mobile transmitter was located in the midpoint of a curve. Foliage on the inside of the curve limited the line of sight to a distance less than 500 feet beyond the transmitter, and foliage appeared to attenuate the 24 GHz SWS signal so that the first alert distance did not exceed much more than 1,022 feet. It was also determined that when a vehicle is close to and between the receiver vehicle and the SWS transmitter, the signal is also attenuated. This can reduce the range at which the
3

first alert and warning message can be received. The amount of attenuation for both cases is presented in Volume I of this report.
Field testing of the SWS with the receiver vehicle approaching the mobile transmitter at the bottom of a hill (below the line of sight), produced a first alert 3,168 feet from the transmitter vehicle. A warning 'message was displayed 1,128 feet from the transmitter vehicle, which was at a point just before the line of sight between test vehicles was established. Detailed rural test results are found in Volume II, Appendix F of this report.
Testing was also conducted in the downtown area of Marietta, Georgia to determine how well the SWS would warn a motorist that a police or emergency vehicle was crossing a "blind" intersection where buildings extended almost to the street at all comers of the intersection. The distance at which the first alert was received prior to the receiver vehicle reaching the "blind" intersection was approximately 500 feet, and the warning message was displayed when the receiver vehicle was approximately 180 feet from the intersection being approached by the transmitter vehicle. Analysis including worst case driver response times showed that the SWS effectiveness for intersection warning would be marginal during wet weather due to the short range at which the warning could be received. More warning time is necessary to stop a vehicle when the highway is wet and tire traction with the wet road surface is less than during dry conditions. A detailed presentation of urban test results are found in Volume II, Appendix G of this report.
Field testing of the mobile transmitter was conducted on Interstate 75 north of Marietta, Georgia, to determine the distance at which the SWS could be detected when the transmitter vehicle was overtaking the receiver vehicle from the rear in the interstate environment. It was found that the first alert warning occurred when the test vehicles were 2,452 feet apart. The warning message was displayed when the transmitter vehicle was an average of 2,389 feet behind the receiver vehicle during tests on the interstate. It was also determined that surrounding traffic affected the propagation of the SWS signal during the overtaking tests. Blockage by a truck behind the receiver vehicle reduced the message reception range to approximately 1,650 feet. However, this was not a problem given the low closing speeds experienced when one vehicle overtakes another at typical interstate speeds.
The fixed site roadside transmitter was also tested in the interstate environment to determine its performance when the receiver vehicle was approaching the fixed site transmitter mounted on an overpass above the highway. The first alert was received from the fixed site transmitter at a range greater than 4,500 feet and the warning message was received at an average range of 3,821 feet. Shadowing and
4

masking of the 24 GHz signal from surrounding traffic was seen in the recorded signal strength data. Detailed interstate test results are fou,nd in Volume II, Appendix H of this report.
Field testing of the SWS was conducted on the interstate to determine if the southbound lanes could be selectively sent an SWS message without the message being received by vehicles in the northbound lanes. It was found that when the fixed site transmitter was elevated and a broad beam antenna was employed to warn southbound motorists, it could be received by northbound motorists over line of sight distances approaching one mile. It was determined that tractor trailers in front of the SWS receiver vehicle reflected the signal back to the receiver. This finding suggests that when a directional lane selectivity is desired, a more narrow antenna pattern is required for the fixed site roadside transmitter.
The field tests described to this point were designed to measure the range at which the SWS first alert and warning message could be received for each of the test situations and this objective was achieved. Next, the test results were analyzed to determine if the SWS performance determined by testing was sufficient to allow the SWS to perform each application.
Human factors testing had been planned to measure driver reaction to the SWS first alert and a driver's reaction to the warning message, as well as the time taken to perceive the warning message. It was further planned to use these human factors test results to determine if the SWS forewarned driver would be able to start collision avoidance procedures sooner than the driver who did not have the forewarning of a hazard. The human factors tests were not performed due to resource limitations.
Without the human factors test results there were no driver reaction data from which to draw conclusions regarding SWS effectiveness as a driver early warning system. A stopping distance model was developed in place of the human factors study to provide braking distances for both dry level pavement and wet level pavement. Worst case driver perception and reaction times were added to the stopping distance, as well as and the additional 5 seconds thought to be a worst case time required for a driver to comprehend the SWS warning message. In a final step, the maximum speed at which the SWS would be useful was calculated for each SWS application that was different from any other.
The primary finding using this approach was that the SWS is capable of alerting an SWS equipped driver to a moving or fixed hazard that is beyond the line of sight in most of the applications that were considered. The amount of time that the SWS could alert the SWS equipped driver to the presence of a hazard was variable. It was determined that the SWS is also capable of providing the driver with a more definitive warning regarding the type hazard to be encountered earlier, in certain cases, than when visual
5

cues are the only warning of the hazard. The only application where the SWS provided a marginal warning was when a police or emergency vehicle was crossing an intersection ahead of an approaching SWS-equipped driver, in an urban area where buildings blocked the SWS direct path signal during adverse weather. One contributing reason for the projected marginal performance was because only worst case assumptions were included in the stopping distance model. Also, the blockage of the SWS signal by buildings on all comers on the intersection was a factor contributing to poor signal propagation. Despite this marginal SWS performance, it was found that the warning time provided by the SWS was still greater than visual detection of the police or emergency vehicle strobe lights.
CONCLUSIONS
During the course of this study, the technical performance of the SWS was tested under almost perfect propagation conditions and also under highway conditions where propagation conditions were highly variable. It was found that the presence of other vehicles in the traffic mix can both reduce and increase the range over which the first alert and the warning message is displayed. Using worst case test results produced during highway testing, the resulting SWS performance data indicated that the system can provide advanced warning of a highway hazard to a driver. The results of driver performance studies taken from technical literature indicate that a forewarned driver will be able to stop his vehicle in a shorter distance than the driver who has not been forewarned. A driver using an SWS-type system and receiving an alarm in advance of encountering a hazard would be cued to the hazard's presence and start the process of braking earlier than the driver who has not been forewarned. The extra time gained from an SWS alarm may allow the forewarned driver to avoid a collision that a non-forewarned driver would not be able to avoid.
The SWS can display anyone of 64 warning messages to a driver. During the development of this report it was determined that the number of applications to which the SWS might be applied could be increased if there were more message texts available. An industry spokesman indicated that the industry was considering the addition of a variable text transmission capability to the system. Also, the industry is considering the inclusion of 64 additional warning texts. The results of this study indicate that a variable messaging capability or the addition of 64 more fixed text messages could provide more flexibility and utility to the system.
A human factors evaluation of the SWS was not performed. Human factors data are essential to the total evaluation of the SWS. Human factors testing that would determine the reaction of drivers of varying ages lv a first alert would validate the hypothesis that an SWS forewarned driver could react more safely
6

(stopping more smoothly, slowing, etc.) than the non-forewarned driver. Human factors testing of the SWS could also determine to what degree in-vehicle signing might improve driver reaction time when the driver is presented with the text from a highway sign in a time critical manner.
A human factors study would also allow problems with the system to be identified, since the manner in which the data are presented to the driver can make a difference in the time required for the driver to perceive the warning message. Human factors testing would also identify the optimum approach to presenting the warning message to the driver, if one exists. A human factors study could also determine if the SWS causes undue distraction from the driving task.
The benefits of using an SWS in terms of the accidents prevented and lives saved was not thought quantifiable within the budgetary and schedule constraints of this study. However, it is thought that this present study provides the foundation for such a benefits study to be conducted.
RECOMMENDATIONS
Additional testing of SWS operation under widely varying traffic conditions should be conducted to further validate the findings of this study regarding the distance at which a warning message can be received under different highway and traffic conditions. The additional testing should be conducted in conjunction with a field test designed to determine the effectiveness of the SWS as an in-vehicle warning system. The supervised field test program could be conducted to determine the effectiveness of the SWS in lowering the peak and average speeds normally driven through a work zone.
A human factors study using a driving simulator should be conducted to measure the driver's response to the first alert and the warning message. Driving simulator testing should be conducted to determine if the SWS causes a driver significant confusion or distraction when the system is in operation.
A benefits study should be conducted to determine how many crashes and deaths might be avoided if an SWS type system is permitted to become operational. Market forecasts supplied by SWSLC should be used in the benefits study to determine the penetration of the system during specific target years. Market forecasts by the mobile transmitter manufacturers should be used to determine the number of police and emergency vehicles that might be SWS equipped during specific target years. The causal factors of crashes of the type that might be impacted by use of the SWS should be extracted from the national accident data bases.
7

Final Report Project A-5285
VOLUME I
Evaluation of Radio and Microwave Technology For Motor Vehicle Warning System
By: E. F. Greneker, Principal Research Scientist
Sensors and Electromagnetic Applications Laboratory Georgia Tech Research Institute Atlanta, Georgia 30332
September 1999

THIS PAGE INTENTIONALLY LEFT BLANK
II

ACKNOWLEDGMENTS
Individuals representing various government agencies, the U. S. military and businesses have contributed to this study during the 14 month period over which it has been conducted. The author would like to acknowledge these contributions.
The Georgia Tech Research Institute
The author is grateful to Mr. B. H. Hudson, Senior Research Technologist, who assisted in several of the tests as driver of the mobile transmitter vehicle, operator of the mobile transmitter vehicle Global Positioning System (GPS) data recording system, and as a consultant to the author. The author also wishes to thank Mr. Evon Braselman, a Georgia Tech student, who assisted the author with the turntable and rural tests.
The Georgia Department ot Transportation (GDOT)
The author wishes to thank Mr. David Jared, GDOT program monitor, for his assistance during the project period as operator of the GPS system in the receiver vehicle during selected tests, his assistance with test planning and coordination of tests with the GDOT's Office of Traffic Operations, and the research and supply of highway design data documentation.
The U.S. Air Force
The author is grateful to General Walter Hatcher, Commander, Dobbins Reserve Air Force Base, Georgia for allowing testing of the SWS on an unused runway at Dobbins. The author is also grateful for the coordination assistance provided during the testing at Dobbins provided by Chief James McCarley, Chief of Operations Security Police, and his officers, Capt. John Hayworth and Officer Dan Adamson, who assisted with safety during testing.
The author appreciates the assistance rendered by Ms. Janice Lee, President, Safety Warning System, L. C. for providing the two experimental fixed site safety warning transmitters used during the test program. The cooperation from the member companies of Safety Warning System L. C. in the form of supplied equipment and consultation is also appreciated.
MPH,lnc.
The author appreciates the efforts of Mr. Fred Perry, President, MPH, Inc. toward arranging the modification of the mobile transmitter to allow it to be controlled during testing, as well as the efforts of Mr. Dan Bowlds, Manager of Microwave Product Development, for his assistance in modifying the mobile transmitter.
BELTRONles L1MITED
While each member company of SWS L. C. contributed to the project, the author is grateful for the special assistance rendered by the BEL-TRONICS L1MITED Engineering department. Mr. Glen Martinson provided data in response to questions regarding the BEL 855 STi radar detector microwave design and Mr. Mark Triska performed the modification of the BEL 855 STi radar detector, allowing it to collect relative signal strength data via an RS-232 interface during the tests.
III

THIS PAGE INTENTIONALLY LEFT BLANK
IV

CONTENTS

Section

Title

Page

ACKNOWLEDGMENTS

iii

EVALUAnON OF RADIO AND MICROWAVE TECHNOLOGY FOR MOTOR VEHICLE

WARNING SYSTEM-

I

1.1 Research Project Purpose and Objectives

I

1.2 The Safety Warning SystemTM

2

1.2.1 Mobile Transmitter

2

1.2.2 SWS Receiver

3

1.2.3 Fixed Site Transmitter

4

1.3 Research Approach

5

1.4 Development a Summary of Operations Concepts ."

6

1.4.1 SWS Message Texts for Mobile Transmitter Application

7

1.4.2 Fixed Site SWS Transmitter Message Set

8

1.4.3 Applications Using Fixed Text Messages

8

1.5 Necessity of Evaluating the SWS

10

1.5.1 Test Plan Development

11

1.6 Testing Activities Conducted to Evaluate the SWS

II

1.6.1 Laboratory Testing of Equipment.

II

1.6.2 Dobbins Reserve Air Force Base Testing

12

1.6.3 Turntable Testing of Receiver Antenna Response

14

1.6.4 Rural Two Lane Road Testing

IS

1.6.5 Urban Intersection Testing

;

17

1.6.6 Interstate Highway Testing

18

1.6.7 Directional Selectivity Test

21

1.6.8 Summary of Test Results

22

1.7 Installation, Operation and Maintenance of SWS Components

22

1.7.1 Receiver

22

1.7.2 Mobile Transmitter

23

1.7.3 Fixed Site Transmitter

24

1.8 Transmitter Types And Public Sector Users

:

25

1.8.1 Number of SWS Transmitters Required

25

v

CONTENTS (CONTINUED)

Section

Title

Page

1.9 Evaluation of the System Performance Under Each of the Previously Defined

Operational Scenarios

26

1.9.1 Evaluation Criteria Development

26

1.9.2 Assessment of SWS Applications

34

1.9.3 Collision Avoidance at Intersection

34

1.9.4 Warn of High Speed Police Chase in Progress

36

1.9.5 Warn of Traffic Stop Ahead

37

1.9.5.1 Duplication of Basic Application

38.

1.9.5.2 Work Zone

39

1.9.5.3 Duplication of Basic Application

39

1.9.6 Transport Vehicle's Pilot Vehicle Equipped with SWS

39

1.9.7 Duplication of Basic Application

39

1.9.8 Train Locomotive Application

.40

1.9.9 Fixed Site Transmitter Applications

.41

1.9.10 Alert Motorist to Detour

41

1.9.11 Deployed in Advance of a Bridge Traffic Gate

.41

1.9.12 Rock Slide Warning

41

1.9.13 School Zone Ahead

42

1.9.14 Advance Warnings and Cautionary Messages Under Fixed Site (Application 14) .42

1.9.15 High Wind Warning

42

1.9.16 Severe Weather Warning

42

1.9.17 Heavy Fog Ahead

43

1.9.18 Temporary Hazard Warning (Application 18)

43

1.9.19 Ice on Highway/ Bridge Warning

43

1.9.20 Duplication of Previous Applications

.43

1.9.21 Emergency Vehicle in Transit

43

1.10 Operation of the SWS in the Rain and Inclement Weather

44

1.11 Recommendations to Enhance SWS Performance

.45

1.1 1.1 SWS Receiver Performance Standard Development..

.45

1.11.2 Transmitter Improvements

.45

VI

CONTENTS (CONCLUDED)

Section

Title

1.12 Implementation of SWS Technology in the Intelligent Transportation System 1.12.1 En Route Driver Information .1.12.2 Traffic Control Including Emergency Vehicle Signal Preemption 1.12.3 Incident Monitoring and Detection 1.12.4 Intersection Collision Warning
1.13 Estimates of Benefits of SWS Technology on Traffic Safety and Traffic Deaths 1.14 Conclusions 1.15 Recommendations

Page
.46 .47 .48 .48 .48 .49 50 52

Vll

THIS PAGE INTENTIONALLY LEFT BLANK
VIII

FIGURES

Figure

Title

Page

Mobile SWS transmitter that mounts on emergency vehicles

3

2

Two SWS capable radar detectors with alphanumeric display

_

.4

3

Prototype fixed site transmitter

5

4

Braking distance of vehicle on a dry and level surface as a function of vehicle speed

30

5

Braking distance of vehicle on wet and level surface as a function of vehicle speed

31

6

Braking distance of a vehicle on dry and level pavement as a function of vehicle

speed with 5 seconds added for SWS display reception time

32

7

Braking distance of a vehicle on wet but level pavement as a function of vehicle

speed with 5 seconds added for SWS display perception time

33

8

Maximum separation distance between vehicle required for SWS to provide adequate

warning of overtaken police or emergency vehicle

35

IX

THIS PAGE INTENTIONALLY LEFT BLANK x

TABLES

Table

Title

The Text of the Mobile Transmitter with Applications

2

Applications for Roadside SWS Transmitter

3

Distance at Which the Two Levels of SWS Warnings Were Presented to the

Driver During Dobbins Testing

4

Distance at Which the Two Levels of SWS Warnings Were Presented to the

Driver During Rural Testing

5

Distance at Which the Two Levels of SWS Warnings Were Presented to the

Driver During Urban Testing

6

Distance at Which the Two Levels of SWS Warnings Were Presented to the

Driver During Interstate Testing

7

Attenuation from Different Rainfall Rates

Page 7 8 13 16 18 19
.44

Xl

THIS PAGE INTENTIONALLY LEFT BLANK
XII

AC ATMS CBD dB FCC FHWA GDOT GPS GTRI ITS LED NHTSA NTIA SWS SWSLC USDOT

ACRONYMS
alternating current Advanced Traffic Management System central business district decibel Federal Communications Commission Federal Highway Administration Georgia Department of Transportation Global Positioning System Georgia Tech Research Institute Intelligent Transportation System light emitting diode National Highway Traffic Safety Administration National Telecommunications and Information Administration Safety Warning System Safety Warning System, L. C. United States Department of Transportation

XIII

THIS PAGE INTENTIONALLY LEFT BLANK
XIV

EVALUATION OF RADIO AND MICROWAVE TECHNOLOGY FOR MOTOR VEHICLE WARNING SYSTEM
Most of the technical aspects of operating a microwave warning and in vehicle signing system were answered by the Georgia Tech Research Institute's (GTRI) testing of the Safety Warning SystemTM (SWS) as the example prototype microwave system. The capability of the SWS to propagate the signal along straight stretches of highway, around curves and over hills was tested and evaluated. The effects of surrounding traffic on the millimeter wave signal propagation was also tested and documented as herein. All tests are described in detail in Volume II of this report, which contains all referenced appendices. Not tested was a driver's reaction to the SWS under various operational conditions, the human interface problems that might occur during operation of the SWS, and the factors that would allow the benefits of using the SWS to be evaluated.
Initially, it had been planned that a human factors evaluation would be conducted as part of this study by an outside contractor. The results of the planned evaluation would be used to determine driver reaction to the SWS, any user problems, and the way that the SWS could be used to allow system benefits to be evaluated. Unfortunately, the human factors part of the study was not conducted by the outside contractor. As a result there are still a number of unknown factors that preclude the assessment and impact of the SWS regarding the human response to the technology and the impact that the technology could have on saving lives on the highway.
1.1 RESEARCH PROJECT PURPOSE AND OBJECTIVES
The subject research project was developed in response to a task from the U.S. Congress to the United States Department of Transportation (US DOT) . The task from Congress to the USDOT was mandated in the National Highway Systems Bill (S440). The guidance for this study originated in the National Highway Safety Bill (S440) that was passed into law by Congress during 1996. Section 358, paragraph c, states in part:
"Section 358. Safety Research Initiatives c. RADIO AND MICROWAVE TECHNOLOGY FOR MOTOR VEHICLE SAFETY WARNING
SYSTEM
I. Study - The Secretary, in consultation with the Federal Communications Commission and the National Telecommunications and Information Administration, shall conduct a study to

develop and evaluate radio and microwave technology for a motor vehicle safety warning system in furtherance of safety in all types of motor vehicles. 2. Equipment - Equipment developed under the study shall be directed toward, but not limited to, advanced warning to operators of all types of motor vehicles of a. temporary obstructions in the highway, b. poor visibility and highway surface conditions caused by adverse weather; and c. movement of emergency vehicles. 3. Safety Applications - In conducting the study, the Secretary shall determine whether the technology described in this subsection has other appropriate safety applications."
The Federal Highway Administration (FHWA) contracted with GTRI, through the Georgia Department of Transportation (GDOT), to perform the microwave warning system assessment study. Once under contract, GTRI chose to assess the SWS as an example of an existing microwave based in-vehicle signing system that is currently being marketed to alert drivers to hazardous roadway conditions.
1.2 THE SAFETY WARNING SYSTEMTM
The SWS was developed by the Safety Warning System, L. C. (SWSLC), a Florida corporation. This system is designed to provide a driver approaching a highway hazard with an advanced warning Of the hazard's presence. Three operational elements comprise the SWS: (1) a mobile transmitter; (2) the vehicular SWS receiver and display unit, and (3) a fixed site transmitter for roadside deployment.
1.2.1 Mobile Transmitter
Figure 1 is a photograph of the mobile transmitter manufactured by MPH Inc., a respected manufacturer of police traffic radars located in Owensboro, Kentucky. The mobile transmitter is designed to mount on the emergency light bar of police and emergency vehicles. The mobile transmitter can be installed to activate when the emergency lights are activated by the vehicle's driver. A radar which is also part of the unit determines if the host vehicle is moving or stationary and selects one of two messages for transmission on the basis of host vehicle status. For example, if the mobile transmitter is installed on a stationary police vehicle, working a traffic accident with e~ergency lights activated, the message transmitted to SWS receivers will be "Stationary Police Vehicle Ahead." When the host vehicle is moving and the emergency lights are activated, the radar senses motion and selects the message text
2

"Police Vehicle in Transit" for transmission to SWS equipped motorists. The texts of the warning messages are programmed to match the type of vehicle using the transmitter. For example, an SWS mobile transmitter used on an ambulance or fire truck would transmit the message "Emergency Vehicle in Transit" when the emergency vehicle is in motion or "Stationary Emergency Vehicle Ahead" when the vehicle is stationary.
Figure 1. Mobile SWS transmitter that mounts on emergency vehicles
1.2.2 SWS Receiver
Figure 2 is a photograph of two implementations of the mobile receiver. The one on the left is manufactured by BEL-TRONICS LIMITED and the one on the right is manufactured by Whistler, Inc. Both receivers present the text warning message to the driver via an alphanumeric display shown extending across the back of the receiver. In addition, the receiver on the right has a voice synthesizer that announces the safety warning message to the driver. Each receiver also issues a warning tone called the first alert, which sounds when the SWS signal is first detected. The first alert presently is the same alert that is received when a police radar is received. It is thought that this type of first alert is more effective to forewarn the driver to a potential hazard than a dedicated SWS first alert. The proposed human factor's study was designed to address this issue. This first alert warning tone continues until the safety warning message appears on the alphanumeric display. The periodic warning tones that occur as a first alert and upon SWS signal reception are designed to raise the situational awareness level of the
3

driver as a precursor to the directed warning message to follow. Currently, SWS compliant receivers are being marketed in the United States by (1) BEL-TRONICS LIMITED, (2) Sunkyong America, Inc., (3) Uniden America Corporation, (4) Whistler, Inc., and (5) A. K. America, with other licensees of the technology pending. In addition to the estimated 1,000,000 SWS compliant receivers sold since 1996, each capable of receiving and displaying a safety warning message, it is estimated that there are 20,000,000 police radar detectors used daily on highways in the United States. The SWS transmitter system will cause any of these existing police radar detectors to alert with the same warning as when a police radar signal has been received. It is calculated that this police radar alert will heighten a driver's situational awareness to the presence of work zones, police and emergency vehicles, as well as any other highway hazard that is also marked with a sign. The activation of the existing base of radar detectors is achieved in a manner consistent with the guidelines developed for drone radar operation established by the National Highway Traffic Safety Administration (NHTSA) and sanctioned by the Federal Communications Commission (FCC).
Figure 2. bvo SWS capable radar detectors with alphanumeric display
1.2.3 Fixed Site Transmitter
Figure 3 shows a prototype fixed site transmitter. This transmitter is designed to be commercialized by a manufacturer under license to SWSLC at a future time. It is capable of being programmed to transmit anyone of 64 fixed text safety warning messages. A full listing of the fixed text messages is found in Volume II, Appendix A of this report. The fixed site transmitter is designed to be mounted on a support
4

beside or over a highway. There are four modes of fixed site transmitter operation: (1) a timer mode allows it to be programmed to be turned on and off up to four times during any 24 hour period; (2) the radar mode allows an associated radar function to sample for 0.5 seconds every 3 to 4 seconds to determine if a vehicle is present or exceeding a selected speed limit before transmitting the safety warning message; (3) the continuous mode of operation allows transmission of the warning message continuously after activation; and (4) the sensor mode allows transmission of the warning rnessage upon activation of an external sensor, such as a fog sensor.
Figure 3. Prototype fixed site transmitter
1.3 RESEARCH APPROACH
The FHWA and GDOT developed a statement of work for the project designed to achieve the research goals and objectives specified by the Congress of the United States. The research approach required to assess the performance of the SWS as a prototype highway hazard warning system involved the performance of thirteen broadly defined tasks, requiring GTRI to:
5

I. Develop summary operations concepts and related system performance requirements; 2. Perform the necessary analysis and field testing to verify the experimental system's ability
to successfully support the performance requirements; 3. Evaluate the signal propagation characteristics of millimeter-wave transmissions; 4. Identify the installation, operation, and maintenance requirements (time, cost, and level of
expertise) of such a system; 5. Identify the number of transmitter devices and types of public sector users; 6. Evaluate the system performance under each of the previously defined operational
scenarios; 7. Identify and assess the significance of potential institutional and legal issues that may arise
from use of the technology/system; 8. Assess the limitations of the technology and systems concept and make recommendations of
changes to either improve/enhance system performance; 9. Coordinate a study of human performance when using/operating the SWS system; 10. Determine the user's response to the technology; I I. Estimate the safety impact measures showing the effectiveness of the technology on traffic
safety; 12. Estimate the benefits of this technology, including the potential number of lives saved or
injuries that would be reduced if this technology were widely implemented, and; 13. Evaluate the findings, and then identify the applicability of the system within the context of
the Intelligent Transportation Architecture, and how it would be best implemented.
The human factors assessment tasks (numbered 9 and 10) were to have been cooperatively conducted by GTRI and another FHWA contractor. These tasks were not conducted due to resource limitations by the other contractor. The sections that follow discuss the findings of each of the remaining I I tasks.
1.4 DEVELOPMENT OF SUMMARIES OF OPERATIONS CONCEPTS
The FHWAlGDOT statement of work requested that a summary of operations concepts be developed for no less than five applications, no less than one per each of the five message categories. The 60 currently existing messages are organized into five basic applications categories as shown in the complete message list found in Appendix A. These five basic message categories are: (I) highway construction or maintenance. (2) highway hazard zone advisory, (3) weather related hazards, (4) travel
6

information/convenience, and (5) fast/slow moving vehicle. Four of the 64 possible messages are blank message slots, unassigned and held for future use. GTRI researchers developed 20 SWS applications for the mobile transmitter and 24 applications for the fixed site transmitter using the text messages which have been determined by members ofthe SWSLC.

1.4.1 SWS Message Texts for Mobile Transmitter Application
Two possible mobile warning messages are programmed at the factory. The message which is transmitted at any given time is determined by whether or not the transmitter's host vehicle is stationary or moving. When the vehicle is stationary, a message warning of a stationary vehicle will be selected for transmission. Likewise, a message warning of a moving hazard will be automatically selected when the vehicle is moving. Per Table I, the text of the warning message that suggests a mobile application of the SWS appears in the left column, and the corresponding suggested application developed by GTRI appears in the right column.

Table 1. The Text of the Mobile Transmitter with Applications

Message Transmitted
Police in Pursuit
Stationary Police Vehicle Ahead Emergency Vehicle in Transit Stationary Emergency Vehicle Ahead Highway Work Crews Ahead
Work Zone Ahead Slow Moving Vehicle (School Bus) School Bus Loading/Unloading Oversized Vehicle in Transit Train Approaching

Mobile Warning Application
1. Collision avoidance at intersections. 2. Warn of high speed police chase in progress. 3. Warn of police car overtaking in same direction. 4. Warn of traffic stop of motorist with SWS. 5. Serve as warning for wreck sites and roadside traffic stops. 6. Alert motorist to drunk driver roadside testing. 7. Alert to police roadblock with police in the roadway.
8. Collision avoidance at intersections. 9. Yield to overtaking emergency vehicle. 10. Warn of wreck sites with ambulance in roadway. 11. Warn of fire hose across roadway during fire fighting. 12. Warn of fire truck blocking the roadway. 13. Marker for slow moving line painting vehicle, grass cutting
vehicles, dump trucks, construction equipment, utility company vehicles, any vehicles in the construction zone and moving slowly. 14. Provide early warning of work zone stationary vehicles used in sign erection, work zone hauling and transport, and utility company vehicles in a work zone.
15. Supplement flashing strobe lights that warn of school bus in transit between bus stops.
16. School bus unloading students with stop sign deployed, stop lights flashing, strobe activated.
17. Oversize load transport vehicle equipped with SWS. 18. Transport's pilot vehicle equipped with SWS. 19. Warn of slow moving oversized farm machinery. 20. SWS transmits warning when any forward motion of
locomotive sensed.

7

1.4.2 Fixed Site SWS Transmitter Message Set
The prototype fixed site transmitter was designed to be programmed by the user to display anyone of the 60 predefined fixed text messages according to the need. The 60 warning messages from which to select contain texts that are applicable to both rural and urban safety warning applications. The messages that can be transmitted by the fixed site SWS transmitters include both the mobile application messages, as shown in Table I, and the 24 messages that relate to the application of the fixed site transmitter.

1.4.3 Applications Using Fixed Text Messages
A total of 24 fixed site SWS transmitter applications have been developed, using applicable selections from the warning message database presented in Appendix A. Each of the 24 applications has been formalized and is presented in Table 2.

Table 2. Applicationsfor Roadside SWS Transmitter

Roadside Message Transmitted
Work Zone Ahead
Highway Work Crews Ahead Road Closed Ahead/Follow Detour Bridge Closed Ahead/Follow Detour All Traffic Follow Detour Ahead All Trucks Follow Detour Ahead
All Traffic Exit Ahead
Right Lane Closed Ahead Center Lane Closed Ahead Left Lane Closed Ahead
Stationary Police Vehicle Train Approaching

Roadside SWS Warning Application
1. The SWS would supply advance warning of a work zone well before a motorist would receive a warning signal from the SWS transmitters used on work zone vehicles (see mobile applications).
2. Warning close to work zone for times when work zone vehicles without personnel are present (working nearby).
3. The SWS would alert motorists to detour. The transmitter could be set up on a short or long term basis by a DOT.
4. Separate trucks and cars in situations where detour route used for cars is not suitable for trucks. Same deployment as 3 above.
5. Marker to advise of end of roadway. This message would serve as a final notice after the detour notice (application 3) was broadcast. SWS transmitter would be deployed by DOT on temporary or permanent basis in advance of roadway end point.
6. Notification in advance of closed lane to start merger to open lane(s) early. The SWS transmitter could be activated by an Automated Traffic Management System (ATMS) center or deployed by a DOT for temporary situations. Utility companies could also use this SWS deployment to advantage.
7. (Used in mobile application only.)
8. An SWS roadside transmitter would be mounted and connected to a rail crossing warning light bar to activate when crossing system activates.

8

Table 2. Applicationsfor Roadside SWS Transmitter (Continued)

Roadside Message Transmitted' Low Overpass Ahead
Draw Bridge Up Observe Bridge Weight Limit
Rock Slide Area Ahead School Zone Ahead Road Narrows Ahead Sharp Curve Ahead Pedestrian Crossing Ahead Deer/Moose Crossing Ahead Blind/Deaf Child Ahead Steep Grade AheadlTruck Use Low Gear Poor Road Surface Ahead No Passing Zone Dangerous Intersection Ahead High Wind Ahead
Severe Weather Ahead
Heavy Fog Ahead

Roadside SWS Warning Application
9. The SWS transmitter would be used in advance of a low overpass. It would be used to provide a general warning to call a high load driver's attention to the situation or the SWS could be triggered when an overheight vehicle sensor alarm was activated by an overheight vehicle.
10. The SWS roadside transmitter would be connected to a bridge traffic gate and light system to send warning anytime that bridge is up and highway gates are down.
11. Transmitter deployed in advance of bridge with signing showing weight limit for bridge. If an in motion weight system were installed, warning would only be issued by the SWS transmitter when a vehicle exceeding the load limit was detected.
12. A portable transmitter would be deployed by police or DOT when a rock slide has occurred or conditions were right for a rock slide to occur.
13. Transmitter mounted on "school zone when flashing" sign and activated when sign is powered as reminder that special school zone speed limits are in effect.
14. These nine messages are grouped under a category of informational types of driver notifications. It is proposed that a fixed site portable SWS transmitter be used for these type messages, only after there is a problem that traditional highway signing has not been successful in solving.
15. An SWS transmitter would be connected to an anemometer on roadways where high winds were seasonal and may have caused accidents or blown large 18 wheel trucks off of the roadway. Warning message would only be transmitted when wind exceeded a velocity threshold, as measured by the anemometer.
16. Application would be in an ATMS control area. The transmitter nearest the severe weather in each direction of travel would be activated by the ATMS so that the severe weather would always be between the pair of transmitters. This SWS function might provide the only alert to the driver that visibility was limited by heavy rain, snow, or other weather induced obscurant. The SWS could be mounted on a variable message sign to warn specifically of tornadoes or hail obscured in the rain ahead. This would provide a motorist a warning before driving into a life threatening situation. The ATMS would use the National Weather Service's emergency broadcast.
17. The SWS can be used anywhere there is a recurring fog problem and fog sensors are installed. Fog sensors would be used to activate the SWS transmitter when the fog was thick enough to impair a driver's vision. An SWS transmitter could also be activated when other signing is activated.

9

Table 2. Applications/or Roadside SWS Transmitter (Concluded)

Roadside Message Transmitted

..
Roadside SWS Warning Application

High Water/Flooding Ahead Blowing Dust Ahead Blowing Sand Ahead Blowing Snow White Out Ahead Ice on Bridge Ahead Ice on Roadway Ahead
Rest Area Ahead Rest Area With Service Ahead 24 Hour Fuel Service Ahead Inspection Station Open Pay Toll Ahead
Congestion Ahead/Expect Delay Expect 10 Minute Delay Expect 20 Minute Delay Expect 30 Minute Delay Expect 1 Hour Delay

18. Each of these hazard warning messages are thought to be of a temporary nature. The SWS transmitter would be deployed on a temporary basis when needed by police agency or a DOT.
19. Both ice conditions can be sensed with a pavement mounted ice sensor system. The SWS transmitter would be activated when the ice sensor determined that bridge or roadway icing was occurring. Police and DOT units could also deploy the SWS roadside transmitter on a temporary basis when icing was determined to exist by the human observer.
20. The SWS transmitter would back up signs that are normally used to alert the motorist to this group of facilities. A facility may be closed from time to time. The SWS would provide primary confirmation that a facility is open while there is time for a motorist to begin planning an exit strategy. The SWS transmitter would be activated by the human attendant upon his or her opening a facility and turned off by an attendant the same way.
21. This group of SWS messages would be activated by an ATMS to advise motorists equipped with an SWS receiver that there are delays ahead. The SWS could be operated in stand-alone mode or in conjunction with an overhead sign to provide this information.

Traffic AlertlTune AM Radio

22. This message would be activated by an ATMS. The SWS transmitter could be used to give a nondirectional AM radio broadcast some specificity by activation of an SWS transmitter to alert those drivers in the affected lanes that a specific message is being directed to them.

Trucks Exit Right Trucks Exit Left

23. These messages would be used when it was required that trucks be separated from the normal traffic mix. The ATMS could control the SWS transmitter.

Emergency Vehicle in Transit

24. Many communities place strobe lights and warning signs outside of fire stations. These signs can be activated prior to the departure of a fire engine or ambulance. An SWS programmed to warn of an emergency vehicle's departure from a firehouse or hospital would be another use of this message that is normally used for a mobile application.

1.5 NECESSITY OF EVALUATING THE SWS
When the SWS applications shown in Table I and Table 2 were conceived, no testing of the SWS had been performed to determine how well the SWS might perform in any application. A test plan was designed to evaluate the SWS in as many rural, urban and interstate highway scenarios as possible given project time and funding for field testing. Next, the SWS would be field tested according to the test plan and the results would be analyzed. Test data analysis, coupled with assumptions regarding perception times and braking distances, would determine the suitability of the SWS for each application. It was

10

determined that worst case SWS performance data would be used in the evaluation to ensure that the applications could be performed under worst case conditions. The preparation of the field test plan early in the program also allowed the GDOT Office of Traffic Operations to determine any safety considerations for the field testing.
1.5.1 Test Plan Development
A test plan was developed and is presented in its entirety in Volume II, Appendix B of this report. The test plan includes a variety of tests that were devised to evaluate the SWS system in the laboratory and also determine the reception range at which the SWS warning message would be displayed under varying street and highway conditions. The varying conditions included highway geometries, transmitter/receiver vehicle speeds, varying topographical features, and both urban and rural settings. The test plan was also designed so that one or more planned tests could be modified or canceled either during testing or prior to testing if necessary without leaving a void in the database. Likewise, there was nothing included in the test plan that prohibited additional tests from being conducted if they were found necessary. It was deemed necessary to conduct two series of tests that were not anticipated when the test plan was prepared: (I) testing at Dobbins Reserve Air Force Base, Georgia, and (2) turntable testing at GTRI. Also, several of the test procedures were modified on the basis of the findings determined during testing.
1.6 TESTING ACTIVITIES CONDUCTED TO EVALUATE THE SWS
Laboratory and field testing of the SWS was conducted from March 1997 through September 1997. The data from a field test could usually be collected in one day, but the large amount of data that a single test produced required several weeks to reduce, analyze, and document. The data collected during field testing included: (I) relative signal strength as a function of distance from the transmitter, (2) Global Position System (GPS) navigation data defining the location of the transmitter and receiver vehicles, (3) video scene data from the receiver vehicle, and (4) a hand log record. Before each test was performed, the equipment was tested in the laboratory set up described in Appendix D, Volume II.
1.6.1 Laboratory Testing of Equipment
Laboratory testing of the receivers and transmitters used in each field test was conducted before the equipment was taken to the field. An off-the-shelf BEL-TRONICS LIMITED0 855 STi radar receiver, shown previously in Figure 2, was used during selected tests to confirm that the message being transmitted by the fixed site or mobile transmitter was the message received and displayed in accordance
II

with the requirements of the statement of work. The BEL-TRONICS LIMITEne off-the-shelf receiver had an alphanumeric light emitting diode (LED) display, as well as a voice synthesizer that would announce the warning message being received.
A modified BEL-TRONICS LIMITEne 855 STi, which had the capability to output relative signal strength data via an RS-232 link to a computer, was used as the example SWS receivers during testing. The BEL 855 STi, with the RS-232 link to a laptop computer, was calibrated before each test using the laboratory test set-up described in Appendix B to determine the linearity of the detector's response to the transmitted signal and ensure that the test receiver was working properly. It was calibrated each time over the five-month period when testing was being conducted, and the calibration results were found to be very stable over time. Appendix D, Volume II, offers an in-depth discussion of the calibration procedures used and the results obtained.
The mobile transmitter and the fixed site transmitter were also tested in the laboratory before being taken to the field test sites to ensure that they were operational, to ensure that they were transmitting the messages that they were programmed to transmit, and to confirm that the transmitted power levels were as specified.
1.6.2 Dobbins Reserve Air Force Base Testing
Before conducting highway testing, it was found necessary to conduct field testing on an 1,800 foot part of an unused runway at Dobbins Air Force Reserve Base, Georgia to collect SWS baseline performance data. This Dobbins baseline data collection effort provided data that was not biased or corrupted by the effects of topography, shadow zones from buildings, and other variables over which GTRI researchers would have no control during highway testing. The test site runway was approximately 150 feet wide, it was level, and line of sight could be maintained between test vehicles during the tests. There were few objects in the vicinity to cause reflections of the transmitter energy, and there were no vehicles to block the transmission of the SWS signal (as experienced during highway testing). The detailed results of Dobbins testing are presented in Appendix E, Volume II.
Table 3 is a summary of Dobbins test results. The data presented in the first column of Table 3 provide a brief description of the tests that were conducted. The third column of Table 3 shows the distance between the transmitter and receiver vehicle when the receiver first alerted the driver with a first alert warning tone or audible alarm signifying there was an SWS transmitter within reception range. The
12

information shown in the fourth column is the range at which the receiver displayed the warning message on the alphanumeric display.

Table 3. Distance at Which the Two Levels of SWS Warnings Were Presented to the Driver During Dobbins Testing

Type of Test

..
Test No. ...

First Alert ..... Warning

Display Warning Message

Receiver overtakes stationary mobile transmitter

(5)

1,310feet 1

1,130 feet 1

Mobile Transmitter overtakes stationary receiver

(8)

1,275 feet 1

1,275 feet 1

Receiver crosses path of mobile transmitte~

(10)

550 feet 1

324 feet

Receiver crosses path of mobile transmitter3 Mobile transmitter vehicle at bottom of a hill
Footnotes:

(12)

811 feet 1

758 feet

(17)

1,510feet 1

1,180 feet

1. This number does not represent the maximum range that can be achieved, instead it is the maximum range that existed between test vehicles at any given time.
2. The path of receiver vehicle was off-set approximately 105 feet in front of transmitter vehicle. 3. The path of receiver vehicle was off-set approximately 5 feet in front of transmitter vehicle.

In summary, the analysis of the data collected during runway testing at Dobbins indicated that the trends observed later during turntable testing (in the section that follows) were valid. The best SWS performance was obtained when the antennas of the receiver and transmitter were pointed at each other. This finding included any scenario where the receiver and transmitter vehicle would be approaching each other, head on, or the receiver vehicle was overtaking the transmitter vehicle from the rear (transmitter vehicle has front and rear pointing antennas). This same trend applies to the case where the fixed site transmitter's antenna is pointed at the receiver's antenna.

The data in Appendix E, Volume II, also indicate that the next best SWS performance occurred when the transmitter vehicle was overtaking the receiver vehicle such as would occur when a police or emergency vehicle was overtaking a motorist. It was found during turntable testing that the SWS receiver used in the test had a high back lobe in the antenna pattern that allowed moderately good reception of SWS signals from the rear. The angular extent of the back lobe response was measured during turntable testing and found to be surprisingly broad. Unfortunately, schedule and budgeting constraints prevented testing of other manufacturer's SWS receiver antenna patterns to confirm that they would also respond to signals from the rear due to the high backlobe in the antenna pattern.

The Dobbins tests also indicated that the next best operational conditions, as reflected by the signal detection range at which the SWS could be received, was when the transmitter vehicle approached the receiver vehicle at a crossing angle of 90 degrees. Two offset distances were used and, as noted in Table 3, the closer to the transmitter vehicle that the receiver vehicle crossed, (i.e., minimum offset distance)

13

the greater the range at which the message could be displayed. This effect was also predicted by the turntable tests. The turntable test data showed that the receiver and transmitter antennas both respond poorly to a' signal from an angle around 90 degrees off main axis. This finding was expected and is totally consistent with the theory of signal propagation that applies to directional antennas.
Testing for the effects of line of sight impairment due to hills was also conducted at Dobbins. The transmitter vehicle was located at the bottom of a hill deep enough so that a line of sight could not be established between test vehicles. When the test vehicles were approximately 1,180 feet apart, the signal strength exceeded the threshold required to cause the warning message to be displayed. Later hill testing at another location showed similar results.
In summary, the analysis of the data collected during runway testing at Dobbins indicated established performance trends that were observed later during turntable testing and during other field tests. The best SWS performance was obtained when the antennas of the receiver and transmitter were pointed at each other. This finding included any scenario where the receiver and transmitter vehicle would be approaching each other, head on, or the receiver vehicle was overtaking the transmitter vehicle from the rear (transmitter vehicle has front and rear pointing antenna). This same trend applies to the case where the fixed site transmitter's antenna is pointed at the receiver's antenna.
1.6.3 Turntable Testing of Receiver Antenna Response
After the test plan shown in Appendix B, Volume II, and Dobbins tests were completed, it was determined that an additional test was needed to determine the approximate antenna response pattern of both the SWS receiver and the mobile transmitter when they were mounted on or inside the test vehicles. These tests were completed using the GTRI turntable facility to rotate the host vehicle while recording relative signal strength as a function of rotation angle. When the SWS receiver is mounted in the test vehicle, the response of the antenna to the transmitter signal is not the same at all angles around the vehicle. Turntable testing allowed the approximate response pattern of both the receiver and transmitter antennas to be measured as a function of angle around the host vehicle. The technique used to make the measurement and the results obtained are presented in considerable detail in Appendix C, Volume II.
The Georgia Tech Research Institute owns a turntable large enough to put an automobile on and rotate it 360 degrees. This turntable facility includes enough property to allow an SWS detector to be mounted in a test vehicle and the SWS mobile transmitter to be located 200 feet away in the far field region. The receiver vehicle was placed on the turntable and rotated around 360 degrees while the signal strength of
14

the transmitter 200 feet away was recorded as a function of rotation angle. Next the transmitter vehicle was placed on the turntable and rotated while the receiver was located 200 feet away recording relative signal strength data as a function of transmitter vehicle rotation angle. Turntable testing provided a measure of relative received signal' strength as a function of aspect angle around the receiver or transmitter vehicle (depending on which is being rotated).
Referring to Appendix C, Volume II, the results of turntable testing indicated that best performance of the SWS would be obtained when the transmitter and receiver antennas were pointing at each other. Turntable testing further indicated that the next best performance would be obtained when the transmitter antenna was pointing at the back of the receiver. Worst case performance would be experienced when the receiver or transmitter vehicle was perpendicular to the other vehicle. These trends indicated by turntable testing were also verified during field testing.
1.6.4 Rural Two Lane Road Testing
Procedures outlined in the test plan were used to test the effectiveness of the mobile transmitter when tested on a straight and curved section of rural two-lane highway, and at the bottom of a hill of a rural two lane highway. The fixed site transmitter was also tested on a straight section of a rural two-lane highway. The detailed test results are presented in Appendix F, Volume II.
The data presented in the first column of Table 4 provide a brief description of the test that was conducted. The third column of Table 4 shows the distance between the transmitter and receiver vehicle when the receiver first alerted the driver with a warning tone or audible alarm that there was an SWS transmitter within reception range. The information shown in the fourth column is the range at which the receiver displays the warning message on the alphanumeric display.
In summary, during testing on the straight part of a two lane rural highway, it was found that the relative signal strength of both the fixed site and the mobile transmitter exceeded the received signal strength threshold level of 20, which is required for three seconds to cause the warning message to be displayed in excess of 3, I00 feet, the approximate distance over which most of these tests were conducted. The fixed site transmitter was observed to produce a higher signal level than the mobile transmitter. This was expected because all power from the fixed site transmitter is directed forward, while the mobile transmitter antenna directs half of the transmitted power forward and half the power transmitted in a rearward direction.
J5

Table 4. Distance at Which the Two Levels of SWS Warnings Were Presented to the Driver During Rural Testing

Type otTest

Test No.

First Alert Warning

Display Warning Message

Receiver approaches stationary mobile transmitter on straiaht stretch of rural two lane road

(8-1 )

3,100 feet 1

3,100 feet 1

Receiver again approaches stationary mobile transmitter on straiQht stretch of rural two lane road

(8-2)

3,400 feet 1

3,400 feet 2

Receiver approaches stationary mobile transmitter on straight stretch of rural two lane road

(8-3)

3,150 feet 1

1,700 feet 3

Fixed site transmitter is substituted for mobile transmitter, transmitter door is open
Fixed site transmitter is substituted for mobile transmitter, transmitter door is closed Mobile transmitter located in midpoint of curve Mobile transmitter located in midooint of curve
Mobile transmitter located at bottom of hill
Mobile transmitter located at bottom of hill Footnotes:

(8-4)
(8-5)
(C-n (C-2) (H-1 ) (H-2)

2,700 feet 1
3,400 feet 1
1,022 feet 3 1,022 feet 3 3,168 feet 1 2,174 feet

2,700 feet 2
3,400 feet 2
642 feet 4,5 564 feet 4,5 930 feet 4,6 858 feet 4,6

1. This number does not represent the maximum range that can be achieved, instead it is the maximum range that existed between test vehicles when receiver was first activated.
2. Receiver displayed warning message when test at limited range began. 3. Line of sight between receiver vehicle and transmitter vehicle blocked by pick-up truck. 4. Adjusted for additional 3 seconds of delay required to activate display. 5. Line of sight from trial starting point to mobile transmitter is blocked by heavy leaf foliage. 6. Line of sight from trial starting point to mobile transmitter is blocked by crest of the hill.

Signal propagation effects were observed during all of the rural tests. The expected multipath peaks and nulls in relative signal level were observed to occur periodically. During one trial a pick-up truck pulled in front of the test vehicle and a spacing of approximately 50 to 70 feet was maintained between vehicles. The line of sight to the transmitter was blocked by the truck for most of the trial. This blockage caused the signal from the transmitter to drop below the levels observed when there was no blockage; however, even with the effects of signal blockage present, the warning time provided by the SWS was adequate to allow the user to make a full stop, had a SWS equipped police or emergency vehicle been stopped in the roadway ahead.

During curve testing, the effects of heavy foliage on the SWS signal were observed. The transmitter vehicle was located 10 feet off of the highway at the "midpoint" of the curve at a break in a line of trees and shrubs. The broadleaf and pine foliage on the trees extended from the ground to a height of approximately 30 feet. The foliage reduced the line of sight between the transmitter vehicle and approaching traffic to a distance of less than 382 feet. The foliage was thought to have attenuated the signal of the transmitter vehicle to the point that the receiver vehicle had to be within a range of between 840 to)62 feet of the mobile transmitter before the relative signal strength was high enough to initiate the warning message display process. The receiver required three seconds to process the signal, extract

16

the message text from memory and display the message. Given the three second delay, the warning message would have been displayed at a range of between 642 to 564 feet. The first alert occurred slightly over 1,000 feet in advance of the transmitter vehicle.
During Hill-l testing, it was found that the SWS first alert warning occurred at the point where the tests started, approximately 3,168 feet from the transmitter vehicle's location at the bottom of the hill. However, the relative signal strength did not exceed the message display threshold of 20 until the receiver vehicle was approximately 1,128 feet from the transmitter vehicle at the bottom of the hill. The receiver vehicle would have been at a range of 930 feet from the transmitter vehicle after the three second delay required for display activation.
During the HillA test, the direction driven was reversed. The receiver vehicle reversed its approach to the transmitter by making an east to west approach (previously west to east). There was a hill further east of the test hill which was used as the test hill during the HillA test. The transmitter vehicle remained in the valley formed by the two hills. The receiver vehicle approached on an east to west path. The first alert warning was received when the range between the receiver and transmitter vehicle was 2,174 feet. The relative signal strength exceeded the display threshold of 20 when the receiver was at a range of approximately 1,056 feet from the transmitter vehicle. The receiver vehicle was at a range of 858 feet from the transmitter vehicle after the three second delay required for display activation.
1.6.5 Urban Intersection Testing
An urban test site was located in downtown Marietta, Georgia. The site layout, test set-up and detailed urban test results are described in detail in Appendix G, Volume II. The urban tests were conducted to determine the range at which the warning message could be received when the receiver vehicle was approaching an intersection and the transmitter vehicle was on the cross street at one of two locations: one block (160 feet) from the intersection and 60 feet from the intersection. Table 5 was developed to summarize the results of the urban tests.
In summary, the data collected during urban testing were very repeatable from trial to trial. A change of one lane width toward the side of the street that the transmitter vehicle was on did not affect the test results. The movement of the transmitter from a location one block (160 feet) east of the test intersection to 60 feet east of the test intersection did not make a difference in the range at which the first alert signal was received during the previous test (U-3). The short warning ranges that were achieved during urban testing were thought to be due to the blockage of most of the signal by the buildings that extended almost
17

to the street where the testing was conducted. Very little of the transmitted signal was reflected off of the fronts of buildings at the comer at an angle of 90 degrees toward the test vehicle.

Table 5. Distance at Which the Two Levels of SWS Warnings Were Presented to the Driver During Urban Testing

Type of Test

First Alert Warning

Display Warning Message

Receiver approaches test intersection and transmitter vehicle is on cross

street 1 block away

(Test No. U-1)

Receiver approaches test intersection and transmitter vehicle is on cross

street 1 block awav

(Test No. U-2)

Receiver approaches test intersection and transmitter vehicle is on cross

street 2 blocks away

(Test No. U-3)

Receiver approaches test intersection and transmitter vehicle is on cross

street 2 blocks away

(Test No. U-4)

Footnotes:

210 feet 1 210 feet 2 500 feet 3 500 feet 3,4

180 feet 1 180 feet 2 180 feet 3 180 feet 3,4

1. Receiver vehicle approaches in right lane of Powder Springs Street. 2. Receiver vehicle approaches in left lane of Powder Springs Street closer to mobile transmitter. 3. Trial starts one block further north of intersection of Roswell and Powder Springs Streets. 4. Transmitter moved from to within 60 feet of Powder Springs Street intersection from the previous location one block away.

1.6.6 Interstate Highway Testing
A detailed account of the tests that were conducted and the results of those tests can be found in Appendix H, Volume II, of this report. Some of the tests were modified from the plan given in Appendix B and the reasons for the modification to the test plan are also presented in Appendix H, Volume II.
For example, a fixed site transmitter was not tested at the side of the interstate as initially planned. No test site within 60 miles of Atlanta on Interstate-75 could be found that provided a 2 mile straight stretch without intervening hills and curves so that line of sight could be maintained during the entire test. The test personnel had recently conducted testing of the fixed site transmitter on a rural roadway with a line of sight distance of over 3,000 feet. It was thought that the data produced during these tests were equivalent to any data that would have been produced during interstate testing.
The first three interstate tests (0-1 through 0-3) conducted were designed to (I) test the range at which the first alert provided by an SWS mobile transmitter could be detected, and (2) determine the range at which warning message would be displayed when the transmitter was overtaking the receiver vehicle. Data from Table 6 show that the average range at which an SWS equipped motorist would receive the warning message from an overtaking police or emergency vehicle would be approximately 1,902 feet. The effects of surrounding traffic, highway geometry (curves preventing maintenance of line of sight), and topography (hills preventing maintenance of line of sight) are thought to be responsible for the differences in the results achieved between the trials.

18

The fourth interstate test (0-4) was conducted in manner similar to the method used in the first three overtaking tests. However, the fourth test had two possible outcomes, as shown in Table 6, and was not included in the average detection distance presented in the previous paragraph. The warning message was displayed briefly at a range of 2,200 feet; however, the display reverted to a first alert before again displaying the warning message seconds later at a range of 1,650 feet. Once the warning message was displayed at the range of 1,650 feet, the warning message was displayed until the conclusion of the trial. Traffic effects, highway topography, and geometry were possible reasons for the reduced performance distance.

Table 6. Distance at Which the Two Levels of SWS Warnings Were Presented to the Driver During Interstate Testing

Type of Test

Test No.

First Alert Warning

Transmitter overtakes the receiver vehicle
Transmitter overtakes the receiver vehicle
Transmitter overtakes the receiver vehicle
Transmitter overtakes the receiver vehicle
Transmitter overtakes the receiver vehicle
Receiver vehicle overtakes transmitter vehicle
Receiver approaches fixed-site transmitter mounted over the interstate
Receiver approaches fixed-site transmitter mounted over the interstate
Receiver approaches fixed-site transmitter mounted over the interstate
Receiver vehicle drives away from fixed-site transmitter on interstate in opposite lane
Receiver vehicle drives away from fixed-site transmitter on interstate in opposite lane
Footnotes:

(0-1) (0-2) (0-3) (0-4) 1 (0_4)2 (0-5) (OV-1)
(OV-2)
(OV-3)
(0-1)
(0-2)

> 2,244 feet 2,567 feet
> 2.500 feet > 2,500 feet > 2,000 feet > 4,000 feet > 4,500 feet
> 4,500 feet
> 4,500 feet
N/A 4
N/A 4

1. Distance when message was first displayed for brief time. 2. Distance when message was displayed for total time until receiver vehicle passed by transmitter. 3. Range achieved before the transmitter vehicle rounded curve and line of sight was lost. 4. Test started at transmitter with warning message displayed/firstalert distance not applicable.

Display Warning Message
1,980 feet 1.877 feet 1,848 feet 2,200 feet 1,650 feet 2,024 feet 3 4,212 feet
3,645 feet
3,608 feet
6,300 feet
4,860 feet

The fifth interstate test (0-5) was performed to determine the range at which the receiver vehicle would display the warning message from the transmitter vehicle when the transmitter vehicle was in front of and being overtaken by the receiver vehicle. A separation range of 2,024 feet was achieved before the tests were interrupted when the transmitter vehicle went around a curve and the signal from the transmitter vehicle was lost.

It was planned that three tests would be conducted with the fixed site transmitter located on an overpass over the three southbound lanes of the interstate for the purpose of determining if a meaningful difference in signal strength would be observed if the transmitter was moved across the three active

19

lanes, one lane at a time. The test plan showed that the transmitter would be moved from the center of the right southbound lane to the center of the center southbound lane and then to the center of the left southbound lane after a sequence of three trials were conducted. The receiver vehicle would approach the fixed site transmitter first in the right lane, then the center lane, and, finally, in the left lane, for a total of three trials before the transmitter would be moved to a new position each time. The original test matrix included a total of nine tests to be conducted to determine the sensitivity of relative signal strength to transmitter location.
The test plan was modified before the overhead tests were conducted because it was found that the overhead tests, as designed (before the experience of testing), would not provide useful information beyond that already collected. The reasoning for the change included the fact that the fixed site transmitter's 3 dB (half-power) beamwidth was approximately 20 degrees in the azimuthal plane and at 1,000 feet would spread over approximately 364 feet, which is a much wider distance than the baseline over which the transmitter would be moved across the three active southbound lanes. This was not the only factor that caused the design of the overhead tests to be reconsidered. Traffic, topography, and geometrical effects had all been observed to bias SWS test results. Thus, the overhead tests, as designed, became a task of trying to determine small scale effects when there was more than one variable that could cause a large effect on the results of the trial.
Given this logic, it was decided that the previous test plan requiring the transmitter to be moved across the three lanes would be abandoned and the transmitter would be located over the center lane for the entire test. Three trials would be conducted to further quantify the effects of the surrounding traffic on SWS signal propagation. Table 6 shows the average range at which the SWS fixed site transmitter's signal prompted the receiver to issue a first alert and warning message to the driver. The warning message was displayed at a distance of approximately 3,822 feet when the results from all three trials were averaged. This distance corresponded to the point where the receiver vehicle came around a slight curve and over a slight rise in the highway and just before establishing a line of sight with the fixed site transmitter.
The overhead tests did not establish the absolute maximum range at which the SWS warning message could be displayed, due to a limitation on the line of sight distance imposed by a curve in the highway. They did show that at the maximum speed limit of 70 miles per hour, an SWS-equipped driver would be provided a warning time of 37 seconds in advance of a hazard given the existing traffic conditions and roadway geometry during testing.
20

The tests did accomplish the secondary goal of showing that traffic affects the strength of the SWS received signal. The reader is referred to Appendix H, Volume II, to the plots showing relative signal strength as a function of distance traveled by the receiver vehicle, The received relative signal strength plots all show the same general trend of increasing as the distance between the receiver vehicle and fixed site transmitter decreases. These plots also show the effects that surrounding traffic has on signal strength of the SWS signal at any given point in time during a trial. Laboratory tests demonstrated that once the SWS warning message is displayed, the signal strength can drop below the display threshold of 20 for up to 10 seconds before the warning message is removed from the display. Thus, the short term traffic and highway geometry effects observed in the SWS signal strength data are not apparent to the SWS user.
1.6.7 Directional Selectivity Test
The Directional Selectivity Test was not planned but was conducted during the period of time that the fixed site transmitter was located on the overpass over the three southbound lanes. The receiver vehicle traveled a reverse track from south to north (north bound lanes) the overpass while the fixed site antenna was pointed north. The purpose of the two directional trials was to determine if motorists using the lanes in the direction away from the fixed site transmitter could receive the SWS signal and if so, how far. The results of the two-directional tests (D-I and D-2) are shown in Table 6 (previously presented). The results achieved were not expected, and additional analysis was conducted to determine why the message was displayed 6,300 and 4,860 feet respectively, beyond the transmitter.
A study of the test site photographs and video tape record showed that line of sight could be maintained between the fixed site transmitter and the receiver vehicle in the northb::mnd lane much further than line of sight existed when the receiver vehicle was in the southbound lane. Also, it was determined that the reflective back doors of the tractor trailer trucks in front of the receiver vehicle extended vertically, at least 13 feet above the pavement and continued to reflect the SWS signal even after the receiver (at an elevation above the highway of approximately 4.5 feet) had dropped below the local horizon. This finding requires that guidance regarding fixed site transmitter placement be provided to users who wish to generate selective directional warnings to only oncoming traffic. Also, a narrow beam SWS antenna must be developed for SWS applications that require directional warnings.
21

1.6.8 Summary ot Test Results
The test results summarized in the previous paragraphs are used in the SWS performance evaluations to be presented in a section that follows. It is suggested that the reader review the appendices that show the test results in detail for a better understanding of how each test was conducted, the detailed results that were achieved during each trial, and the non-repeatable effects caused by the road geometry, surrounding traffic, and other highly specific conditions.
1.7 INSTALLATION, OPERATION AND MAINTENANCE OF SWS COMPONENTS
1.7.1 Receiver
The SWS receiver, previously shown in Figure 2, is presently an "after market" accessory purchased by a motorist from a retail supplier or through a wholesale catalog outlet. The industry plans to offer two types of receivers: (I) a unit with the standard police radar detector and SWS functions built into a single unit, and (2) a unit with only the SWS function and no radar detector function. The SWS only unit without a radar detector capability could be used in the classes of trucks that are currently prohibited from using radar detectors. The installation takes less than one minute to mount the receiver on the windshield using suction cups. It operates on 12 volts DC, and plugs into a vehicle's cigarette lighter outlet. Several models are battery powered. The receiver is designed to be mounted on a vehicle's windshield approximately 4 to 6 inches above the dashboard.
The features that are offered on an SWS receiver are determined by the price paid. For example, the BEL 855 STi that was used in the tests offers the option of turning on or off the SWS feature, controlling the volume of the first alert signal, and the option of disabling the voice synthesizer by using a combination of push button switches. All communications with the driver occur by messaging on an alphanumeric display, an announcement from the built in synthesizer, or by alert tones, which are different for each alert status.
The least expensive receivers use alert tones and illumination patterns of light emitting diodes (LEOs) to communicate with the driver. The low cost receivers may use a pattern of LEOs to communicate the type of SWS warning to the driver. For example, one orange LED might indicate that the warning is from an SWS transmitter, and the lighting of two adjacent red LEOs may indicate that the source of danger is a stationary hazard. A moving hazard might be signaled by the orange LED indicating that the warning is an SWS transmitter, and one red LED may further indicate that the source of danger is a moving hazard.
22

The advertised retail price for the top of the line BEL 855 STi that was used in the tests is approximately $280. However, some catalog outlets sella variant of this top end. detector for $150. One member of the industry has estimated that an SWS only detector could be developed to sell for less than $100 retail. Several prototype SWS receivers without the radar detector function have been developed. Once engineered for production, the "warning only" receiver would be more sensitive than the SWS with police radar detector, because the SWS receiver would cover only K-band. Single band coverage allows the sensitivity to be optimized. Also, the time required to find and decode the SWS message would be shorter after optimization.
Maintenance is not required on an SWS receiver. There are no user repairable parts in the SWS receiver. If the receiver fails and it is under warranty, it will be repaired or replaced for a nominal charge.
1.7.2 Mobile Transmitter
The mobile transmitter, described in a Section 1.2.1 and shown in Figure 1, was developed by MPH, Inc., a leading manufacturer of police speed radar equipment, under the license of SWS, L.c., the developers o! the SWS. The two warning messages that the mobile transmitter is capable of transmitting are programmed by the factory. The choice of the texts of the messages to be transmitted is determined by the type of service for which the transmitter is designed, as defined in a previous section.
Installation of the mobile transmitter is not difficult on police and emergency vehicles with an existing emergency light bar. It is assumed by the manufacturer that the agency or individual purchasing the transmitter has access to an equipment installation shop such as a city-operated radio shop responsible for installing radios, roof mounted emergency light bars, and other equipment on the city owned veh,icles. The mobile unit installation instructions provide guidance regarding how to install the transmitter on the emergency light bar for best signal radiation characteristics. MPH, Inc. has also discussed providing the installation shops with a test set to use during final system check out. The test set would offer the installer a way to ensure that the power being transmitted was within the FCC limitations, and would stimulate the transmitter, once activated, to transmit the two warning messages and also monitor and display the transmitted message to the installer.
The mobile transmitter installed as described will be activated any time the host vehicle's emergency lights are activated. It is also possible to activate the mobile transmitter without wiring it into the light bar circuit. Then, the transmitter can be on a separate switch under vehicle driver control. The text of the warning message is selected automatically by the mobile transmitter's associated radar unit, which
23

determines if the host vehicle is in motion or stopped. The mobile transmitter stops transmission when power is removed from the emergency light bar and when the lights are turned off. The automated message selection and operation relieves the host vehicle's driver from the requirement that he or she remember to tum on or off the mobile transmitter or select the proper message to be transmitted.
There are no user serviceable parts in the SWS mobile transmitter. However, the operating agency or individual should test the transmitter periodically to ensure that the proper message is still being transmitted and that the power output continues to be within tolerance.
1.7.3 Fixed Site Transmitter
No manufacturer is presently under license from the SWSLC to build the fixed site transmitter. A prototype unit is shown in Figure 3. However, it is assumed that most of the features of the prototype unit will be included in the design when a manufacturer is selected and the prototype design is converted to a production model.
The prototype is capable of being programmed to transmit anyone of the existing 60 of 64 possible fixed text safety warning messages. A full listing of the fixed text messages is found in Appendix A, Volume II, of this report. The fixed site transmitter is designed to be mounted on a support beside or over a highway. There are four modes of fixed site transmitter operation: (I) a timer mode, which allows it to be programmed to be turned on and off up to four times during any 24 hour period; (2) the radar mode, which allows an associated radar function to sample for 0.5 seconds every 3 to 4 seconds to determine if a vehicle is present or exceeding a selected speed limit before transmitting the safety warning message; (3) the continuous mode of operation, which allows transmission of the warning message continuously after activation; and (4) the sensor mode, which allows transmission of the warning message upon activation of an external sensor, such as a fog sensor.
Conceptually, the fixed site transmitter could be carried In a car trunk by traffic engineers, police officers, construction personnel, utility workers, or anyone else authorized to operate the transmitter under the applicable rules of the FCC. Anyone of the 60 fixed text warning messages could be programmed into the transmitter's message memory circuit using a handheld programmer. The system could be installed on a temporary tripod, a vehicle, an overhead sign, or any other structure that could serve as a safe support system. The prototype unit is housed in a waterproof container and can operate up to approximately 15 hours on a self contained 7 amp hour battery before charging is required. The use of solar cells or an alternating current (AC) current source to charge the battery could allow 24-hour
24

operation of the transmitter with a potential of 15 hours of battery operation in the case of a cloudy day or a power failure.
It is thought that once the fixed site transmitter is in production there will be no user serviceable parts except a gel-cell battery. If the gel-cell battery is used in the production model, it is recommended that it be inspected every three months for cracks in the plastic case, a condition that will allow leaking of electrolytes to occur. The battery should be charging when not in use and should be replaced annually.
It is possible that the fixed site transmitter could be programmed remotely by an automated traffic management center via telephone or satellite link, in order to allow the text of the transmitted message to be changed in response to real time traffic conditions. It is also possible that the radar function could send data back to the automated traffic management center regarding traffic flow speeds and traffic counts.
1.8 TRANSMITTER TYPES AND PUBLIC SECTOR USERS
At the time that this report was generated an action was before the PCC that would allow the SWS to be operated under Part 90 of the FCC rules by the following services:
I. Local Governments 2. Police Agencies 3. Fire Departments 4. Highway Maintenance 5. Forestry Conservation 6. Railroads
It was requested that the FCC allow railroads to operate the SWS under the provisions of Part 90 to warn approaching motorists of the presence of trains approaching a rail crossing. Another function would allow an SWS to be used for preemption of traffic lights by police and emergency vehicles.
1.8.1 Number of SWS Transmitters Required
The purpose of this section is to present a discussion on the number of mobile and fixed transmitters required to cover a small metropolitan area or sections of an interstate highway. During field testing of
25

the SWS, the maximum range at which the SWS warning message can be received was not determined because a straight and level highway in the north Georgia area that allowed line of sight over at least a two mile path could not be identified. On the basis of test results, it is estimated that when the transmitted signal is directed toward the receiver vehicle and a line of sight exists without intervening traffic, the SWS will receive and display the warning message at a range of at least one and a half miles from the transmitter. It was also determined that when an SWSequipped receiver vehicle is overtaken by a transmitter vehicle, the range is not as great as the approaching case, unless traffic ahead of the SWS equipped vehicle reflects energy back to the receiver vehicle. It was also found that thick foliage on trees that interrupt the line of sight between the receiver and transmitter vehicles may shorten the operating range of the system. It was further determined that when the SWS is operated in an urban environment for the intersection warning function, the range over which the first alert and actual warning message can be very short.
Given these findings from test data, exact numerical estimates regarding the number of fixed site transmitters required to cover a small metropolitan area or interstate can only be made after conducting a site survey in the area of installation. It is reasonable to assume, on the basis of findings during the tests, that each host vehicle having an emergency light mounted on it could enhance its early visibility to those drivers who are equipped with an SWS. Thus, as a starting point every police and emergency vehicle in a community might be SWS equipped.
1.9 EVALUATION OF THE SYSTEM PERFORMANCE UNDER EACH OF THE PREVIOUSLY DEFINED OPERATIONAL SCENARIOS
The purpose of this section is to evaluate the performance of the SWS In each of the operational scenarios previously defined in Tables I and 2. The required actions of the operator of the transmitter device and the driver using the SWS when he or she receives an SWS message will also be defined. The effects of signals intended for other drivers will also be discussed as will the effects of multiple transmitters sending different signals.
1.9.1 Evaluation Criteria Development
One purpose of this study is to evaluate the performance of the SWS as an in-vehicle signing or alerting system. Its design purpose was to first to alert a driver to the presence of an SWS equipped vehicle or a fixed site transmitter, each being associated with a potential highway hazard. Following the precursor first alert provided by the receiver when an SWS transmitter is detected, a more specific warning
26

message is provided to the driver. The warning may be in the form of a displayed or voice synthesized alphanumeric message, the text of which is- found in Appendix A, Volume II. On less expensive receivers, the type hazard to be encountered will be indicated by a combination of LED indicators lit in a meaningful sequence. Drivers equipped with a radar detector without an SWS signal processor and message extractor (older models) will receive the same warning as if a police radar were ahead, an action calculated to raise situational awareness to any highway signing or emergency lights that are associated with the highway hazard.
Another goal of this project is to define applications of the SWS and, through testing, determine if an application is practical and could (if implemented) reduce vehicle accidents. One approach initially planned to achieve this goal was to measure the reaction of an SWS equipped motorist to the SWS warning. The measurement of driver reaction to the SWS warning required that a human factors test program be conducted using a driving simulator. Driving situations would be presented to the test subject while the SWS would present warning messages regarding hazards via the SWS alphanumeric display and/or voice synthesizer. Effectiveness would be determined by recording the time that the test subject takes to perceive the message and react to the message. Warning message effectiveness would also be determined by recording the action (if any) taken by the test subject after the message was perceived.
It was reasoned that if the total time for driver perception/reaction and braking was greater than the maximum time that could be provided by an SWS warning, then the SWS would be judged to fail to meet the requirements for an application.
This research approach to determine SWS effectiveness is greatly simplified. Other factors such as the confidence that a driver has in the SWS to supply the correct information, the other tasking that the driver is involved in when the SWS warning is presented, and the ease of message interpretation were also important factors and were scheduled to be evaluated during driver simulator testing.
GTRI had scheduled to use a driving evaluation system being developed by the FHWA to obtain the human factors data to be used in this study. Resource limitations precluded the collection of the human factors data to determine driver reaction times to the SWS warnings. The lack of driver reaction data from human factors testing requires that the SWS performance be based on literature pertaining. to general ized dri ver performance data.
27

The book Highway Engineering, by Paul H. Wright and Radnor J. Paquett, (fourth edition), published by John Wiley & Sons, New York offers a discussion on driver reaction times. First, it is pointed out that reaction times are variable:
I. The speed of reaction varies from one person to another and from time to time in the same person.
2. Reaction time changes gradually with age, very young and very old people being slower in their reaction.
3. People generally react more quickly to very strong stimuli than to weak ones. 4. Complicated situations take longer to react to than simple ones. 5. A person's physical condition affects his reactions. For example, fatigue and alcohol tends to
lengthen a persons reaction time. 6. DistraGtions increase the time of all reactions except to reflexes.
Highway Engineering also provides the results of a study provided by Johanssen and Rumar of 321 drivers who had some degree of braking expectation. The median brake-reaction time was 0.66 seconds and the range was 0.3 to 2.0 seconds. They also conducted a second study to determine the brake reaction time to a completely unexpected signal. The report stated that when the requirement to brake was not anticipated in advance, the brake-reaction time increased by a factor of 1.35. It was also found that a person responds most quickly to the stimulus of touch and that hearing requires only slightly more response time for response to visual stimuli. Response times for the kinesthetic and vestibular stimuli are considerably longer. It was recommended that, for the purposes of computing stopping sight distances, the use of a combined perception-reaction and brake reaction time of 2.5 seconds be used.
The first alert of the SWS cues the driver that a signal has been detected, always well in advance of the transmitter location and usually at least three seconds before the warning message text is displayed. Thus, the SWStext based warning message is not unexpected after the first alert. As previously discussed, the median brake reaction time is 0.66 seconds to an expected stimulus. The shortest time from SWS first alert to warning message display is approximately three seconds and, if the message is not read by the driver the first time, two additional seconds may be required to read the second display of the message. Thus, a total worst case of five seconds may be' required for the driver to perceive the warning and start to react by braking his or her vehicle. While this is thought to be a worst case warning perception time, this is the time that will be used in the applications analysis that follows.
28

According to Highway Engineering, the braking distance of a vehicle determined by:

v2

d=-

(1)

2fg

where d = braking distance in feet V = velocity of the vehicle in feet per second when the brakes are applied g = acceleration due to gravity f = coefficient of friction between tires and roadway

The coefficient of friction varies as a function of road surface, tire condition, tire pressure, and the environmental conditions affecting the road surface such as ice, snow and rain. The acceleration due to the gravitational constant is dependent on the grade of the highway. Stopping is shorter on the upgrade than the downgrade.

Figure 4 shows the braking distance of a vehicle on dry and level pavement usmg the results of Equation (1). The coefficient of friction varies as a function of vehicle speed. A coefficient of friction ranging between 0.53 and 0.62 was used in the calculation. Vehicle speed versus stopping distance is shown. These data represent best case stopping distance and will be used in a I.ater section when SWS application effectiveness is being assessed.

Figure 5 shows the braking distance of a vehicle on wet but level pavement using the results of Equation (I). A coefficient of friction ranging between 0.27 and 0.38 was used in the calculation due to the fact that rain is assumed to have wet the highway. The coefficient of friction also varies as a function of vehicle speed. Vehicle speed versus stopping distance is shown. These data represent worst case stopping distance during rain and will be used in a later section when SWS application effectiveness is being assessed.

Figure 6 shows the braking distance of a vehicle on dry and level pavement usmg the results of Equation (I) with the worst case SWS perception-reaction time of 5 seconds added to the total reaction time. All other factors used to generate previous Figure 4 are were kept the same.

Figure 7 shows the braking distance of a vehicle on wet and level pavement usmg the results of Equation (I) with the worst case SWS perception-reaction time of 5 seconds added to the total reaction time. All other factors used to generate previous Figure 5 are were kept the same.

29

;~o-~}

Rtwi,'

u~ 15U

Z

,..

~
!12

~

w
0

Q Q 20;) ~

~

at

=tX

fWJ

SPEED {tAPfi}
Figure 4. "Braking distance of vehicle on a dry and level surface as afunction of vehicle speed

~.
trl

wo

.l.4..;........

W

t)

"Y'
:t

~ ~~

40~

Q

z~

~

W

~.

{t: :j()j

~

Figure 5. Braking distance o/vehicle on wet and level surface as a/unction o/vehicle speed

i!i)O

7t){)

wr-

~
6..

t~)~

w u Z
4J( ~ Ui
~
Q

0

eN N

~

~ 1i:(

4tYJ

~

to

~Qc::

SPEEl> (MPH~
Figure 6. Braking {listance of a vehicle on dry and level pavement as afullction of vehicle speed 'rt'ith 5 seconds addedfor SWS display reception time

:Oo>: , '

':.

.Oo"

::.:::
. '. :;":-:,,, :.;.. ::::,

10
Figure 7. Braking distance of a vehicle on wet but level pavement as a function of vehicle speed with 5 seconds added for SWS display perception time

Figure 8 on page 35 was generated to show the minimum distance at which an overtaking police or emergency vehicle would have to be detected to allow an SWS equipped driver time to take evasive action. The vehicle separation distance includes the distance traveled by overtaking police or emergency vehicle during the 5 seconds required for perception-reaction time.
1.9.2 Assessment of SWS Applications
There were 20 operational scenarios defined in Table I for the mobile transmitter and 24 operational scenarios defined in Table 2 for the fixed site transmitter. Subsections 1.9.3 through 1.9.21 utilize the data and assumptions developed in the previous sections to determine if the SWS can perform an application.
The application is shown as a subheading. The same application is not analyzed mUltiple times for each type of service. For example, when an application is analyzed for an application that applies to one service, such as police, the analysis will not be presented again when the application is used in another service such as an emergency vehicle, unless there is some difference in the way that the SWS would be used in each service. The applications are first taken from the "Mobile Warning Application" column of Table 1 and then from the "Roadside (Fixed Site Transmitter) Warnbg Application" heading of Table 2.
1.9.3 Collision Avoidance at Intersection
The mobile transmitter sends the message "Police III Pursuit" when approaching and crossmg an intersection. During urban testing, the SWS provided a first alert at a range of 500 feet from the test intersection and the warning message was displayed 180 feet from the intersection. Referring to Figure 6, the highest speed that the motorist could maintain and be assured of stopping prior to entering the intersection on dry and level pavement with 5 seconds of perception-reaction time included is 48 miles per hour when a first alert range of 500 feet is provided by the SWS. Referring to Figure 7, the highest speed that allows braking upon receipt of the first alert message is 43 miles per hour when the pavement is wet from rain or snow and the road is level. This worst case, worst case scenario also assumes that the emergency vehicle is no more than 30 feet from the intersection when first detected.
Assessment: The SWS will successfully provide an adequate warning of an intersection crossing police or emergency vehicle when the police or emergency vehicle is within a worst case distance of 30 feet of the intersection, during dry weather on level pavement, to an SWS equipped motorist at any speed up to
34

,::'
~
1:
).;..:,,0....

W
~

2S:;}:

~

~

Jl~

6

~ 200

w
Ul

52
II(
ecn:

1~

Figure 8. Maxi/11illll separation distance between vehicle requiredfor SWS to provide adequate ~t'ar/li/lg ofovertaken police or emergency vehicle

48 miles per hour. The SWS will provide an adequate warning for this scenario at speeds up to 43 miles per hour when there is rain on the roadway. Better performance can be achieved if the emergency vehicle is farther from the intersection when the SWS equipped motorist approaches.
1.9.4 Warn of High Speed Police Chase in Progress
The mobile transmitter sends the message "Police in Pursuit" when in motion on a highway or interstate. During a high speed chase, the police vehicle would overtake the SWS equipped motorist. It was assumed that the motorist's reaction would be to either pull to the side of the road or change lanes, and thus a full stop is not required.
The interstate tests database was consulted. During interstate testing of the overtaking scenario, the very worst case range at which the warning message was displayed was 1,650 feet with a first alert being received at much greater ranges. First, applying interstate data to an urban situation, if the SWS equipped motorist was driving the speed limit of 35 miles per hour and the overtaking police car was driving at 70 miles per hour, the closing rate between vehicles would be 35 miles per hour. Based on the data shown in Figure 8, the SWS would be effective in alerting the SWS equipped motorist of the police vehicle's approach if the warning message was displayed at any time of approach before the police vehicle reached a distance of approximately 250 feet from the SWS equipped motorist. If another IO seconds (assuming a closing speed of 51.3 feet per second) were allowed for lane change, the SWS equipped motorist would be required to detect the police vehicle no closer than 763 feet (513+250) .
At interstate speeds, the equivalent situation would be the SWS equipped motorist driving 55 miles per hour and the overtaking police or emergency vehicle closing at 90 miles per hour, which also produces a closing speed of 35 miles per hour. Again, if IO seconds were allowed for the motorist to change lanes, the overtaking police vehicle must be detected no closer than 763 feet.
Another application shown in Table I which is similar to the overtaking scenario is the notification of an SWS equipped motorist that he or she is being stopped for a traffic violation. The SWS would be used to get the violator's attention. The police vehicle would pull behind the motorist and activate the emergency lights and SWS mobile transmitter to transmit the message "Police Vehicle in Pursuit." This function could easily be performed by the SWS given the close proximity that most police cars maintain behind the subject's vehicle during a traffic stop.
36

Assessment: The SWS would provide an effective warning of an overtaking police or emergency vehicle under all conditions where the typical speed limit was being observed by the SWS equipped motorist.
1.9.5 Warn of Traffic Stop Ahead
A police vehicle making a traffic stop would be transmitting the SWS warning message "Stationary Police Vehicle Ahead." The police vehicle would be on the side of the highway behind the stopped motorist, but in the worst case, the rear end of the police car would be extending into the roadway, requiring the oncoming motorists to stop or collide with the rear of the police vehicle. The SWS equipped motorist would be approaching the police vehicle from the rear. The police vehicle would transmit the SWS signal both forward and rearward.
During rural testing along a straight section of highway, the worst case range at which the SWS could be received was 1,700 feet. This worst case performance occurred when the mobile transmitter's signal was blocked by a pick up truck 70 feet in front of the receiver vehicle during the rural tests. Normally, the range between vehicles when the SWS message was received was twice this distance; however, this worst case situation could occur. Referring to Figure 6, the maximum speed that could be driven by an SWS equipped motorist on a dry and level highway with a need to come to a full stop to avoid colliding with the police car would be in excess of 80 miles per hour, which is thought to be excessive speed for a rural highway. Referring to Figure 7, the maximum speed that could be driven by an SWS equipped motorist required to stop on a wet but level highway would still be over 80 miles per hour.
If the police car was located in the midpoint of a curve with heavy foliage blocking the line of sight beyond 500 feet, as occurred during the rural tests, the maximum range at which the SWS message would be displayed was 642 feet with the first alert displayed over 1,000 feet. Attenuation by tree foliage was thought to be the reason for this poor SWS performance and, had the test vehicle been in the roadway instead of 12 feet off of the edge of the roadway, the SWS performance would have been better. Referring to Figure 6, the maximum speed that could be driven by an SWS equipped motorist who does not react to a first alert but instead requires the warning message to start braking, on a dry and level but curved highway with a need to come to a full stop to avoid colliding with the police car, would be slightly less than 60 miles per hour given the curve scenario. Referring to Figure 7, the maximum speed that could be driven by an SWS equipped motorist required to stop on a wet but level highway would be approximately 52 miles per hour. It should be noted that the speed limit was 45 miles an hour on the portion of highway where the rural tests were conducted.
37

The rural and Dobbins tests also included several trials where the simulated police vehicle was parked at the bottom of a hill. The worst case range at which the warning message was displayed during the Dobbins hill trials was I, I80 feet and, during rural trials, the worst case hill trial produced a warning message at a range of 858 feet.
Referring to Figure 6, a vehicle could travel 74 miles an hour, receive the SWS warning message, and still come to a full stop on a dry and level highway stop to avoid colliding with a police car at the bottom of a hill. The fact that in the example case, the receiver vehicle is going downhill increases the gravity term in Equation (1), so the maximum speed would be closer to 70 miles per hour. Using the hill warning message display range of 858 feet and referring to Figure 7, the maximum speed that could be driven by an SWS equipped motorist required to stop on a wet but level highway would be approximately 62 miles per hour. The fact that the vehicle is going down hill would increase the gravity term in Equation (1) and further reduce the worst case speed to approximately 55 miles per hour. These speeds are slightly above or at the legal speeds allowed on the section of rural highway where the hill tests were conducted. Also, this scenario assumes that no action was taken by the driver when the first alert was received because the danger could not be seen.
Assessment: The SWS would provide an effective warning of an overtaking police or emergency vehicle under all straight highway conditions. The SWS would provide effective warning to SWS equipped motorists traveling no faster than 62 miles per hour around a curve in dry weather when there was heavy foliage blocking the SWS signal reception beyond the curve. In the rain, the SWS would provide an adequate warning to motorists traveling approximately 52 miles per hour or less under the same conditions. When a police car was located at the bottom of a hill, the SWS would provide a warning to motorists traveling 70 miles per hour or less in dry weather. During wet weather, the effective speed would be reduced to approximately 55 miles per hour. This assessment is thought to be very worst case because the driving assumption was that the SWS equipped motorist must come to a full stop to avoid colliding with the police vehicle in the roadway.
1.9.5.1 Duplication of Basic Application
Mobile applications 5, 6, 7, 8,9, 10, 1I, and 12 are thought to. be applications that have already been addressed by the previous analysis.
38

1.9.5.2 Work Zone The previous analysis in Section 1.9.5 is thought to apply to work zones with several exceptions, notably construction equipment in a work zone or a line painting vehicle that moves slowly. The work zone is very well marked in advance as are line painting trucks, at least in Georgia. Further, if a piece of . construction equipment will be moving into traffic, a flagman is usually required to control traffic around the construction vehicle. Assuming that vehicles in a work zone or a line painting vehicle moves at 15 miles per hour, the maximum speeds at which the SWS would be effective is 15 miles per hour greater than the overtaking examples. Thus, the performance of the SWS would be better than the previous case where the police vehicle was stopped in the highway and a non-anticipated full stop was required to avoid a collision. Thus, the SWS would effectively warn an SWS equipped motorist overtaking a 15 miles per hour work zone vehicle at speeds up to 65 miles per hour.
Assessment: The SWS would provide an effective warning to any vehicle obeying the speed limit while approaching and driving through a work zone. The SWS should also be effective in warning of the presence of a line painting vehicle or other slow moving construction vehicle.
1.9.5.3 Duplication of Basic Application Mobile applications 14, 15, 16, and 17 are thought to be applications that have already been addressed by the previous analysis.
1.9.6 Transport Vehicle's Pilot Vehicle Equipped with SWS The overtaking case which has been analyzed covers the case of the SWS equipped vehicle approaching an oversized load's forward and trailing escort vehicles. The mobile transmitter would be mounted on the top and in the middle of the oversized load transmitting both forward and rearward. This scenario would apply to both two lane and interstate scenarios. The rural tests results would also apply to the scenario where the SWS equipped motorist would approach and meet the oversized load's lead pilot vehicle on a two-lane road.
1.9.7 Duplication of Basic Application Mobile application 19 is thought to be an application that has already been discussed.
39

1.9.8 Train Locomotive Application
No testing of the installation of a mobile transmitter on a locomotive to warn of the locomotive's approach to a highway crossing was conducted. The performance of the SWS transmitter on the locomotive could best be estimated by using the worst case performance of the receiver crossing the path of the mobile transmitter at Dobbins. This analogy assumes the crossing is in the open (as in the Dobbins case) and the locomotive crosses tangentially to the motorist's path. The maximum message display range achieved during the Dobbins test was 758 feet. To extrapolate the Dobbins test results to a railroad application, the locomotive is required to be very close to the rail crossing as the SWS vehicle approaches (a worst case scenario). Referring to Figure 6, the maximum approach speed to the rail crossing of the SWS equipped vehicle could be any speed up to approximately 63 miles per hour when the roadway is dry and level. Referring to Figure 7, the maximum approach speed would be approximately 55 miles per hour on a wet roadway.
Another approach used to implement the rail warning application using an SWS would be to install a fixed site transmitter at the crossing, align it to transmit along the crossing roadway, and activate it when the crossing warning equipment was activated. The worst case rural trial, which included signal blockage by a pick-up truck, determined that the SWS equipped driver would have received a warning 1,700 feet from the transmitter even with signal blockage. In the case of a rail crossing, signal blockage would not be as severe as the rural case if the transmitter were elevated above the traffic on the crossing safety equipment supports. The problems of signal reduction due to blockage by tractor trailer trucks was not quantified during testing but could be a potential problem. Referring to Figure 6, the SWS transmitter installed at a crossing so that it is aligned with the roadway would have provided a warning message to the SWS equipped motorist approaching on a dry level road at a maximum approach speed exceeding 80 miles per hour. Referring to Figure 7, the maximum allowable wet road approach speed would also have been in excess of 80 miles per hour.
Assessment: The use of the SWS for railroad warning has not been tested. There are several approaches to implementing the warning system at a rail crossing. The optimum method would be to mount a fixed site transmitter on the crossing safety equipment and align the transmitter's antenna so that it illuminates the approaching traffic. The application that would provide the shortest warning range would be to mount a mobile transmitter on the locomotive. This method of SWS deployment would not require that crossing safety equipment be installed at a crossing for the method to be successful.
40

1.9.9 Fixed Site Transmitter Applications
The applications for the fixed site transmitter are found in Table 2. Many of them have previously been analyzed, given that the signal strength of the fixed site transmitter is slightly better than the mobile transmitter. The most important finding in the assessment of the fixed site transmitter is that the transmitter can always be placed in advance of the hazard when warning times would be too short if it was located at the hazard. It is thought that fixed site applications 1 and 2 have already been addressed.
1.9.10 Alert Motorist to Detour
The detour warning application does not require the SWS fixed site transmitter to be mounted at the point where the detour occurs. The SWS fixed site transmitter would provide supplemental warnings in addition to existing signs warning of the detour. Instructions received by the SWS may also be easier to interpret than highway signing when the motorist is in heavy traffic or between large trucks that could block his or her view of the detour sign. The fixed site transmitter would be deployed in a similar manner for applications 4, 5, 6, 7, 8 and 9. Based on all of these applications, proper warning can be provided by the SWS applications for all of these types at any reasonable speed within the posted speed limit.
.1.9.11 Deployed in Advance of a Bridge Traffic Gate
The drawbridge warning application does not require the SWS fixed site transmitter to be mounted at the bridge, although by, mounting the transmitter at the bridge the complexity of the wiring required to link to the bridge tender's traffic warning system would be minimized. The placement of the fixed site transmitter in advance of the bridge would offer adequate warning to an SWS equipped motorist at any reasonable speed. Most drawbridges are usually located in areas that experience heavy fog during certain parts of the year. During heavy fog conditions,_ the SWS might also provide a longer range warning than the fog obscured warning lights. Thus, the SWS fixed site transmitter could provide proper warning to any motorists obeying the speed limit while approaching a bridge traffic gate.
1.9.12 Rock Slide Warning
The warning functions for a rock slide does not require the SWS fixed site transmitter to be mounted at the hazard. Thus, by proper placement of the fixed site transmitter, adequate warning could be provided at any reasonable speed. However, the rock slide area would most likely have become a work zone by
41

the time that the fixed site transmitter was deployed. Therefore, the analysis presented earlier for a work zone also may apply to the rock slide case.
1.9.13 School Zone Ahead
The warning functions for a school zone do not require the SWS fixed site transmitter to be mounted at the start of the school zone. Thus, by proper placement of the fixed site transmitter, adequate warning could be provided at any reasonable speed.
1.9.14 Advance Warnings and Cautionary Messages Under Fixed Site (Application 14)
The warning functions shown in application 14 do not require the SWS fixed site transmitter to be mounted at the hazard. Thus, the fixed site transmitter could be mounted in advance of the hazard to warn oncoming traffic. Thus, by proper placement of the fixed site transmitter, adequate warning could be provided to any motorist obeying the speed limit.
1.9.15 High Wind Warning
The high wind warning function could also be in advance of the area where the wind was a danger to the highway traffic. The fixed site transmitter could be located at the anemometer in situations where the line of sight distance to the approaching traffic is not obstructed by a hill or curve. In cases where an obstruction exists, the fixed site transmitter could be located in advance of the high wind corridor. Thus, by proper placement of the fixed site transmitter, adequate warning could be provided to any motorist obeying the speed limit.
1.9.16 Severe Weather Warning
This application assumes that there would be multiple fixed site transmitters located in a network to provide continuous coverage along an interstate highway, given that severe weather could cross a highway at any point. It is further assumed that the Advanced Traffic Management System (ATMS) would activate the transmitters on either side of the location of the severe weather. It is also assumed that several "downstream" transmitters would be used to warn of the severe weather ahead. Thus, the fixed site transmitter network should provide an adequate warning of severe weather ahead of the SWS equipped motorist obeying the speed limit.
42

1.9.17 Heavy Fog Ahead
The location of heavy fog areas is generally known and fixed site transmitters could be placed in advance of the fog area to warn any SWS-equipped motorist driving any reasonable speed. In situations where fog occurs randomly along a corridor, a network of SWS transmitters could be installed and operated similar to the concept used to warn of severe weather crossing the highway. Input from a roadside fog sensor would be used to activate the SWS fixed site transmitter.
1.9.18 Temporary Hazard Warning (Application 18)
A warning regarding the temporary hazards shown in application 18 could be provided via a temporary fixed site transmitter installed in advance of the hazard area or a network of transmitters activated as in the previous two applications. Either solution should provide a sufficient warning to an SWS-equipped driver.
1.9.19Ice on Highwayl Bridge Warning
This application would be handled in a manner very similar to the previous three applications. Either solution should provide a sufficient warning to an SWS-equipped driver.
1.9.20 Duplication of Previous Applications
Applications 20, 21, 22 and 23 each share a common approach to fixed site transmitter use similar to previous examples. Given that the fixed site transmitter can be located in advance of the conveniences and hazards being marked, it is reasonable to assume that an SWS equipped motorist can be successfully informed of the conveniences as well as the hazard category situations.
1.9.21 Emergency Vehicle in Transit
This fixed site transmitter application is intended to be used to warn of an emergency vehicle leaving a fire house or hospital area. The fixed site transmitter could be aligned to favor the approach streets that the emergency vehicle may cross or tum on to, a process that will improve the range over which the' warning message can be received. The fixed site transmitter can also be located above the traffic so that blockage by intervening traffic can be avoided. Thus, the fixed site transmitter used in this application could warn any SWS equipped driver of the departure of an emergency vehicle from its base of operation.
43

1.10 OPERATION OF THE SWS IN THE RAIN AND INCLEMENT WEATHER
The SWS was tested during periods of no rain. However, rain will provide some attenuation to the SWS . signal. The distribution of rain drop sizes is very important to the calculation of the amount of attenuation that will be produced during a rain event. The drop size distribution and the rain rate (usually in units of millimeters per hour) determine the attenuation that will be produced. The Fifth Edition of Reference Data For Radio Engineers, published by Howard W. Sams and Company, Inc., was consulted to determine the attenuation produced by various rainfall rates at the SWS operating frequency of 24 GHz. Table 7 was developed from the data in the reference.

Table 7. Attenuationfrom Different Rainfall Rates

Rainfall Rate (mmlhour)
0.25 1.0 4.0 16.0

Attenuation (dB/kiiometer)
0.02 0.1 0.5 2.0

Description
Drizzle Light Rain Moderate Rain Heavy Rain

A kilometer is approximately 3,280 feet, and this distance approximates the maximum direct path range at which the fixed site transmitter was tested during the rural straight road testing and the overhead testing on the interstate. The calibration of the BEL 855 STi showed that one decibel (dB) corresponded to approximately two units of relative signal strength. Referring to Table 7, light drizzle and fog will have negligible effect on the SWS signal over the kilometer signal path. The operation of the SWS over a kilometer path in moderate rain would change the signal level one dB or two relative signal strength units. Heavy rain would change the signal level two dB or four relative signal strength units.
A review of the signal strength plots presented in the appendices was undertaken to determine if the loss of four relative signal strength units would have caused the SWS to fail to perform in an application. It was found that the momentary variations in relative signal strength that occur, before the 15 point moving average was applied, are greater than the 2 dB attenuation induced by rain. As a result of this data review, it is felt that the attenuation caused by heavy rainfall will not appreciably reduce the effectiveness of the SWS system over the one kilometer or less operating range at which most applications were evaluated.

44

1.11 RECOMMENDATIONS TO ENHANCE SWS PERFORMANCE
1.11.1 SWS Receiver Performance Standard Development
It is recommended that a performance standard for the SWS receiver be developed. The BEL Model 855 STi radar detector that was used for testing provides excellent sensitivity to SWS signals and reliable decoding of the SWS modulation. It is recommended that SWSLC develop a detector performance standard that will apply to all companies building an SWS-capable receiver under SWSLC license. This will ensure that those receivers will perform as well as the BEL 855 STi detector that was used during testing by GTRI. The proposed minimum performance standard might include provisions to require:
I. A minimum sensitivity threshold at which the receiver provides a first alert; 2. A minimum sensitivity threshold at which the receiver provides an alphanumeric message to the
motorist; 3. A minimum antenna response as a function of aspect angle to the transmitter; 4. A maximum response time to provide a first alert once the signal level has reached the detection
threshold; 5. A maximum response time to provide the message to the motorist once the signal level has
reached the message decode threshold; 6. A maximum time that the message will be displayed after loss of the SWS signal; 7. A limitation on the display of a false message versus signal strength; 8. A minimum illumination level of the alphanumeric display (if used); 9. A minimum sound level produced by the voice synthesizer (if used); 10. Standardization of LED patterns presented to the motorist using a non alphanumeric display; I I. Message text standardization; 12. Allowable techniques for message text display (scrolling versus other presentations); and 13. Allowable techniques for first alert notification (tone versus clicks or other audio notification).
1.11.2 Transmitter Improvements
It is recommended that a performance standard also be developed for both the mobile and fixed site transmitters. The FCC will determine which part of their rules and regulations the SWS will operate under. Certain performance standards, in addition to those required by the FCC, should be developed by
45

SWSLC to ensure that their licensees who build both the mobile and fixed site transmitters meet certain minimum standards.
It is recommended that a narrow beam width antenna be developed for use on the fixed site transmitter for use when the transmitted message must be confined to a single direction of traffic flow.
It is recommended that a modem be designed into the fixed site transmitter. Although the fixed site transmitter's final form has not yet been determined, a human is currently required to be present at the fixed transmitter site to program the warning message. The addition of a modem to the fixed site transmitter's design would allow the transmitted message to be programmed by radio, land line, or satellite link. The remote programming capability would allow an ATMS to program and broadcast any one of the current 60 fixed text messages in near real time. This capability would be extremely useful for rural applications of the SWS, where the fixed site transmitter may be some distance from the automated traffic management center.
It is recommended that the SWSLC investigate the addition of a variable text messaging capability to the fixed site transmitter. The ability to transmit a variable text message combined with the previously suggested modem capability would allow any message to be programmed and broadcast by a remote ATMS.
It is recommended that software be developed to allow the SWS fixed site transmitter to be operated as an autonomous system for rural safety applications. The fixed site transmitter contains a radar which can determine an approaching vehicle's speed and range. This remote sensing capability, if utilized, would allow vehicle count rates and rural highway occupancy rates to be sensed in real time. These data could be used dynamically to change the message being transmitted in response to traffic congestion conditions that are sensed in real time.
1.12 IMPLEMENTATION OF SWS TECHNOLOGY IN THE INTELLIGENT TRANSPORTATION SYSTEM
According to information from ITS America, the only national public/private organization established to coordinate the development and deployment of ITS in the United States, "Traffic congestion costs the American people $100 billion dollars a year in the form of lost productivity. In 1993 traffic accidents claimed the lives of 40,115 people and injured an additional three million people. Vehicle emissions are a major cause of air pollution. Trucks, buses and automobiles idling in traffic emit tons of pollutants each year and H'aste billions of gallons offuel."
46

Given these statistics and the Congressional mandate to improve conditions, ITS America has developed a plan to increase the capacity of highways without the requirement that new highways be built. In addition, the plan is designed to improve highway safety. The product of this plan is the Intelligent Transportation System (ITS). ITS America predicts that by the year 2011, a projected $209 billion dollars will be invested in ITS technology, with 80 percent of that investment coming from industry. There are many technologies planned to achieve the goals of ITS. The SWS fits into several of the ITS technology areas. The areas where the SWS fits into the ITS plan include:
I. En route driver information including in-vehicle signing and driver advisory; 2. Traffic control, including emergency vehicle signal preemption; 3. Incident monitoring and detection; and 4. Intersection collision warning.
1.12.1 En Route Driver Information
Traditionally, an en route driver information system has been thought of as a system that will provide a driver information regarding the navigation information (directions) required to reach an unfamiliar destination and also provide the traffic conditions along the chosen route. An element of the en route information system in the ITS context is the issuance of in-vehicle signing and driver advisories. The SWS can perform in these two areas.
The SWS message set contains 60 total messages (see Appendix A for a complete listing) that allow standard highway signing to be presented to the SWS-equipped motorist. Several of the more important messages that the SWS would provide to an SWS-equipped driver include:
1. Work Zone Ahead 2. Highway Crews Ahead 3. Utility Work Crews Ahead 4. Drawbridge Up 5. Low Overpass Ahead 6. Road Narrows Ahead
47

7. Accident Ahead 8. Ice on Bridge Ahead 9. Ice on Road Ahead 10. Heavy Fog Ahead
If more information is required to warn the motorist of a more complex condition, the message "Traffic AlertfI'une AM Radio" is provided in the SWS message set the SWS transmitter would complement the AM broadcast messaging system when used in this way. In the case of fog, the SWS message may be the only alert that the motorist gets if the fog obscures the roadway signs that warn of fog.
1.12.2 Traffic Control Including Emergency Vehicle Signal Preemption
The mobile transmitter can be used in conjunction with an SWS receiver located at a traffic signal to preempt the red light cycle in the direction of approach of the police or emergency vehicle. An interface between the receiver and traffic signal controller will be required.
1.12.3 Incident Monitoring and Detection
The fixed site transmitter contains a radar that can determine the speed of the traffic flow passing its location. With the proper algorithms and a modem to send data to an automated traffic center, the fixed site transmitter can serve as an incident detection system in addition to its function as a safety message transmitter. Research shows that an incident can be detected in one lane of a three-lane expressway by the fixed site transmitter's radar. Given this capability, it would be a simple matter to program the fixed site transmitter to call the ATMS to report when an incident occurs. The incident report could specify the exact section of highway where the incident occurred. For example, if there was a fixed site transmitter located at an interval of everyone half-mile along a roadway, the incident could be located to the nearest one half mile.
1.12.4 Intersection Collision Warning
As previously suggested, the SWS could provide the SWS-equipped motorist with a warning during an approach to a "blind" intersection that a police or emergency vehicle equipped with a mobile transmitter is crossing his or her path. While this application does not warn the SWS-equipped motorist of every
48

potential source of a collision at an intersection, it does address the one potential source of collision that is a high probability threat.
1.13 ESTIMATES OF BENEFiTS OF SWS TECHNOLOGY ON TRAFFIC SAFETY AND TRAFFIC DEATHS
A meaningful prediction regarding the number of deaths and collisions that will be avoided by the deployment of SWS technology has proven very hard to make. The author conducted a limited literature search to determine what data other researchers have used to make a forecast regarding the project benefits of a transportation hazard warning system. Najm and Burget of the Volge National Transportation Systems Center and the National Highway Traffic Safety Administration, respectively, authored a paper entitled "Benefits Estimation for Selected Collision Avoidance Systems." This paper appeared in the Proceedings of the 4th World Congress on Intelligent Transportation Systems, Berlin, Germany, 1997. This paper documented that a benefits estimation study requires, as a minimum, the following information and data: (1) proportion of market penetration of the device being evaluated, (2) proportion of device utilization during relevant crash hazard, (3) effectiveness of the driver using the safety device in preventing a crash situation, and (4) other quantifiable factors. The computation of probabilities that a police or emergency vehicle or a motorist will be SWS-equipped requires that the results of an SWS market penetration study must be available to the analyst. The SWS is a new concept, and market penetration studies have not yet been conducted to determine the number of transmitters that will be placed in service each year over the next 10 years. Also, because the SWS concept is new, there is no past history of sales volumes on which to base future projections.
There are no data available on driver reaction to the SWS warnings because the planned the human factors study was not conducted. The data produced by the planned human factors study would have helped define typical driver reaction times with and without a prewarning from an SWS first alert for various driving scenarios. For example, the human factors study would have determined if the SWSequipped driver reacts to the first alert signal or would he or she have waited for the warning message to be displayed to start to react? The medium used to display the warning message to the driver is thought an important associated issue. There are no data available to the author that provide a breakdown of the percentage of detectors that will have full alphanumeric display or voice synthesizer capability versus the percentage of those with simple LED displays.
There were indications during the study that the SWS fixed site transmitter could be effective in providing in-vehicle signing relating to a hazard in advance of the hazard. The fixed site transmitter can
49

be placed any reasonable distance prior to the hazard so that an SWS-equipped motorist driving at the posted speed limit will receive the warning message in time to react, regardless of signal propagation factors. Assuming that, upon receipt of the warning, the driver reacts as instructed in the warning message, the basic objective of utilizing an SWS to increase safety can be achieved in the same manner as if he or she had obeyed a warning sign external to the vehicle. The argument that increased safety would be achieved is supported by the fact that a transportation department places a warning sign (or SWS) anywhere along a highway with the expectation that it will improve a driver's chances of collision avoidance. However, any further extrapolation of this logic to derive the number of collisions that could be avoided using the fixed site transmitter is not possible at this time.
The mobile transmitter is different from the fixed site transmitter. It is designed to be installed on a moving hazard. The warning range provided by the mobile transmitter is measured from the vehicle (hazard) itself. Field testing of the SWS showed that the first alert always occurred well in advance of the mobile hazard. The results of a study referenced in a Section 1.10.1 showed that a pre-warned driver performs braking over a shorter distance than a driver who has not been pre-warned. Prewarning a driver to a hazard using the SWS can increase a driver's reaction time. The driver without the capability to be pre-warned must wait until he or she visually detects the strobe lights on the emergency vehicle or aurally detects the electronic siren before reaction can begin.
Tests results indicated that the SWS can provide several seconds of prewarning time to a driver's reaction if the first alert is acted upon. Field test results of the mobile transmitter also showed that the text of the warning message was usually displayed a minimum of several seconds before a line of sight was established with the transmitter vehicle. Thus, it could be argued that even if the driver did not react to the first alert, the display of the warning message would give the driver a second chance to begin his or her response. In most of the test scenarios, the non forewarned driver would still not have detected the strobe lights visually even at this point in the warning sequence.
In summary, while the number of deaths and collisions that the SWS would avert cannot be quantified, an argument can be made that the system can provide a driver additional warning time regarding the presence of a hazard that could be the source of an accident.
1.14 CONCLUSIONS
During the course of this study, the technical performance of the SWS was tested under almost perfect propagation conditions and also under highway conditions where propagation conditions were highly
50

variable. It was found that the presence of other vehicles in the traffic mix can both reduce and increase the range over which the first alert and the warning message is displayed. Using worst case test results produced during highway testing, the resulting SWS performance data indicated that the system can provide advanced warning of a highway hazard to a driver. The results of driver performance studies taken from technical literature indicate that a forewarned driver will be able to stop his vehicle in a shorter distance than the driver who has not been forewarned. A driver using an SWS-type system and receiving an alarm in advance of encountering a hazard would be cued to the hazard's presence and start the process of braking earlier than the driver who has not been forewarned. The extra time gained from an SWS alarm may allow the forewarned driver to avoid a collision that a non-forewarned driver would not be able to avoid.
The SWS can display anyone of 64 warning messages to a driver. During the development of this report it was determined that the number of applications to which the SWS might be applied could be increased if there were more message texts available. An industry spokesman indicated that the industry was considering the addition of a variable text transmission capability to the system. Also, the industry is considering the inclusion of 64 additional warning texts. The results of this study indicate that a variable messaging capability or the addition of 64 more fixed text messages could provide more flexibility and utility to the system.
A human factors evaluation of the SWS was not performed. Human factors data are essential to the total evaluation of the SWS. Human factors testing that would determine the reaction of drivers of varying ages to a first alert would validate the hypothesis that an SWS forewarned driver could react more safely (stopping more smoothly, slowing, etc.) than the non-forewarned driver. Human factors testing of the SWS could also determine to what degree in-vehicle signing might improve driver reaction time when the driver is presented with the text from a highway sign in a time critical manner.
A human factors study would also allow problems with the system to be identified, since the manner in which the data are presented to the driver can make a difference in the time required for the driver to perceive the warning message. Human factors testing would also identify the optimum approach to presenting the warning message to the driver, if one exists. A human factors study could also determine if the SWS causes undue distraction from the driving task.
The benefits of using an SWS in terms of the accidents prevented and lives saved was not thought quantifiable within the budgetary and schedule constraints of this study. However, it is thought that this present study provides the foundation for such benefits study to be conducted.
51

1.15 RECOMMENDATIONS
Additional testing of SWS operation under widely varying traffic conditions should be conducted to further validate the findings of this study regarding the distance at which a warning message can be received under different highway and traffic conditions. The additional testing should be conducted in conjunction with a field test designed to determine the effectiveness of the SWS as an in-vehicle warning system. The supervised field test program could be conducted to determine the effectiveness of the SWS in lowering the peak and average speeds normally driven through a work zone. A human factors study using a driving simulator should be conducted to measure the driver's response to the first alert and the warning message. Driving simulator testing should be conducted to determine if the SWS causes a driver significant confusion or distraction when the system is in operation. A benefits study should be conducted to determine how many crashes and deaths might be avoided if an SWS type system is permitted to become operational. Market forecasts supplied by SWSLC should be used in the benefits study to determine the penetration of the system during specific target years. Market forecasts by the mobile transmitter manufacturers should be used to determine the number of police and emergency vehicles that might be SWS equipped during specific target years. The causal factors of crashes of the type that might be impacted by use of the SWS should be extracted from the national accident data bases.
52

Final Report Project A-5285
VOLUME II APPENDICES
Evaluation of Radio and Microwave Technology For Motor Vehicle Warning System
By:
E. F. Greneker, Principal Research Scientist Sensors and Electromagnetic Applications Laboratory Georgia Tech Research Institute Atlanta, Georgia 30332
September 1999

THIS PAGE INTENTIONALLY LEFT BLANK

APPENDIX A TEXT OF THE 60 SAFETY WARNING MESSAGES
WITH POSITIONS FOR 4 MORE SHOWN
A-I

THIS PAGE INTENTIONALLY LEFT BLANK A-2

TEXT OF THE 60 FIXED TEXT MESSAGES

1

WORK ZONE AHEAD

2

ROAD CLOSED AHEADIFOLLOW DETOUR

3

BRIDGE CLOSED AHEADIFOLLOW DETOUR

4

HIGHWAY WORK CREWS AHEAD

5

UTILITY WORK CREWS AHEAD

6

ALL TRAFFIC FOLLOW DETOUR AHEAD

7

ALL TRUCKS FOLLOW DETOUR AHEAD

8

ALL TRAFFIC EXIT AHEAD

9

RIGHT LANE CLOSED AHEAD

10

CENTER LANE CLOSED AHEAD

11

LEFT LANE CLOSED AHEAD

12

FOR FUTURE USE 1

13

STATIONARY POLICE VEHICLE AHEAD

14

TRAIN APPROACHING/AT CROSSING

15

LOW OVERPASS AHEAD

16

DRAWBRIDGE UP

17

OBSERVE BRIDGE WEIGHT LIMIT

18

ROCK SLIDE AREA AHEAD

19

SCHOOL ZONE AHEAD

20

ROAD NARROWS AHEAD

21

SHARP CURVE AHEAD

22

PEDESTRIAN CROSSING AHEAD

23

DEERIMooSE CROSSING

24

BLIND/DEAF CHILD AREA

25

STEEP GRADE AHEADITRUCK USE LOW GEAR

26

ACCIDENT AHEAD

27

POOR ROAD SURFACE AHEAD

28

SCHOOL BUS LOADINGIUNLOADING

29

NO PASSING ZONE

30

DANGEROUS INTERSECTION AHEAD

31

STATIONARY EMERGENCY VEHICLE AHEAD

32

FOR FUTURE USE 2

33

HIGH WIND AHEAD

34

SEVERE WEATHER AHEAD

35

HEAVY FOG AHEAD

36

HIGH WATER/FLOODING AHEAD

37

ICE ON BRIDGE AHEAD

38

ICE ON ROAD AHEAD

39

BLOWING DUST AHEAD

40

BLOWING SAND AHEAD

41

BLOWING SNOW WHITE OUT AHEAD

42

FOR FUTURE USE - 3

43

REST AREA AHEAD

44

REST AREA WITH SERVICE AHEAD

45

24 OUR FUEL SERVICE AHEAD

46

INSPECTION STATION OPEN

47

INSPECTION STATION CLOSED

48

REDUCED SPEED AREA AHEAD

49

SPEED LIMIT ENFORCED

50

HAZARDOUS MATERIALS EXIT AHEAD

51

CONGESTION AHEAD/EXPECT DELAY

52

EXPECT 10 MINUTE DELAY

53

EXPECT 20 MINUTE DELAY

54

EXPECT 30 MINUTE DELAY

55

EXPECT 1 HOUR DELAY

56

TRAFFIC ALERTITUNE AM RADIO

57

PAY TOLL AHEAD

58

TRUCKS EXIT RIGHT

59

TRUCKS EXIT LEFT

60

FOR FUTURE USE 4

61

EMERGENCY VEHICLE IN TRANSIT

62

POLICE IN PURSUIT

63

OVERSIZE VEHICLE IN TRANSIT

64

SLOW MOVING VEHICLE

HIGHWAY CONSTRUCTION OR MAINTENANCE HIGHWAY HAZARD ZONE ADVISORY WEATHER RELATED HAZARDS TRAVEL INFORMATION/CONVENIENCE FAST/SLOW MOVING VEHICLE

A-3

THIS PAGE INTENTIONALLY LEFT BLANK A-4

APPENDIX B SAFETY WARNING LABORATORY AND FIELD TESTING
B-1

THIS PAGE INTENTIONALLY LEFT BLANK B-2

APPENDIX B SAFETY WARNING LABORATORY AND FIELD TESTING
Testing of the Safety Warning SystemTM (SWS) is the next phase of the Federal Highway Administration (FHWA)/Georgia Department of Transportation (GDOT) sponsored research program. During this test phase, both the mobile and the roadside version of the Safety Warning Transmitter (SWT) will be tested to determine the effective range over which the safety warning message can be received under varying highway conditions. There will be two phases to the overall test program: (1) laboratory testing and calibration of the systems to be used in the field tests, and (2) field testing of the SWS system.
LABORATORY TESTING AND CALIBRATION
The following primary equipment that will be used during laboratory and field testing:
1. Hewlett-Packard Spectrum Analyzer model 8562B to record the frequency and signal level as a function of time, using an HPffi interface and laptop computer.
2. BEL-TRONICS Model 855 STi radar detector with special factory installed RS-232 port to provide frequency and signal level data output as a function of time to a laptop computer data collection system.
3. MPH, Inc. experimental mobile transmitter modified by the factory to transmit, upon command, a continuous wave (CW) carrier or one of two safety warning vehicular messages. Also, a vehicular roof top MPH, Inc. transmitter mounting bracket that uses bungee cords and suction cups to hold the unit on a vehicle. This apparatus has been built to allow the transmitter to simulate emergency-lightrack mounting of the transmitter on any vehicle used during testing.
4. Two experimental roadside transmitters obtained from Safety Warning System, L. c., each capable
of transmitting anyone of 64 fixed text messages or a CW signal. The roadside transmitter can be mounted on signs or on an overhead structure. Each contains a wet cell battery with approximately 14 hours of operational capacity.
5. Custom software developed by the Georgia Tech Research Institute (GTRI) to interface with the Hewlett-Packard 8562B spectrum analyzer. The custom software requests data from the spectrum analyzer, and stores the received data in a file that can be retrieved for analysis after testing.
6. Custom software developed by GTRI that interfaces with the BEL-TRONICS Model 855 STi and provides signal strength information, as well as the index number of the selected safety warning message.
7. A stock radar detector modified to provide the 10.7 MHz intermediate frequency (IF) output and the output of the quadrature frequency modulation (FM) discriminator signal of the SWT. The radar
B-3

detector is tuned manually to the SWT signal. The output of the IF amplifier can be fed to the spectrum analyzer or to a high speed digitizer for modulation signal capture and examination.
8. A laptop computer for recording test data.
9. A small absorber lined box that serves as an anechoic chamber and a 24 GHz calibrated signal injection system (to be described) that allows a signal to be radiated and controlled. This system allows a radar detector's sensitivity to be measured and calibrated, the mobile SWT transmitter's power to be measured very accurately, and the radiating systems to be tested in a controlled environment.
10. Over 30 radar detectors manufactured by BEL-TRONICS LIMITED, Sanyo Tecnica USA, Inc., Sunk-young America, Inc., Uniden America Corporation, and Whistler, Inc. Many of these are SWS capable systems.
11. A video camera mounted on the vehicle carrying the SWR and spectrum analyzer.
12. Two recording Global Positioning Systems (GPS) units to determine the range between the vehicle carrying the SWT and the SWR and the range between the roadside transmitter and the vehicle carrying the SWR.
LABORATORY CALIBRATION OF EQUIPMENT
Calibration of the equipment to be tested will be performed before any equipment is taken out of the laboratory for field testing. The size of the GTRI test chamber requires that all testing and calibration be performed in the near-field of the antenna.
TEST SET-UP
Figure B-1 shows the laboratory set-up that will be used to test the sensitivity of any radar detector supplied as an off-the-shelf product from any of the five manufacturers of SWS capable detectors. The laboratory setup shown in Figure B-1 will allow samples from each of the radar detector manufacturers to be tested before FHWAlGDOT field testing is conducted to determine best and worst case detector sensitivity on the basis of current product performance. After the laboratory sensitivity measurements have been made and field testing has determined the absolute transmitter signal level versus range that can be reliably expected, the performance of each manufacturer's detectors will be extrapolated from the resulting data.
Referring to Figure B-1, the test setup is designed to conduct both single and dual SWS transmitter signal injection testing. Single transmitter testing will be conducted using the setup shown in Figure B-1 to calibrate receivers and conduct comparative evaluations of receiver sensitivity and operability. Dual
B-4

SWS signal injection testing allows the combined effects of two SWS transmitters, operating in close proximity to each other, to be evaluated. This is necessary to simulate anticipated real world situations of interest. For example, when a police vehicle and an emergency vehicle are in close proximity (such as at an accident site) and both are operating their SWS transmitters, it is suspected that there will be mutual interference. The dual transmitter test setup shown in Figure B-1 will allow the mutual interference effects to be evaluated.

Radar Detector Under Test

1 24.1 dB Standard Gain

-63.4 dB Space Loss

Hom

.~.....---A1c2ro3s.s 1th9ecPmath

~

Camera

Absorber Lined Microwave
Test Chamber

i .

.

.~. /'

.

.~~ ~iIIII~!-....

K-band External Mixer Assembly

Television Monitor

Hewlett-Packard Model 8562A Spectrum
Analyzer

Modulator from SWS L. C. Transmitter

Hewlett-Packard Model 83640A Signal Generator

Figure B-1. Laboratory setup used to conduct single and dual SWS transmitter testing

SINGLE TRANSMITTER TESTING

The single transmitter sensitivity testing will be conducted using the Hewlett-Packard Model 83640A synthesized signal generator. The Hewlett-Packard signal generator is adjusted so that the maximum deviation, during modulation, is 5.0 MHz around the center frequency of 24.100 GHz. The use of the Hewlett-Packard signal generator allows the frequency to be changed without affecting the amount of modulation deviation and symmetry. This cannot be reliably achieved using a mechanically tuned Gunn device transmitter. The ability to easily change signal source frequency allows the sensitivity of the receiver under test to be determined at the SWS band limits (25 MHz from the 24.1 GHz center frequency).

B-5

The 24.1 GHz radio frequency signal is fed from the signal generator to a WR-42 wave guide transition. Measurements show that approximately 3.4 dB is lost between the signal generator and the WR-42 waveguide. The transition feeds a 20 dB sidewall coupler. The coupler's direct path goes to a second 20 dB sidewall coupler and then into a 24.1 dB standard gain hom antenna. A 50 dB attenuator and SWS L. C., Inc. transmitter are connected to the hom antenna through the 20 dB arm of the last sidewall coupler in the chain (used only during dual transmitter testing). The external mixer of a Hewlett-Packard Model 8562A spectrum analyzer is connected to the 20 dB arm of the second sidewall coupler to monitor the spectrum and levels of the signal(s) being transmitted. The energy from the hom antenna illuminates a test chamber that is lined with radar absorbing material. The measured path loss across the 123.19 centimeter (cm) long test chamber (to the receiver test port) is approximately 63 dB 2 dB. Measurements must still be conducted to determine the absolute power density at the receiver test port.
The receiver port is a square hole cut into the test chamber at a point that is aligned with the standard gain hom on the other side. The receiver will be placed on a plexiglas support during testing. The receiver's antenna will be placed approximately 1 centimeter inside the chamber during testing. A low light level television camera will be focused on the display of the receiver that is under test. A television monitor will display the enlarged receiver display to the test operator to allow the reaction of the receiver to signal stimulus to be monitored at a distance. The Hewlett-Packard signal generator has an electronic attenuator that is calibrated to the nearest tenth of a dB. Attenuation in excess of 110 dB can be introduced using the front panel attenuation control.
During detector sensitivity testing, the test operator will determine the SWS signal level at which the radar receiver first alarms (but does not display the safety warning message) by decreasing the signal generator's attenuator, starting from a signal level of -80 dBm, until the minimum sensitivity at which the receiver first alarms is determined. The signal generator's output signal level will be increased from the point of first detection by decreasing the signal attenuation in half dB steps. Next, the signal level at which the SWS warning message is displayed reliably by the receiver will be recorded. A second test will be conducted to ensure that no false messages are triggered at the minimum signal level required to cause the SWS to be activated. (This would be evident by the presentation of a warning message or triggering light emitting diodes LED indicating SWS signal reception.) Next the sensitivity testing will be performed at the SWS band edges of 24.075 GHz on the low side and 24.125 GHz on the high side to ensure consistent operation over the entire SWS frequency band. Other testing will be performed to ensure that a given manufacturer's receiver is operating properly and providing the correct warning messages.
B-6

DUAL TRANSMITTER TESTING
Referring to Figure B-1, two additional components in the test set-up will be used to determine the susceptibility of a receiver to the effects of two SWS transmitters operating in close proximity to each other. The signal from a fixed frequency, fixed power output (+ 14.5 dBm) Gunn device SWS transmitter
supplied by SWS L. c., Inc. will be fed into the input of a 50 dB adjustable attenuator. The output of the
attenuator will be coupled into the 20 dB arm of the sidewall coupler and, after being further attenuated by the coupler, the signal will be radiated by the standard gain horn antenna. The spectrum analyzer information will be used to monitor the amplitude and frequency of each signal, both the absolute values and the corresponding ratios. The combined effects of system isolation provided by the adjustable attenuator and the sidewall coupler will reduce the frequency pulling of the Gunn device transmitter by
the impedance mismatch resulting from this set-up. The SWS L. c., Inc. transmitter will be programmed
to transmit the message "Stationary Police Vehicle Ahead" while the modulation driver for the HewlettPackard signal generator will be programmed to cause the signal generator to transmit the message "Stationary Emergency Vehicle Ahead".
The power fed into the test chamber from the SWS L. c., Inc. transmitter will then be adjusted by
varying the attenuator until a reference signal level is reached that causes the safety warning message to be reliably displayed. Next, the signal from the Hewlett-Packard signal generator will be introduced starting at a signal level of -80 dBm. Eventually, the receiver under test will be affected by the second signal, and the absolute and ratios of signal levels at which this interference occurs will be noted. The level of both signal sources will be monitored by the spectrum analyzer. Testing occurs at several power levels so that results can be noted.
SINGLE TRANSMITTER RECEIVER SENSITIVITY TESTING
Receivers supplied by the four SWS manufacturers will be given the single transmitter test to determine the reliable SWS reception range that might be expected from off-the-shelf receivers and the quality control applied to each receiver in areas that affect performance. Each receiver manufacturer will be requested to pull receivers from their production runs without testing them before shipping. SWS capable receivers will be supplied by BEL-TRONICS LIMITED, Sunkyong Limited, Uniden America, Inc., and Whistler, Inc. Once calibrated in the laboratory, the range at which they reliably receive the SWS signal can be estimated when the field test data have been analyzed.
B-7

MOBILE SWT CALIBRATION
The mobile safety warning transmitter (SWT) manufactured by MPH, Inc. is a sealed unit. The output power port of the Gunn oscillator cannot be accessed for direct measurement of power output using the waveguide power meter thermistor mount. The antenna gain of the SWT antenna is also not known with certainty. Thus, a measurement of estimated radiated power will be made. Given that the space loss between the ports of the test chamber is known, the mobile SWT estimated radiated power can be measured before each field test. The SWT will be set up on the test pedestal normally used to test the radar detector. The SWT antenna will illuminate the chamber; the power meter will be attached directly to the standard gain hom antenna normally used to transmit on the opposite end of the chamber. The space loss, determined during initial testing, will be subtracted from the measured power reading obtained with the power meter. This procedure should produce the SWT estimated radiated power. The transmitter will be rotated 180 degrees to test each of the two antennas to ensure each is supplying power and each is transmitting the same power level.
ADVANTAGES OFFERED BY LABORATORY CALIBRATION
Once calibrated, the absolute measurements made on the SWS capable detectors in the laboratory can be compared to measured field test results. For example, if a specific detector was found through laboratory testing to alarm and display the safety warning message at a known power level, the range at which the detector would alarm and display the safety warning message can be estimated. This estimate would use the laboratory calibration data combined with the field test results and alleviate the requirement of testing each detector in the field. If laboratory measurements show one detector to be several decibels more or less sensitive than other detectors, the loss in detection range can be estimated using the combined laboratory measurements and field test measurements.
The laboratory calibration also allows certification that the test transmitters are functioning at a known performance level before being tested in the field. This determination allows test results to be validated with confidence.
FIELD TESTING
Three field test scenarios have been developed. The three test scenarios will provide SWS performance characterization when the SWS is used in a vehicle traveling on:
B-8

I. A rural, two lane highway, 2. An urban street in the central business district where high buildings come to the corner of
intersections, and 3. An expressway with a capacity of at least three lanes in each direction.
RURAL HIGHWAY TESTING
The first tests will be conducted on a rural two lane highway or equivalent paved surface that simulates a rural two lane road. Line-of-sight (LOS) will be maintained during testing, except where specified. The best SWS performance is predicted to occur during the rural two lane tests. The tests will determine the maximum range at which the SWS message can be received and displayed to the driver with and without intervening traffic during the following types of encounters:
1. In a head-on encounter where the SWT will be mounted on a vehicle simulating a police or emergency vehicle-The SWT antenna and the Safety Warning Receiver (SWR) antennas will be pointed at each other on a two lane rural road with few and, preferably, no intervening vehicles between the SWT and the SWR. This test scenario is shown in Figure B-2. Both mobile and fixed transmitter configurations will be tested, and LOS will be maintained between the SWT and SWR at all times. This test will determine the maximum range at which the SWS will operate under the very best conditions. A minimum of three trials will be conducted.
Figure B-2. Set-up for best case test of SWS to be conducted on rural highway
2. In an encounter where the SWT will be mounted on a vehicle simulating a police or emergency vehicle-The overtaking scenario shown in Figure B-3 will be tested. The mobile SWT equipped vehicle will be overtaking a vehicle equipped with a forward pointing SWRon a two lane road with a minimal number of intervening vehicles between the SWT and the SWR. Both mobile and fixed transmitter configurations will be tested and LOS will be maintained between the SWT and SWR at all times. This test will determine the maximum warning range between a mobile transmitter on an emergency vehicle overtaking a vehicle with a forward pointing SWR. A minimum of three trials will be conducted.
B-9

..

..

Figure B-3. Set-up for second test of SWS to be conducted on rural highway
3. In an encounter where the SWT will be mounted on a vehicle simulating a police or emergency vehicle-As shown in Figure B-4, the SWT equipped vehicle will be located around a curve so that LOS does not exist with the SWR equipped vehicle until the SWR equipped vehicle comes around the curve. The SWT equipped vehicle will be located at least 10 feet to the side of the active roadway (on the shoulder) for safety purposes, with emergency flashers activated. A minimum of three trials will be conducted with the SWT equipped simulated emergency vehicle located at different distances from the maximum point of curvature in the highway.

TRIAL 1

LINE OF SIGHT

SWR

....----:::,......::::""----=----------- _....- - - -

TRIAL 2 LOCATION SWT ,
Figure B-4. Test of SWT reception range around a curve at m'o trial location
4. In an encounter where the SWT will be mounted on a vehicle simulating a police or emergency vehicle-As shown in Figure B-5, the SWT equipped vehicle will be below the crest of a hill, so that LOS does not exist to the approaching vehicle with the SWR. Only after the SWR equipped vehicle passes over the crest of the hill will the SWR equipped vehicle establish LOS with the SWT equipped vehicle at the side of the roadway. The antenna of the SWT will point rearward and the antenna of the SWR will point forward. This test simulates a police or emergency vehicle at an accident site in a blind area caused by shadowing from a hill. The SWT equipped vehicle will be located at least 10 feet off the pavement of the active roadway (on the shoulder) for safety purposes, with emergency flashers activated. A minimum of three trials will be conducted with the SWT equipped simulated emergency vehicle located below the crest of the hill.
B-lO

1RIAL I LOCATION

... LINE OF SIGHT

SWR

t
CREST OF HilL

Figure B-5. Test set-up where SWT vehicle is below LOS over crest in hill
5. In tests that duplicate rural tests 3 and 4, except the roadside transmitter will be substituted for the vehicle mounted SWT-The roadside SWT will be mounted on a fixed tripod to simulate the mounting technique that would be used to mount the roadside transmitter on an existing sign or other roadside structure. A minimum of three trials will be conducted at each test location.
6. In an encounter on a two lane rural roadway where the roadside SWT will be mounted on a tripod on the opposite side of the road and ahead of the vehicle equipped with the SWR-The purpose of this test will be to. determine the distance over which the SWT will be detected when the message transmission is not intended for the traffic in the lane used by the vehicle carrying the SWR. This test scenario is shown in Figure B-6. The antenna of the SWT will be pointing in the same direction of travel of the vehicle carrying the SWR (180 degrees away from the SWR). A minimum of three trials will be conducted.
SWR

AN1ENNA POINTING ANGLE

swr

Figure B-6. Testing of roadside transmitter beam shaping in opposite direction
7. In an encounter on a m'o lane rural roadway where the roadside SWT will be mounted on the same side of the road and ahead of the vehicle equipped with the SWR-This scenario is shown in Figure B-7. There are two purposes for this test: (1) the test will determine the distance over which the roadside SWT will be detected when the message transmission is intended for the traffic in the approaching lane, and (2) the test will determine the distance past the transmitter that the SWT's signal can be received and displayed. A minimum of three trials of each test will be conducted.

B-l1



ANIFNNA POINIlNG ANGlE

swr

. .SWR
...

Figure B-7. Test of roadside SWT when SWR and SWT antennas are boresighted
8. In an encounter where the SWT will be mounted on a vehicle simulating a police or emergency vehicle-As shown in Figure B-8, the SWT equipped vehicle will cross the intersection at a 90 degree angle to the SWR equipped vehicle. LOS will be maintained between vehicles at all times. A minimum of two trials at each of three ranges will be performed.

I

I SWT I

I

I

I

CLEAR AREA

I

I

I

I

---I I I I

TRIAL 1

TRIAL 2

TRIAL 3

I

SWR

SWR

SWR

I

I

I

I

I

Figure B-8. Test geometry to test the capability ofa SWT r;m a vehicle crossing an intersection

B-12

RURAL TEST DATA ANALYSIS
The results of each of the previously described tests will be analyzed. A narrative will be prepared that discusses the results of each of the rural tests. The narrative discussing the test results will contain a figure that shows signal strength of the SWT plotted as a function of distance between SWR and the transmitter for each of the tests. Notations will be made showing the time at which initial SWT signal detection occurred, and the distance (time) at which the message was first received and displayed.
CANDIDATE RURAL TWO LANE ROAD TEST SITE SELECTION
Before a test of the SWS performance can be conducted on a rural two lane road, sites must be located that allow the types of tests described herein to be conducted as defined in a safe manner. Three criteria for each test site were established for the candidate rural roadway where straight line testing will be conducted. The site selection criteria are:
1. A level and straight two lane roadway. 2. At least 2,000 feet of unobstructed view (LOS) between the vehicles used during testing. 3. Light traffic so that there will be times when a test can be conducted and there will be no
intervening traffic between the vehicle carrying the SWT and the vehicle carrying the SWS (or the roadside transmitter). 4. Presence of a clear and level place for test vehicles to pull on to the shoulder of the roadway far enough from the active roadway to minimize the possibility of creating a traffic hazard. After the development of the four site selection criteria, a search was undertaken to find a typical two lane rural roadway for testing that meets the established criteria.
RURAL TWO LANE ROADWAY STRAIGHT LINE TEST SITE SELECTIONCANDIDATE SITE ONE
The first candidate location for the mobile to mobile and mobile to fixed site transmitter LOS testing is shown in Figure B-9. This map shows a section of the State Route (SR) 120 connector located in west Cobb County, Georgia commonly known as Drag Strip Road. The test area begins at the intersection of SR 120 connector and Georgia Highway 92 and proceeds approximately 3,174 feet to the intersection of SR 120 connector and Citizens Square (Lucille) Road.
B-13

Figure B-9. Candidate location of straight line testing along SR 120 connector
Figure B-1 0 is a photograph showing the view from the proposed transmitter location at this intersection (as annotated in Figure B-9) with a view to the northwest back toward the intersection of SR 120
Connector and Georgia Highway 92. The mobile to mobile LOS testing would occur over this stretch of
the two lane roadway. The fixed site transmitter tests would locate the transmitter in the grassy area from where the photograph was taken. The highway curves several hundred feet in front of this point. The location of the fixed site transmitter at this point ensures a LOS between the fixed site transmitter and the . vehicle carrying the SWR as if the SWT was in the center of the test lane.
B-14

Figure B-10. Photograph showing view from transmitter site back along candidate test area
RURAL TWO LANE ROADWAY STRAIGHT LINE TEST SITE SELECTIONCANDIDATE SITE TWO
A second site for two lane road LOS testing was identified as a section of Dixie Road in Smyrna, Georgia, starting at the intersection of R1chardson Road and Dixie and continuing to the intersection of Dixie Road and Dixie Way Road. Figure B-l1 shows the approximate test area along Dixie Road as annotated with the heavy black line. The LOS test distance is approximately 2,700 feet. As in the previous scenario, the test transmitter would be located on the opposite side of the road from the vehicle's travel. The slight curve in Dixie Road places the SWT in the center of the lane used by the oncoming SWR equipped vehicle. Figure B-12 is a photograph that shows the view from the transmitter site along Dixie Road back to the intersection of Richardson Road.
B-15

Figure B-11. Candidate location of straight line testing along Dixie Road in Smyrna, Georgia
CURVE TESTING AT CANDIDATE SITE ONE
The previous Candidate Site One features a curved section that will suffice for conducting testing of the SWT around a curve, as described in scenario 3 of the Rural Highway Testing description. An embankment in the apex of the curve will eliminate any LOS transmission path between the SWT (fixed or mobile) and the vehicle carrying the SWR. Figure B-13 shows the curved section of SR 120 connector from the point where the transmitter will be installed during the straight line testing. The map shown in Figure B-9, shows the relationship of the curve to the proposed transmitter location at Candidate Site One.
B-16

Figure B-12. View from proposed transmitter site along candidate Dixie Road test site
Figure B-13. Curved test area that is an extension of Candidate Site One B-17

CURVE TESTING AT CANDIDATE SITE TWO
Figure B-14 shows the curve on Dixie Road. There are railroad tracks located on the embankment on the left side of the photograph. The SWT would be located on the same side of the roadway as the approaching SWR, but around the curve so that LOS could not be established between the SWT and SWR until the vehicle carrying the SWR was very close to the SWT.
Figure B-14. Curved test area that is an extension of Candidate Site Two CANDIDATE SITE HILL TESTING Both the SR 120 connector and the Dixie Road candidate sites were chosen because of their level grade and, therefore are unsuitable as a hill test site. A candidate test site for hill testing was found along SR 381 in Paulding County, Georgia. The hill test run would start at the intersection of Industrial Parkway and SR 381 and continue along SR 381 to Acorn Tree Road, -as shown in Figure B-15. Referring to Figure B-15, there is an approximate 600 foot section of the two lane roadway between the crest of the hill and the point on the side of the road where the SWT would located (for both mobile and fixed site units).
B-18

1996 DeLonne Street Atlas USA
Figure B-15. Candidate site location of hill testing along SR 381 in Paulding County, Georgia Figure B-16 shows the proposed point where the fixed site transmitter will be located and the vehicle on which the SWT will be mounted. Per Figure B-16, the area is paved, and the test vehicle can be located approximately 10 feet from the approaching lane. The SWR maintains a LOS with the test vehicle only during the last 600 feet of the total 3,145 feet of the proposed test run. Otherwise, the vehicle is beyond the hill crest.
Figure B-16. View showing proposed location of SWT during hill testing along SR 381 B-19

URBAN TESTING

An urban arterial highway was to be selected for testing the mobile SWT deployed on a simulated police or emergency vehicle. The proposed test scenario is shown in Figure B-17. A series of two tests were to have been conducted. The tests were to be designed to determine the capability of the SWSto detect the approach of an emergency vehicle crossing at an intersection across an urban arterial highway. Rather than require both vehicles to move simultaneously, only the SWT equipped vehicle would have been required to move. The SWR equipped vehicle was to have been stationary, near the roadway, with emergency flashers activated. At least two trials at each of three ranges were to have been performed. However, due to safety concerns, this test sequence was abandoned. Elements of this plan are present in the Central Business District (CBO) test plan.

I

I SWT I

SINGLE-STORY

I I

SINGLE-STORY

BUILDINGS

I

BUILDINGS

I

I

I

I

- - - --~-------------- I ---------------

I

I

TRIAL I

TRIAL 2

I

TRIAL 3

I

SINGLE-STORY

SWR

SWR

SWR

I

BUILDINGS

I

I

I

I

Figure B-17. Test of SWT at an urban arterial highway intersection with single story buildings around the intersection of interest

CENTRAL BUSINESS DISTRICT TESTING
An urban street in the central business district (CBD) of Atlanta will be selected for testing the mobile SWT deployed on a simulated police or emergency vehicle. The test scenario is shown in Figure B-18. Testing will be conducted during light traffic conditions for safety purposes. The tests will determine the capability of the SWS to detect the approach of an emergency vehicle while. no LOS exists between the test vehicles and both vehicles are in the "canyon" areas formed by the buildings in the CBD. Rather than
B-20

require both vehicles to move simultaneously, only the SWT equipped vehicle will be moving. The SWR equipped vehicle will be stationary, near the roadway. A minimum of two trials at each of three ranges will be performed. During each trial, the SWR equipped vehicle will be moved closer to the intersection. The safety of both the SWT and SWR equipped vehicles will be the overriding consideration during testing, given the expected traffic volumes in the urban environment.

I

MULTI-STORY BUILDINGS

II SWT
I I I

MULTI-STORY BUILDINGS

I

I

I

- - - - - - -II- - - - - - - - - - - - - - - -

I

I

TRIAL 1 SWR

TRIAL 2 SWR

I

TRIAL 3 SWR

I I I

MULTI-STORY BUILDINGS

I

I

I

Figure B-18. Test set-up to be used during testing in the central business district area

URBAN CBD TEST ANALYSIS
Analysis will be performed on the CBD intersection test results. A narrative with illustrated plots and figures will be presented. The plots will show detection range versus signal strength of the SWT signal. When the CBD testing has been completed, three sets of data collected under progressively better microwave propagation conditions will have been collected. Performance for each test scenario will be contrasted and the viability of the SWS will be assessed for each change in the test scenario.

EXPRESSWAY TESTING
Two basic types of testing will be conducted to test SWS performance in the expressway environment: (1) the effective communications range will be determined when an SWT equipped vehicle is overtaking an SWR equipped vehicle, with three trials in the test sequence, and (2) the effective communications range between a roadside transmitter and vehicle equipped with an SWR. The first test simulates a police or emergency vehicle overtaking a motorist on the expressway when the SWR equipped vehicle is
B-21

in each of the three lanes of the expressway. The second test scenario simulates communications from a roadside transmitter mounted above the expressway surface or at the side of the expressway on a sign.

EMERGENCY/POLICE VEHICLE OVERTAKING SIMULATION

Figure B-19 shows the testing of SWS performance when the SWR equipped vehicle is being overtaken by the vehicle equipped with the SWT. Under the test condition shown in Figure B-19, the SWR antenna points away from the SWT antenna. During trial one, the SWT and SWR will be in the same lane with no intervening traffic. The SWT equipped vehicle will overtake the SWR equipped vehicle at SWR position L During the second trial, the SWT equipped vehicle will remain in the left lane of the expressway, while overtaking the vehicle carrying the SWR at SWR position 2 in the center lane. Traffic will be behind the SWR equipped vehicle so that some signal shadowing exists. The vehicle carrying the SWT will overtake the vehicle equipped with the SWR. During the third trial, the vehicle carrying the SWT will remain in the left lane of the expressway, while overtaking the vehicle carrying the SWR in SWR position 3. Signal blockage and multipath will be more severe than during trials at positions one or two due to the shadow zone caused by intervening traffic.

SWR POSITION 3 ~----------------------

POSSWIRTION2---I ~~-.-.- ~ - - - . - - -

~_._--------~-----_.- SWR
POSITION 1

-----------------------
SWT~

Figure B-/9. Expressway test set-up simulating an overtaking emergency vehicle

ROADSIDE/OVERHEAD IN-VEHICLE SIGNING TESTING IN INTERSTATE ENVIRONMENT
In-Vehicle Signing: Effectiveness of Overhead Transmitter Mounting
A total of nine trials will be conducted to determine the effectiveness of the roadside transmitter when mounted over an expressway to simulate an in-vehicle signing system function. The roadside transmitter will be mounted on a tripod that will be set up in the emergency lane of an overpass extending over the expressway, as shown in Figure B-20. The tripod will be located next to the overpass guardrail. The elevation of the antenna will be on the horizon for maximum transmission range. A student will deploy
B-22

the roadside transmitter at one of the three positions within the three expressway lanes designated for each trial, ensure that it is operation, and ensure that it is not removed by unauthorized persons during periods when the test personnel are on the interstate highway testing the SWS. Each of the three roadside transmitter positions will be evaluated with respect to each of the three SWR positions.

OVERPASS

ROADSIDE

POSITION 1

- ----------/-- ROADSIDE
POSITION 2

SWR POSITION 1

- - - .- _ -

-SwRPOSITION 2

---------._-------------

----------------~-

ROADSIDE POSITION 3

SWR POSITION 3

Figure B-20. Test set up to test roadside transmitter on overpass simulating an overhead sign

The vehicle containing the SWR will begin taking data from SWR position I during approach to the roadside transmitter while traveling toward the roadside transmitter that is in roadside position 1, and will continue to take data until going under the overpass and losing the signal of the roadside transmitter. The roadside transmitter will be moved to "roadside position 2" while the vehicle containing the SWR sets up for the next trial at the location shown as SWR position I. The vehicle containing the SWR will begin taking data upon approaching the roadside transmitter, while traveling toward the roadside transmitter in position 2, and will continue to take data until going under the overpass and losing the signal of the roadside transmitter. The roadside transmitter will then be moved to "roadside position 3." The vehicle equipped with the SWR will begin taking data upon approaching the roadside transmitter in SWR position I, and will continue to take data until going under the overpass and losing the signal of the roadside transmitter. The conclusion of this test sequence will end trials one through three.
Similarly, data will be taken for SWR positions 2 and 3 traveling toward each of the three roadside transmitter positions sequentially. Trials 4 through 7 will be for SWR position 2, trials 8 through 10 will be for SWR position 3.

B-23

IN-VEHICLE SIGNING: EFFECTIVENESS OF ROADSIDE TRANSMITTER MOUNTING
The roadside transmitter will be moved to the side of the expressway, as shown in Figure B-2!. It will be mounted on a tripod approximately 5 ft above the surface of the expressway and will be set back from the pavement by the same distance that a sign support would normally be set back from the edge of the emergency lane. The antenna elevation will be set for the horizon.

II ....

ROADSIDE

.+ -- -/- TRANSMITTER NORMAL SET-BACK DISTANCE FOR SIGNS

SWR POSITION 1

-

-----------

-

- - .- .- . . - - . -SwRPOSITION2
-----------------------
------------------~SWR POSITION 3

Figure B-21. Roadside transmitler setup

A total of three trials will be conducted to determine the effectiveness of the roadside transmitter. A student will be deployed with the roadside transmitter. The student will ensure that it is operational and will ensure that it is not removed by unauthorized persons during periods when the test personnel are on the interstate highway testing the SWS. The roadside transmitter will be located at the same roadside position during all of the roadside trials.

The vehicle containing the SWR will begin taking data upon approach to the roadside transmitter in position I, and will continue to take data until going past the roadside transmitter and losing the signal. During trial 2, the vehicle containing the SWR will begin taking data while traveling toward the roadside transmitter in position 2 and will continue to take data until going past the roadside transmitter and losing the signal. During trial 3, the vehicle equipped with the SWR will begin taking data while traveling toward the roadside transmitter in position 3, and will continue to take data until going past the roadside transmitter and losing the signal. This will end trials one through three of the roadside transmitter simulation of a roadside sign mounted in-vehicle signing system.

B-24

CANDIDATE SITE FOR EXPRESSWAY FIXED SITE OVERHEAD TRANSMITTER TEST
A section of the Interstate 75 expressway in Cobb County, Georgia has been selected that will allow easy entry to the expressway by the test vehicles at the start of each test run and during tum-around. Also, the section that is selected will have an overpass that has a safe area that can be used as a simulated overhead sign for the roadside transmitter by placing the transmitter near the rail of the overpass so the antenna illuminates the oncoming traffic from above. The candidate test site selected for expressway testing is Exit 123 on Interstate 75 between Chattanooga, Tennessee and Atlanta, Georgia, as shown in Figure B-22. Exit 123 was chosen as a test site because there is a southern tum around point at Exit 122 within 2 miles of Exit 123. Likewise, within 2 miles to the north of Exit 123 is Exit 124, which can also therefore, which message to display. There could be scenarios where two or more possible messages are serve as a turn around point. In addition, Exit 123 has an accessible emergency lane which allows direct access to the guard rail by the support vehicle.
Figure B-22. VieYt-' of candidate test site on Interstate 75 at Exit 123
B-25

Figure B-23 shows a vehicle parked in the emergency lane next to the guard rail where the fixed transmitter would be mounted. It is proposed that the vehicle would be parked so that the transmitter operator would be shielded from the same lane traffic coming across the overpass by the parked vehicle.
Figure B-23. View of emergency lane at Exit J23 Figure B-24 shows the geometry of the southbound lanes of Interstate 75 passing under the overpass. This is the view that the test transmitter would have of oncoming traffic. As shown in Figure B-24, the roadway is three lanes. The distance from the overpass to the peak of the hill is approximately 0.5 miles, a distance that should be sufficient for the testing to be conducted. No other expressway candidate site has been selected. EXPRESSWAY DATA ANALYSIS The expressway data will be analyzed to determine the range at which the SWT signal is acquired and the range at which the signal-to-noise (SNR) ratio would be high enough to cause the safety warning message to be displayed for the expressway scenarios. The effects of the various traffic configurations that were encountered during testing will be evaluated. Signal level as a function of SWR range from the SWT will be evaluated. The signal strength statistics will also be evaluated to determine the fade characteristics of the communications link as a function of time. This information will be useful when predictions are made in a later reporting stage regarding the effectiveness of a variable text messaging system.
B-26

Figure B-24. View of southbound lanes of Interstate 75 at Exit 123
SWT MUTUAL INTERFERENCE TESTING
The SWR scans a 50 MHz region of the SWT band, centered on 24.1 GHz, to find the SWT signal. When the SWT signal has been located, the detector logic stops the scan and decodes the safety warning message for display. When only one SWT is transmitting, there is no problem with mutual interference. However, in the case where there may be an SWT equipped police car and an SWT equipped emergency vehicle working an accident together or some other instance where two or more SWT equipped vehicles are in close proximity, the detector may become confused as to which transmitter to "lock" on to and, being transm"itted by two or more SWT units, each in a different service (police, ambulance, and fire truck, for example).
Laboratory testing will help to defiI1e if a mutual interference problem does exist and, if so, the extent of the problem. After the laboratory testing has been conducted and a conclusion regarding mutual interference issues has been reached,. field testing may be conducted to determine the extent of the problem that can occur during SWT deployment, if there are doubts about the validity of the laboratoiJ' test results. The mutual interference field test scenario will utilize two SWT roadside transmitters transmitting the same message and different messages. Minimum spacing between transmitters will also be determined.
B-27

COMMUNICATIONS Each participant in these tests will be equipped with a two-way radio. Two-way radios will allow all parties participating in these tests to maintain communications at all times. The communications function is required for safety and test coordination purposes. SAFETY CONSIDERATIONS When testing is conducted on an active highway or expressway, there are always concerns for the safety of the test v~hicles, the vehicle occupants, and the equipment that will be used during the test. In addition, liability issues regarding injuries to other motorists are of concern. If it is determined that safety or liability is an issue in any of the test scenarios, either before a test is conducted or during the conduct of a test, the test may be modified to address the safety or liability concerns. If safety concerns cannot be addressed, the test scenario will be dropped from the test sequence. TEST REPORT A test report will be prepared after the tests are concluded as one of the phase deliverables. The report will be distributed to the Georgia Department of Transportation and the Federal Highway Administration through the established sponsorship chain. A description of each test will be developed and the results of each test will be presented in the report. This report will represent a major effort toward determining the effectiveness of the SWS system to serve as an in-vehicle signing system to alert a user of highway hazards.
B-28

APPENDIX C BELTRONICS 855 STi ANTENNA PATTERN MEASUREMENTS AND
MPH, INC. MOBILE TRANSMITTER ANTENNA PATTERN MEASUREMENTS
C-I

THIS PAGE INTENTIONALLY LEFT BLANK C-2

APPENDIX C BELTRONICS 855 STi ANTENNA PATTERN MEASUREMENTS AND
MPH, INC. MOBILE TRANSMITTER ANTENNA PATTERN . MEASUREMENTS
BACKGROUND
A factory modified BEL-TRONICS 855 STi Safety Warning System (SWS) capable detector is used to record signal strength measurements during field testing of the SWS. This detector was modified by BEL-TRONICS LlJ\1ITED, the manufacturer, to provide the signal strength output via an RS-232 attached to a port on the body of the modified detector. The signal strength data are recorded using a lap top computer as the data storage system. Before conducting field testing, the response of the BEL 855 STi detector antenna to SWS signals (as a function of aspect angle around the receiver vehicle) was characterized. The response of the mobile transmitter's forward and rear pointing antennas was also characterized as a function of aspect angle. The first part of Appendix C discusses the measurements obtained during the detector antenna pattern characterization, while the second part of Appendix C discusses the characterization of the mobile transmitter's antenna pattern.
PURPOSE OF BEL 855 STi ANTENNA CHARACTERIZATION
During field testing, the SWS system is tested in scenarios that require the transmitter vehicle to approach the receiver vehicle from the front, side and rear. Normally, when a receiving hom antenna, typical of the one used in the BEL 855 STi, is illuminated from the side and rear by a transmitter source, the response of the antenna to the transmitted signal is much lower than when the transmitter and receiver antennas are operated at or near boresight (i.e., receiver and transmitter vehicles face each other). There can also be effects to the receiver caused by the windshield, support structures for the windshield, and, in the case of illumination from the rear, headrest and other structures within the vehicle. These structures can absorb or attenuate the signal and also reflect the signal. Reflections introduce a phase change between the direct path and reflected path signals that result in an effect known as multipath. Multipath effects cause peaks and nulls to occur in the signal over time. Natural occurring lobes in the antenna pattern also introduce peaks and nulls in the received signal as the aspect angle between the transmitter and receiver chan"ges in azimuth (around the vehicle). Given these realizations, a test was developed to further characterize the BEL 855 STi test recei ver' s antenna pattern, as well as the MPH, Inc. mobile
C-3

transmitter when it 1S mounted on or within the test vehicle 10 accordance with the manufacturer's instructions.
PURPOSE OF THE MPH, INC. MOBILE TRANSMITTER ANTENNA CHARACTERIZATION
The mobile transmitter, manufactured by MPH, Inc., was used in all of the field tests as the SWS mobile transmitter. It is designed to be roof mounted on the light bar of an emergency or police vehicle. The mobile transmitter has two beams: one forward pointing and one pointing rearward. Half of the transmitter power is radiated forward and half, is radiated rearward. The mobile transmitter's antenna response was tested as a function of aspect angle in the same manner as the SWS detector to determine antenna pattern characteristics, the depth of the null that occurs to the side of the emergency or police vehicle due to effects between the two antennas, and any other effects that may affect field test results.
TURNTABLE TEST FACILITY
The Georgia Tech Research Institute (GTRI) has a large turntable that can be rotated 360 degrees in azimuth. The turntable is capable of rotating large military vehicles, thus the rotation of a passenger car size vehicle was easily accomplished. First, a test to determine the rotation rate of the turntable with the receiver vehicle mounted on it was conducted. The rotation tests determined that a maximum rotation rate, over exactly 90 degrees in azimuth, required exactly one minute and forty eight seconds to complete. Normally, the data collection system is synchronized with the turntable through a servo system that allows the turntable angle to be recorded with each data point. Connection of the turntable servo system to the laptop computer used to record signal strength data was not an option. There was not sufficient time in the schedule nor money in the budget to develop the special interface and software to support the interface.
Further testing proved the turntable rotation rate to be very repeatable, and additional testing demonstrated that the detector signal strength recording system could be manually synchronized to start within I second of turntable start. Analysis showed that a one second error in synchronization shifted the position of where the data were taken 0.83 degrees out of synchronization with the actual turntable position. Given the trade-off of spending time and money on a synchronization interface, the use of manual synchronization was used. The good results obtained using the manual synchronization method support this decision.
C-4

THE MEASUREMENT RANGE
Initially, it appeared that the SWS antenna characterization tests could not be accommodated due to a conflict with another project in scheduling. However, a day long break in the radar test program provided the window of opportunity needed to conduct SWS antenna characterization. The short notice regarding turntable availability did not allow time, however, to remove all other vehicles from the test facility. Military vehicles and other man made signal reflecting sources were present. A reflective tree line around the turntable, combined with the man made objects, resulted in less than perfect measurements of the antenna pattern using the method to be described. One alternative to turntable testing was to characterize the receiver and transmitter antennas using the GTRI far field antenna range. However, this range does not allow testing of a system mounted in or on a vehicle because the receiver turntable is only capable of rotating several hundred pounds. In addition, that facility is approximately 70 feet up on top of a tower. Thus, the decision was made to proceed with turntable testing during the short window of opportunity.
Figure C-I shows the turntable test lay-out. The receiver vehicle was driven onto the turntable and centered so that the radar detector was centered over the center of the turntable. There were objects on the range that could generate multipath effects. The antenna range tower has a chain link fence around its base which reflected some energy. A vehicle was parked at the corner of the fence during the time that the receiver was tested on the turntable. A large semi-stationary military vehicle was located within 50 feet of the boresight path between the transmitter and receiver antennas. A stand of trees formed a tree line extending approximately 180 degrees around the turntable and at a distance of approximately 50 feet from the turntable. Plots of signal level versus rotation angle taken during testing do show some effects attributed to multipath from this range clutter when the received signal levels are low. However, even given these precautionary statements, the data produced by turntable measurements are thought to be very useable to assess the antenna response of both the receiver and transmitter.
The SWS detector was mounted in the receiver vehicle approximately 4 inches above the dashboard in the center of the windshield, as shown in Figure C-2. The MPH, Inc. mobile transmitter veh, icle was located 228 feet away. The mobile transmitter was mounted, as shown in Figure C-3, in the center of the roof of the transmitter vehicle on a suction cup platform, which allows an emergency or police vehicle to be simulated. The forward pointing transmitter antenna was aligned as closely as possible with the receiver vehicle on the turntable.
C-5

'"rranSln itter Vehicle

n

-----,... - L~argc Mjlitary Vehicle

....

I

0\

Path Length 228 l~'eet

--

Cjrass Covered Tllrlltable Antenna Test Range Tower

Receiver Vehicle - - _......
Figure C-I. Test layout of the turntable, receiver vehicle and transmitter vehicle '

Figure C-2. View from turntable showing transmitt~r vehicle and SWS mounting position
Figure C-3. Mobile SWS transmitter manufactured by MPH, Inc. C-7

Figure C-4 is a photograph taken from the test tower showing the receiver vehicle on the turntable. The turntable has been covered with sod as a result of the radar test measurements that were interrupted to allow the SWS to be tested. Figure C-5 is another photograph from the test tower showing the test vehicle 228 feet from the turntable. with the MPH, Inc. mobile transmitter mounted in the center of the roof.
Before the start of the tests, the unmanned mobile transmitter was turned on and allowed to stabilize to a frequency near 24.1 GHz. With receiver and transmitter operating, the turntable was rotated until the point of peak received signal strength was found. This point represents boresight between the receiver and transmitter and was referenced as zero degrees in azimuth. With the receiver and transmitter at boresight, SWS signal strength data were collected for approximately 90 seconds to allow the stability of the range environment to be characterized over time. Figure C-6 shows the result of this measurement. A trendline, comprised of a moving average of 15 samples, was laid over the raw data. Over the 90 second period during which over 1,800 samples of signal strength data were collected, an instability over time in the signal was noted to occur. The range over which the signal amplitude varies, peak to peak, is approximately 15 signal units or approximately 7.5 dB. The source of the instability is unknown.
The turntable operator and the receiver vehicle SWS operator were each equipped with two-way radios. The data collection and turntable motion synchronization was accomplished by the turntable operator providing a countdown to the start of turntable rotation. At one second from end of countdown, the operator in the receiver vehicle would tum on the BEL 855 STi which would start data collection at a rate of 15 signal strength samples per second. No turntable run exceeded 90 degrees of rotation. It was found the buffer memory of the data collection system could be exceeded if the data collection effort lasted much longer than 150 seconds. Also, the limitation of turntable rotation to 90 degrees per data collection run allowed the data collected during anyone run to be checked before going to the next run.
The rotation of the turntable was clockwise, referenced to a viewpoint looking down on the turntable froin above. Figure C-7 shows the BEL 855 STi antenna response from 0 degree~ to 90 degrees in azimuth. A trendline has been added to smooth the data by applying a moving average of 15 points. Referring to Figure C-7, and also referring to Appendix D on system calibration, a signal strength drop of 6 points equals a drop of power received by 3 dB. The half power or 3 dB beamwidth of the BEL 855 STi was of interest. The 3 dB point was found to occur on the side of the antenna main beam at 4 degrees off main beam axis. This number is normally multiplied by two and the resulting value, 8 degrees, is the
C-8

Figure C-4. View from GTRI antenna range tower of receiver vehicle on turntable
Figure C-5. View from GTRI antenna range towe~ of translnitter vehicle C-9

100

140
no

120

110

100

90

-s
~

f 70

(')

I

0

60

I=Setle61 ' . r......,,.........." ..1
L._~~~:.~:..'::~~ ..(~~~~~l.i

$ample pqtnts
Figure C-6. Boresight signal strength measurement to test propagation environment

150

14i}

130
un

110

100

9Q
II.>
'5 80
,~
I 70
n I 5()

SD

40

30

20

10

,,)

.- w
'<"

F1

..:> N

,,) ~1

~

;

&

P~9r~e$ or R(,latiofl

Figure C-7. BEL 855 STi antenna pattern over the interval of 0 to 90 degrees

C-12

3 dB beamwidth of the antenna. However, given the size of the receiver antenna, it was suspected that the real 3 dB beamwidth was wider than 8 degrees. Unfortunately, as will be documented later, the beam was not symmetrical around the beam peak and was without symmetry between the right and left sides of the beam, so the normal 3 dB beamwidth can not be specified with accuracy. The other observation regarding the data plotted in Figure C-7 was that the signal decreased in amplitude as the receiver vehicle is rotated from the boresight position of 0 degrees to 90 degrees, as expected.
Figure C-8 shows the antenna response of the BEL 855 STi when the turntable is rotated from 90 to 180 degrees. There is a noticeable peak in the antenna response at the turntable angle of approximately 130 degrees. The reason for this peak is not known. As the rotation angle reaches 164 degrees, there is a dramatic increase in signal level. It was first thought that reflections from the tree line located approximately 50 feet away could have been the cause of this signal enhancement in the back lobe of the antenna. However, Figure C-9 shows that the enhancement continues beyond 180 degrees to approximately 196 degrees rotation. The enhancement begins 16 degrees before the 180 degree point is reached and ends 16 degrees beyond the 180 degree rotation point. The appearance of this symmetry in the antenna response pattern indicates that the observed effect is not induced by reflections from the tree line but is a real response of the antenna to signals from the rear. This response to signals from the rear increases the probability that the SWS warning from a same lane overtaking emergency or police vehicle will be received earlier than if the signal arriving from the rear were rejected more by the antenna. When a rotation point of 270 degrees is reached, the signal level is approximately the same as that received when the turntable was tangential to boresight at an angle of 90 degrees. This result was expected and confirms that range effects are not biasing the results unreasonably.
Figure C-I 0 shows the antenna pattern over the angular increment of 270 to 360 degrees. This side of the pattern is not symmetrical with the 0 to 90 degree side of the pattern, but is much better behaved than the
o to 90 degree side of the pattern. If this side of the pattern was used to compute the half power antenna
beam angle, a beamwidth of 18 degrees would have been computed. While this measurement was closer to the 3 dB beamwidth thought to be correct, Mr. Glen Martinson, Chief Engineer, BEL-TRONICS LTD. was consulted to determine the measured half power beamwidth of the antenna used on the BEL 855 STi to confirm this estimate. He supplied a measured antenna pattern generated for the BEL 855 STi prototype antenna by an antenna test facility. He pointed out that the measured beamwidth shown for his prototype antenna was for the antenna in free space before the addition of the detector chassis board, infrared detector assembly, and plastic radome mounted over the antenna.
C-12

120. 130

100

00

.,
." 00
::l
1::
f

n

I

60

$0

40

1030
20

{)
St ~



,

r'---S1er5 ielpOer1. Mo\". AvS. (Serie&lj :( L .._~W WM~_._ W

W N

_

Figure C-8. BEL 855 STi antenna pattern over the interval of 90 to J80 degrees

is.:}

140

130

120

110

H)()

00-
.
ao "0
~ ,~
11.
,~ 70

n

60

I

.jO>.

50

4()

30
20-

10

()

- - .. i1

~

.-
Ql

~

N
~

t'o-
~

I")
N

................................ , ..... i--Seriesl I
l==~~_~f_~?~-~_~9. f~~~~lL

Figure C-9. BEL 855 STi antenna Pattern over the interval of 180 to 270 degrees

1~}

140

130

120

110

100

90

~ 80
::l
J 70

n

I

00

VI

50

40

30

~'O

16

()

~

~
N

~

~
N

N

I--

t!3 Fl

<"l
~

--Serie$l ~l~P.e.r ~~. ~V.9 (Senesl)

Figure C-IO. BEL 855 STi antenna pattern over the interval of270 to 360 degrees

The half power beamwidth of the BEL 855 STi prototype antenna, as measured in the antenna test facility, was approximately 24 degrees, a difference of approximately 5 degrees over that measured using the turntable technique and one side of the antenna beam. It was further speculated that, in addition to the effects of the mounting board, infrared detector assembly and radome, vehicle effects may have contributed to the non-symmetrical antenna pattern obtained using the turntable method of measurement. This conclusion indicated that the choice of making antenna measurements with the detector installed in the vehicle was prudent and these vehicle measurements may be more indicative of the actual response that will be experienced in the real world.
TRANSFERENCE OF BEL 855 STi RESULTS TO OTHER DETECTORS
It is in the interest of this study to determine if the findings regarding the performance of the BEL 855 STi test detector can be transferred to the other SWS compliant detectors being marketed by other manufacturers. Figure C-ll shows the antenna placement in a BEL 855 STi and two competing detectors. The BEL 855 STi test detector, packaged in its plastic enclosure, is shown in the bottom right for comparative purposes. The detector on the upper left is manufactured by Sunkyong and uses a horn antenna that is smaller than the one used on the BEL 855 STi detector, shown without its cover in the upper right. The detector in the bottom left is manufactured by Whistler. The antennas on the BEL and Whistler are very similar and have almost the same dimensions. It is speculated that the forward gain and half power beamwidth of the BEL and Whistler detectors are very similar. It is further speculated that the Sunkyong detector's antenna gain is lower and the half power beamwidth is greater that of the BEL and Whistler detectors. However, even with the differences in antenna design, the difference in performance between detectors is not thought to be significant enough to warrant field testing all detectors on the market (there is no RS-232 output available on the other detector).
In summary, the antenna pattern measurements that were described in the previous paragraphs provide the response of the SWS receiver as a function of angle around the receiver vehicle. The results obtained and presented are thought to represent the approximate response of a typical radar detector's antenna. These measurements can be used as a predictive tool in advance of field testing to determine what the measurements should demonstrate. For example, the longest range at which the SWS can be received will be when the transmitter and the receiver antennas are facing ~ach other (boresighted) over a line of sight path. The next best performance will be obtained when the SWS transmitter is overtaking a vehicle equipped with an SWS compliant detector. The next best performance of the SWS will be when the receiver vehicle is approached from either side by an emergency vehicle, such as when an emergency
C-16

vehicle crosses an intersection. Exceptions to these generalized rules will be due to unique site effects that are encountered while conducting field testing.
Figure C-II. Three different SWS compliant detectors and their respective antennas MPH, INC. MOBILE TRANSMITTER TURNTABLE TESTING The transmitter vehicle and the receiver vehicle changed places during mobile transmitter testing. The transmitter vehicle was moved to the turntable. Referring to Figure C-12, the mobile transmitter was secured to the transmitter vehicle's roof, using the suction cup mount in the approximate position where it would be mounted on an emergency or police vehicle. The transmitter and receivers were boresighted as before so that maximum signal was received. The point where the maximum signal occurred was referenced as 0 degrees, the starting point of rotation. A comparison of Figure C-13 with Figure C-6 shows that, during boresight testing, the maximum signal level shifted 20 signal strength units (10 dB) between the time that boresight testing was performed with the receiver vehicle on the turntable and the transmitter vehicle was transferred to the turntable.
C-17

Figure C-12. Mounting Point ofMPH, Inc. mobile transmitter on transmitter vehicle The reason for the change in the signal level is not known but an increase of 20 units of signal strength is significant and should be explained, if possible. The change may have occurred when the transmitter was moved to the turntable. A vehicle near the antenna test tower fence was moved at the same time. The movement of this source of reflections m!ly have resulted in the 20 unit change in signal level that was observed. The impact on the turntable measurements was next assessed. The BEL STi 855 antenna patterns were examined to determine if the shift occurred during those tests. If the shift occurred during the antenna characterization of the BEL 855 STi, as previously described, the test results would be invalid. Figure C-7 shows that when the detector rotation started, the peak signal level was 95. Figure C-IO shows that an ending boresight signal level of 95 was recorded. This observation provided reassurance that the shift in signal level did not occur during the detector rotation tests. Either a multipath null generated by reflections from the vehicle that was moved caused the level shift or the 20 signal unit peak was the result of better antenna alignment due to the swap in vehicle locations.
C-18

tt>o

'.4(}

1:30

1?O

110

10D

90

41 'Q

80

,~

Ii

~ If]

n

60

I

\0

50

40

30
20

10

()

--S~~':P~?

-

15 pt":r !vlov. A."9. (~.::(~,:~~~( 1)

Sample Point$
Figure C-13. Boresight data taken with transmitter vehicle on turntable

MOBILE TRANSMITTER ANTENNA CHARACTERIZATION
Figure C-14 shows the effects of rotating the mobile transmitter vehicle from 0 to 90 degrees. The main beam of the forward antenna is radiating the signal that is received during this rotation. A trendline moving average of 15 points is also laid over the raw data, as was done in the previous detector signal strength plots. The main beam half power beamwidth is approximately 22 degrees, assuming that the beam on the other side is symmetrical. This was found not to be the case and this effect is discussed in a later paragraph. Figure C-15 shows the response of both the forward and rear pointing antenna during rotation from 90 to 180 degrees. As the signal level from the forward pointing antenna decreases near 180 degrees, the rear pointing antenna is approaching boresight and the signal level climbs back to the starting signal strength level of 120.
Figure C-16 shows the response of the rear pointing antenna when the rotation is between 180 and 270 degrees. Referring to Figure C-17, the rotation extends from 270 to 360 degrees and the forward pointing beam becomes the predominant signal source. The signal strength peak returns to a level slightly below 120 when the point of rotation reaches 360 (0) degrees.
This operational impact caused by the non-symmetrical transmitter antenna pattern is thought to be minor. Slightly more power will be radiated to the left front of the emergency or police vehicle than toward the right front. Likewise, slightly more power will be radiated toward the right rear of the vehicle than toward the left rear of the vehicle.
In summary, the SWS mobile transmitter antenna's response as a function of aspect angle was measured and the on-axis peak signal strengths produced by the forward looking and rearward looking beams were almost identical. Although there is some non-symmetry around the main beam axis of each antenna, the difference in power radiated by the right and left lobes of either antenna is not of great concern. These measurements can be used as a predictive tool in advance of field testing to determine what the measurements will show. For example, when the transmitter is operated in a "clean" environment, relatively free of reflecting bodies, the longest range at which the SWS can be received will be when the emergency vehicle or police car is being approached by an SWS equipped motorist from the front or rear. The next best performance will occur when the police or emergency vehicle is overtaking the SWS equipped motorist. The next best performance of the SWS will be obtained when the emergency or police vehicle is crossing the SWS equipped motorist's path at a 90 degree angle.
C-20

150

140

1]<)

120

110

100

00

"iii 80
;:)

E

~
~

70

n

60

I

N

50

40

30

20

10

--Senes1 :-:--15pet~~}\...~(~~rJe,"1 )

>

i;:i>

r

,"'" ,.,"

'.if.

~

~

C)

&3

~

t'!

2

:g

Oe9,eei {If Rotatkm

Figure C-/4. Turntable transmitter antenna pattern for cut 0 to 90 degrees

150

'\/' }.'<i" .,:",';"".;:' ./ '. '>

}.:".'} I. "'.,.

',}"'}"."', / .... ',,}/.

\}

140

130

120 < ".,.'
.".. <I
110 ..',".:'"',''. '"
100 .....:.,

'" .

.'

'

00
':':.:,:<:,

'.....

~
'0 00
iE
I 70

n

I
tv

00

tv

,,-,

"<,;

40

:l<:l

20

10 /".

}
."".':

,<: "'.",

0
88

...........

--SeJies1 - 1 5 p"", Mo'i A.\l9 (S"'llesl)'

Figure C-15, Turntable transmitter antenna pattern for cut 90 to 180 degrees

150 14{)

1.30 120

110 100

00

..

~ &1

} 70

n

I

N
\,;.)

60

50

4IJ

30 20

10

-Slolfl#lil - 1 5 poer. MQ\I A\lg (Senes1)
Figure C-/6. Turntable transmitter antenna pattern for cut /80 to 270 degrees

150 ~Tr27?'> .

. ' . 1\,' .:".

"'. y

...'1).'\.,:

[<.L,y; I. ", >'\>'J":
. 110 I,: :.:
,", i" ....
100 l:;~'

I."'.

00 L:,> r

-.g.

n' ./",'
60 .

I1> 7 0 , ' c, L

~

n

. ,.

I

N

.j:::..

'. ro.; ..,. . .,.. ,. 1.<
,.'
4VrU' 6r!~
';.:'

10

'.'

::'

>'

,':

...
~

I'-
~

N
F1

~

g

<'
,.... ,~.

'."','"
... . ~.. ',.,
"i'C>.

"",.; L >ii:.

"-.. ~

'::'~

--$er1e51 ~~~pe{."~V~~~iS~n..s t)

".,."':,"'... "
.,:: ,

Figure C-/7. Turntable transmitter antenna pattern for cut 270 to 360 degrees

APPENDIX D CALIBRATION OF TEST RADAR DETECTOR
D-l

THIS PAGE INTENTIONALLY LEFT BLANK D-2

APPENDIX D CALIBRATION OF TEST RADAR DETECTOR
BACKGROUND
When the Safety Warning System (SWS) test program was begun, it was determined that the SWS would require testing on the highway. It was envisioned that the highway tests would involve the use of (I) a roadside SWS transmitter, (2) a mobile transmitter mounted on a transmitter vehicle and (3) a receiver system, capable of measuring signal strength of the SWS as a function of time (distance traveled) installed in a receiver vehicle. The roadside transmitter was provided for the tesi:s by Safety Warning System L. C. (SWS L. C.) The mobile transmitter was provided by MPH, Inc., a manufacturer of police radars with a license from SWS L. C. to manufacture and market the SWS mobile transmitter.
There were two choices of receivers to be used during the collection of signal strength data during testing. The first choice for a receiver was a GTRI owned Hewlett-Packard model 8562-A spectrum analyzer, complete with external mixer, that allowed operation of the analyzer at 24.1 GHz. This analyzer had the capability to very accurately determine signal strength in several units of measure and output the data to an external device through a digital interface. The interface between the analyzer and a lap top computer was achieved using a PCM-CIA computer interface card plugged into the laptop computer and a special cable that connected the analyzer to the PCM-CIA card. The PCM-CIA card was purchased and software was developed to control the analyzer and capture the data on the lap top's hard drive. The link speed was found to be extremely slow; the maximum rate at which the signal strength could be sampled was 2 samples per second. This sample rate was determined to be too slow to capture the rapid signal fades due to multipath in the highway environment.
The second option (and the one chosen) was to use the factory modified BEL 855 STi to provide signal strength data at a rate of 15 samples per second, or one sample every 66 milliseconds. There were other advantages in using a radar detector as the receiver other than the high sample rate that could be achieved. First, the use of an SWS capable radar detector as the receiver duplicated the type of system that was being evaluated. The modified BEL 855 STi detector's response to SWS signals is very close to the response expected from most other brands of SWS capable detectors. The STi 855' s linearity of signal strength measurements, the dynamic range over which the SWS signal could be measured, and the stability of the measurements over time were unknown but of concern. Laboratory testing of the modified STi 855 detector was conducted to quantify these unknowns.

LABORATORY TESTING
Laboratory measurements were made on the modified BEL 855 STi, using the laboratory test set-up described in the section that follows, entitled "Test Set-Up." First, the point at which the SWS signal was detected by the BEL 855 STi was determined by placing the detector in the laboratory test set-up. The attenuator of the Hewlett-Packard signal generator was set so that the radio frequency (RF) output level, before system losses, was -80 decibels referenced to I milliwatt (0 dBm). The power output was increased by lowering the attenuator setting until the detector alarmed and displayed a signal level 1 on its built-in light emitting diode (LED) display. It was found that the minimum signal generator output level required to activate the detector was -74 dBm. The detector alarmed at this level but did not display the SWS message. The signal generator attenuation level was lowered further in 0.5 dB steps until the SWS message was displayed at a level of -64 dBm. Thus, it was found that the detector will alert the driver at a power level IO dB lower than the signal level at which the,SWS signal is displayed. The signal level being output to the laptop computer was also monitored to determine the relative signal strength output reading to the computer by the detector when the warning message was displayed. This signal strength measurement at which the SWS message was displayed was found to be 20. During field testing of the system, the signal strength is plotted as a function of distance traveled by the test vehicle in motion. This measurement shows that any time the plotted signal level is above 20, the warning message would be displayed after 3 seconds required for the detector to decode and display the message.
CALIBRATION CURVE GENERATION
Next, a calibration curve was developed to determine the dynamic range and linearity of the detector. The received signal strength value that was sent from the detector to the laptop computer was recorded on the hard disk of the computer while the attenuator on the signal generator was stepped in 5 dB steps, starting at a level of -65 dBm. Per Figure 0-1, the attenuator was held at -65 dBm for 10 seconds. It was found that a signal generator output level of -65 dBm corresponds to a detector signal strength output level of approximately 15. After 10 seconds, the signal generator's attenuator was stepped to a level of60 dBm increasing the output power of the signal generator an additional 5 dB. It was observed that the signal level reading produced by the detector was approximately 28. The test continued by reducing the attenuation in 5 dB steps for IO seconds until the attenuator setting of 0 dBm was reached.
0-4

1t/)

140

l.j<)

UO

110

l 100

lOi
-I

'S ~) &
;,

C!

C (Y.)

~

tJ I
VI

..~ 10
'P
.. 00
IX
...,

l- SO

40

XJ

:<'0

1()

Figure D-l Calibration curve showing radar detector received signal level as a function ofsignal generator injection level

The data shown in Figure D-l supports the statistical theory of signals and noise. Specifically, as the signal power from the signal generator is increased and the signal moves further from the receiver noise floor, the standard deviation of the signal amplitude around the mean value decreases. - This effect is normal but demonstrates that the lower the signal level, the greater the uncertainty of the actual instantaneous signal level. For this reason, it was decided that when the data collected in the field were analyzed, a trendline would be overlaid on the data. The trendline would be a moving average using an averaging window width of IS samples. This corresponds to the averaging of approximately one second of receiver data. The use of the trendline overlaid on the data allows the very rapid fluctuations of the signal around the mean value to be observed while simultaneously providing a "smoothed" signal level so that the mean value of the signal can be easily determined.
The signal strength versus attenuator setting data in Figure D-2 was generated by interpolating between the signal steps of 5 dB that were plotted in Figure D- I. This interpolated data shows the response curve of the detector in a more easily interpreted format. Referring to Figure D-2, the data shows that the detector could respond in a linear fashion over a range of approximately 50 dB before signal compression started at a signal generator attenuator setting of approximately -15 dBm. These data also show that the signal was heavily compressed when the signal generator attenuator setting was - I0 dBm and higher.
The calibration curve is also useful for converting relative signal strength readings from the radar detector to dB, a more common unit used when characterizing signal levels. Referring to Figure D-2, the signal level increases approximately 10 units in signal strength for every 5 dB of change in signal generator attenuator setting. This relationship generally holds true over the linear portion of the signal output curve. This relationship also defines the resolution to which the signal can be determined in units of decibels. For every two counts of signal strength over which the signal output chqnges, the signal level has changed 0.5 dB.
A measurement of absolute power density received by the BEL 855 STi was not a goal of the calibration effort. Absolute calibration of the test chamber though, would provide a calibration curve that relates incident power levels to signal generator power output and could be accomplished by measuring the test system's gains and losses and the free space losses in the test chamber. To date, time has not been expended to develop an absolute calibration procedure because -the use of absolute power density is thought not to be meaningful to the primary audience of this report.
D-6

1y.) 0

140.0

130.0

1;).),)

1100

., 1000
?to<-l

'S 000
a~

~ 80.0
'

o

8 70 ()

I --.)

~

..'0
a-:~:

600

.~... 500

400

JOG

2(j{)

\0 ()

OJ)

~5

-80

.5S

.51.)

45

Signlll G~I~rilt(Jf InJ~1iml l~t\lel dBm

Figure D-2 Calibration curve showing radar detector received signal level as afunction of signal generator injection level

The remaining question to be answered by laboratory testing was, "Is the calibration curve stable over time?" The stability of the calibration curve of the STi 855 was determined by repeating the test a week after the first test was conducted. An almost identical response curve to the one shown as Figure D-l was produced one week later. However, in accordance with good practice, it was determined that the detector would be calibrated in the test chamber before or after every field test.
SUMMARY
In summary, the linearity, dynamic range, and stability of the modified BEL 855 STi detector was found to be totally adequate to allow the detector to be used as the primary receiver for the test program. Several highway tests of the system also demonstrated that the sample rate of a 15 signal strength reading per second was adequate to capture the rapid fluctuations in signal level that occur when multipath effects are present. These findings indicate that a high level of confidence can be placed in the results that are obtained during field testing.
Test Set-up Documentation
The laboratory set-up, discussed in the previous section of this appendix, is used in the testing of all elements that comprise the SWS, including the roadside and mobile transmitters and other manufacturer's SWS capable detectors.
The primary equipment available to the project during laboratory and field testing consists of:
1. Hewlett-Packard Spectrum Analyzer model 8562B to record the frequency and signal level as a function of time, using an HPIB interface and laptop computer (slow data rates only).
2. BEL 855 STi radar detector with special factory installed RS-232 port to provide frequency and signal level data output as a function of time to a laptop computer data collection system.
3. MPH, Inc. experimental mobile transmitter modified by the factory to transmit, upon command, 'a continuous wave (CW) carrier or one of two safety warning vehicular messages. Also, a vehicular roof top MPH, Inc. transmitter mounting bracket, that uses bungee cords and sucl;on cups to hold the unit on any vehicle. This set-up has been built to allow the transmitter to simulate emergency-lightrack mounting of the transmitter on any vehicle used during testing.
4. Two experimental roadside transmitters obtained from Safety Warning Foundation L. c., each
capable of transmitting anyone of 64 fixed text messages or 'a CW signal after field programming. The roadside transmitter can be mounted on signs, or on an overhead structure. Each contains a wet cell battery with approximately 14 hours of operational capacity.
D-8

5. Custom software developed by the Georgia Tech Research Institute (GTRI) to interface with the Hewlett-Packard 8562B spectrum analyzer. The custom software requests data from the spectrum analyzer, and stores the received data into a file that can be retrieved for analysis after testing. A data rate of two samples per second is typical.
6. Custom software that has been developed by the GTRI to interface with the BEL 855 STi to provide signal strength and the index number of the safety warning message.
7. A stock radar detector that has been modified to provide the 10.7 MHz intermediate frequency (IF) output and the output of the quadrature frequency modulation (PM) discriminator signal of the safety warning transmitter (SWT). The radar detector is tuned manually to the SWT signal. The output of the IF amplifier can be fed to the spectrum analyzer, or to a high speed digitizer for modulation signal capture and examination.
8. A laptop computer for recording test data.
9. A small absorber lined box that serves as an anechoic chamber and a 24 GHz calibrated signal injection system (to be described) that allows a signal to be radiated and controlled. This system allows a radar detector's sensitivity to be measured and calibrated, the mobile SWT transmitter's power to be measured very accurately, and the radiating systems to be tested in a controlled environment.
10. Over 30 radar detectors, many SWS capable, manufactured by BEL-TRONICS LIMITED, Sanyo Tecnica USA, Inc., Sunk-young America, Inc., Uniden America Corporation, and Whistler, Inc. Many are SWS capable systems.
II. A video camera mounted on the vehicle carrying the safety warning receiver (SWR) and spectrum analyzer.
12. Two recording Global Positioning Systems (GPS) to determine the range between the vehicle carrying the SWT and the SWR, as well as the range between the roadside transmitter and the vehicle carrying the SWR.
LABORATORY CALIBRATION OF EQUIPMENT
Calibration of the equipment to be ,ested is performed before any equipment is taken out of the laboratory for field testing. The size of the GTRI test chamber requires that all testing and calibration be performed in the near-field of the antenna.
Test Set-up
Figure D-3 shows the laboratory set-up that is being used to test the sensitivity of any radar detector supplied as an off-the-shelf product from any of the five manufacturers of SWS capable detectors. The laboratory set-up shown in Figure D-3 allows samples from each of the radar detector manufacturers to be tested before Federal Highway Administration/Georgia Department of
D-9

Transportation (FHWAlGOOT) field testing is conducted to determine best and worst case detector sensitivity on the basis of current product performance. After the laboratory sensitivity measurements have been made and field testing has determined the absolute transmitter signal level versus range that can be reliably expected, the performance of each manufacturer's detectors cari be extrapolated from the resulting data.

Radar Detector Under Test

24.1 dB Standard Gain

-63.4 dB Space Loss

Hom

Across the Path 123.19 cm - -.~.~

Camera

1 Absorber Lined

Microwave Test Chamber

I~~

SWS L. C. Transmitter
I
50 dB attenuator
~MAto WR-42 Transition / t

K-band External Mixer Assembly

Television Monitor

Hewlett-Packard Model 8562A Spectrum
Analyzer

Modulator from SWS L. C. Transmitter

Hewlett-Packard Model 83640A Signal Generator

Figure D-3. Laboratory set-up used to conduct single and dual SWS transmitter testing

Referring to Figure 0-3, the test set-up is designed to conduct both single and dual SWS transmitter signal injection testing. Single transmitter testing can be conducted using the set-up shown in Figure D-3 to calibrate receivers and conduct comparative evaluations of receiver sensitivity and operability. Dual SWS signal injection testing allows the combined effects of two SWS transmitters, operating in close proximity to each other, to be evaluated. For example, when a police vehicle and an emergency vehicle are in close proximity, such as at an accident site, and both are operating their SWS transmitter, it is suspected that there will be mutual interference. The dual transmitter test set-up shown in Figure 0-1 will allow the mutual interference effects to be evaluated. However, the dual transmitter testing is not discussed further in this appendix and the reader is referred to Appendix B for a more complete discussion of dual transmitter testing.

D-lO

SINGLE TRANSMITTER TESTING
The single transmitter sensitivity testing is being conducted using the Hewlett-Packard Model 83640A synthesized signal generator. A SWS. modulator is connected to the external modulation of the signal generator. The Hewlett-Packard signal generator is adjusted so that the maximum deviation during modulation is 5.0 MHz around the center frequency of 24.100 GHz. The use of the Hewlett-Packard signal generator allows the frequency to be changed without affecting the amount of modulation deviation and symmetry, a feature which can not be reliably achieved using a mechanically tuned Gunn device transmitter. The ability to easily change signal source frequency allows the sensitivity of the receiver under test to be determined at the SWS band limits (25 MHz from the 24.1 GHz center frequency).
The 24.1 GHz radio frequency signal is fed from the signal generator to an SMA to WR-42 wave guide transition. Measurements show that approximately 3.4 dB is lost between the signal generator and the WR-42 waveguide. The transition feeds a 20 dB sidewall coupler. The coupler's direct path goes to a second 20 dB sidewall coupler and then into a 24.1 dB gain standard horn antenna. A 50 dB attenuator and SWS L. C. Inc. transmitter is connected to the horn antenna through the 20 dB arm of the last sidewall coupler in the chain (used only during dual transmitter testing). The external mixer of a Hewlett-Packard Model 8562A spectrum analyzer is connected to the 20 dB arm of the second sidewall coupler to monitor the spectrum and levels of the signal(s) being transmitted. The energy from the horn antenna illuminates a test chamber that is lined with radar absorbing material. The measured path loss across the 123.19 centimeter (cm) long test chamber (to the receiver test port) is approximately 63 dB 2 dB. Measurements must still be conducted to determine the absolute power density at the receiver test port.
The receiver port is a square hole cut into the test chamber at a point that is aligned with the standard gain horn on the other side. The receiver under test is placed on a plexiglas support during testing. The receiver's antenna is placed approximately one centimeter inside the chamber during testing. A low light level television camera is focused on the display of the receiver that is under test. A television monitor displays the enlarged receiver display to the test operator so that the reaction of the receiver to signal stimulus can be monitored at a distance. The Hewlett-Packard signal generator has an electronic attenuator that is calibrated to the nearest tenth of a dB. Attenuation in excess of 110 dB can be introduced using the front panel attenuation control.
D-Il

APPENDIX E TESTING OF SWS AT DOBBINS RESERVE
AIR FORCE BASE, GEORGIA
E-I

THIS PAGE INTENTIONALLY LEFT BLANK E-2

APPENDIX E TESTING OF SWS AT DOBBINS RESERVE
AIR FORCE BASE, GEORGIA
DESCRIPTION OF SWS TESTING AT DOBBINS
The Safety Warning System (SWS) was tested at the Dobbins Reserve Air Force Base, located in Marietta, Georgia on 28 May, 1997. Dobbins testing was not a planned activity to be conducted under the present contract. However, when a task shifted away from SWS human factors evaluation at Oak Ridge National Laboratory, testing at Dobbins became an preferred optional task for three reasons: (1) additional time was made available to allow more thorough SWS testing in general, (2) the SWS could be tested on a 1,600 foot section of unused runway without concern for the safety of personnel testing the system (since there was no highway traffic) under low speed conditions that allowed high resolution signal strength data to be collected, and (3) the environment was clear of vehicles which could cause unknown effects on the data being taken. In addition, Dobbins testing allowed all of the equipment to be used during highway testing to be checked out and the reliability of each system determined. After test results were analyzed, it was found that the Dobbins tests were a good investment because they provided a performance baseline against which the effects of testing the SWS under highway conditions can be measured.
Two types of tests were conducted at Dobbins. The first set of tests (to be described later in greater detail) were conducted on an abandoned runway for the purpose of determining the free space performance characteristics of the SWS, given that there were few objects to reflect the SWS signals and the runway was flat so that line of sight could be maintained between the receiver vehicle and the transmitter vehicle. A second set of tests was performed on a Dobbins patrol road to determine how well the system would operate when the transmitter vehicle was at the bottom of a hill and the receiver vehicle was approaching. Elevation relief maps with contour profiles of two feet were available to allow site documentation to be developed.
TEST SITE AT DOBBINS RESERVE AIR FORCE BASE
The test site at Dobbins is a 1,900 foot long, ISS foot wide section of unused concrete runway. The runway is approximately 30 degrees off of a north and south alignment. The heading of the test vehicle was approximately 330 degrees when making a run from south to north and approximately 150 degrees
E-3

when the test vehicle was making a run from north to south. A Dobbins fire department training facility is presently located at the south end of the runway and that facility takes approximately 300 feet of the 1,900 foot existing runway. To the north, an active taxiway intersects the runway. Georgia Tech Research Institute (GTRI) personnel were asked to stay south of the active taxiway in order to avoid an incursion with an aircraft. Given these boundaries, there remained an approximate distance of 1,600 feet that was clear of obstructions and available for use during testing. Trees have been cleared approximately 500 feet away from each side of the runway, a feature that allowed two global positioning satellite (GPS) receivers to be operated even when the satellites in the constellation are near the horizon.
TEST INSTRUMENTATION
Two vehicles were used during both tests: (1) a transmitter vehicle, and (2) a receiver vehicle. The transmitter vehicle, driven by Mr. B. H. Hudson, Senior Research Technologist, GTRI, was equipped with a mobile transmitter as shown in Figure E- I. A flat plate mount, equipped with suction cups, was developed by GTRI project personnel to allow the SWS mobile transmitter, manufactured by MPH, Inc., to be mounted on the top of the vehicle at the same approximate point where a transmitter would be mounted on a police vehicle. The antenna on the mobile transmitter radiated half of the 30 milliwatts output power forward and half rearward (15 milliwatts in each direction). The mobile transmitter antenna gain was approximately IS dB. The antenna radiation pattern was broad, as shown in more detail in Appendix C of this report. The transmitter operating frequency was 24. I GHz. The transmitter vehicle was equipped with a GPS receiver linked to a laptop computer which was used to record the GPS data to determine the transmitter vehicle's location during testing.
The receiver vehicle, driven by the author, was equipped with an a BEL-TRONICS LIMITED Model 855 STi SWS capable radar detector modified by the manufacturer to provide signal strength data to a laptop computer via an RS-232 link from the radar detector. The laptop recorded approxifYJately 15 samples of signal strength data per second. The radar detector was calibrated in the laboratory before testing and the calibration data are contained in Appendix D of this report. Figure E-2 shows the radar detector mounted in the receiver vehicle in the center of the windshield, approximately four inches above the dashboard, a position that is typical of where drivers normally mount their radar detectors. The laptop computer used to record the signal strength data was positioned on the front seat.
E-4

Figure -1. Mobile transmitter manufactured by MPH, Inc. mounted on transmitter vehicle
Figure -2. Radar detector mounted in the receiver vehicle in the center of the ~t'indshield, approximately four inches above the dashboard, data collection system in passenger seat
E-5

In 'addition to the radar detector and the data recording system, an 8-millimeter video tape recorder (VTR) camera was mounted outside of the receiver vehicle on the rear window as shown in 'Figure E-3. The camera recorded the test scene, normally in front and ahead of the receiver vehicle. The VTR was on a swivel mount that could be turned when necessary to record the approach of the transmitter vehicle from the rear. A microphone was mounted on the driver's seat and was connected to the audio channel of the VTR, so that comments regarding the tests could be recorded on the audio track of the video tape. Also, when the radar detector was turned on, it was possible to hear the detector's beep that occurs at the point where it begins producing data. These video scene data and the soundtrack recording provided valuable information regarding the test run identification number, problems encountered during the test run, and the actual point at which the radar detector began producing data.
Figure -3. VTR camera mounted olltside of the receiver vehicle on the rear window The receiver vehicle also carried a GPS receiver connected to a laptop computer used to record GPS ,derived location data as a function of time. This system, shown in Figure E-4, was operated by Mr. David Jared of the Georgia Department of Transportation (GDOT), who assisted with the tests. The recording of GPS data was started when the radar detector began. recording data and was stopped when the radar detector was turned off at the end of the run.
E-6

Figure -4. CPS receiver connected to a laptop computer used to record CPS derived location data as afunction of time
GLOBAL POSITIONING SYSTEM TRACKING DATA
GPS data, recorded during the Dobbins Test 5 (the first test) is shown in Table E-l as an example of the type of GPS data collected by each vehicle. Referring to Table E-l, the date of the test appears in the leftmost column followed by the time of observation, as derived by the GPS in the receiver vehicle. The time entry is followed by the receiver vehicle's latitude and longitude. The receiver vehicle was moving during Dobbins Test 5 and the speed of the receiver vehicle (in knots) appears in the next column followed by the vehicle's heading referenced to true north. A calculation was performed to convert the speed data to the receiver vehicle's rate of travel in feet per second. The next three columns show the data collected by the transmitter vehicle's GPS. The time that the transmitter vehicle's latitude and longitude was sampled is shown first. Next the latitude corresponding to the time is shown, followed by the longitude.
The GPS data were recorded once every two seconds. After the tests were conducted and the GPS data were reduced, it was found that the time synchronization between the GPS in the receiver vehicle and the GPS in the transmitter vehicle was different. For example, the receiver vehicle's GPS sent data to the recorder on the odd second and the transmitter vehicle recorded data on the even second. A better GPS synchronization process will be the subject of experiments before future highway testing is undertaken.
E-7

Table -1. Receiver Vehicle and Transmitter Vehicle GPS Files

Receiver Vehicle

Transmitter Vehicle

Date Time Latitude Longitude Knots Heading Feet/Sec Tima Latitude Longitude

280597 191521 3354.393 8430335 97 325.6 HU714 1\:11522 3354.569 843CL453

280-597 11'1523 3354.398 8430338 9.8 326A 16.5402 191524 3354.569 8430453

280597191525 3354404 enO.342 10 327.6 16.8778 191526 3354.57 8430452

280597 19'\527 3'354.403 8nO.346 10 327.6 15.8778 191528 3354.57 8430.453

280597 191529 3354.413 8430.349 9.9 329. 1 16.709 191530 3354.57 8430452

280597 191531 3354.419 8430.352 10

331 16.8778 191532 3354571 84~m453

280597 '191533 3354.424 8430.355 99 3316 16.709 191534 3354571 8430..453

28058'1 191535 3354.429 8430.358 9.8 331.8 16.5402 191536 3354.571 8430452

28.0587 191537 3354.434 8430.361 9.6 33t.5 "to.2027 191538 3354.57 8430.452

280597 191539 3354438 8430.365 93 331.2 156963 191540 3354.57 8430,452

280S97 191541 3354.443 8430,368 9.1 :530.7 153588 Hl1542 2354.57 8430451

280S97 191543 3354.447 8430.371 91 330.6 153588 191544 3354.571 8430A52

280597 191545 3354.4~,2 8430.374
2:80597 191547 3354.456 8430.377

9 330.3 9 330.3

1519 15.19

191546 3354.571 8430.452 191548 3354.572 8430,452

280597 191549 3:154.461 8430.38

9 330.3 15.19 191550 3354.572 8430,452

280597 191551 3354454 8430382

9

331

t5.19 191552 3354.572 8430,452

280597 191553 3354.469 8430.385

9

331 15.19 181554 3354.573 $430452

230597 191555 3354.472 8430388 8.7 330,8 14.6837 191556 3354.573 8430452

280597 191557 3354.476 84:30 391 8.7 330,8 14.6837 191558 3354,575 8430452

280597 191559 3354.481 8420.394 8.7 330.8 146837 191500 3354S74 8430452

2:80597 191601 3354.485 8430.397 $.6

331 145149 19 i 602 3354.573 13430.453

280597 191".303 3354.488 84:30399 8.6 332.4 14.5149 191304 3354.573 8430453

280597 191605 3354,491 8430.402 8.4 332.3 14.1773 191606 3354.572 8430.454

280597 191607 3354,495 B4:30.404 8.4 332.3 14,1773 191608 3354.57 8430.455

250597 191609 3354.499 84:30.407 8.4 332.3 14.1773 191610 3,354571 ~W30455

280597 '191611 3354.503 8430409 8.4 332.3 141773 191612 3354.57 84:~O.456

280597 191613 3354.507 8430.412 8.4 :332.3 141773 191614 3354.571 8430.456

280597 191615 3354 51 8430.414 83 332.2 140086 191616 3354.572 8430.456

280597 191617 3354.514 8430417 83 3322 140086 1916~a 3;354572 8430.457

280597 191619 33545t9 8430419 8 :3 332,2 14.0086 191620 3.3540,573 3430458

280597 19~621 3354521 843042.2 82 3:32,1 13.8398 191622 3354573 5430453

28059'7 19,623 3354.525 84?,O.424 82 332.1 13.8398 191624 33S4574 e.,430A58

280597 191625 3354.528 84?,O.428 8A 332.5 141773 191626 3354574- e.430A58

230597 191627 3354.532 8430.431 8.4 332,5 14.1773 191628 3354.575 8430458

2805-97 '~9'1629 3354.537 8430.433 8.4- 332,5 14.1773 191630 3354.575 2.430.451

2~,O597 Hii631 3354.539 8430.438 8.6 3326 145149 191632 3354.575 8430.45'7

2>30597 \91633 3354543 343044 86 3326 14,5149 191634 3354.575 8430.457

2.90597 1(1635 3354547 8430443 8.7

333 146837 191636 3354.575 8430.457

280597 191637 33.54 551 8430.446 87

:533 '14683'7

191538 3354.575 84304~>7

280597 19i639 33,54555 8430,449 87

333 ~46837

19~540 3354.576 843045S

280597 191641 3354 550 M30.452 8.9 3133 15.02"12 Fd642 3354.576 8430.458

280597 19,643 3354.562 8430.455 8.9

~"'>I,~","-'>"

.....
,J

150212

191644 3354575 8430.457

280597 191545 3354566 8430.458 8.9 32,3.3 15,0212 191646 3354575 8430.457

280597 191647 3354.571 8430.461 B.9 333.3 150;"12 191648 3354,574 8430.457

280597 191649 3354.575 8430.463 B.9 :~33.3 150212 191650 3354.574 8430458

280597 19H351 3354.579 8430467 89 3336 1502:12 1';1652 3354574 8430.458

280597 191653 3354.583 843CL47 8.9 333.6 '15.0212 191654 3354.573 8430458

Sum ;:

706.504 A\(~rage ::t 3354,573 8430.455

Total "" 141301

E-8

The GPS data were recorded to allow the location of the moving vehicle to be determined in relation to the other vehicle, also to determine the absolute location of both vehicles in relation to the runway. Also, the stability of the GPS system could. be determined by observing the drift in the GPS location of the stationary vehicle during each test.
GPS receivers used in civil applications are not capable of receiving and decoding the high resolution Department of Defense (DOD) GPS signals that provide a location accuracy to 1 meter. As a result, the errors in absolute location of a non DOD GPS receiver can be as great as 100 meters. The amount of positional wander or positional drift over time is dependent on a number of factors including: the number of satellites visible to the GPS receiver at the time of measurement, the signal to noise ratio of each received satellite's signal, and factors related to GPS signal propagation. The Dobbins test allowed typical GPS errors to be characterized by examining the GPS data file of the stationary receiver to determine the rate and magnitude of positional drift. This same positional drift was included in the GPS data collected by the moving vehicle, but was not as apparent due to the vehicle's motion induced change in latitude and longitude. For example, using the data for the transmitter vehicle (stationary) shown in Table E-l, the wander in the positional data derived by GPS can be shown by plotting the latitude of transmitter vehicle over time. Figure E-5 shows latitude wander (in feet) over a period of approximately 92 seconds when Dobbins Test 5 was being conducted. Figure E-6 shows the corresponding drift in the longitude data collected by the GPS for the same period of time.
The error in absolute position of the stationary vehicle derived by GPS was determined by plotting the GPS position on a 1 inch to 100 foot scale topographic map furnished by the Dobbins Civil Engineering Department. At no time were errors as large as 100 meters found in the GPS derived absolute position. However, errors in absolute position of the stationary vehicle on the order of 100 feet were experienced during several of the test runs. Comments regarding GPS system stability during testing are included in the test analysis that follows.
RUNWAY TESTING SCENARIOS
The runway tests were designed to determine the technical performance of the SWS receiver in a relatively "clean" environment without other vehicles in the traffic mix causing severe multipath and attenuation of the SWS signal. Three types of tests were conducted on the runway. The first measurement, Dobbins Test 5, was conducted to determine the range at which an SWS receiver would detect and decode the safety message when the receiver vehicle approached and passed the transmitter
E-9

45

41)

:15

'

: :J(J
...
"~> ;l5
C-

~

Q

V) 2() ...

.~

tTl
..I...
0

g~. 15

i..l.

l.l..

,6

CI 10

1:
Cl

;<:;

()

f~

Figure -5. Latitude wander (in feet) over a period ofapproximately 92 seconds when Dobbins Test 5 was being conducted

r~'&e;~1 ~
L .....
tTl
....I........
Figure -6. Corresponding drift in the longitude data collected by the GPS in Dobbins Test 5

vehicle parked on the side of the roadway. The second measurement, Dobbins Test 8, was designed to determine the range at which the SWS warning message could be detected and decoded when the transmitter vehicle overtakes the receiver vehicle, simulating the overtaking of an SWS equipped motorist by an emergency vehicle in the same lane. The third test, Dobbins Test 10, was designed to determine how well the SWS receiver would respond when the emergency vehicle is crossing the path of the SWS receiver vehicle at an offset distance of approximately 105 feet. The fourth measurement, Dobbins Test 12, was a repeat of Dobbins Test 10, except the offset crossing distance was reduced to less than 20 feet.
DOBBINS TEST 5
Figure E-7 is a diagram of the test layout for Dobbins Test 5. Referring to Figure E-7, the transmitter vehicle was stationary on the east side of the runway heading approximately 330 degrees. The transmitter vehicle was radiating a continuous wave (CW) signal in both the forward and rearward direction. The receiver vehicle started the approach at a distance approximately 1,310 feet from the transmitter vehicle on a south to north path with a GPS derived heading of 332 degrees. A small off-set in the test path was established so that when the receiver vehicle reached the transmitter vehicle, the separation between vehicles was less than 10 feet. This small offset distance allowed the receiver and transmitter beams to remain almost at boresight until the receiver vehicle passed the transmitter vehicle. The receiver vehicle approached and passed the transmitter vehicle and continued for a distance of approximately 103 feet beyond the transmitter vehicle before the navigation and signal strength data collection was stopped.
Table E-l, presented earlier, shows the nayigation data developed by both the receiver and transmitter vehicles. Referring to Table E-l, the run time was from 19:15:21 hours until 19:16:53 hours, a span of 99 seconds. During this period, the receiver vehicle approached the transmitter vehicle at a speed that varied between 8.2 and 10 knots. The speed data were converted to rate of travel data. The rate of travel varied between 13.8 to 16.9 feet per second. Next, the rate data were summed over time and it was determined that the vehicle had traveled a distance of 1,413 feet; the summed rate data were multiplied by two to account for the fact that the rate data is sampled every two seconds. The GPS derived test run distance of 1,413 feet was found to be close to the distance of the test run determined from physically measuring the length of runway between the start point and the ending point for Dobbins Test 5.
E-12

Dnbbins Test .5

N

rt

i

Stational'v 'f'ransrnitter .,;
Vehicle

'Totall~un 1,,t(11"."1'

1~3 10'

.....__ Section of Unused Run\vay at [)obbins

I I
i
I
" ....

rVlovirH! Receiver Vehicle ..(,.",. '

Figure -7. Diagram of the test lay-out for Dobbins Test 5

Laboratory tests on the ST 8551 detector show that it will display the warning message after 3 seconds of receiving a level of 20 in signal strength units. Those same tests also show that once the message has been detected and decoded, the message will continue to be displayed 3 seconds after the signal level has dropped below the threshold of 20 in signal strength. Figure E-8 shows receiver signal strength displayed as a function of distance traveled by the receiver vehicle. The receiver sampled the signal strength approximately IS times per second, a rate fast enough to ensure that all multipath effects could be recorded. A trendline was inserted onto the data. The trendline represents a moving average of IS points and has the effect of smoothing the data. Referring to Figure E-8, the starting signal level is 75 at the starting range of approximately 1,310 feet from the transmitter vehicle. This signal level represents a signal to noise ratio of approximately 23 dB (see the calibration chart in Appendix D) above the signal level of 20 required to first initiate the SWS message detection, decoding and display to the driver. Multipath does not effect the signal level until the receiver vehicle has traveled approximately 1,025 feet toward the transmitter vehicle, at which point a very deep multipath null occurs and the signal drops below the level where the SWS can detect and decode the SWS message. However, this first null is below the message detection and display threshold of 20 signal strength units for less than three seconds which is not enough time to cause the SWS message to go blank on the display. The severity of the multipath nulls that occur even closer to the transmitter vehicle are off-set by the fact that the signal strength has increased to a level where the signal does not drop below the detection and display threshold value of 20 again. When the receiver vehicle is passing the transmitter vehicle after traveling a total distance of approximately 1,310 feet, the signal drops very rapidly as the receiver vehicle enters the sidelobes of the transmitter vehicle's antenna and the transmitter moves into the backlobes of the receiver's antenna. The receiver vehicle continues to take data an estimated 103 feet after passing the transmitter vehicle.
In summary, Figure E-3 shows that there is a signal to noise ratio of approximately 23 dB above that required by the receiver vehicle's receiver to detect and decode the SWS message when the receiver vehicle starts the tests at a range of 1,310 feet from the transmitter vehicle. Multipath was not severe over the test run, indicating that when an emergency vehicle using an MPH Inc. transmitter is parked on the roadside and line of sight is maintained, the SWS warning will be available to an approaching motorist a minimum of at least 1,310 feet in advance of reaching the emergency vehicle. At highway speeds of 55 miles per hour, this represents a warning time of 13 seconds, even when an additional 3 seconds is required for detection, decoding and display of the warning message by most radar detectors.
E-14

100

Ji lilQ 80

1

..,2'
It.!

i'Q

i 00

It:

50

40

30

l~s~$l
1-15 :. ,..... per Mov. A~-g IS~H~ri1}'.~

Figure -8. Receiver signal strength displayed as afunction ofdistance traveled by the receiver vehicle

These data predict that this minimum range and warning time will increase dramatically during highway testing when the distance over which testing can be conducted increases to several miles.
DOBBINS TEST 8
Dobbins Test 8 was designed to evaluate how well the SWS detector would detect the transmitter vehicle overtaking the receiver vehicle, a scenario that simulates a police or emergency vehicle overtaking an SWS equipped motorist. Figure E-9 shows the layout used in this test. Referring to Figure E-9, the layout is the same as the previous test, except that the receiver vehicle takes the place of the stationary transmitter vehicle. The heading of the stationary receiver vehicle is approximately 330 degrees. The receiver vehicle is approached from the rear and is overtaken by the transmitter vehicle on a GPS indicated heading of approximately 330 degrees. Table E-2 was developed to allow the receiver and transmitter vehicles combined GPS data to be studied. The distance traveled by the transmitter vehicle was computed as 1,251 feet, a run distance slightly shorter than during Dobbins Test 5. This run distance was determined by summing the rate information from the transmitter vehicle's GPS file. It was determined that because approximately six seconds of data are missing from the transmitter vehicle GPS file, the summation process would produce inaccurate results and as a result was not used to compute the test run distance. Instead, the GPS derived longitude and latitude data from the receiver vehicle's GBS were used to plot the start and end points on the 1 inch to 100 foot scale map supplied by the Dobbins Civil Engineering Department. Using the plotted location, it was estimated that the total test run distance was 1,340 feet, which is in close agreement with the previous run distance calculation; however, signal strength data were recorded over only 1,275 feet of the run, as recording was stopped when the transmitter vehicle was parallel with the receiver vehicle.
Figure E-l 0 shows the signal strength that was recorded during the Dobbins Test 8 run over a distance of approximately 1,275 feet. The signal level generated when the transmitter is located approximately 1,275 feet from the receiver vehicle is much lower than in the previous run, with justification. The antenna of the SWS receiver is pointing forward, and away from the transmitter vehicle that is overtaking from the rear. A trendline has also been added to Figure E-lO that averages 15 SWS signal samples for each point plotted on the trendline. The signal level indicated by the trendline remains 10 points above the SWS signal detection and display threshold of 20 units of signal strength until the transmitter has traveled approximately 725 feet toward the receiver vehicle. At that point the signal drops below the detection and display threshold of 20 units of signal strength. This drop is thought to be caused by multipath effects.
E-16

N
t
I
I
-t'I1 I
-..l

Stationary H.eceiver Vehicle

1,340~

l,275'

~.~_. _ Section of Unused Runway at Dobbins

1 --- Moving c-.f'ransnlitter

4Jl(

155'

\lehicle

Figure -9. Dobbins Test 8 test layout

Table -2. ReceiverlTransmitter Vehicles Combined CPS Data

Receiver Vehicle GPS Data

TIme latitude Lt:l11glttld.

:80597. t93~~ 3364.567 8430.401

26'"A;587 ~S33S() 3354.507 8430Ac1

2'5rJ5G7 ~83352 33f>4568 flA30.46

2SOf;.&7 ~ G;33.,.')4 3-.-:e>4. 566 84.30046

2SQ5.9'! ~ S"J3$ 33-."4 51')$ 8430 4$

280597 t93358 3354fllS6 84:lC4B

2sos..<17 t93400 3354565 842C.46

25C697 193402 3354585 8430.4.."'9

280S97 tS34G4 33f>4.584 843C4S9

1&3400 3354563 84.30459

280597 1934f1:l 3364562 84]{i459 ZOC:5S7 193410 3354,502 ~30.4Se

28469-7 193412 :n:::4.5$1 E'A304::.::, 28t-"<F.A/7" ~g34~4 2.::54561 8430458

:80697 t9~jA'B 3354,56 84:3O.4">8

2B05&l 18:.;4 H~ 3-354 50 8430A5S

2"Bl.'i?~7 193420 3:i54.559 6'A30.4::S

2~C597 1S3422 :3.304.555 84.30.457

2e,o:::~97 !93424 3364.558 !2,430.457

280587 193426 3J5455l 8430.457

2&J...~7 193428 ;;}354.550 8430457

2OC~S7 183430 33f.>4555 8430.457

2BC~&7 183432 :;,3-.".4. 55<t 84J0.457 2~vS7 19'3434 3354.554 5430457

280597 19:>438 335ot553 42D,457 280597 19:3433 3354552 S4~-\O.457

260597 193440 3;354.551 8430.457

2&:$7 '~93442 3354.~~ 8430457

193444 3354.548 8430.456

19:~ 3354.547 &t:30,45$

193448 3354.545 8430,456

28G'6fEfl 183450 33.54.545 l:430A55

280597 193452 3354. 543 8430454

21:<:)$7 WJ4-.'".4 3354543 5430454

i.BfJ5S7 19345 3354 542 842D.454

l~..s7 19345-9 3354 541 8430452

2OCf:G7 193500 33f>4 539 8430.452

2OC,E,97 19350Z 3354 539 8430.452

2~{)t:@l 1S35t\4 3354 53.9 8430.451

2e,0597 1935C 3354.538 8430.451

280597 19354)8 33C''tJ.537 &43045

19351D 3354,536 2;4'30.45

193512 3354.535 &430 45

:~:S;':S87 1S3514 33S4.5J4 &43045

26CS97 193516 J354.S;:I:} 8430449

193518 33f~ 532 3430.449

183520 3354.532 ,3430.448

2:306('17 19:?iSZ2 7n.S4.53 843<)449

28C597 183524 3.354.52'0! 8430449

.2&J~597 \9352$ 3354528 8430449

280597 1835;;~g ~lf>4.52e i'l43QA48

230597 193530 3354,.5::'>6. 843C.44S

2S0597 19:3532 3354. 525 8430449

230597 '93534 3:3.54.524 843C44E~

2B0597 193536

843C.4~1

~OO597 193538 3354.521

Tra.nsmitter Vehicle OPS Data
nm. L.atll:l.ld. LClflgiwd Knots Headln9

3304.381 843034~

329,~

9.114

tS3358 3354.387 8430339

3313.0 10.:~S44

~93358 33f~391 8430341

3313.5 102'#:'44

~93+~OO 3354394 8430342

3385 10.295M

~S3402 ):354.39S 8430.342

T37 1~ 81444

~S3404 2.Je4,4C3 8430:-;;44

7

~83400 3.3.54.407 8430.345

,"

337 ~ '.8~:444 337 11.31444

!934fJ8 336441 8430347

7

\S:34~O 3354413 8430.35 75

337 1if.l1444 3325 11,oc:-e~:n

:3:354,418 $430353 7.4 3318 12.48908

~3J4~A 3-3PA.42 8430355 7.5

':~j:3A ~,6 ;~;').f:4423 8430.359

74

330,6 12.B5S33
3:302 12.413956

\934'13 3354.42? g4:jlJ3tH

12.489~~

103420 3.354.43 8430 364 7.4

330 12.48956

'8:3422 :3304.433 8430.31'$6 L]3A24 33[>4.4.3.7 $43{L'l$

:l30S 1198322

7

;330.9 .~ 1.-e~ 444

181426 .:nS4.439 >9430.3'1

330.7

$$54442 8430 373 6.9 3Y.JS t~ .64567

1~i3430 3.254.448 8430 376 6.9 3:108 11.64567

1334J2 :;,304,449 84","C3?9

3.30.'7 '1:1 fd~;5t31

1934-1.4 3354452 84:'l(J38 6,B
183436 3354.456 3430.382 sa

33'1,2 ~ 1.476139 330,8 11.47f,s9

1934~ 3354459 84X,385 67'

3.30.7 11,3C81~

193440 :3:3S4.4B2 80C387 1;;7 :n12 ~t3(l81'1

193442 3::'>54.486 f'A30392 6.8 3307 ~~: .4768&

12'3444 33544'7 8<;30 300 S3 32<0,7' 1 ~ 476:S8

193446 3354.474 8430,398 68

330.7 1~645-97

19:;,448 3354.477 5430.403 S.8 330.8 11.47689

193450 3354.48 8430AG7 5.8 3304 11.47MJ:!

1S345~' 3354.483 8430.41 B.9

1S::W:>4 :';'3(>4.489 84:i<}411

7

330,3 11S4E~,7
3~1.~ ne,l444

19:>.45$ .3354.482 8430.415

7 3302 11.81444

193458 3354495 8430.419

7

~328,9 11e1444

19350Cl 3354 499 8430.421

328$1 11.SQ44

183512 3354502 8430425

3285 11.81::<322

~S3~J4 3354505 84;:K)429 7.~

:~2$3 11:)$322

19:35C16 3354.508 8430432 7,.';

3283 1188322

~93f;..Q8 33545'2 843(l434 19-3510 335t.5~5 84::'\0 436

.,,

328.9 1181444 329.3 11 ;5~444

182<512 3354.S~9 84.30.439

7

3295 ~~.e~M4

1'93615 3354. 52~, 8430.44 ~

7

~ 1.e~444

~S(i517 3354525 843(1443

7

329 11e~444

193519 3354 528 3430.446

7

328.8 ~ ~ .81444

183521 3354531 8430447

37",,1 ~ 1tl$322

1~~:352'3 ~:C:l54. 534 8430 451

7.1

1$3525 3354.5:38 84~'l(l453

7

328.5 3:.>9 , 18~444

193527 3354541 843OA$5 ~93529 33.."'4 545 8430.458

3~g.4 1~ .64:567
3;:'~.2 ~~-.476a$

1S353t 3'.154549 8430461 183533 X354,5-'52 8430 464

328,9 ~'2!i8 11 S14~~4

1')3535 3.354.556 84:30 400

32~U3 1~:,a1444

19:3537 :J3S45S9 8-<K,'\C,4QS

325B 11.31444

tS:J539 33'4::'.<6'2 i;4:::'lI:}4$9

3281

~9:~C-....4t :;-3[-..4 .7.'4 ~4':1C~'. 47

322.2

Sumfl'li R"tliI ::: 825. S<:;{',s

Ciswnc$ "

1251.284

E-18

1~

14(i

130

120 l1l}

100

g$Ii
,to 90

liE <. 00 ;;

l::
..C!;
.Ii 7Q

-tT1 I
\0

,~

"i
iii

W

at.

SO

4V
J(l
;;'{)

1(}

......,
."

,.,.....'.

.., ,"."' >

.... :..

.....

< ..
..".".,...:. "'..,,...'.\,}.'>, .. .
""</ ...,.
"}\"
.,..,.....:.

...........
i.'.,.<....',
", '" , '.,.
. " "., ..,...~
>
11
..~

.........................
-=='--Serlesl
j 1S.~'e.r.~~~ .~"9..J~e.I.I~!11

.....
Figure E- Jo. Signal strength recorded during the Dobbins Test 8 run over a distance of 1,275feet

At a highway speed of 55 mph (80.6 feet per second) the detector would have required 3 seconds to activate the display of the warning message. Thus, the warning message would have been displayed to the driver at a point 242 feet into the run. Deep multipath nulls that drop below the detection and decode threshold of 20 units of signal strength occur at 725,840,940,970, 1,100 and 1,140 feet into the run. Several of these stay below the detection threshold for several seconds at the low speed at which the Dobbins' tests were concluded. However, at highway speeds the SWS receiver will continue to display the SWS warning message because any of the drop in signal level below 20 signal units would last less than 0.5 seconds. The signal peaked to a signal strength of approximately 115 before data collection was stopped. The transmitter vehicle continued northward until it had traveled approximately 1,340 feet; however, collection ofsignal strength data ended after approximately 1,275 feet of transmitter vehicle travel.
The analysis of Dobbins Test 8 data indicates that while the signal to noise ratio is not as great as it was in Dobbins Test 5, it is sufficient to cause the SWS message to be first displayed at an equivalent distance as in Dobbins Test 5, at the speed being driven during the test. Highway testing of the SWS will introduce blockage of the signal by other vehicles on the highway. However, additional signal loss due to blockage and shadowing effects may be offset by increased signal levels resulting from reflection of the signal off of objects in front of the receiver vehicle. Highway testing will detennine the true effects that surrounding traffic has on the operation of the SWS.
DOBBINS TEST 10 Dobbins Test 10 was designed to evaluate the range at which the SWS detector would detect and decode the SWS message when the receiver vehicle was crossing the path of the transmitter vehicle. Figure E-ll is a diagram of how the test was laid out. This scenario simulates a police or eccrgency vehicle crossing a rural intersection just before the intersection is crossed by an SWS equipped motorist. Normally both vehicles would be in motion and crossing the intersection at a 90 degree angle to each other. However, in order to keep the test simple, the transmitter vehicle was kept In a stationary position and the receiver vehicle crossed the projected path of the transmitter vehicle. Referring to Figure E-ll, the test was run from north to south. The transm,itter vehicle was parked heading west approximately
E-20

Dobbins Test I0
745'
tTl
--r
N

~
! I I I I I I I I

I

I

I

I

I

I

I

...

... I

105' I

I

I

..I
155'

Stationary Transtnitter Vehicle
Section of Unused
H,un\va.v at [)obbins
Moving Jleceiver Vehicle

Figure -11. Dobbins Test 10 test layout

515 feet from the start point of the run. The receiver vehicle run continues approximately 230 feet beyond the transmitter vehicle before the SWS signal is lost.
Table E-3 shows the GPS navigation data that were generated as a result of the test. The GPS data recorded by the transmitter vehicle are stable. The distance between the starting point of the receiver vehicle's run and the end of the receiver vehicle's run is approximately 745 feet, based on the summation of the receiver vehicle rate data. The receiver vehicle crosses the path of the transmitter vehicle at an offset distance of approximately 105 feet to the front of the transmitter vehicle. The offset was used. to examine the signal strength of the SWS mobile transrnitter at two ranges from the test path. During Dobbins Test 12 (which follows) the offset distance is reduced while all else remain the same.
Figure E-12 shows the signal strength as a function of range between vehicles. Referring to Figure E-12, signal lobing is present from the start of the run. The peaks and nulls in the signal level shown in Figure E-12 are not totally the result of multipath effects. Because the receiver vehicle is approaching the transmitter vehicle from the side, the sidelobe nulls and peaks in the receiver's and transmitter's respective antenna patterns are being detected along with the peaks and nulls from multipath effects. The antenna patterns for both the SWS receiver and transmitter are shown in Appendix C.
The receiver vehicle passed the transmitter vehicle (the intersection) at a point approximately 550 feet into the test run. Approximately 180 feet beyond this point where the receiver vehicle passes the transmitter vehicle, the signal level drops below the SWS detection and decoding threshold of 20. The rapid signal decrease seen once the receiver vehicle passes the intersection is a result of the receiver moving into the sidelobes of the transmitter antenna and the transmitter's position moving into the backlobes of the receiver antenna.
It is uncertain if the SWS would have detected and decoded the SWS message from the start of the run while traveling the first of 226 feet of the run. However, beyond that point, it is believed that the warning message would have been displayed until the receiver vehicle passed the point (the intersection) at which the transmitter vehicle's path was tangent to the receiver vehicle's path. At a highway speed of 55 mph the warning time would have been I second If the detector began processing the SWS signal at a point 226 feet into the run and assuming 3 seconds processing time. It is suspected that this performance will improve when objects that reflect energy are added to the scenario during urban testing.
E-22

Table -3. GPS Navigation Data Generated as a Result of Dobbins Test 10

Receiver Vehicle GPS File

Date Time Latitude longitude Knots Heading FeetfSe~

280597 194834 335453 8430.457 9.5 145.3 15.0~389

280597 194836 335452 8430453 95 147.8 16.03389

280597 "HN838 3354.52 8430..449 S.O 146.7 16.20267

280597 19-4840 3354.51 8430445 9A 147.5 15.86511

280597 1St4842 335451 8430.441 ':1.3 147.7 15.69633

280597 1B4844 3354.5 8430A38 92

148 15.52756

280597 194846 3354.5 8430.435 9-.2

148 1552756

280597 194848 3354.49 8430-432 92 147.1 15.52756

280597 194850 3354.49 8430.429 92 147.1 1552756

280597 194852 3354.43 8430.425 92 147.1 15.52756

280597 194854 3354-48 8430.42 8.4

12.8 14.17733

280597 194856 3354.46 8430.417 8.2 137.6 13.83978

280597 194358 3354.48 8430.413 7.8 142.4 1316467

280597 194900 3354.47 8430.411 76 143.9 12.82711

280591 194902 3354.47 84300407 7.4 145.3 1248956

280597 194904 3354.47 8430A05 7.3 147.7 12.32078

280597 194906 3354.46 8430.403 7.5 148.9 1U15833

280597 194908 3354.46 8430A01 75 148.9 12:.65833

280597 194910 3354.46 8430.398 7.5 148.5 12.65833

280597 194912 335445 8430.396 7.$ 148.2 12.82711

ZS0597 194914 3354A5 84:)(}.39~ 7.6

148,2 1262711

280597 194916 3354.44 8430.391 1.8 146.8 13.16467

28C59? 194916 3354.44 8430.388 7.8 146.8 13.16467

250597 194920 3354.44 8430.385 7.8 146.8 13.16467

2eC59i '94922 3354.43 8430.38 10 137.2 16.87778

2SCS97 194924 3354042 8430.378 S.7 139.1 16.371

Transmitter Vehicle GPS File Tim. latitude Longitude 194834 3354.482 8430,394194836 3354.462 8430.394 194838 3354A63 8430:394 194840 3354,463 8430.394 194842 3354A64 8430.394 194844 3354.464 3430.394 194846 3354.464 8430.394 194848 3354.465 8430.393 194850 3354.465 54:}O,393 194852 :>354.466 8430.~93 194854 3354.466 8430.393 194856 3354.466 3430392 194858 3~54.4tn 84:)0392 194900 3354467 3430392 194902 3354.467 5430.392 194&04 3354.468 2;430.392 194905 3354.4G8 5430391 194908 335-4.47 8430.:}92 194910 3354.47 8430.391 194912 3354.47 8430.391 194914 335447 8430.391 194916 3354.471 3430391 194918 3354,471 843CU91 194920 3354.471 8430.39l 194922 3354.472 843039 19492. 3354.472 843039t

Feet lit

745.3227

E-23

15<:)

140

HO

1 ~1)

HO

H10

,:;;,

.teil ~1
~

I'f")
':i

00

l::

t:l'1

tTl I

W... to
.~

IV
+:-

~
~

60

IX

~~)

40

:~O

20

10

0
<;3-

~

\i) ~~.

- ... 9'(

~

*,'. ~

~~

~0

~
(')

~
(')

'"

~..,

~~

Dh.t"'ll(l~ Traveled by RlM:.iv~f Yehk..

r~==5~;j~;1 ~ - 1 5 Pf.'f M,)v:l\vp {S(lHe~1.1

Figure -12. Signal strength as afunction of range to the point that represents the simulated intersection

DOBBINS TEST 12
Dobbins Test 12 was conducted in a similar manner to Dobbins Test 10, except the track of the receiver vehicle was moved approximately 100" feet further east, very close to the transmitter vehicle, as shown in Figure E-13. The purpose of Dobbins Test 12 was to simulate an emergency vehicle very close to the intersection while the receiver vehicle crossed the simulated intersection.
Table E-4 shows the GPS navigation data that was developed as a result of Dobbins Test 12. The distance of the run was computed by integrating the receiver vehicle's rate of travel as done in the previous tests. A run distance of 811 feet was derived, a distance slightly longer than the previous intersection test (Dobbins Test 10). The starting and ending latitude and longitude positions of the receiver vehicle were also plotted on the 1 inch to 100 foot scale map provided by the Dobbins Civil Engineering Department. Reasonable agreement was found between the two methods of determining the test start point, end point, and run distance. Referring to the GPS data in Table E-4, there was a slight drift to the south in the GPS data. There was little drift to the east or west in the GPS derived position of the transmitter vehicle.
Figure E-14 shows signal strength displayed as a function of run distance when the moving average trendline is laid over the raw data. The reduction in the offset distance of the test path toward the transmitter vehicle shifts the starting received signal level up by almost 10 units over the starting signal level received during" Dobbins Test 10, when the offset distance was 105 feet.
The signal level exceeds the detection and decode threshold of 20 signal units by 10 points at the start of the test run. This starting signal level would have caused the SWS to display the warning message within three seconds of the start of the test run. The nulls caused by antenna sidelobes and multipath drop the signal below the display threshold after the receiver vehicle has traveled approximately 350 feet; however, the duration of these signal nulls are too short to cause the SWS message display to blank due to loss of signal. The signal level is high until the receiver vehicle passes the transmitter vehicle after traveling approximately 533 feet, at which point the signal peaks. Beyond this point, the signal drops as in Dobbins Test 10 for the same reasons as in Dobbins Test 10.
Given the high signal to noise ratio at the beginning of the run, the SWS message would have been displayed within the first three seconds. The warning message would have continued to have been displayed until the receiver vehicle passed the tangent point (intersection) with the transmitter vehicle.
E-25

I.".)(.)1)1"HilS. "I'.es.t'1l .')~

I

I

I

I

I

1

I

I

I

811 '

RI1~1

I

tTl I

I

tv

0\

!

I

!

!

Stationary T'ransrn itter
Vehicle
....__. Section of Unused
RUl1vvay at I)obbins

'. I"'--~"---f-""""'- Moving Receiver V ellicle

'1 'S' ~ .-dItt.,_...",....."....

~
. -.;)W,

- - - I I.......

Figure -13. Dobbins Test 12 with the track of the receiver vehicle moved approximately 100 feet east

Table -4. GPS Navigation Data Developed as a Result of Dobbins Test 12

Rtu::ei'At( Vehicle GPS Data

Date Time Latitude Longitude Knots Heading Feet/Sec

280597 195913 3354.537 8430.428 11 150.5 18.56556

280597 195916 3354.528 843QA21 12.3 150.5 20.75967

280SG7 195918 3354522 -8430.416 12.2 150.6 20.59089

280597 195920 3354.516 8430.412 11.7 150.4 19.747

280597 195922 3354.51 8430A08 11.3 151.4 19.07189

280597 t95924 3354.506 8430.404 11 1515 18.56556

280597 195926 3354.501 8430.401 106 15t.4 1789044

280597 195928 3354.495 8430.398 10.2 150.9 1721533

280597 195930 3354.492 8430.394 9.1 1509 15.35878

280597 195932 3354,488 8430.392 fJ.7 150.5 14.68367

280597 195934 3354.484 8430.389 a.5 150.9 14.34611

280597 195936 3354.48 8430387 8.3 150.1 14.00856

280597 195938 3354.477 8430.386 7.9 150.9 13.33344

280597 195940 3354.473 8430.384 7.9 150.9 13.33344

280597 195942 3354.47 8430.38t .7,..8,
280597 195944 3354.46$ 8430.378 ( ./

150.8 13.16467 150.9 12.99589

28D597 195946 3354,462 8430.315 7.6 151 ..2 12.82711

280597 195948 3354.458 8430.373 7.5 151.2 12.65833

280597 195950 3354.455 8430.371 7.4 151.3 12.48956

280597 195552 3354.452 8430.37 7.4-

152 12.48956

280597 195954 3354.448 8430.367 7,5 152.5 12.65833

280597 195956 3354.444 8430.364 7.6 152.3 1282711

280597 195958 3354A41 8430.362 7.6 152.3 12.82711

280597 200000 3354.437 8430.359 7.9 152.3 13.33344

280597 200002 3354.433 8430358 7.9 152.6 13.33344

280597 200004 3354429 8430.355 7.9 152.6 13.33344

280597 200006 3354.425 8430.353 7.9 152.6 13.33344

Sum'" 405.7416

Total '" 811.4836

Transmitter VehlcJe GPS Data
Time Latitude Longitude 195915 3364.466 8430376 195917 3354.465 8430.376 195919 3354,465 8430ne 195921 3354.464 8430376 19Sn3 3354,464 3430.376 195925 3354.464 8430376 195927 3354.464 8430.376 195929 3354.463 8430.377 195931 3354463 8430376 195933 3354,463 8430376 195935 3354.464 S4;~O.376 195937 3354.463 8430.376 195939 3354.453 8430.376 195941 3354.463 8430.376 195943 3354.462 84:.W.3'77 195945 3354.462 8430377 195947 3354.462 3430.376 195949 3354.462 8430.376 195951 3354.463 8430.3'76 195953 3354463 8430.376 195955 3354.462 8430376 195957 3354.463 843D.376 195959 3354A62 84:10.376 200001 3354.482 8430.376 200003 3354.462 8430.377 200005 3354.483 84:30.377 200005 3354,463 E430377

E-27

lSi)

141.1

13!)

14c'Q

110

100

.~ 00
ii

t;)
-;;

an

c

"~

~ 70

t'I1 I
tv"

iii

III

00

Ii> 60

IX

50

4Q

Vi

20

10

!--Sitfit>11 "-1Sp!r M<'.N A~ (St>n~l)
Figure E- J4. Signal strength displayed as a function of run distance

TESTING THE EFFECTS OF OPERATING SWS AT THE BOTTOM OF A HILL
After the conclusion of runway testing, the test vehicles were moved to a controlled access patrol road on the south side of Dobbins Reserve Air Force Base. Figure E-15 shows the test layout from an aerial perspective. The test run was from west to east over a distance of approximately 2,200 feet, as measured on the map, the GPS derived distance of 2,086 feet was not used due to the large positional drift in the data. The stationary transmitter vehicle was located at a point approximately 1,510 feet from the starting point of the run at the bottom of a hill. The transmitter vehicle was heading east, in a position that placed the left wheels of the transmitter vehicle on the pavement approximately 2 feet beyond the side of the roadway. The MPH, Inc. mobile transmitter was mounted on the transmitter vehicle in the same position as previously described. The receiver vehicle approached the transmitter vehicle from the west and passed the transmitter vehicle.
The transmission from the mobile transmitter was III the CW mode and the antennas radiated 15 milliwatts ERP in the forward direction and 15 milliwatts ERP rearward in the same manner as during the runway tests. The mobile transmitter is approximately 55 inches above the ground and the SWS detector in the receiver vehicle is approximately 45 inches above the ground.
Figure E-16 is a contour plot of the topography of this site. This plot was generated from the 1 inch to 100 foot scale map, supplied by the Dobbins Civil Engineering Department which provided 2 foot elevation contour data. The elevation shown in Figure E-16 is in feet, mean sea level (MSL), and is shown along the 'Y' axis. The elevation interval is 10 feet, in Figure E- I6. The distance traveled by the receiver vehicle is shown along the 'X' axis. It should be noted that because the 'X' axis was not generated from points closer together than 100 feet, the slope is somewhat exaggerated at a point near the transmitter vehicle's location, approx:mately 1,510 feet from the start of the run.
Table E-5 shows the GPS data that were generated by the receiver vehicle and the transmitter vehicle during the test. When the stationary transmitter vehicle GPS latitude and longitude data were averaged, it was found that the GPS derived position of the transmitter vehicle was approximately 90 feet south of the actual transmitter vehicle's location and approximately 90 feet further east than the actual position obtained from the 1 inch to 100 foot map of Dobbins. Next, the path of the receiver vehicle was plotted and the path was found to parallel the roadway on which tests were conducted, but there was an error of approximately 110 feet to the south. Further analysis of the GPS data showed that the signals from several previously available GPS satellites were blocked due to the trees located very close to the patrol
E-29

n" I) "l) 'I' ! .,' t, . 1t.)1' l ~;) . t.,..~;,t .''7
N

Total Distance of Run =:; 2,200
-, ........ .~,~-~~~- -----------~--I
trJ I
Wo

t ~~D-ist-an-ce-::=-J-,5-1 0-'-... .

iv10vingReceiver \!ehiclc

Stationary TranSlnitter 'Vehicle

Figure -15. Test layout from an aerial perspective

1010

1010

HX\l


!t. 4i
.ill
.J 900
41
.<Il
'~"
;:i:

tTl

I
W

ue.;)

970

96:)

~ ~ <:3

8....

~

~

~

n~

~

8
!'-.

~

~

,...

~

~







8
t'.

~

....

8
N

...<*

fe~f TrllVllkd ity Rec~ivllr "<lWei"

Figure -16. Contour plot of the topography of the site located on the south side ofDobbins Reserve Air Force Base

Table -5. GPS Data Generated by the ReceiverlTransmitter Vehicle During Dobbins Test 16

Receiver Vehicle GPS

Transmitter Vehicle GPS

Date Time Latitude Longitude Knots Head!n Feet/Sec.

280597 204955 3353.824 8431.03 10.5 92.2 17.72167

280597 204957 3353.825 8431.023 10.5 92.2 17.72167

280597 204959 3353.824 8431.016 10A 93.3 17.55289

280697 205001 3353824 8431009 103 94.1 17.38411

280597 205003 ~353.a24 8431.002 10.3 94.1 17.3841'1

280597 205005 3353,1323 8430.995 10,4 93.5 17.55289

280597 205007 3353823 8430.988 10.1 95.2 17.04656

280597 205009 3353.822 6430.982 10.1 95.2 17.04656

280597 205011 335382 8430.976 9.9 94.2 ~6.709

280597 205013 3353821 8430.97 9.4- 92.2 15.8651 t

28.0591 205Q15 3353.821 8430963 9.1 93.4 15.35878

280597 205017 3353.821 8430.957 8.9

93 15.02122

280597 205019 3353.82 8430.951 8.9 94.2 15.02122

280597 205021 3353.82 8430.945 8.9 94.4 1502122

280597 205023 3353.82 8430.S39 8.8 94.4 1485244

280597 205025 3353.819 8430.933 8.9 94.3 1502122

280597 205027 3353.818 8430.927 e.9 94.3 15.02122

28059'7 205029 3353.818 8430.922 8.9 94.8 15.02122

280597 205031 3353..81$ 8430,S16

9 95,8

15.19

280597 205033 3353817 8430.91

9 95.3

15.19

280597 205035 3353.816 8430.903 3.8 95.8 14.85244

280597 205D37 3353.816 8430.898 8.8 95.8 14.85244

280597 205039 3353.815 8430.892 8.8 95.8 14.85244

280597 205041 3353.816 8430.885 88

fJ7 14.85244

280597 205043 3353.815 8430.88 8.8

97 14,85244

280597 205045 3353.816 8430.874- 8.7 97.2 14.68367

28CS97 205047 3353.815 8430.888 8.7 97.2 1468367

280597 205049 3353.814 8430.863 8,7 97.2 14.68367

280597 205051 3353814 8430.857 8.7 97.2 14.68367

280597 205053 3353813 8430.851 8.7 97.2 14.68s67

280597 205055 3353.813 8430.84$ 807 96.8 14.68367

280597 205057 3353.812 8430.839 8.7 96.~ 1468367

28CS97 205059 3353.811 8430.834 e.7 96.8 1468367

260597 2Q5W1 3353.811 8430.827 8.9 &6.4 1502122

280597 205103 3353.81 8430822 8,9 96,4 15.02122

280597 205105 3353.81 8430816 8.8 96.2 14.85244

28CS91 205107 3353.81 8430.81 8e 96.2 14.85244-

280591 205109 3353809 8430804 8.8 962 14.8:;244-

280591 205111 3353809 8430198 88 962 14.85244

2805Bl 20-5113 3353.809 8430.792

9 96.1

15.19

280597 205115 3353.808 8430.786

9 96.1

1519

280591 205117 3353.808 8430.779 9.3 955 1569633<

280597 205119 3353.807 8430.773 9.3 95.5 15.69633

280597 205121 3353.808 8430766 9.5 95.2 16.03389

280597 205123 3353.807 8430.759 9.5 95.2 1603389

280597 205125 3353.807 8430.753 9 ..l.:; SS.2 1603389

Time Latitude LongitUde 204954 3353.799 8430.'118 204956 3353799 8430.717 204958 3353.799 8430.717 205000 3353.798 8430.716 205002 3353.798 8430.716 205004- 3353.798 8430715 205006 3353.798 8430.715 205008 3353.798 8430.715 205010 3353.198 84~<O.715 205012 3353.797 e.43C.714 205014 :3:353.798 8430.714 205016 3:353.7$7 $430713 205018 3:353.797 8430.713 205020 3353.787 f}430]13 205022 3353.796 8430,7'~2 20$024 3353.796 8430.712 205026 3353.796 &430712 205028 :1353.796 8430.712 205C30 3353,795 8430.711 205032 3353.796 8430.71 205034 3353.796 8430.71 205036 3353.797 8430,71 205038 3353.798 B430.71 205040 3353.798 8430.71 205042 3353.799 8430.71 205044 3353,799- 843\:1-71 205046 33538 3430.709 205048 3353.501 8430.709 205050 3353.802 8430.71 205052 3353.802 8430.71 205054 :D53.802 8430709 205056 3353.802 8430709 2.05058 3353.802 8430709 2051GO 3353.802 84307CS 205102 3353,B02 8430708 205104 3353802 842~O,.7G8 20:J106 3353.802 8430.701 205103 3353.802 8430707 205110 :B53.802 8430707 205H2 335'3.802 8430.706 205114- 3353.802 8430.707 205116 3353502 8430706 205118 3353.302 8430.706 205120 3351.802 8430,706 205122 3353.803 8430,"706 205124 335-3..305 8430.705

E-32

Table -5. GPS Data Generated by the ReceiverITransmitter Vehicle During Dobbins Test 16 (Concluded)

Receiver Vehicle GPS

Transmitter Vehicle G?S

Date Time Latitude Longitude Knots Headin Feet/Sec.

280597 205131 3353806 8430.731 95 949 16.D3389

280597 205133 33-53805 8430.724 9.3 95.1 15.69633

280597 205135 3353.804 8430.717 92 95.7 15.52756

280597 205137 3353804 8430.711 91 95.9 15.35878

280597 205139 3353)303 8430.103 8JJ
280597 205141 3353.803 8430.696 a9

96.2 14.85244 96.4 15.02122

280597 205143 3353.804- 8430.691 89 962 15.02122

280597 205145 3353.803 8430.682 91 96.8 15.35878

280597 205147 3353.802 8430.675 92 96.8 15.52.756

280597 205149 3353.802 8430,067 9.3 97.1 1-5.69633

280597 205151 3353.803 8430.66 9.5 97.2 16.03389

280597 205~53 3353.803 8430.652 9.5 96.9 16.03389

2130597 205155 3353.803 8430.645 96 97.1 16.20267

280597 205157 ::'353.804 8430.641 96 966 16.20267

280597 205159 3353.803 8430.634 9.5 962 16,03389

280597 205201 3353,003 8430.628 9.5 96.2 16.03389

280597 205203 3353.802 8430.621 9.5 96.2 16.03389

280597 205205 3353,802 8430.615 9.4 95.7 15.86511

280597 205207 3353.801 8430.609 93 95.4 r5.69633

280597 205209 3353.801 8430.603 93 95.4 15.69633

280597 205211 3353801 8-430.597 9.1 94.4- 15.35878

Sum- 1043.041

Feet :II 203&.093

Time latitude Longitude 205130 3353807 8430.706 205132 3353807 8430.707 205135 3353.807 8430.707 205137 3353.8013 8430.707 205139 3353J~01 6430.707 205141 3353.808 8430.707 205143 3353.807 8430.708 205145 3353.807 8430.707 205147 3353.807 8430iQ7 205149 3353.807 M3t1708 205151 3353.807 8430.707 205153 3353.807 8430]07 205155 3353.807 34307C7 205157 3353.807 8430.707 205159 33S::'L807 B430JC 205201 335:-L807 8430700 205203 335:1.807 8430.7D6 205205 3353.807 8430 'fOe 205207 3353.807 8430.706 205209 3353.806 a..nO.7Q6 205211 3353.806 843Q.700 Avg 3353.8015 3430.70945

E-33

road. It is thought that the loss of these satellites in the GPS constellation contributed to the larger errors observed. This problem forced more reliance on the video tape record to determine the location of the test vehicles.
DOBBINS TEST 17
Figure E-17 shows signal strength plotted as a function of distance traveled by the receiver vehicle. The signal was received from the start of the run and peaked above the threshold of 20 signal strength units required for message display at a distance of approximately 330 feet from the start of the run and approximately 1,170 feet from the transmitter vehicle. Referring to Figure E-16, there were two hills between the receiver vehicle and the transmitter vehicle at this point. The very rapid and large swings in signal strength around the trendline indicate that the source of the signal was most likely reflected energy from the tops of the trees that were set back approximately 50 feet from the test road. There was also a 10 foot high chain link fence that paralleled the road 50 feet to the south. The amount of reflected power received from the chain link fence, when there was no line of sight between the receiver vehicle and the transmitter vehicle, is unknown.
When a line of sight was established between the receiver and the transmitter, first at a point approximately 1,100 feet from the start of the run, the signal jumped to a trendline average level of 105. After going through a deep multipath null, the signal went higher to a trendline signal level of I 12. At a point approximately 1,340 feet beyond the starting point, the signal dropped due to the fact that line of sight was lost for approximately 4 seconds because of a dip in the roadway. The signal drops to a trendline signal level of approximately 62 briefly before line of sight is reacquired. The signal next peaks to a saturated level of 135 and then drops to a trend line level of 115 as the receiver vehicle passes the transmitter vehicle. The signal then peaks again to a trendline level of 140 before dropping However, the amount of signal loss after passage of the transmitter vehicle by the receiver vehicle is not the extreme drop below a level of 20 witnessed during runway testing when the receiver vehicle passed beyond the transmitter vehicle. A study of the video tape indicates that the slope of the hill was such that it may have reflected much of the transmitted energy back toward the receiver vehicle. Also, the trees may have also served as a reflector of the energy back toward the receiver vehicle.
E-34

150
14<;
no
1::'{}
1W
100 ~ ~ 90
J60
"iii
,I2::"
41) 70
~'il 00
0:' 50
40
::lO 2(t
10
n
Q

.". ...
~

f~

t...,,
(

lY)
;,t

J,
<i)

~

r:ff."s

~

t,
8l

8
~,

0r>,

p:j
l.'"-!

~

S1

r~~
~

i'-l
f.::

f.l'~
~

~

:g
N

Oiiitan~iIl Tfillleled {~ti

r~:;';i~~;~l
:.-.o='~J"!~~)~'A.v~ (Seril;>!il)

Figure -17. Signal strength plotted as afunction ofdistance traveled by the receiver vehicle

SUMMARIZATION OF DOBBINS TESTING Dobbins testing demonstrated several principles that were also observed during urban and highway testing (see Appendices G and H): 1. When the receiver and transmitter are boresighted, the maximum range of the system will be
achieved (assume line of sight exists). 2. The test scenario that simulated the overtaking emergency vehicle demonstrated the next best range
performance of the system (assume line of sight). 3. The scenario that simulate the crossing of intersection by an emergency vehicle provided the shortest
detection distance of the three basic scenarios. 4. The scenario that simulated the emergency vehicle at the bottom of the hill produced interesting
preliminary results on the basis of limited data. It appears that if there are any objects near or over the roadway that will reflect the radiated energy from the transmitter, the detection range will increase over the range obtained in an object free environment. Also, it appears that energy is being reflected form the forward pointing beam, to a point high up the hill in front of the transmitter vehicle, which produces a high signal level line of sight is established between vehicles. 5. The Dobbins tests provide a baseline against which unexpected highway results can be evaluated.
E-36

APPENDIX F RURAL TESTING
F-l

THIS PAGE INTENTIONALLY LEFT BLANK F-2

APPENDIX F RURAL TESTING
RURAL TEST SCANERIO
Testing of both the SWS mobile and fixed site transmitters was conducted in a rural setting to determine their effectiveness to provide a timely warning. There were three test scenarios completed during the rural tests: (1) straight rural highway testing with line of sight established; (2) curved rural highway testing with tree foliage blocking the line of sight, and (3) hill testing with the crest of the hill blocking the line of sight.
TEST SITE FOR STRAIGHT ROAD TESTING
The approaching straight road and curve tests were conducted along the State Route (SR) 120 Connector located approximately 10 miles west of Marietta, Georgia. Figure F-1 shows the section of the SR 120 connector that was used for the straight road testing, hereafter called "Straight". The receiver vehicle transited from the intersection of SR 92 and SR 120 connector heading southeast, passing the transmitter site at the intersection of SR 120 Connector and Lucille Drive and continued until the transmitter signal could not be received.
Figure F-2 is a photograph that shows the view from the intersection of SR 92 and SR 120 where each test began. The receiver would be activated as the receiver vehicle moved through this intersection. Usually, relative signal strength data collection would begin before the receiver vehicle passed 300 feet beyond the intersection toward the transmitter location, except when curve testing was performed. During curve testing the mobile transmitter signal could not be received at the intersection due to tree foliage blockage of the signal.
Figure F-3 is a photograph showing the mobile transmitter mounted on the transmitter vehicle at a point east of the intersection of Lucille Drive and SR 120. The transmitter position was located on the outside of a curve that started in front of the transmitter vehicle. This location for mobile and fixed site transmitter operation was chosen for safety. Referring to Figure F-3, it can be seen that locating the transmitters at this point was equivalent to placing the transmitter at the side of the approaching lane, had the SR 120 Connector continued in a straight line.
F-3

Lo..-tioILmo..il<Olilc:d _ fixed !;iia k1US~ttc[ &t !'Oil>, 'l'1'",."'maLtLy ~,361
(1.'d.(E'CC'I~&tActine;ptJ;n1
Figure F-i. Map ofState Route i20 connector where rural straight road testing was conducted
Figure F-2. intersection of SR i20 and SR 92 looking southeast toward transmitter location F-4

Figure F-3. View from transmitter site, down SR 120 connector
STRAIGHT TESTS INSTRUMENTATION AND SET-UP
The receiver was a BEL-TRONICS Model 855 STi radar detector with RS-232 interface capability used to transfer relative signal strength data to a laptop computer where it was stored for later analysis. A forward facing video camera was mounted on the back window outside of the receiver vehicle so that the scene data could be recalled during analysis. Notes regarding observations made during the test were recorded on the audio track of the video tape recorder. Mr. David Jared of the Georgia Department of Transportation (GDOT) operated a Global Positioning System (GPS) receiver that recorded navigation data to a laptop computer. The GPS provided navigation data that showed the receiver vehicle's location as a function of time and this GPS data was used as ground truth data during the analysis of the received relative signal strength data.
STRAIGHT-1 TRIAL
The receiver vehicle approached the mobile transmitter vehicle along the roadway as shown previously in Figure F-l. The receiver and transmitter antennas pointed at each other, a best case SWS performance geometry. Figure F-4 is a plot of relative signal strength versus distance traveled by the receiver vehicle.
F-5

The receiver vehicle maintained a speed of 37 miles per hour during the entire trial. A trend line moving average of 15 points, shown as a dark line, has been laid on top of the relative signal strength plot. Referring to Figure F-4, the signal is above the threshold of 20 required to activate the display of the warning message during most of the trial, except for the very short time when the vehicle was approximately 600 feet from the starting point. At this point the receiver vehicle overtook and briefly approached close to the rear of a pick-up truck (in the same lane) that was slowing to tum ~ff of the SR 120 connector. The truck momentarily obscured the line of sight between the receiver and transmitter vehicle. Blockage of the signal by approaching traffic is also apparent when the receiver vehicle was approximately 1,000 and 1,200 feet from the transmitter. However, at the point where this blockage from approaching traffic occurs, the relative signal strength was well above the display threshold of 20 and the safety warning message would have continued to have been displayed through the temporary fades in signal strength. A multipath induced fade did occur in the plotted relative signal strength at a distance of approximately 2,375 feet from the starting point. The SWS relative signal strength drops very quickly once the receiver vehicle passes the location of the transmitter vehicle.
STRAIGHT-2 TRIAL
Figure F-5 is the relative signal strength plotted as a function of receiver vehicle travel distance from the intersection of SR 93 and SR 120 extension. The initial relative signal strength is higher during this trial than the previous trial because there is no vehicle in the same lane in front of the receiver vehicle as in the previous test. The receiver vehicle passed a vehicle parked on the side of the road at a distance of approximately 850 feet from the starting point that caused a signal cancellation null to occur. Traffic induced drops in the signal also occurred at a point approximately 1,400 feet from the starting point. Another traffic induced effect was observed to occur at approximately 2,239 feet from the starting point. A large dump truck approached and passed between the test vehicles when the receiver vehicle was approximately 200 feet from the mobile transmitter's location. A slight drop in relative signal strength is noticed during the dump truck's passage.
F-6

1fiO

140

130

120

110

1IX)

.<:.

Bc> 00

t

Vi
<ci

50

a>

~ 70

OJ
2:
~
n~: tiC>

'T,1

-J

~-o

40

30

10

10

r":"

h~

(}islam;:/! Tr<lvl1ll1{! (Feel)

Figure F-4. Relative signal strength displayed as afunction of distance traveled during straight-l trial

150

140
no

120

1\0

100

~
.."c&:> 90;)

Vi e<:1
c:
'".'
.<Il 7iJ

'T,l

.~,.

00

1n1:: 00

50

--Senesi - 1 5 pel Mo>, Avg (!>etw~l f

Figure F-5. Relative signal strength plotted as afunction ofdistance traveled during straight-2 trial

STRAIGHT-3
The Straight-3 trial was conducted in a manner similar to the previous two trials. However, it provided a unique opportunity for the effects of same lane traffic to be observed when the same lane traffic was in front of the receiver vehicle and very close to the receiver vehicle. Figure F-6 shows a plot of relative signal strength as a function of receiver vehicle distance traveled. A 15 point moving average has been laid over the plotted signal level. During the Straight-3 trial, a full size pick-up truck pulled out approximately 70 feet in front of the receiver vehicle from a side street when the receiver vehicle was approximately 300 feet from the starting point and remained in front of the receiver vehicle for the entire trial. The data in Figure F-6 show several effects from signal blockage caused by the truck. First, there is considerable blockage of the mobile transmitter signal by the truck's presence. The fluctuations in relative signal strength around the moving average value appear to be more frequent and somewhat higher in amplitude than when a vehicle is not introducing blockage.
In the data produced during both the Straight-l and Straight-2 trials, the relative signal strength is normally above the display threshold value of 20 most of the time. In the data shown in Figure F-6 the relative signal strength stays much lower than the more normal previously cited cases. It is interesting to note that after the receiver vehicle passed the transmitter vehicle there was a short term increase in signal strength over that normally observed after the receiver vehicle had passes the transmitter vehicle. This increase in relative signal strength is caused by the signal from the transmitter being reflected off of the back of the pick-up truck. The results of the Straight-3 trial show the effects that signal blockage from near-by traffic can cause to both negatively and positively affect SWS operational performance.
STRAIGHT-4
The mobile transmitter was moved and the fixed site transmitter was placed on a tripod at the same location. The receiver vehicle was driven the speed limit of 45 miles per hour during this trial. The Straight-4 trial was designed to determine the difference in signal strength generated by the fixed site transmitter and the mobile transmitter, also to determine any differences in operating the fixed site transmitter with the door open and closed.
Figure F-7 shows the relative signal strength produced by the fixed site transmitter with transmitter door closed, plotted as a function of distance traveled from the start of the test. The 15 point moving average
F-9

1()()

140

)Y)

120

110

1(X)

..::

&ce. 00

Vi Icii

50

Q>

(.0.. 70

-"T1 I
0

~ ~
~ 60
OC

:,0

40

30
~:()

10

--$elml;l
- l o p e r Mo\!. Avg (:\eri~ll

~
N

.-
?,. lfl

H'S,<'
~'l

~.
,~ ..

.~)
'l

..'r."<-

8~.-.l

.,(-.
n<)0

~

~
r

~
~
""

....
f~
~

0,-:;>' f~.-..)'

L1lstam:;tl b'<\\I~I~<l (feet)

Figure F-6. Relative signal strength of SWS signal when receiver vehicle follows a shadowing pick-up truck

00
50
40
30
20
10 i"
{)
i~

Oist..o(;~ Traveled (f~t~

Figure F-7. Relative signal strength produced by the fixed site transmitter with front cover closed

trend line has been applied to the data. The travel distance of the receiver vehicle is shorter than in the previous tests because the data collection effort did not begin until the receiver vehicle was further past the intersection of SR 93 and SR 120 Connector than during other tests. When data collection was begun the relative signal strength was 85, a much higher level than previously recorded due to the fact that all power transmitted by the fixed site transmitter is directed in the forward direction (the mobile transmitter directs half its power forward and half rearward) and the fixed site transmitter emits slightly higher power than the mobile transmitter. There are two multipath nulls that reduce the signal level. The first occurs at a point approximately 1,925 feet from the starting point of the trial. Another multipath null occurs at a point approximately 2,454 feet beyond where the trial began. The maximum relative signal strength at transmitter passage is approximately 132 compared with the maximum peak: relative signal strength level of 118 observed during previous trials using the mobile transmitter as the signal source. This difference in relative signal strength between the mobile and fixed site transmitter is approximately 6 to 7 dB.
STRAIGHT-5 TRIAL
The Straight-5 trial was conducted to determine if the fiberglass door on the transmitter, thought to be transparent to radar signals, caused attenuation of the SWS signal. Figure F-8 shows relative signal strength produced by the fixed site transmitter with the enclosure door open, plotted as a function of travel distance from the start of the test. A 15 point moving average trend line has been applied to the data. The relative signal strength data that were produced are almost identical to the results previously shown in Figure F-7 with several exceptions. The relative signal strength shown in Figure F-8 is 10 units higher at the start of the Approaching-5 trial. The two slight decreases in relative signal strength at a point 550 and 850 feet from the starting point of the trial were induced by two vehicles momentarily passing through the fixed site transmitter and the receiver vehicle's line of sight. The very deep nulls that occur at a point approximately 3/4 way through the data in both Figures F-7 and F-8 are from multi path effects.
The peak relative signal strength values monitored as the receiver vehicle passed the transmitter shows that there is very little difference in the relative signal strength -produced by the fixed site transmitter whether the front fiberglass door radome is open or closed. For example, the peak signal value in Figure F-7 is approximately 132 and the peak: value shown in Figure F-8 is also 132. The data also show that the signal from the fixed site transmitter was lost immediately aftex the receiver vehicle passed the
F-12

120

110

, tOO
.c; i. ...

.'5.c,>
Z"")

00 ...'..

1\
[j

1...JK ,..

00

C

0'>

<Ii 70

-'Tj I
w

.'.":!:

I. i

iil

r

00 .......

~ I>i .

50 .>

rr..

...

40

'i

:m > )/
i> :"'0 ...... .....
1{}

. F==s~~~i""H'
!~t5~~~~~~~H(~

,.-l,O
N

,~,~',1

~
to

,.0n.....

-Ie
{..". )

,n~,'"

m~1

~

,l.,O.,

&

'!f

*,-

N

N

g~1

~.,
<">

wr",,.,
<'>

Disti>1m:e l'f"\tele4lf~l}

Figure F-8. Relative signal strength offixed site transmitter with front cover open

fixed site transmitter due to the high degree of forward directivity of both the fixed site transmitter and detector's antenna beam. CURVE TESTING The SWS mobile transmitter was tested to determine its effectiveness in warning a motorist that an emergency vehicle is around a curve either stopped in the active lane or on the side of the highway. Figure F-9 is a map of the roadway where the curve performance was evaluated. The degree of curvature of the test curve is 3 degrees according to the construction plans for the SR 120 Connector supplied to the author by the GOOT.
Figure F-9. Map of curve test layout
TRANSMITTER VEHICLE SET-UP Referring to Figure F-I0, the transmitter vehicle was headed toward the northwest approach path of the receiver vehicle and parked off-set from the approach lane appro~imately 11 feet on the shoulder of the road. There was extremely heavy foliage that precluded a line of sight being established between test vehicles until the receiver vehicle was very close to the transmitter vehicle. This heavy foliage was very effective in blocking the propagation from the mobile transmitter, as will be shown in the data.
F-14

Figure F-lO. Photograph ofset-up of the transmitter during the curve tests
CURVE-1 TRIAL
The receiver vehicle approached the transmitter from the starting point of the intersection of SR 93 and SR 120 Connector. No signal was received from the SWS mobile transmitter until the receiver vehicle was several hundred feet west of the previous transmitter site at the intersection of Lucille Drive and SR 120 Connector at an approximate distance of 1,022 feet from the mobile transmitter. There was no line of sight between the test vehicles at this point. Figure F-ll shows the relative signal strength plotted as a function of distance traveled from the point of first signal detection.
Referring to Figure F-ll, a trendline representing a moving 15 point average has been laid on the relative signal strength data. The transmitter was located at a point 1,022 feet from where the signal was first received and a corresponding signal peak is found at this distance plotted on Figure F-ll. The relative signal strength oscillates very rapidly around the mean value even during periods when line of sight has been established with the transmitter vehicle. The signal strength does not rise above the display activation threshold of 20 until the receiver vehicle has traveled approximately 497 feet beyond the point
F-15

.,.
:.
f,() f.,

I=-:=S~fl~S1
i~l!.1Wf MOl: A'V,!~"I*sj)

o
(;)
Figure F-ll. Relative sigl'al strength plotted as afunction of the distance at which the SWS signal was first detected

at which the signal is first detected. Line of sight is established between the test vehicles approximately 745 feet after first signal detection at which point the relative signal strength begins to rise very rapidly. The receiver vehicle passes the transmitter vehicle at a point approximately 1,022 feet after initial signal detection. The signal continues to be detected above the display activation threshold of 20 until the receiver vehicle reaches a point 1,740 feet past the point where the SWS signal was first detected, a distance of approximately 718 feet beyond the transmitter vehicle. The signal from the transmitter vehicle was received beyond the transmitter vehicle because the mobile transmitter antenna radiates equal power both forward and rearward and there is some response to a signal from the rear by the SWS receiver.
CURVE-2 TRIAL
Figure F-12 shows relative signal strength plotted as a function of distance traveled by the receiver vehicle since first signal detection. The trendline moving window average applied over 15 points has been overlaid on to the plotted data as a black line. The relative signal strength exceeds the display threshold of 20 after the receiver vehicle has traveled approximately 497 feet beyond the point at which the signal was first detected. As in the Curve-1 trial, line of sight is achieved between test vehicles at a point approximately 650 feet after the transmitter is first detected or when the test vehicles are approximately 382 feet apart. The signal peaks approximately 1,100 feet from the point at which the SWS signal was first detected, a range approximately 78 feet different than achieved in the Curve-l trial. The signal level does not drop below the display threshold of 20 until the receiver vehicle is approximately 1,525 feet beyond the point where the signal was first detected, a distance 425 feet beyond the transmitter vehicle.
SUMMARY OF CURVE TESTING
The curve testing demonstrated that heavy foliage that reaches the ground will absorb (attenuate) the Safety Warning Signal to the point that a direct path signal can not be propagated through the tree line to allow long range detection of the signal. During Curve-I and Curve-2 testing, a worst case situation was observed due to the canopy of the trees extending to the ground. If the trees had been mature pine or hardwood the foliage canopy would have been above the transmission path. The trunks of the trees would have attenuated and reflected the signal some, but not as much as the thick canopy attenuated the
F-17

.c
.;'5er.:;:>;. 00 6<i

~

I:;l)

~., 70

'TI

~..

..I....

'is W

00

IX

SO

40

--Seriesf -=lSpef Mov A'ig {Se/i<!~l)

Figure F-12. Relative signal strength plotted as a function ofdistance at which the signal was first received during curve testing

signal in the two curve trials that were conducted. The forward transmission mode that allowed the signal to be received beyond the line of sight between test vehicles is thought to be forward scatter from the trees on the opposite side of the road that can be seen in Figure F-lO. This site was intended to be used as a location to test the seasonal effects on the SWS transmission capability. While a very elaborate test would be required to develop a chart of absolute attenuation per linear foot of tree line, the relative seasonal variation in signal strength for the curve test location was to be determined by conducting a second curve test during late fall when the leaves were not on the trees. This test, however, could not be conducted within the scope of the project.
HILL TESTING
Previous to the rural tests being conducted, hill testing was conducted at Dobbins Reserve Air Force Base. Given the results from the Dobbins hill tests, two series of hill tests were conducted. The hill trials described in this appendix were conducted on Georgia State Route (SR) 381, a section of highway approximately one mile west of New Hope, Georgia, a small community that is northeast of Dallas, Georgia. A map of the test site is shown as Figure F-13. Referring to Figure F-13, the test began at the intersection of Industrial Way and SR 381 and continued in an northeasterly direction along SR 381. SR 381 construction plans supplied by the GDOT showed that the elevation ofthe highway at the crest of the hill was 1,040.5 feet above mean sea level (MSL) and the elevation at the bottom of the hill where the transmitter vehicle was located was approximately 1,033 feet MSL. The distance from the crest or peak of the hill to the point where the transmitter vehicle was located was approximately 1,100 feet. The transmitter was located approximately 5 feet above ground level on the rooftop of the transmitter vehicle the receiver was located at an approximate level 8 feet above ground level. Either the "as built" design was different from initial design plans or the presence of the hill is an optical illusion.
Several visits to the test site were made to try and resolve the fact that the difference between the elevation of the hill's crest and the bottom of the hill is only 8 feet as shown in the highway design drawings completed in 1940. During the initial test site survey, the elevation between the crest of the hill and the bottom of the hill was thought to be approximately 40 feet on the basis of an approaching
F-19

driver's perceived line of site to transformers on power poles installed at the bottom of the hill. Apparently, the effect of a deep hill is caused by an optical illusion created by the low elevation at the starting point for each trial. Figure F-14 shows a picture of the portion of the highway over which the testing was conducted viewed from the intersection of Industrial Way and SR 381. The highway plan shows that the elevation at this intersection is approximately 1,000 feet MSL which is 40. 5 feet below the crest of the hill and 36.5 feet below the transmitter site on the other side of the crest of the hill. Figure F-14 does show a depression in the roadway so that a driver's line of sight would be elevated during approach to the hill.
Figure F-13. Map showing location oj hill testing and point at which line-oj-sight between test vehicles occurred
Figure F-15 shows the view of the hill from the transmitter vehicle perspective (concrete truck not present during testing). It is conceivable that if the camera had been leveled to the elevation of the mobile transmitter, the top of the mobile transmitter would have intersected the hill at a point 8 feet below the crest. The photograph also shows that the tops of the trees drop away rapidly beyond the crest of the hill further supporting the theory that the high degree of sl~pe at the approach to the hill had the effect of making the hill appear deeper.
F-20

Figure F-14. Starting point at intersection of Industrial Way and SR 381, with view toward test hill
Figure F-15. View from transmitter vehicle location, with view toward hill used in the Hill-4 trial F-21

The data that were collected and presented in the paragraphs that follow show that until line of sight was established, the message display threshold of 20 was not exceeded. Thus, any attempt to develop relationships between hill crest and depression ratios is unnecessary. Line of sight is the rule except where there are objects above the hill that will reflect the SWS signal (a scenario which will be discussed).
SYSTEM SETUP
Figure F-14 showed the view from the starting point of the trial, at the intersection of Industrial Way and SR 381 with a view toward the northeast. The vehicle in the foreground is just approaching the top of the hill that was used as the test hill. The receiver vehicle began the test by turning left, entering the right lane and proceeding northeast along SR 381. Figure F-15 showed the transmitter vehicle offset from the side of the roadway approximately 12 feet. The point at which the transmitter vehicle is located was considered a safe location close to the bottom of. the hill. The large concrete mixer truck that appears in the photograph was not present on the highway when the hill trials were conducted. The mobile transmitter was mounted on a suction cup platform on top of the transmitter vehicle driven by Evon Braseleman, a GTRI student. The receiver vehicle approached the transmitter vehicle from behind down the test hill shown in the background. The relative signal strength was recorded as a function of time via the RS-232 port of the modified BEL-TRONICS detector previously described in the rural tests.
HILL-1 TRIAL
The receiver vehicle was driven by the author at a constant speed of 45 miles per hour. At this speed, the receiver vehicle required approximately 48 seconds to reach the transmitter vehicle 3,168 feet from the starting point. The trial run continued for an additional 21 seconds or 1,386 feet after the receiver vehicle passed the transmitter vehic.Ie. Line of sight was established between the receiver vehicle and the transmitter vehicle 34 seconds after the receiver vehicle left the starting point. The distance between the receiver and transmitter vehicle when the line of sight was established was approximately 1,122 feet.
Figure F-16 is a plot of received relative signal strength as a function of distance traveled by the receiver vehicle. As in previous tests the BEL 855 STi detector with RS-232 interface to the laptop computer was used to record data. The trendline of a 15 point moving average was laid over the data to provide a mean value. The signal from the mobile transmitter located at the bottom of the hill is received during the
F-22

--Selie$l -==t5f~. t.-1<J. A'i~I.Sl,!(i:Sld
Figure F-16. Relative signal strength plotted as afunction ofdistance traveled by receiver vehicle

entire trial, however the signal level does not exceed the display activation threshold of 20 until the receiver vehicle reaches a point approximately 1,950 feet from the starting point or until the receiver vehicle was 1,128 feet from the transmitter vehicle. This was the point where line of sight was established between test vehicles. The warning display would have remained activated until the receiver vehicle had passed approximately 1,500 feet beyond the transrrutter vehicle at the speeds driven.
HILL-2 AND HILL-3 TRIALS
Two additional trials were conducted in the same manner as the Hill-l trial. The results were the same as shown in Figure F-16, with the exception of several nulls in the signal strength caused by reflections from traffic in the approaching lane. Because the data are almost identical, the additional two figures are not presented.
HILL-4 TRIAL
During the Hill-4 trial, the receiver vehicle approached the transmitter vehicle from the opposite direction from the top of the hill further east. Figure F-17 is a photograph that shows the hill to the east. The elevation of the crest of this hill was shown on the highway design drawing as 1,080 feet MSL, 39.5 feet elevation above the point where the transmitter vehicle was located. The receiver vehicle approached at a constant speed of 45 miles per hour. rfhe SWS signal was received for 18 seconds or 1,118 feet prior to the peak: of the hill being reached by the receiver vehicle, at which point a line of sight was established between test vehicles. The receiver vehicle was parallel with the transmitter vehicle 16 seconds later after traveling an additional 1,056 feet from the peak of the hill.
Figure F-18 shows the relative signal strength plotted as a function of distance traveled after the ~ignal was first detected. Again, the signal exceeds the display threshold of 20 when a line of sight is established at a point approximately 1,162 feet, between vehicles. There are several nulls in the plotted signal due to traffic in the approaching lane passing between the receiver and transmitter vehicle. The relative signal strength drops very quickly to a relative signal strength of 50 very quickly after the receiver vehicle passes the transmitter vehicle at the bottom of the hill. Once above 20, the relative signal strength did not drop below 20 for any period greater than 10 seconds until the receiver vehicle was approximately 950 feet beyond the transmitter vehicle.
F-24

Figure F-17. View from transmitter vehicle location with view toward hill used in the Hill-4 trial SUMMARY OF HILL TEST RESULTS The hill test results again confirmed that the SWS signal can be received beyond the line of sight, but generally a line of sight must be established before the signal level is high enough to cause the warning message to be displayed. During testing at Dobbins Reserved Air Force Base there was an exception to this finding. It was observed that a mechanism was present which reflected enough signal to exceed the display threshold before a line of sight was established. This caused the warning message to be displayed before line of sight was established between test vehicles. This reflective mechanism contributing to early detection was thoue# to be the upper tree canopy.
F-25

150

1M)

t )(}

t :c'O

110

1W

.c.

.'c5. 00

,:;;

(!)
~

80

C.

..0>
J) 70

;>

~

~ 00

OC

f,.O

4fi

j()

:::'0

10

r==$~~;;;~i

"j

I 1 -H> ~r M<iV. Avg i ~efit!s1)

I.

,

_J

'fJ

N.

~

~

r~-

;:
a)

'0,"..

&'J
'<"

1'(~'<

('4
U'it'

~
~..,
~.

';j r-

~
a)

g~

... ~

<'"j

N~

."
~
'"

V
~

Ol lI) 1', N

~<.;~

8..l

0il<1.Ul(;e Travel,Hi (fli'~tl

Figure F-18. Relative signal strength plotted as a function of distance traveled by receiver vehicle since first signal reception

APPENDIX G URBAN TESTING OF THE SAFETY WARNING SYSTEM (SWS)
MOBILE TRANSMITTER AND RECEIVER
G-l

THIS PAGE INTENTIONALLY LEFT BLANK G-2

APPENDIX G URBAN TESTING OF THE SAFETY WARNING SYSTEM (SWS)
MOBILE TRANSMITTER AND RECEIVER
The functionality of the SWS mobile transmitter, when used to warn SWS equipped motorists of the presence of police and emergency vehicles crossing an urban intersection, was evaluated in the urban environment of downtown Marietta, Georgia. These urban trials measured the range at which the safety warning message, transmitted by the mobile transmitter, could be received when the receiver vehicle was approaching a blind intersection surrounded by multi-story buildings, a worst case test for the system. The Central Business District Test Plan found in Appendix B was followed when setting up the urban test.
GENERAL TEST GEOMETRY
Figure G-l is a schematic view layout of the urban test area. The primary test intersection was located where Anderson and Powder Springs Streets intersect. Figure G-2 is a photograph of the intersection of Anderson and Powder Springs Streets. Powder Springs Street is a two lane street, 40.5 feet wide curb-to-curb, with one-way traffic flow to the south. This photograph was taken from a point south of Anderson Street looking north along Powder Springs Street. Figure G-3 is another photograph of the same intersection taken from a point north of Anderson Street looking south along Powder Springs Street.
Figure G-I shows the sidewalk widths within the test area by utilizing numeric index codes on the figure. Definition of these index codes may be found in the legend at the upper left of the figure. Eleven index numbers are shown in the sidewalk width legend in the upper left comer of Figure G-I. The first number that appears in the legend is the index location reference number. The number that follows the equal sign is the width of the sidewalk (feet) at the point in Figure G-I where the index reference number appears on a sidewalk. Referring to Figure G-I, the curb width north of Anderson Street at index number 2 was 11.5 feet and at index point I the curb width was 7.5 feet. South of Anderson Street, the curb width at index number 10 was 11.5 feet and at index number 3 the curb was 21.5 feet. The building setback data were taken for analysis purposes should test results have shown that there was a detection range sensitivity to the setback distance of buildings from the street. Analysis of the detection range data
G-3

Sidewalk Width

1= 7.5' 2= 11.5' 3= 21.5' 4= 7'

5 = 7' 9 = 7.5'
6 = 5' 10 =11.5'
7 = 7' 11 = 7.5'
8 = 7.5'

Marietta Town Square
Roswell Street 138'

L

Powder

Springs

Commercial Mu Iti-story
Building

Street
.. 1
40.5'
2 ....

Commercial Multi-story Building

5

Commercial

E

Multi-story Building

t 24.3'
-------I

4

~2~ 1'

Anderson Street

I---------~~-~r_--------___,

.24.3'

10

Commercial

Multi-story

Building

461'

Commercial Multi-story
Building
3:=====:::;
Commercial Multi-story
Building

Old Library 9 461'

_ _ _ _ _ _ _ _ _ _ _ _w_a_V_er-'IY:...S_t_re.::..e;;..:t....:s;...w....:....-

t

Atlanta
- -, Street

I
I

Winters Street

Figure G-l. Layout of urban test site in downtown Marietta, Georgia

Figure G-2. The intersection ofAnderson and Powder Springs Streets looking north along Powder Springs Street
Figure G-3. The intersection ofAnderson and Powder Springs Streets looking south along Powder Springs Street G-5

showed that there was no detectable sensitivity to building set-back distance from the street. South of Anderson Street, the right lane of Powder Springs Street is a right turn only lane into Waverly Street to the west. The left lane of Powder Springs Street is a left turn only lane into Waverly Street to the east. The intersection of Anderson and Atlanta Streets served as a second test site. Figure G-4 shows this intersection from a point south of the intersection looking north toward Roswell Street. This test intersection did not have any buildings on the southeast corner and provided a chance to determine if the lack of buildings on all four corners would affect the range at which the safety warning message could be received. It was found that the detection results were not measurably sensitive to the presence or absence of a building, assuming that the absence of a building did not allow line-of-sight (LOS) to be established between the receiver and transmitter vehicles.
Figure G-4. The intersection ofAnderson and Atlanta Streets looking north towards the intersection URBAN TEST SITE SELECTION CRITERIA The intersection of Powder Springs and Anderson Streets was chosen as the primary urban test site because it is very similar to many intersections in an urban area and is similar to those intersections found in the central business district (CBD) of most cities. A multi-story building was located on each corner of the intersection. The buildings precluded LOS being established between the transmitter vehicle and the receiver vehicle, except when the receiver vehicle was in the middle of the intersections
G-6

of Anderson and Powder Springs Streets or Anderson and Atlanta Streets. In addition to the test location being typical of most urban intersections, the traffic using Anderson and Winters Streets was very light, which allowed the transmitter vehicle to operate safely from two stationary positions (to be discussed). In addition, light traffic ensured that the signal radiated by the mobile transmitter would be propagated toward the intersection without blockage by transiting vehicles.
TEST INSTRUMENTATION
A BEL Model 855 STi SWS capable radar detector was mounted on the windshield of the test vehicle 4 inches above the dashboard in the standard manner shown in figures included in the turntable and Dobbins test section. A video camera was mounted on a tripod in the back seat of the receiver vehicle so that the display of the radar detector could be recorded on the video track of the video tape. The author drove the test vehicle and used a head mounted boom microphone to record situational comments regarding receiver vehicle location on the audio track of the video tape. No high resolution signal strength data were recorded using a laptop computer, as was done in previous tests, because the test run was very short and multiple traffic lights would hold the receiver vehicle stationary for long periods of time. The stops precluded the receiver vehicle from making the test run at a constant speed, a requirement anytime high resolution relative signal strength data have been collected. The primary goal of measuring the distance from the test intersection at which the safety warning signal could be detected and the warning message displayed was met by determining the location of the test vehicle when the warning message first appeared on the detector display. The video tape supplied this information, also information regarding the receiver vehicle's location when the safety warning message was replaced by the stand-by message "Highway." This video record, coupled with the voice notation on the audio track of the video tape, was sufficient to allow the urban test goals to be met.
URBAN-1 TEST RESULTS
The mobile transmitter unit, manufactured by MPH Inc., was mounted to the top of the transmitter vehicle as shown in a figure in the section on Dobbins testing. Mounting the transmitter in this location simulates the manufacturer's suggested mounting point on an emergency vehicle or police car's emergency light bar. The driver of the mobile transmitter vehicle was Mr. B. H. Hudson of the GTRI technical staff. Figure G-5 shows the view from the transmitter vehicle when looking west along Anderson Street toward the intersection of Anderson and Powder Springs Streets. When located at this point, the transmitter was located approximately 202 feet from the center of the intersection of Anderson
G-7

and Powder Springs Streets. Figure 0-6 shows the view from the transmitter looking east toward the intersection of Anderson and Atlanta Streets. During Urban-I testing, the transmitter vehicle was approximately 150 feet from the center of the intersection of Anderson and Atlanta Streets. The mobile transmitter signal was radiated in both directions along Anderson Street by the forward and rearward pointing antennas of the mobile transmitter. When the Urban-l test began, the mobile transmitter was turned on and allowed to stabilize for approximately five minutes before testing was begun. The tests were coordinated by two-way radios in each vehicle. When the transmitter was thought to be stable and Anderson Street was clear of traffic, the receiver vehicle started the test run.
Figure G-5. View from the transmitter vehicle looking west towards the intersection ofAnderson and Powder Springs Streets
The starting point of each trial in the test series began at the intersection of Atlanta and Waverly Streets, as shown in Figure 0-7 (double line trail of the receiver vehicle). Referring to Figure 0-7, the receiver vehicle first displayed the safety warning message "Stationary Emergency Vehicle Ahead" approximately 150 feet from the starting point. The point at which the message was displayed is shown as the point where the double line receiver vehicle trail was replaced by the solid black line on Figure 0-7. The safety warning message continued to be displayed as the receiver vehicle continued north on Atlanta Street through the intersection of Anderson and Atlanta Streets. The safety warning message display returned to the stand-by message "Highway" when the receiver vehicle reached a point approximately 20 feet from the intersection of Atlanta and Roswell Streets, as shown in Figure 0-7.
0-8

Upon reaching the intersection of Roswell and Atlanta Streets, the receiver vehicle turned left onto Roswell Street and proceeded west along Roswell Street. There was no detection of the safety warning signal during the two block transit of Roswell Street.
Figure G-6. Viewfrom mobile transmitter looking east towards the intersection of Anderson and Atlanta Streets (truck was not present during testing)
When the receiver vehicle reached the intersection of Powder Springs and Roswell Streets a left turn was made. Figure G-8 shows the scene looking south along Powder Springs Street at the point where the receiver vehicle turned left into the right lane of Powder Springs Street from Roswell Street. The right lane transit through the test intersection of Powder Springs and Anderson Streets was specified to detennine if there was a lane based sensitivity that would affect test results. (The Urban-2 test was conducted from the left lane for comparative purposes.) When the receiver vehicle turned left onto Powder Springs Street from Roswell Street, the receiver vehicle speed was maintained at 20 miles per hour. The safety warning message was displayed starting at a point 35 feet south of the intersection of Roswell and Powder Springs Streets. The mobile transmitter safety warning message continued to be displayed until the receiver vehicle was approximately 150 feet from the intersection of Powder Springs and Waverly Streets. After the receiver vehicle turned left on to Waverly Street, there was no reception of the SWS message or a detector alert during the receiver vehicle's transit along Waverly Street, even
G-9

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _:-1

Commercial Multi-story

t

Marietta Town Square

N

Winters Street

Left Lane --.... .-

........

"

"'1--- Roswell Street
IA ~

Commercial

Commercial

Commercial

a I
o

Multi-story Commercial

Multi-story
~
Commercial

J~

Multi-story
~
Old Library

...._..._- - Anderson Street

Multi-story

Multi-story

Commercial Multi-story

Commercial Multi-story
"

II Transmitter Vehicle I
Legend

n II
I --

Starting Point Waverly Street (one-way East)

Powder Springs Street (one-way South)

Antenna Beam Direction Receiver Vehicle Path Where Message Displayed

Atlanta Street (one-way North)

Figure G-7. Urban-l test layout with detection results plotted as shown in the legend

though Winters Street divided the two city blocks and offered a possible path for safety warning transmitter signal propagation. The total distance over which the mobile transmitter's warning signal was displayed by the detector in the receiver vehicle was thought to be much shorter than that observed and shown in Figure G-7. During the calibration of the detector used in the urban tests, it was noted that the BEL 855 STi detector would continue to display the safety warning message for a period of 9 to 10 seconds in duration after the SWS signal was turned off because of a long time constant programmed into the display logic. This finding indicates that the reception of the safety warning signal was most likely lost immediately after the receiver vehicle passed through the intersection of Anderson and Powder Springs Streets even though the safety warning message continued to be displayed after the receiver vehicle passed the intersection. It is suspected that the detector manufacturer incorporates the display latency factor to ensure that the safety warning message does not return to the stand-by message during brief periods of signal drop-out. More on the subject of the effects of display latency will be presented in the urban test summary section which follows.
Figure G-8. Scene looking south along Powder Springs Street towards test intersection URBAN2 TEST RESULTS Referring to Figure G-9, the same test path was followed during the Urban-2 test as during the previous Urban-l test, except the test vehicle turned down the left lane of Powder Springs Street instead of the right lane. It was reasoned that more signal might be present when the receiver vehicle was closer to the buildings on the east side of Powder Springs Street. Referring to Figure G-9, the message was lost on the
G-Il

Commercial Multi-story

t

Marietta Town Square

N

Winters Street

RjghtLane~

-.-

V

....E - - - Roswell Street
j~

a
..I....
tv

Commercial Multi-story

Commercial Multi-story
~

..

Commercial Multi-story

~

..._...._. --Anderson Street

Commercial Multi-story

Commercial Multi-story

J~ Old Library

Commercial Multi-story
"

D Transmitter Vehicle I

Legend

Powder Springs Street

Antenna Beam Direction

(one-way South)

Receiver Vehicle Path

Where Message Displayed

=

ri
I
t

Starting Point Waverly Street (one-way East)

Atlanta Street (one-way North)

Figure G-9. Receiver vehicle test path during Urban-2 test

point at which SWS message was displayed was the same as for the Urban-I test. The point at which the display of the detector was also the same as for the Urban-I test. These results showed that the displacement of the receiver vehicle by 20 feet toward the transmitter did not make a noticeable difference in the results obtained. Tht? detection results obtained during Urban-I tests at Anderson and Atlanta Streets were also obtained during Urban-2 tests at the intersection of Anderson and Atlanta Streets. This result was expected because the path driven by the receiver vehicle remained the same and the transmitter vehicle's location was not changed. The consistent results achieved between the Urban-2 and Urban-3 tests also indicate that traffic effects were not biasing the test results.
URBAN-3 TEST RESULTS
It was thought desirable to determine if there might be enough power propagated at a high angle over the tops of the multi-story buildings so as to allow detection of the safety warning signal at a distance beyond the immediate test area further north on Powder Springs Street. This possibility was tested by moving the test path further north as shown in Figure G-IO. Referring to Figure G-IO, the receiver vehicle crossed the intersection of Atlanta and Roswell Streets, continuing north until reaching the intersection of Atlanta and Lawrence Streets. The receiver vehicle turned left onto Lawrence Street and traveled along the north end of the Marietta, Georgia, City Square until reaching the intersection of Lawrence and Powder Springs Streets, at which point the receiver vehicle turned left and headed down Powder Springs Street. At a point approximately 200 feet from that intersection, the detector alarmed and the display indicated that that a K-band signal had been received; however, the test safety warning message was not displayed. The detector continued to provide an indication of a K-band signal (but no safety warning message) until the receiver vehicle was crossing the intersection of Roswell and Powder Springs Streets. At a point approximately 30 feet south of the center of the intersection of Roswell and Powder Springs Streets, the safety warning message was displayed and remained activated until the receiver vehicle reached a point approximately ISO feet from the intersection of Waverly and Powder Springs Streets. After the receiver vehicle reached that point, the display returned to the stand-by mode.
It is thought that the brief detection of the safety warning transmitter signal, without a message display, that occurred midway along the Marietta City Square was the SWS signal being reflected off vehicles passing through the intersection of Anderson and Powder Springs Streets. However, because the energy was reflected for such a short time, the detector was not able to decode and display the safety warning message. This test indicated that the safety warning signal was as effectively blocked by the multi-story
G-I3

Commercial Multi-story

t

Marietta Town Square

III

N

Winters Street

,j

Lawrence Street

Left Lan~

...~-- Roswell Street

"

-a I

Commercial Multi-story

Commercial Multi-story
~

.. Commercial Multi-story ~

Anderson

Commercial

Commercial

~~

Old Library

Street

Multi-story

Multi-story

Commercial Multi-story
"

D Transmitter Vehicle I

Il
I --

Starting Point Waverly Street

Powder Springs Street lone-way South)

Legend
Antenna Beam Direction Receiver Vehicle Path

(one-way East)
t

Where Message Displayed Warning Beep Only

Atlanta Street (one way North)

Figure G-lO. ModifiNl test path to test for high angle radiation coming over the tops of multi-story buildings

buildings on each corner of the test intersection as if the buildings had been much higher (like the building found in the CBD). This finding also indicated that perhaps CBD testing was not necessary.
URBAN-4 TEST RESULTS
The test plan' being followed called for the mobile transmitter vehicle to be moved closer to the intersection of Powder Springs and Anderson Streets and further from the intersection of Anderson and Atlanta Streets. Figure G-II shows the modified test set-up where the transmitter vehicle was moved further west along Anderson Street to a point approximately 60 feet from the center of the intersection of Anderson and Powder Springs Streets. Also, for safety purposes, the transmitter vehicle was moved to the left side of the street in a parking area. Figure G-12 is a photograph that shows the Anderson and Powder Springs Streets intersection from the point where the mobile transmitter was located. There was some concern that the two telephone poles in the scene might affect test results, but test results showed that no noticeable effect from the telephone poles could be detected.
Referring to Figure G-II, the UrbanA test was conducted along the same test path as that used during the Urban-3 test. The receiver vehicle traveled north on Atlanta Street to the northeast corner of the square, turning west on Lawrence Street and then south when the corner of Lawrence and Powder Springs was reached. The detector alerted that a K-band signal had been detected at a point of receiver vehicle travel two hundred feet from the point at which the turn onto Powder Springs Street was made. The safety warning signal level was briefly higher at that point than at the same point during the Urban-3 test, but the signal strength was not high enough to cause the safety warning message to be displayed. This higher signal strength reading indicated that slightly more signal was received as a result of moving the mobile transmitter closer to Powder Springs, but the small increase in signal strength was not significant.
When the front of the receiver vehicle entered the cross walk on the south side of Roswell Street at Powder Springs Street, the safety message was displayed. This is the same point at which the message was displayed during the Urban-3 test. Better results had been expected when the transmitter vehicle was moved closer to the test intersection. It was reasoned that the proximity of the transmitter vehicle to the intersection of Anderson and Powder Springs Streets would allow more power from the mobile transmitter to propagate north up Powder Springs Street. This anticipated effect did not occur, although the signal strength reading at a point half way down the square was stronger than the signal strength reading that occurred during the Urban-3 test.
G-15

Commercial Multi-story
f LeftLane~

t

Marietta Town Square

N Winters Street

lit

Lawrence Street
J~
....:----- Roswell Street

Commercial Multi-story

Commercial Multi-story

Commercial Multi-story

(;)
...I....

.-.I 11II'I

0\

Commercial Multi-story

~ Commercial
Multi-storv

~
Old Library

Anderson Street

Commercial Multi-story

Transmitter Vehicle

"

Powder Springs Street (one-way South)

Legend
Antenna Beam Direction Receiver Vehicle Path

Dn
I --
t

Starting Point Waverly Street (one-way East)

Warning Beep Only Where Message Displayed - - - -

Atlanta Street (one-way North)

Figure G-ll. Transmitter is moved much closer to the test intersection at Anderson and Powder Springs Streets

Figure G-12. Scene viewed by mobile transmitter when it was moved closer to Powder Springs Street
URBAN TESTING SUMMARY
The results obtained during urban testing were anticipated on the basis of results documented during earlier testing of the SWS on the turntable and during testing at Dobbins Reserve Air Force Base, as well as on the basis of the physics of radiowave propagation at the SWS operating frequency 24.1 GHz. The SWS mobile transmitter uses directional antennas beams that focus the radiated energy in a forward and rearward direction from the transmitter vehicle. The transmission frequency of 24.1 GHz will not penetrate the brick wall of a building at the low power levels that are radiated by the transmitter. As a result, the only propagation path that can be established is either over a direct LOS path or by reflected ray path. Figure G-13 shows one possible explanation as to how the short range propagation was achieved. Referring to Figure G-13, the main lobe of the transmitter vehicle's signal was propagated west and east down Anderson Street toward the two test intersections. (The situation at the intersection of Anderson and Atlanta Streets is not discussed further because the discussion applies to both cases.) The buildings along Anderson Street blocked propagation at most angles other than ahead and behind the transmitter vehicle. However, it is suspected that some energy was reflected from the front of the building at the intersection of Powder Springs and Anderson Streets. The physics of signal scattering from objects is very complex and is the subject of many books, technical journals and scholarly papers,
G-17

Multi-Story Building

Multi-Story Building

Multi-Story Building

a

Main Antenna Lobe

I

00

Containing Most of Signal

Anderson Street

Multi-Story Building

Powder Springs Street

Multi-Story Building
Multi-Story Building

Mobile Transmitter

Figure G- J3. Suspected propagation mechanism that allows SWS signal to be detected during urban intersection cross tests

and is not easily simplified. However, it is suspected that the ray reflected off of the building fronts propagated toward the intersection of Roswell and Powder Springs Streets as multiple reflections as shown in Figure G-13. It is also suspe~ted that vehicles passing through the intersection of Anderson and Powder Springs Streets reflected energy north along Powder Springs Street.
The short range at which the mobile transmitter could be detected once the receiver vehicle had passed the test intersection was discussed in an earlier paragraph. The only reason that the safety warning message was displayed after the test intersection was because 10 seconds were required after the mobile signal was lost before the display returned to stand-by. The fact that the signal is lost almost immediately past the test intersection is a desirable feature of the SWS. The fact that the SWS will not respond to emergency or police vehicles crossing an intersection behind the receiver vehicle allows an SWS equipped motorist to assume that when the safety warning message is displayed, the emergency vehicle is either overtaking or approaching on the street of travel or is crossing in front of the receiver vehicle.
In summary, the following guidelines were developed regarding the SWS performance when used by an emergency or police vehicle to warn of their crossing of an intersection:
I. The distance at which an SWS equipped motorist will be warned of an emergency or police vehicle approaching an intersection will be at a range of a typical urban block. The early warning time that the SWS provides will be dependent on the speed of the SWS equipped vehicle as it approaches the intersection;
2. The physics of signal propagation at 24.1 GHz prevents direct path propagation through buildings between the receiver and transmitter vehicle;
3. Propagation of the SWS signal over any building that blocks the LOS between the receiver and transmitter vehicle will normally not occur unless a reflecting object provides a path of propagation; and
4. The distance at which the safety warning signal can be detected from a transmitter vehicle crossing behind the receiver vehicle is extremely short.
G-19

APPENDIX H INTERSTATE TESTING
H-l

THIS PAGE INTENTIONALLY LEFT BLANK H-2

APPENDIX H INTERSTATE TESTING
TEST SEQUENCE ORDER
Interstate testing was conducted in a manner defined in the SWS Test Plan included in this report as Appendix B, with several exceptions to be discussed in the paragraphs that follow. The first interstate test that was performed was designed to test the range at which an emergency vehicle equipped with a mobile SWS transmitter could be detected as it overtook a motorist equipped with an SWS receiver (both were traveling in the same direction). The second test was conducted to determine the range at which the mobile transmitter could be detected when the receiver vehicle was the overtaking vehicle (both vehicles moving in the same direction). The third interstate test was conducted to determine the range at which the fixed site transmitter could be detected when mounted over the traffic lanes of the interstate. A variation on the test plan found in Appendix B was used because large scale traffic effects on SWS signal propagation were thought to be so great as to totally obscure the small scale effects the tests were designed to explore. The fourth test was designed to determine the range at which the safety warning signal would be first displayed when the receiver vehicle was not in the lanes intended for reception of the warning message. In the process of conducting this last test, the magnitude of signal strength variations caused by reflections and blockage from interstate traffic was further defined.
INSTRUMENTATION FOR THE INTERSTATE TESTS
The SWS mobile transmitter built by MPH, Inc. was mounted on a suction cup platform in the middle of the roof of a passenger car vehicle driven by Mr. B. H. Hudson, a member of the GTRI staff, as shown in Appendix E. The mobile transmitter radiated the safety warning signal both straight ahead and behind the vehicle on which it was mounted. Mounting the mobile transmitter in the center of the transmitter vehicle roof simulated the way that the MPH, Inc. transmitter will be mounted on the light bar of an emergency or police vehicle. During interstate testing, when the transmitter vehicle was in motion, the message "Emergency Vehicle in Transit" was transmitted. When the test vehicle was stationary, the message "Stationary Emergency Vehicle Ahead" was transmitted.
The battery powered fixed site transmitter, borrowed from the Safety Warning Foundation, L. C, was a tripod mounted unit capable of transmitting anyone of 64 fixed text safety warning messages. The
H-3

message "Slow Moving Vehicle Ahead" was selected as the test message. The fixed site transmitter radiated the safety warning message only in one direction.
The receiver vehicle was driven by the author. Two BEL 855 STi detectors were used during the tests. One of the detectors was a BEL 855 STi SWS compliant unit supplied by BEL-TRONICS which had been taken from off-the-shelf stock. It was used to confirm that the laboratory measured thresholds discussed in the next section were an accurate indicator of signal level at which the warning message would be displayed. It would display the safety message on an alpha numeric display when the moving mobile transmitter's signal was received. A second model BEL 855 STi detector, modified by BEL-TRONICS to output relative signal strength data via an RS-232 link, was also used. During the tests the two detectors were alternated on occasion.
The ideal situation would have been to operate both detectors in the receiver vehicle at the same time. However, it was found that when the detectors were in close proximity there was a danger that leakage from either detector's sweeping local oscillator could interfere with the other detector's reception of the SWS signal. Thus, the stock detector and the modified detector with RS-232 output were operated in alternating fashion during several of the tests.
METHOD USED TO ALLOW THE TWO DETECTORS TO BE COMPARED
Because both detectors could not be operated at the same time and both were used in alternating fashion, laboratory tests were conducted on the stock BEL 855 STi for the purpose of determining at what relative signal strength (display threshold) the message would be activated also to characterize the delays involved in activating and deactivating the warning display as a function of display threshold signal level. The laboratory test showed that a relative signal strength threshold of 20 had to be exceeded to enable the display logic. Next, the detector required that the threshold of 20 be maintained or exceeded at low signal levels for 3 additional seconds before the warning message would be displayed. It was also determined that the warning message would remain on the display 10 seconds after the relative signal strength dropped below the display threshold of 20. These findings were used to analyze the plots of relative signal strength data produced by the RS-232 capable detector to determine when the warning message would be displayed. It was concluded that detector eq~ivalence between detectors could be achieved by subtracting 3 seconds from the time that the display threshold of 20 was first exceeded. Testing also demonstrated that the RS-232 capable detector was capable of collecting approximately 15 relative signal strength samples per second. Given this information, the elapsed time (seconds) over which any test was conducted could be determined by dividing the total number of samples by 15.
H-4

TEST DESIGN AND SETUP
The first trials were designed to detennine the distance over which the safety warning transmitter would cause the safety warning message to .be displayed when the transmitter vehicle overtook the receiver vehicle in interstate traffic in a manner defined in this appendix. During the initial test setup, both test vehicles started at the same point on the interstate. The receiver vehicle was driven the speed limit of 70 miles per hour while the trailing transmitter vehicle was driven at a speed of 55 miles per hour; a procedure designed to put distance between the receiver and transmitter vehicles while both were traveling in the same direction. The detector in the receiver vehicle continuously displayed the warning message when each trial was begun due to the close proximity of the test vehicles. The receiver was continuously monitored to detennine when the warning message could not be received. After the signal from the transmitter vehicle was lost and not displayed for one minute, the overtaking trial was begun.
Just prior to the beginning of each trial, the receiver vehicle took the lane position specified in the test plan and reduced its speed to 55 miles per hour. The transmitter vehicle took the position prescribed in the test plan and increased its speed to the speed limit of 70 miles per hour so that there was a speed differential of 15 miles per hour between the two vehicles. The transmitter vehicle driver started recording the Global Positioning System (GPS) derived latitude and longitude position of the transmitter vehicle every two seconds using a lap top computer as the recording system. The transmitter vehicle driver provided a time mark to the receiver vehicle driver at the start time via two way radio communications between the two vehicles. Later, during data analysis, the start time mark was used to correlate the GPS derived transmitter vehicle location to its actual location on Interstate 75 to determine where the overtaking test began. The detection range was computed by measuring the time taken for the transmitter vehicle to overtake the receiver vehicle from the point where the safety warning message was first displayed to the point where the transmitter vehicle passed the receiver vehicle. A constant overtaking rate of 15 miles per hour was assumed for the time distance calculations.
QVERTAKING-1 TEST
The procedure outlined in the previous paragraph was used to set up the receiver and transmitter vehicles for the Overtaking-l Test. The starting time of 12:35:30 placed the transmitter vehicle at the starting point shown in Figure H-l, a location approximately 34 degrees 4.925 minutes north latitude and 84 degrees, 37.911 minutes west longitude. Both vehicles were in the northbound lane of Interstate 75, and the receiver vehicle was in the right lane as a starting position for the test sequence.
H-5

Figure H-l. Mapshowing starting point of Overtaking-l trial Fifteen seconds after the starting mark at 12:35:30, the safety warning message was received and displayed on the BEL 855 STi detectors light emitting diode (LED) display. The transmitter vehicle passed the receiver vehicle 90 seconds after the safety warning message was first received and displayed. The differential speed of 15 miles per hour was maintained. The distance between test vehicles when the safety warning message was first detected was approximately 1,980 feet. No signal strength data were recorded during Overtaking-l testing because the stock off-the-shelf detector with no RS-232 output provisions was used.
OVERTAKING2 TEST
The set-up for the Overtaking-2 test was the same as in the previous test, except that the test vehicle was in the center lane of the 3 lane interstate and the RS-232 capable receiver was substituted for the stock receiver. The mark time that the test was begun was I :02:30 PM. At this time the GPS data placed the transmitter vehicle at the point shown in Figure H-2, which was approximately 34 degrees, 11.917 minutes north latitude and 84 degrees, 45.864 minutes west longitude. The part of the interstate where the test started was straight. The speed of the receiver vehicle was set at 60 miles per hour, since tractor trailer trucks closely followed the receiver vehicle when it attempted to maintain a speed of 55 miles per hour in the center lane. The transmitter vehicle's speed was 70 miles per hour. As a result the closing speed was reduced to 10 instead of 15 miles per hour.
H-6

Figure H-2. Map showing starting point ofOvertaking-2 trial
The previously discussed method of using relative signal strength data to determine the time at which the warning message would have been displayed was used in the analysis process. Figure H-3 shows the recorded relative signal strength data plotted as a function of elapsed time in units of seconds. Referring to Figure H-3, a plot representing a moving average of 15 points has been overlaid as a solid black line on top of the time variant SWS signal to provide the reader with a median signal reference point.
The SWS signal was received for approximately 153 seconds from the time of first detection until the transmitter vehicle passed the receiver vehicle. The first time that the signal exceed a relative signal strength of 20 was approximately 13 seconds into the trial. However, the signal did not remain above the display threshold for the required 3 seconds to enable the display. Referring to Figure H-3, the relative signal strength next exceeded the display threshold value of 20 at a point 22 seconds into the test and remained above 20 for the three seconds required to activate the display. The signal dropped below 20 for between 60 and 69 seconds, but the display would most likely have continued to display the warning message. There was another short drop in the signal level at approximately 110 seconds, but the time was too short to cause the warning message not to be displayed.
Given this analysis, the warning message would have been displayed for 131 seconds before correction. The previously derived correction factor of 3 seconds was subtracted from 131 seconds, making the total time that the warning message was present as 128 seconds. The range at which the warning message was first reliably displayed was 1,877 feet, given a closing speed of 10 miles per hour (14.7 feet per second).
H-7

110
::r: I
00

o ,.: : , , . ~. ,~r',,

.:.':::>
-t\.j'

t.
~""';j

,~':: (~-:>

,;'

'~<"~I'.

<.:~
G

hc

{'i

>,::::
~-~

(.::;

~;~ ,.-

q
~.

t
~t

Figure H-3. Plot of relative signal strength versus time shown the effects of other vehicle on the SWS signal

There were signal blockage and propagation effects present in the data. The relative signal strength dropped 60 seconds into the test due to a pick-up truck blocking the signal by pulling close to the rear of the receiver vehicle. The pick-up truck changed lanes at a point approximately 87 seconds into the test and the relative signal strength increased again to a level well above 20. After 107 seconds had elapsed, the relative signal strength again dropped to due to blockage by another vehicle changing lanes behind the receiver vehicle. After approximately 3 seconds of blockage, the obscuring vehicle moved into the left lane and the relative signal strength once again began to increase as the transmitter vehicle closed on the receiver vehicle. QVERTAKING-3 TEST The Overtaking-3 trial was not conducted from the left lane as planned. A mix-up in communications resulted in the transmitter vehicle not immediately following the receiver vehicle when Overtaking-3 test was begun. As a result of this problem, the receiver vehicle had to wait for the transmitter vehicle to turn around and find the receiver vehicle. The receiver vehicle took the fixed position shown in Figure H-4 for approximately 5 minutes before the transmitter vehicle was first detected approaching the receiver vehicle from behind. The location shown is on the side of the southbound ex.it ramp of Interstate 75 at Exit 123.
Figure H-4. Location of receiver vehicle during the Overtaking-3 trial
H-9

Figure H-5 shows the signal level that was recorded during the approach of the transmitter vehicle when the receiver vehicle was stationary at the bottom of the southbound ramp at Exit 123. The total elapsed time between the point where the signal level exceeded a threshold of 20 and the point at which the transmitter vehicle passed the receiver vehicle was approximately 22 seconds. The short time between first detection and passage occurs because the closing speed of the transmitter vehicle is 70 miles per hour.
Once detected, the safety warning message was present for 22 seconds above the display threshold of 20 before the transmitter vehicle passed the receiver vehicle at 70 miles per hour in the right lane. When the detection time was decreased to 18 seconds due to the correction factor, the range at which the safety warning message was displayed during the Overtaking-3 test was calculated to be 1,848 feet.
Figure H-5 also shows the effects of surrounding traffic on the SWS signal level. The 15 point moving average is plotted as a black line over the time varying SWS signal. At an elapsed time of approximately 24 seconds, the signal level oscillates around the median value established by the applying the trendline moving average. These oscillations are sinusoidal in nature, very large by comparison to changes in signal strength as a function of time seen previously, and thought to be caused by the traffic surrounding the transmitter vehicle.
QVERTAKING-4 TEST The Overtaking~4 Test began at Exit 122 on Interstate 75 southbound, and continued south on Interstate 75. Referring to Figure H-6, the warning message was displayed intermittently until the transmitter vehicle reached latitude 34 degrees, 5.180 minutes north, longitude 84 degrees, 41.611 minutes west, at which point the safety message was displayed continuously until the transmitter vehicle passed the receiver vehicle. The receiver vehicle was driven at a constant speed of 55 miles per hour, and (he transmitter vehicle was driven a constant speed of 70 miles per hour. The stock BEL 855 STi detector was substituted for the BEL detector with the RS-232 relative signal strength output. Signal strength was not recorded as a function of time as in the previous two tests. Instead, the video tape record which showed the message activity on the detector's LED display was used to determine when the SWS message was and was not displayed. No SWS signal was det~cted when the test began. The first detection of the SWS signal occurred 16 seconds into the test; however, the detector did not decode or display the SWS message even though the signal strength peaked to a level 3 on the
H-IO

1:

Wi

1,,0

120

110

1(l()

.s::.

0.
c

~tO

'l> ~

if)
i\j

flO

c

Ol

iii 70

co

::r:
..I....

.;~
0; 60

ll:

&0

40

:W

20

1D

0

0

.2

Eiapsed Time (Sflconds}

- - Skrlt::,l

- 1 5 p~~f. toAov A\-\l {Sf;r;os'l}

.." "

-,

"

.

Figure H-5. Plot of relative signal strength versus time showing signal level variations during Overtaking-3 trial

indicator. Recall that a level 3 signal is sufficient to cause the message to be displayed. It is suspected that the effects of surrounding traffic may have caused the signal strength to vary rapidly so that the decoder could not decode and display the message. At 52 seconds into the test, the warning message "Emergency Vehicle in Transit" was displayed for the first time and was held on the display until a point 65 seconds into the test, at which point in time the detector's display showed the message "Highway", the stand-by message displayed when the detector is not receiving a signal. At a point 77 seconds after the test had begun, the message "Emergency Vehicle in Transit" was again displayed and continued to be displayed until the transmitter vehicle passed the receiver vehicle 169 seconds after the test had begun.
Figure H-6. Location of transmitter vehicle at start of Overtaking-4 trial when warning message was received
Because there were two periods that the warning message was present, Overtaking-4 produced two results. If the effective warning range of the system were to be measured from the point in time when the SWS message was first presented to the driver, 100 seconds would have elapsed between display of the first message and the time that the vehicles passed. No correction factor was required using the stock detector. Given the speed differential of 15 miles per hour (22 feet per second) between transmitter vehicle and receiver vehicle multiplied by 100 seconds elapsed time provides an estimate that the vehicles were 2,200 feet apart when the safety warning message was first displayed. The reason that the signal was lost after 13 seconds was because the receiver vehicle went around a curve and the message was not displayed again for 25 seconds. If it were determined that the criterion for a successful test required that the message be displayed constantly until the transmitter vehicle passed the receiver vehicle, the time over which the message was constantly held would have been 75 seconds instead of 100 seconds. This more rigid specification of SWS performance would have reduced the effective detection range to 1,650 feet. The selection of one
H-12

or the other detection criteria would have made the difference of 550 feet in range performance in the test results.

SUMMARY OF THE OVERTAKING TESTS
It was during the Overtaking test that it was realized that the interstate tests results were strongly influenced (biased) by the surrounding traffic conditions. The effects of SWS signal blockage by large tractor trailer trucks, pick-up and passenger vehicles could cause the relative signal levels to drop and then rise rapidly, an effect that masked the small scale effects that were expected by moving the receiver vehicle to a different lane after each trial. It was also determined that the receiver vehicle could not safely drive at 55 miles per hour in the left hand lane where traffic was heavy and the legal speed limit was 70 miles per hour. These two factors made it impossible to determine if there were observable effects caused by the receiver or transmitter vehicle changing lanes after each Overtaking trial.

The ranges at which the warning signal was first displayed during each of the trials is shown in Table H-l. Referring to Table H-l, while the detection ranges vary due to traffic effects and topography along the interstate, the safety warning display range averaged over all 4 trials (also using both results obtained during trial 4) is 2,103 feet. This warning distance is realistic assuming that topography does not totally block the emergency vehicle and the motorist who is being overtaken.

Table H-I. Maximum Display Range in Overtaking Trials

Trial Identification

Elapsed Time

Differential Velocity

Overtaking-1

90 seconds

22 Feet per second

Overtaking-2

137 seconds

15 Feet per second

Overtaking-3

26 seconds

103 Feet per second

Overtaking-4 Overtaking-4

100 seconds 75 seconds2

22 Feet per second 22 Feet per second

1. Assumes that first display of warning message was proper criteria. 2. Assumes that continuous display of warning message was proper criteria.

Display Range 1,980 Feet 1,887 Feet 1,848 Feet 2,200 Feet 1,650 Feet

TRAILING TEST
The trailing test (not included in the test plan) was designed the day of testing to determine the range over which the SWS message could be received when the receiver vehicle was trailing the transmitter vehicle. Again, it was found that test results were highly sensitive to the amount of signal blockage from surrounding traffic and the effects of topographical features (trees along a curve) blocking the signal. The trailing test was comprised of only one trial.

H-13

A time mark was noted when the transmitter vehicle passed the receiver vehicle at the end of the Overtaking-4 test. The transmitter vehicle driver was instructed to continue to maintain a speed of 70 miles per hour while the receiver vehicle continued to maintain a speed of 55 miles per hour. The SWS message was displayed for 92 seconds after the transmitter vehicle passed the receiver vehicle. At a speed differential of 22 feet per second (15 miles per hour), the range between the test vehicles was 2,024 feet when the warning message was lost from the display. It was noted on the video tape record that when the warning message ceased to be displayed, a large truck had come between the receiver and transmitter vehicle. The receiver was moved to another lane to avoid the blockage effects caused by the large truck, but by that time the transmitter vehicle had passed around a curve and there was no longer a line of sight between the vehicles. The warning message was not acquired again.
ANALYSIS OF TRAILING TEST RESULTS
It is suspected that the range at which the warning message could have been received would have been greater had not the line of sight been first blocked by a tractor trailer truck and later blocked by a tree line in a curve of the interstate. This prediction is made because the radar detector antenna and the mobile transmitter antenna were pointing at each other, a factor which should have produced a message display at a distance approaching a mile if line of sight could have been maintained between the test vehicles.
FIXED SITE TRANSMITIER OVERHEAD TESTING
The Overhead test procedures outlined in Appendix B were developed to address two issues: (1) the operational range of the fixed site transmitter when it is elevated over the roadway, and (2) to determine if there is an effect on SWS performance as a function of which lane of a three lane interstate highway the fixed site transmitter is mounted over.
The test plan suggested that the fixed site transmitter would first be located over the right southbound lane and the receiver vehicle would conduct three total trials. During the first trial, the receiver would be driven southbound in the right lane. During trial two, the receiver vehicle would drive in the center lane. During trial three, the receiver would drive in the left lane. At the end of trial three, the SWS fixed site transmitter would be moved to a position over the center lane. A trial in each lane would again be conducted after which the fixed site transmitter would be moved over the left lane.
When the test plan was developed it was reasoned that repeatable results could be achieved during each test and that the anticipated small scale effects due to the transmitter placement over each lane could be
H-14

seen in the data. These assumptions were reasonable when the test plan were written before actual experience with the SWS had been gained through field testing in the interstate environment.
After the overtaking tests were completed and prior to the overhead tests being conducted, it was realized that the results obtained during each overtaking trial were different, and that there was no repeatability in the results, when the randomly spaced surrounding traffic caused unpredictable effects on results from trial to trial. The non-repeatable random results witnessed during overtaking testing were thought to be caused by the reflections and blockage of the signal path by the presence of other vehicles between and around the test vehicles. These traffic induced SWS signal propagation effects were observed to be so large that any measurable small scale effects caused by moving the fixed site transmitter one lane width would be hard to determine. Given this situation the author modified the overhead test procedures in the field to allow the magnitude of traffic effects on signal propagation to be explored while still measuring the relative distance over which the SWS would reliably display the safety warning message.
The modified test procedure to explore traffic effects on SWS propagation required that the fixed site transmitter location be optimized and that it not be moved between trials. Movement between lanes would have introduced yet another variable in the data. Also, it was determined that the receiver vehicle would stay in a selected lane during each of three trials so that traffic effects on signal propagation would be the only effect being measured between trials.
QVERHEAD-1 TEST
The fixed site transmitter was located over the center of the middle southbound lane of Interstate 75 on the north side of Exit 123 as located in Figure H-7. The elevation angle of the antenna was adjusted so that the beam center of the antenna was on the center of Interstate 75 at the horizon. The fiberglass door on the SWS transmitter was closed and latched, forming a radio frequency transparent radome over the antenna. The fixed site transmitter was programmed to transmit the message "Slow Moving Vehicle Ahead." The BEL 855 STi detector that had been modified at the factory to record signal strength data via RS-232 output was mounted 4 inches above the dash board of the test vehicle, and the laptop computer was used to record the data during the trial.
H-15

Figure H-7. Location offixed site transmitter during the overhead trials Figure H-8 shows the relative signal strength that was recorded as a function of signal sample index number along the 'X' axis. The trendline moving average of 15 points has been overlaid on the data. Referring to Figure H-8, a strong SWS signal was detected while the receiver vehicle was still out of the line of sight of the fixed site transmitter. It is thought that this first detection was due to reflections of the SWS signal off of traffic in the northbound lane because there was no line of sight between the transmitter and receiver vehicle at this point. At sample point 501 the relative signal strength began a steady upward increase. Multipath peaks and nulls can be seen in the relative signal strength data as the receiver vehicle moves closer to the fixed site transmitter. Normally, multipath effects will be present from the road surface, but the multipath peaks and nulls seen in the data plotted in Figure H-8 are thought to be caused by vehicles passing the receiver vehicle, as well as by reflections from several vehicles in front of the receiver vehicle. Several of the deep nulls were also thought to be caused by traffic blocking the line of sight between the receiver vehicle and the fixed site transmitter. The maximum distance over which the warning message was displayed during the Overhead-l trial is presented in a later section.
H-16

U,,)

14D l:YJ

1:'0

110

100

00

fY1

/0

-:::r:: I
-..l

t~J

y)

<10

Ji)

~O

: ,...._.~"..... Se-!te51
( - 1 S per. ~y1i~:. ~~'a ..~,~~.~~lf~'i;~ 1~ 1

Figure H-8. Relative signal strength data plotted as afunction of sample index number for Overhead-l trial

OVERHEAD-2 TEST The next test was conducted the same way the Overhead-1 test was conducted, but 100 less data points were collected, due to the fact that the receiver vehicle was driven at a speed of 60 miles per hour instead of 55 miles per hour in order to maintain a safe speed in the traffic. This difference in the number of data points collected constitutes a difference of 6 seconds in run time between the two tests. Comparing Figure H-9 to previous Figure H-8, the first peak from the fixed site transmitter reflecting off of the northbound traffic appears at approximately the same point in time. However, the peak does not last as long and the signal rises and falls around the peak average value, possibly indicating that the signal is reflected.
Referring to Figure H-9, the signal begins to peak starting at sample point 610 and continues to peak at a higher slope angle than previously in Figure H-8. Multipath peaks and nulls are present in Overhead-2 test data, however they are not the deep nulls seen in the data plotted in Figure H-8. This indicates that all else being equal, traffic must be responsible for the deep nulls in the relative signal strength. The receiver vehicle passes under the fixed site transmitter at approximately sample point 1,275 and the signal is not received again. The safety warning display distance achieved during Overhead-2 is presented in a later section.
OVERHEAD-3 TEST The Overhead-3 trial was conducted the same way the previous two overhead tests were conducted. Referring to Figure H-I 0, the plotted data from the Overhead-3 trial more closely match the data from the Overhead-2 trial than the Overhead-1 trial and the effects of traffic can be seen in the data. A trendline moving average of 15 points has also been laid on top of the data in Figure H-IO to provide an indication of a median value for the data. At data sample 60 I, the relative signal level begins to rise as in the previous two trials. Very rapid fluctuations occur in the signal level starting at sample point 650 and continue until sample point 725. This effect is thought to be traffic induced. Multipa:'1 signal lobing is apparent and the lobing pattern is very similar to that seen in the previously presented Overhead-2 trial data.
H-18

:_..._ - ser~~1 :--15per M<iV A;;\J. (~;t!1''''51)!
-::r:: I
\0
Figure H-9. Overhead-2 trial showing relative strength data plotted as afunction ofsignal index number

--Seri;'l;l -15per M.W.A'ig(S~'1Wt>I):
::r:: I
otv
Figure H-IO. Overhead-3 trial showing relative signal strength data plotted as a function ofsignal index number

MESSAGE DISPLAY RANGE
The range at which the relative signal strength exceeded 20 units (the message display threshold) was sample number 605 in Figure H-8, sample point 615 in Figure H-9, and sample point 625 in Figure H-lO. Previous laboratory tests showed that approximately 15 samples of data were produced each second of system operation. Using this factor as timing information, the range at which an SWS message would have been displayed is computed by first dividing the total number of sample points above the display threshold of 20 by 15 and next multiplying the result by the receiver vehicle velocity in units of feet per second. The data then have the 3 second correction factor subtracted from it. Table H-l shows the results of these calculations.
The data in Table H-2 show the variability in display range resulting from propagation effects induced by surrounding traffic. The results between Overhead-2 and Overhead-3 are very consistent, even though there were traffic effects observed in both data sets that would have prevented the small scale effects of moving the transmitter one lane length from being observed. It also appears that traffic effects actually increased the range at which the display could be activated during Overhead-I. The range at which the display can be activated by averaging the three trials is 4,572 feet. It should be noted that the maximum distance over which the warning signal can be displayed was not determined, because no section of interstate could be found in north Georgia that met all of the test site criteria and allowed line of sight to be maintained for one or more miles.

Table H-2. Detection Rangefor Each Overhead Trial

Test Scenario

Seconds Threshold of 20 is Exceeded*

Overhead-1

52

Overhead-2

45

Overhead-3

41

*Time after subtractIon of three seconds.

Velocity of Receiver Vehicle (FPS)
81 81 88

Message Display Range (Feet)
4,212 3.645 3,608

DIRECTIONAL SELECTIVITY TEST
The directional selectivity of the SWS operated from an overhead mounting position was measured during a test planned the day that the tests were conducted. The purpose of the directional selectivity test was to determine if the overhead fixed site transmitter could be detected by an SWS equipped motorist traveling in the northbound lane of the interstate and, if so, for how far. The fixed site transmitter was set

H-21

up as it was in the previous overhead test series so that the signal was being radiated north along and over the southbound lanes. The receiver vehicle traveled north along the northbound lanes.
DIRECTIONAL-1 TRIAL
The receiver vehicle approached Exit 123 from the south in the right northbound lane. The BEL 855 STi radar detector with the RS-232 output was used to record the data. Figure H-ll shows the relative signal strength recorded as a function of sample index number. Referring to Figure H-ll, a trendline moving average of 15 points has been imposed over the time varying signal data. The SWS relative signal exceeds the display threshold of 20 as the receiver vehicle passes under the overpass on which the fixed site transmitter was located at sample index number 90 and remains above the display threshold until going below the display threshold at sample index number 1105.
The receiver vehicle's speed was 57 miles per hour (84 feet per second). The time that the signal was above the display threshold of 20 was approximately 68 seconds. Because the vehicle was going away from the transmitter, the equivalent time that the warning message would have been displayed was calculated in a different manner than the approaching warning time was calculated. Three seconds were subtracted to compensate for the display delay and 10 seconds was added to compensate for the delay between the loss of the message and the delay in time when the message status returns to standby. After correction, the equivalent time the message would have been present was 75 seconds. Given these data, the calculated maximum range at which the safety warning message would have been displayed was 6,300 feet.
DIRECTIONAL-2 TRIAL
The Directional-2 trial was conducted In the same manner as the Directional-I trial. Referring to Figure H-12, a trendline comprised of a 15 point moving average was laid over the time varying relative signal strength data as was done in each of the previous plots. The relative signal strength exceeds the display threshold of 20 at sample index number 1SO and stays above the threshold until sample index number 950, resulting in the relative signal strength being above the display threshold of 20 for 53 seconds. After 7 seconds were added as the correction factor, this brought the total to 60 seconds. The receiver vehicle's speed was 55 miles per hour (81 feet per second). The calculated range at which the safety warning message would have been displayed was 4,860 feet past the overpass in the northbound lane..
H-22

ISU

14t)

130

120

110

10<)

-5

Q) 90

ii

w~
Iei

80

0>

<n 10

::r:

";..
:e
l

I
N W

11 60
tr

50

40

....
j-s"n"'~i1

.

i..

l==.~~~~_~~'c.v._~~lt.~.??!~~ 1.1.1

"..
0

,..
g

g,"

(;
'<t

& &5

;:) f'.

g

r
~

8

0

~.

Sample Index Number

~

0m

Figure H-II. Directional-l trial data plotted to show relative signal strength as afunction ofsample index number

140

130

~20

110

100

-
Ol

00

i

b

til 80 iri:i:

0\

in 'lO

C>

::,r::

~'"
"it 00

tv
+:-

l:

,-",.)"h~

40

30
1 1 1 _ _

-i r==s~~i~;i"HH'"

i - 1 5 p.>r Me>\'_ Avg tS~ri(l:f;1}:

1.

~.~ M U"'

,...,.,,""'

S~mple lndel( Number
Figure H-12. Directional-2 trial data plotted to show relative signal strength as function ofsample index number

ANALYSIS OF DIRECTIONAL TEST RESULTS
The range at which the warning message would have first been received during the Directional-l trial exceeds any display range obtained during any other trial of the SWS system. Yet, the receiver vehicle was moving away from the fixed site transmitter with the detector antenna pointed forward; a nonoptimum alignment. Also, there was a difference in the calculated message display ranges between the Trial-l and Trial-2 of 1,470 feet. Both of these factors were considered unusual and indicate that conditions were not the same between trials.
Referring back to Figures H-l1 and H-12, it can quickly be seen that there are no similar features in the data comprising the two plots, thereby supporting the theory that something unusual had occurred during one or both trials. A review of the video tape record showed that an 18 wheel tractor trailer truck was in front of and moving away from the receiver vehicle when the data shown in Figure H-ll were generated. Referring to Figure H-ll, the signal remains constant for some time before dropping toward the noise floor of the receiver. It is most probable that the received energy was reflected off of the rear of the truck during this trial. This explanation is plausible given that the back of the truck extends much higher than the receiver vehicle and would have a line of sight to the fixed site transmitter over a much longer distance. The video tape record also showed that there was line of sight back to the fixed site transmitter over a much further distance in the northbound lane than in the southbound lane where the overhead trials were conducted.
Figure H-12 also shows the results of traffic on the SWS signal. Although the SWS signal did exceed the display threshold and remained above the threshold, there are many rapid peaks and nulls, indicating that surrounding traffic may have been the source of multipath and possibly reflection points of the signal back to the recei ver.
It was concluded that using an overhead fixed site transmitter would ensure long range safety warning message reception to SWS equipped vehicles approaching the transmitter, the use of the overhead transmitter could also cause the message to be received by motorists traveling away from the fixed site transmitter when traffic conditions were moderate to heavy. This finding indicates that additional ways to operate the fixed site transmitter must be developed if signal localization to the target audience is a goal of operation in the interstate environment.
H-25