VFILELf) DOT/FAA/PM-87/14 Traffic Alert and(0) Collision Avoidance System4")

Program Engineering &Maintenance Service TCAS-111

()0 Washington, D.C. 20591

Allied-SignalBendix Communications Division1300 East Joppa RoadBaltimore, Maryland 21204

April 1987

Final Report

This document is available to the publicthrough the National Technical InformationService, Springfield, Virginia 22161.

U.S. Department of Transportation

Federal Aviation Administration D T ! -

CO~t,41ELECTEAll * D l* ' SEP"' 11987 f

This document has been approved for pubic raeease and sale;its distribution is unlimited.

Best Available Copy

NOTICE

This document is disseminated under the sponsorship of theU.S. Department of Transportation in the interest of informationexchange. The United States Government assumes no liability forthe contents or use thereof.

Technical Report Documentation Page

1. Report No. 2. Government Accessicn No. 3. Recipients Catalog No.

DOT/FAA/PM-87/14 Lf 8 *c64. Title and Subtitle 5. Report Date

TRAFFIC ALERT COLLISION AVOIDANCE SYSTEM APRIL 1987•L.•(TCAS 111) 6. Performing Organization Code

FINAL ENGINEERING REPORTSS. Performing Orgattizotion Report No.

7. Autho,'s) BCD-TR-157E.A. Hecht, J.E. Reed9. Performing Organizatiop Name and Address 10 Work Unit No. (TRAIS)

Allied-SignalBendix Communications Division 11. Contract or Gront No.

1300 E. Joppa Road DTFAOI-82-C-10019Baltimore, Maryland 21204 13. Type of Report and Period Ct.ered

12. Sponsoer j Aaency Name and Address

U.S. Department of TranLsportation Final ReportFederal Aviation AdministrationProgram Engineering and Maintenance Service 14. Sponsoring Agency Code

Washincton, D.C. 20591 APM-330

15. Supplementnry Notes

* ~ F16.AlhstractThe Task-l contract on TCAS-II produced a syszem concept and a DesignSpecification that show the 743 Aircraft L. A. Basin Model with peakdensities well in excess of 0.4 aircraft per square nautical mile canbe met. This Task-2 contract is for the design, construction, and testingof two Enhanced TCAS-II (now called TCAS III) Engineering units.

This report contains a description of the TCAS III engineering model andresults from various flight *tests of the system. System #1 has neeninstalled on the FAA's 727 since 1983. Testing shows that angularaccuracies in the order of 1 to 2 degrees are feasible and LhatHorizontal Resolution Advisories can be given on intruders. System #2has been delivered to the FAA Technical Center and installed on aConvair 580. Both systems will continue to fly as test beds for TCAS 1II.

ton.ý T10 be In 1)31r- ,

17. Ke.7 Words •8. Distribution StatementBeacon-based Collision Document is available to the U.S.Avoidance System, BCAS public through the Nacional Tech-TCAS-III Traffic Alert Collision nical Information ServiceAvoidance System Springfield, Virginia 22161.

19. Security Class,,. (of this report) 20. Security Classf. (,>( this page) 21. No. of Pages 12. Pr-ce

Unclassified Unclassified

Form DOT F 1700.7 (8-72) Reproduction of completed page authorized

i

....... A

I

$- uu

S 3

N'- -N3

__ �A �X * -

n* * ** �

S T#

� �

* . k -'-'

�

N- � �

w .�' '- �

-

>�

- S S . . )�\- -�

/� �fjS ,�. -

*

K-.N

�

* � S

-0

<\

� *i ,

- � .N:Kk;•:

�9t� *- -S * a m � - -.-- �--- - � --p.c.., -V � - w 0� * 0 5 -.

- . . . -- N. N, �N. .�\'.p...-.,.' .'� s'�**-�.- � -� ,-% * .- **.-

I ON

08.'3-S..2

PREFACE

The three color photographs on the preceding pages showthe i.rnstallation of the TCAS III engineering model cockpit dis-play in the Federal Aviation Administration's test aircraft, aBoeing 727. The display is an Electronic Flight InstrumentSystem (EFIS) display, mounted above the throttle quadrant. Thefirst photograph was taken during a test flight into DullesInternational Airport. There is a single target on display; itsposition relative to the TCAS airczaft is marked by the yellowcircle. This target is an actual aircraft that generated a traf-fic alvisory from the TCAS III equipment. The photograph wastaken at 11:45:32 when the TCAS aircraft was flying at 1200 feetMSL at 190 degrees heading. The target was approximately 1.5nauti-al miles in slant range, -30 degrees relative bearing, and500 feet below the altitude of the TCAS aircraft.

The second photograph was taken at 11:47:03, just one andone-half minutes later than the first. The TCAS aircraft was onfinal approach to Dulles at the time. The target shown in theprevious photograph has now moved behind the TCAS aircraft and isno longer threatening (as shown by the open diamond symbol)b-.cause the traffic advisory aircraft, a small private plane,turned left so it would pass to the left and just below the TCAS!27. The TCAS aircraft continued its approach so the target at-:47:03 is 4 nautical miles in slant range and 180 degrees rela-t.ve bearing from the TCAS aircraft. The target's position isdisplayed with an open diamond with "-01" above the symbol. Thismeans that the target was flying at an altitude 100 feet belowthe TCAS aircraft. There ý%re four additional targets on the dis-play, all as open diamonds •hich indicate non-threateningaircraft. Two of the four d'1 not have numbers above the symbol,indicating that those targets had non-altitude encoding tran-sponders. Only the range and bearing of these targets are known.

The third photograph is a close-up of the cockpit displayduring a simulated test run. The TCAS aircraft was flying a' 500feet at a heading of 210 degrees which was set up by the simula-tor. In this simulation, there were two targets. The one to theleft of the TCAS aircraft is shown as an open diamonO. It is notthreatening because it is below the threshola used tu indicatethat it is on the ground and therefore is not considered an air-borne intruder at this point. It is flying 500 feet below theTCAS aircraft and ascending as evidenced by the "-05T" mark bythe target symbol. The second target is dlsplaya&' as a redsquare; the color and shape of the symbol indicating that th•target was determined to be a resolution advisory. This mean~sthat the target is threatening and that an advisory is posted onthe display for the pilot. In this case, the adviso..y is "TurnRight", shown as a green arrow pointing to the right in the upperleft corner of the display.

v

For all displayed targets, current relative position ismarked by the target symbol. For all targets generating anadvisory, 25-second prediction vectors are also displayed. Thehead of the arrow points to the target's estimated postion in 25seconds based on current position and rates. A more detaileddescription of the relative motion display can be found insection 2.7.

Accession For

MTIS GRA&IDTIC TABUnannouncedJustifioati on---- (

ByDistribution/

Availability Code4Avail and/or

Dist Special

vi

TABLE OF CONTENTS

PREFACE1.0 INTRODUCTION 12.0 TCAS III SYSTEM DESCRIPTION 2

2.1 TCAS III Features 22.2 Surveillance 32.3 Garble Reduction 62.4 Beam Sharpening 82.5 Monopulse Measurements and Receive Sidelobe 18

Suppression

2.6 Stabilized Coordinate System 102.7 TCAS III Relative Motion Display and Aural Alarm 122.8 Alarm Reduction 142.9 Threat Detection and Resolution 172.10 Engineering Test Unit Hardware 19

3.0 TCAS III CONTRACT MODIFICATIONS 233.1 Cockpit Display 233.2 Whisper/Shout Techniques 233.3 Automatic Antenna Performance Monitoring 253.4 Effective Antenna Beam Width 253.5 Altitude Decoding Errors 253.6 Additional Aural Advisory Capability 283.7 Target Splitting 293.8 Horizontal Threat Detection Program 3T3.9 Horizontal Threat Detection and Resolution Logic 30

4.0 FLIGHT TEST DATA 314.1 Angular Accuracy Tests 314.2 TCAS Encounter Flights, Feasibility of 33

Horizontal Maneuvers4.3 Los Angeles Basin Flights 37

4.3.1 Density Measurements 484.3.2 High Density Encounters 484.3.3 Targets of Opportunity 58

4.4 Performance Statistics 645.0 CONCLUSION 70

APPENDIX A

LIST OF FIGURES

Figure 2.1-1 TCAS III FEATURES 3Figure 2.2-1 SURVEILLANCE - DYNAMIC THREAT DRIVEN LOGIC 5Figure 2.3-1 TCAS III REDUCES SYNCHRONOUS GARBLE 7Figure 2.4-1 INTERROGATE MODE OF OPERATION 9Figure 2.5-1 REPLY MODE OF OPERATION 9Figure 2.6-1 STABILIZED COORDINATE SYSTEM 11Figure 2.7-1 TCAS III COCKPIT DISPLAY 12Figure 2.7-2 COCKPIT DISPLAY WITH RESOLUTION ADVISORY 15Figure 2.8-1 MISS DISTANCE FILTERING 16Figure 2.9-1 3-D RESOLUTION ADVISORIES 18Figure 2.10-1 ENGINEERING MODEL DIRECTIONAL ANTENNA 20Figure 2.10-2 TCAS III ANTENNA MOUNTED ON FAA'S BOEING 20

727

vii

LIST OF FIGURES (Continued)

Figure 2.10-3 BLOCK DIAGRAM OF TCAS III ENGINEERING MODEL 22Figure 2.10-4 TCAS III ENGINEERING MODEL INSTALLATION ON 22

FAA'S BOEING 727Figure 3.2-1 THE FOUR-LEVEL WHISPER/SHOUT USED BY ETEU#1 24

SHOWS EXCELLENT SEPARATION OF TARGETS INTHE L.A. BASIN

Figure 3.2-2 THE SHOUT LEVEL #4 STATISTICS ARE EXTENDED 26BY COMPUTING 'I E NUMBER OF REPLIES FROMTHE DIFFERENT ATTENUATION LEVELS

Figure 4.1-1 ANGULAR COVERAGE PERFORMANCE 32Figure 4.2-1 EVALUATION OF TCAS III HORIZONTAL MISS 34

DISTANCE ESTIMATEFigure 4.2-2 340-KNOT, 90-DEGREE ENCOUNTER 35Figure 4.2-3 810-KNOT, HEAD-ON ENCOUNTER 38Figure 4.2-4 415-KNOT, 15-DEGREE ENCOUNTER 40Figure 4.2-5 HORIZONTAL FAKEOUT, TARGET CROSSES BELOW 42

TCASFigure 4.2-6 HORIZONTAL FAKEOUT, TARGET CROSSES IN FRONT 44

OF TCASFig-ire 4.2-7 HORIZONTAL FAKEOUT, TARGET CROSSES BEHIND 46

TCASFigure 4.3.1-1 TYPICAL RECEIVED TARGET REPORTS, LOS ANGELES 49

FLICHT TESTFigure 4.3.2-1 AIRCRAFT ENCOUNTER GEOMETRIES 50Figure 4.3.2-2 TYPICAL ENCOUNTER DISPLAYS, LOS ANGELES 51

FLIGHT TESTSFigure 4.3.2-3 HEAD-ON ENCOUNTER 54Figure 4.3.2-4 45-DEGREE ENCOUNTER 56Figure 4.3.2-5 VERTICAL FAKEOUT 59Figure 4.3.2-6 HORIZONTAL FAKEOUT 61Figure 4.3.3-1 TARGET OF OPPORTUNITY, #1 65Figure 4.3.3-2 TARGET OF OPPORTUNITY, #2 67Figure 4.4--i TCAS III PERFORMANCE STATISTICS 69

LIST OF TABLES

Table 1.0-1 TCAS III CONTRACT MILESTONES 2Table 4.3.2-1 LOS ANGELES BASIN ENCOUNTER STATISTICS 63

viii

BCD-TR-157

GLOSSARY

OF

ABBREVIATIONS AND ACRONYMS

A/D Analog to DigitalALT AltitudeANG AngleARTS Automated Radar Terminal SystemATC Air Traffic ControlATCRBS Air Traffic Control Radar Beacon System

BCAS Beacon Collision Avoidance System

CAS Collision Avoidance SystemCL ClimbCPA Closest Point of Approach

DABS Discrete Address Beacon SystemdB DecibelDEG DegreeDES DescendDMTL Dynamic Minimum Threshold LevelDP Data Processor

EFIS Electronic Flight Instrument SystemEL ElevationETEU Engineering Test Unit

FAA Federal Aviation AdministrationFAATC FAA Technical CenterFPM Feet Per Minute

GC General Correlation

ID IdentificationIFPM In-Flight Perfoxmance Monitor

IP Interrogator Processor

KTS Knots

MHz MeqahertzMode C ATCRBS ModeMode S Modified DABSMSL Mean Sea LevelMTL Minimum Threshold Level

nmi Nautical MileNS Negative Slope

PS Positive Slope

ix

BCD-TR-157

GLOSSARY

OF

ABBREVIATIONS AND ACRONYMS (Continued)

QV Quantized Video

RA Resolution AdvisoryRF Radio FrequencyRSLS Receiver Side-Lobe Suppression

SBC Single Board ComputerSEC SecondSQV Synchronized Quantized Video

TA Traffic AdvisoryTCAS Traffic Alert and Collision Avoidance SystemTL Turn LeftTR Turn Right

W/S Whisper/Shout

3-D Three-dimensional

*1

TCAS III FINAL ENGINEERING REPORT

1.0 INTRODUCTION

The TCAS III (Traffic. Alert and Collision AvoidanceSystent) was developed by the Bendix Communications Division ofthe Allied-Signal Corporation under contract to the FederalAviation Administration, U.S. Department of Transportation. Acontract (DTFA01-81-C-10041) for concept development and computersimulation of the TCAS III was awarded in January 1981. This wasfollowed by a contract award (DTFA01-82-C-10019) for the develop-ment and test of two Engineering Models in December 1981. Thisreport covers the major features and test results of the TCAS IIIEngineering Models.

TCAS III assesses possible airborne collision threats byactively interrogatir.g transponder-equipped aircraft using anelectronically steered phased array antenna. The accurate bear-ing measurement of other aircraft allows for horizontal maneuvercapability and an accurate relative motion target display inaddition to reducing the alarm rate and reducing interferencesusceptibility. These are the principal features thatdistinguish the TCAS III from the TCAS II.

The original design of the TCAS III was developed andvalidated by computer simulation. The computer program logic anda model of the hardware were tested in a Simulation Test Bed thatsimulated flight through the 1982 Los Angeles Basin traffic envi-ronment model. The model has 743 aircraft within an area 120miles on a side. The results of the Factory Performance Testingshowed targets being detected and tracked in high density trafficareas. Approximately 100 targets were successfully trackedwithin 20 nautical miles of Own aircraft. This level of surveil-lance activity was maintained through a minute of simulatedflight time. During specific tests, conflict situations weresuccessfully detected and reported. Accurate bearing was shownto be most effective in the correlation and association of data.Even in the 743 aircraft case where the peak density of targetsis 0.47 aircraft per square nautical mile in the annulus betweensix and seven miles, tracks were initiated and maintained.

Once the design concept was validated, the fabrication ofthe two Engineering Models began. The first unit was deliveredto the FAA Technical Center in March 1983. It was installed onthe FAA's Boeing 727 in May 1983. This first TCAS III Engineer-ing Model was put through a variety of tests - static testing onthe ground, flight testing over various terminal areas, orbits,and encounters. In addition, this unit was tested twice in theLos Angeles Basin area for operation in a high density environ-ment. Details on the flight tests can be found in section 4. InSeptember 1984, a modification to the contract was added to focus

eftorts in several problem areas such as altitude decoding and

target splitting. Also, upgrades to the cockpit display andaural advisory capability were identified as priority items. The

second TCAS III Engineering Model was delivered to the FAA Tech-nical Center in May 1986. This unit was installed on the FAA'sConvair 580 and flown on several test flights. Flight testing ofboth units will be continuing. Important contract milestones aresummarized in Table 1.0-1.

TABLE 1.0-I. TCAS III CONTRACT MILESTONES

Dec 1981 Contract begins to design, fabricate, arý test twoEngineering Models.

Dec 1982 Factory Acceptance of first unit.

Mar 1983 Delivery of first unit to FAA Technical '.enter.

May 1983 Installation of first unit in FAA's Boeing "127.Testing begins.

Dec 1983 Flight testing of first unit with TCAS II logic(vertical maneuvers only) in Los Angeles.

Jan 1984 TCAS TII logic installed (vertical and horizontalmaneuvers). First flight test with horizontaladvisory displayed.

Mar 1984 Flight tests in Atlanta area for high densityoperation.

Sep 1984 Modification number 0014 to contract.

Feb 1985 Flight tests in Los Angeles Basin.

Oct 1985 Upgrade of cockpit display and aural advisory

capability. Demonstrations in Ottawa, Canada.

Apr 1986 Factory Acceptance of second unit.

May 1986 Delivery of second unit to FAA Technical Center.Installation on FAA's Convair 580.

2.0 TCAS III SYSTEM DESCRIPTION

This section gives a brief overview of the TCAS III systemand emphasizes the differences of this system versus the TCAS IIsystem.

2.1 TCAS III FEATURES

As shown in Figure 2.1-1, the TCAS III accurately deter-mines the relative position of all ATCRBS and Mode S equippedaircraft operating within its surveillance limits by measuringthe relative bearing and range of each aircraft and extractingits encoded altitude from the transponder reply. From thesedata, the track of each aircraft is generated and its futurecourse projected. This three-dimensional track allows estimatesof horizontal miss distance to be made which are used in theevaluation of alternative resolution maneuvers. Through the useof relatively narrow antenna patterns1 synthetically narrowed

2

OMNI COVERAGE

SEL•RNICALLy S-rEF

ATCRBS

2 NARROW SECTORIZED BEAMS WITH ACCURATE BEARING DATA

8 THREAT DRIVEN LOGIC-EFFICIENTLY USES RESOURCES

X MISS DISTANCE FILTERING-REDUCES ALARMS

U POSITION WITH RELATIVE MOTION-EFFECTIVE TRAFFIC ADVISORIES

i . U HORIZONTAL RESOLUTION-ADDITIONAL MANEUVER OPTIONS

FIGURE 2.1-1. TCAS III FEATURES

sectorized iiiterrogations, and monopulse angle measurements, theTCAS III is able to accomplish its surveillance functions in thehighest traffic densities.

The threat driven logic adaptively controls the system toconcentrate its operations on collision threats and potentialcollision threats. It dynamically controls its surveillanceparameters (angle, range, power, and transmission rate) toachieve very high surveillance efficiency (in terms of the pro-ceasing activities used in obtaining the information it needs)and low RF interference to other ATC functions sharing the beacoiispectrum.

The accurate bearing measurements of other aircraft iscritical for the use of a horizontal maneuver. The results ofi•azimThseactrhase beerng met.Thssurvmeillanoter caabiircaty

flight tests show that the design objectives of one to twodegrees, one-sigma bearing accuracy in the forward 180 degree•...__:azimuth sector has been met. This surveillance capability

enables a horizontal maneuver to be safely executed if needed.The track data may also be used to display the relative positionand motion of traffic to pilots.

2.2 SURVEILLANCE

TCAS III is designed to operate in traffic densities ashigh as 0.4 aircraft per square nautical mile (or approximately

3

.1 AI A I . TA r t V ) IL -r k I)z2ý j d T

32 aircraft within 5 nautical miles of the TCAS III aircraft)

with targets having closing speeds of 1200 knots. The design of

the system includes not only the necessary hardware but also a

surveillance control strategy based upon the use of a dynamic,

adaptive, real-time, threat-driven logic, which operates within

the same environmental limitations established for the TCAS II.

ATCRBS equipped targets are interrogated using the Mode S

format identified as Mode C only All Call (to which Mode Ctransponders will reply but to which Mode S transponders will not

reply). Mode S equipped aircraft are acquired and tracked using

their Mode S squitter and replies to discrete interrogations. Inboth cases, a narrow antenna beam is used which can be instantan-

eously steered to any bearing angle by the system. For Mode C

equipped aircraft, all such targets receiving the interrogationwill reply to the Mode C interrogation. This results in someprobability that the replies from different transponders willoverlap in time and produce what is known as synchronous garbleat the TCAS receiver, which makes it difficult to identify and

track individual aircraft. To minimize the synchronous g:rbleproblem, several provisions are made in the TCAS III.

The most significant of these added provisions is that thenormal angular width of the zone in which transponders will reply

is "synthetically sharpened" from the free space antenna beam-width (which is 64 degrees) to 22.5 degrees. Also the angularposition of the narrowed sector can be stepped (from one inter-rogation to another) in increments of 5.625 degrees, a total of64 equally spaced positions.

Each of these 64 angular positions is visited often enoughto ensure a 0.98 probability of detection for any aircraft(traveling at the maximum closing velocity one can anticipate atthat bearing) before it can pass through a 10,000 foot detectionzone and enter the outer boundary of the TCAS III protected zone,shown in Figure 2.2-1. The outer boundary of the protected zoneensures that the system then has at least 45 seconds remainingbefore the point of closest approach (assuming a worst case situ-ation). The 45 seconds provides sufficient time to assess thethreat, coordinate with the other Lerget if it is also TCASequipped, select the proper advicory, display the advisory to chepilot and provide him sufficienc time to react and then time forthe aircraft t3 accelerate and depart from its original flighttrajectory and provide a safe sepdration distance. The worstcase, defined for each bearing angle in terms of the maximumpossible relative closing velocity that can exist for two air-craft traveling near Mach 1, results in a protected zone havingboandaries as shown in Figure 2.2-1. For the head-on conflict(-1200 knots relative closing velocity), the range at which atarget must be detected is significantly greater than it is forthe tail chase situation (with a maximum speed differential of-350 knots).

4

22-1/2 DEGREE SHARPENED BEAMSFB-VG1762AATCRB$ SURVEILLANCE

FB-VG-1762A 20 NMI POWER TAPERED & RANGE GATED

15 NMI

/ 45 SECOND PROTECTED

BONACKROUDMD

I W----..__10,00D0 FT. ATCRBSS' DETECTION ZONE

OWNTCAS

BACKGROUND MODE

SQUITTER DETECTION64 DEGREE BEAM

FIGURE 2.2-1. SURVEILLANCE - DYNAMIC THREAT DRIVEN LOGIC

Q - 5

2.3 GARBLE REDUCTION

In a high density traffic environment, the over.wirgreplies of closely spaced ATCRBS transponders to each interroga-tion make it difficult for the receiver to detect individualtargets. In a density of 0.4 aircraft per square nautical mile,some 32 aircraft lie within 5 nautical miles of range of theTCAS equipped aircraft. Because the duration of each reply is21 microseconds (or 3.4 nautical miles in range), as many as 26replies can synchronously overlap at the TCAS receiv3r if anomnidirectional antenna were used,

As shown in Figure 2.3-1, the most effective mechanism forreducing synchronous garble is to prevent generation of overlap-ping replies by dividing the space surrounding the interrogatorinto angular regions using directional antenna radiationpatterns. Such patterns are used to limit the angular sectorwithin which transponders are interrogatea Using the 22.5degree synthetically sharpened pattern for the transmission ofinterrogations, an effective improvement factor of about 12 to 1is obtained. (Variations in individual transponder reply thresh-olds spread the effective beamwidth about 30 percent, therebyreducing the improvement from its theoretical value of 16.) Thistechnique reduces the 26 overlaps that an omnidirectional receiv-ing system would have in a worst case environment to about 2overlaps, assuming uniformly distributed traffic.

To further reduce the incidence of overlapping replies, asimplified (fouL-level) form of the whisper-shout technique usedin the TCAS II is applied to each 22.5 degree sector that theTCAS III interrogates. Since individual transponders differsomewhat in sensitivity, a low power interrogation transmnissionwill elicit replies from some transponders but not from others.When such a transmission is tollowed by a S0/Pl suppression pairnear the same power level, those transponders replying to thefirst interrogation will be prevented from replying to succes-sively higher power interrogations. If immediately thereafter ahigher power interrogation is transmitted, another group oftransponders will reply. In the example shown in Figure 2.3-1,an SO pulse preceding the P1 pulse by 2 microseconds at 3 dBlower power than the preceding interrogation provides suppressionof those transponders which replied to that previous interroga-tion. This process is repeated four times in each 22.5 degreeangular sector in the TCAS III using interrogation power levelsseparated by 18 dB, 14 dB, 10 dB, and 0 dB, respectively. Thisproduces a net improvement of about 2.5 to 1 in reducing syn-chronous garble compared to a single high power interrogation.

The total improvement due to the combined use of bothsectorized interrogation and four-level whisper-shout is about 30to 1 in reducing synchronous garble at the receiver terminals. Areply processor is then used, which reliably decodes as many asthree overlapping signals, thereby producing a high probabilitythat the TCAS III will resolve individual target replies in hightraffic density environments.

6

DIRECTIONAL INTERROGATIONSUPPRESSED

TARCETS REDUCED BY ANGLE-EFFECTIVE BEAMWIDTH (EBW)IMPROVEivEh;T .. 30°'"SRELATVE TO OM. 30-)'•W- 12/1 FOR 22-1/2 DEGREE BEAM

REEF[LIE S RATVTOMNI E8By

WHISPER SHOUT Pi P3 s i P3

U 1 U3P1 0

± TIME-"

TARGETS RESOLVED BY POWER LEVELS

IMPROVEMENT t 2.5 FOR 4 LEVELS

DIRECTIONAL INTERROGATION a WHISPER SHOUT

TARGET REPLYING TO INTERROGATION IS ISOLATEDTO A SMALL CELL

3602,TOTAL = Esw A !W/S ; 30/1

U.

FIGURE 2.3-1. TCAS III REDUCES SYNCHRON,2c.eS GARBLE

7

2.4 BEAM SHARPENING

The technique used to sharpen the 64 degree 3-dB beamwidthto 22.5 degrees is illustrated in Figure 2.4-1. All ATCRBStransponders include Side Lobe Suppression circuitry. Thistransponder feature allows comparison of the amplitude of the P1pulse (which is normally transmitted on a omni-directional anten--na) so that it can determine when it is located within the mainbeam of the interrogator. If it detects the amplitude of the P1pulse at a substantially higher amplitude than that of the P2pulse, the transponder will reply provided it receives a P3 pulseat the proper time. If on the other hand P2 is greater than P1,the transponder is caused to "suppress". In this suppressionstate, the transponder will not respond to any interrogation fora period of about 35 microseconds.

Because TCAS cannot employ a large aperture antenna havingthe directivity of a ground based interrogator, the techniqueemployed by TCAS III to limit reply zone width uses two speciallytailored antenna patterns having the shapes shown in Figure2.4-1. The P1 and P3 (interrogation) pulses are transmitted onthe sum beam, while the P2 (suppression) pulse is transmitted onthe difference beam.

A transponder located near the center line (or boresightaxis) of the two beams will receive the P2 pulse at considerablylower amplitude than the P1 pulse because it is in the null ofthe difference pattern. Therefore it will zeply. As the trans-ponder moves in bearing away from the sum beam center, the P2pulse amplitude will increase wh:le that of the P1 decreasesuntil a point is reached where the P2 pulse power exceeds the P1and the transponder is then suppressed.

Using the antenna patterns shown in Figur:e 2.4-1, togetherwith an adjustment of the total power in each beam (the sum beampower iE reduced about 5 dB from that in the difference beam), a22.5 degree effective beamwidth is achieved. This width willvary to a small degree with individual transponders due tovariations in individual design.

2.5 MONOPULSE "V\SUREMENTS AND RECEIVE SIDELOBE SUPPRESSION

The same F a and difference patterns used for transmissionare alse used on the reception to estimate the bearing angle off-set of each target from the antenna boresight axis. This tech-nique is known as monopulse since an angle estimate can be madeon a single -pIse. It contrasts with existing angle measurementtechniques emfloyed by ground ATCRBS interrogators, which requirea series of intercgations and replies as the beam traverses thetarget (hence many pulses) that is ther, processed to estimate thecenter point. Monc.ooulse measurements exhibiting 1 to 2 degreestandard deviation trrors are obtainable with TCAS III. Figure2.5-1 illustrates .Ž ;.-,fopulse measurement technique. A mono-pulse curve, ;imilaz to that shown at the lower part of the

8

FB-VG-1544 AIRCRAFT

WITH ATCRBS

TRANSPONDER

AAJ-.\" < , 22.50 EFFECTIVE

BEAMWIDTH

AIRCRAFT WITHP 2 ON A ATCRBS TRANSPONDER

SUPPRESSED

PI - P 3 ON 2E

P2

P2 > PI

FIGURE 2.4-1. INTERROGATE MODE OF OPERATION

FB-VG-1545f\ /

; •, / • 20.3 MICROSECONDS .... i

S\ I

A PATTERN F1 F2 REPLY DATA

AT 1090MHz

Z: P".TTERN "it,,!•/..•.

AIRCRAFT WITHATCRBS TRANSPONDER

MONOPULSE VIDEO SUPPRESSED

MLOG (NLS,-)- LOG (Z- 0) §GIVES ANGLE OFF IBORESIGHT - , +0•

FIGURE 2.5-1. REPLY MODE OF OPERATION

A f% VKvX M U k I 'J A -J r1^ PJUAM .r A i -fA . %XJ'AJ A i , - - 7 .

figure, is digitally stored in the data processor for each beamposition so that any variations in voltage ratio due to physicaldifferences in antennas are corrected. Each monopulse curve isgenerated from sum and difference patterns taken at each beamposition at 10 degrees elevation. Ten degrees was chosen as anaverage value between 0 degrees and 20 degrees.

Bearing measurements are made on both A"'•RBS and Mode Sreplies. Measurements are made on each ATCRBS ly using up tofour different pulses and the results averaged. II only onepulse is received in the clear (not overlapped by another pulse)whenever a bracket decode is declared, the reply monopulseestimate is made on that one pulse. Sixteen chips from each ModeS reply are averaged to produce the monopulse estimate.

Receive sidelobe suppression (RSLS) is also implemented in

TCAS III to eliminate replies (called fruit) appearing fromangles outside the beam sector of interest. The circuitry com-pares the ratio of sum to difference pattern voltages for eachpulse. In the sector of interest where the sum pattern exceedsthe difference pattern, the ratio has a large positive value.When the bracket pulse voltages exceed the RSLS threshold, abracket decode is enabled and the reply decoded. Replies fromother angles are rejected because the voltage ratio is low. Thisreduces the amount of fruit that must be processed by a factor of12.

2.6 STABILIZED COORDINATE SYSTEM

To maintain bearing angle estimates accurate to one degree

on intruding aircraft, it is necessary to keep the track file ina north referenced stabilized system. Flight records show thatwhile flying a nominally straight flight path, the test aircraftvaries +/- several degrees heading in 20 seconds. Then, tocompute an accurate bearing rate in the order of 0.1 to 0.2degrees per second as needed for a miss distance calculation, Ownaircraft's stable reference must be read four times per second.

Measurements made with the directional antenna must betransformed from Own aircraft's coordinate reference frame to theearth stabilized (north reference, lccally level) coordinateframe to correct the angle for pitch, roll, and heading. Whentfacking a Latyt:L, the interrogation angle is computed by trans-forming back from the stabilized coordinate to Own aircraft'scoordinates. Although these ccordinate transformations requireconsiderable computational power, accurate bearing angle esti-mates cannot be maintained without stabilizing the track filereference fraRe.

A Because a stabilized reference frame is required to retainangle accuracies and aircraft movements are reasonably linear inthe local horizontal plane, TCAS III track equations takeadvantage of the orthogonal stabilized X-Y coordinate frame asshown in Figure 2.6-1. The stabilized coordinate frame is

10

EARH -TA~LIEASORDNTET .,

FIGUREE2.6H1.SSTABILIZEDCCOORDINATE SYSTE

+11

centered in the TCAS III aircraft with the +X axis pointingnorth, the +Y axis pointing east, and the +Z axis down. Ownaircraft's coordinate system has own aircraft's nose representedby the +x axis, +y axis passing through the right wing, and +zthrough the floor. in Figure 2.6-1, the remaining parameters are*(own aircraft's heading), P (own aircraft's roll), 0 (ownaircraft's pitch), f (target's bearing in own aircraft's coordi-nate system), B (target bearing in the earth-stabilized system),and E (target elevation angle in the earth-stabilized system).Measurements are made in range and bearing in the aircraft coor-dinate because the directional antenna is fixed to the aircraftskin. Vertical tracking is accomplished independently in the Zcoordinate using the non-linear tracker developed for the TCAS IIby the MITRE Corporation.

2.7 TCAS III RELATIVE MOTION DISPLAY AND AURAL ALARM

The TCAS III relative motion display on the firstengineering model was upgraded in October 1985. This cockpitdisplay consists of a modified Electronic Flight InstrumentsSystem (EFIS) display and symbol generator. Figure 2.7-1 is anillustration of the cockpit display. It has the followingfeatures:

The location of Own aircraft is designated as an airplanein the center of the display. Own heading is toward thetop of the display so all angle positions relative to Ownaircraft can be easily discerned.

V FLO052 1 1 ±2000

S21 i 2ONm

.'..

+114,•

5

S08:33:11

FIGURE 2.7-1. TCAS III COCKPIT DISPLAY

12

The display range is selectable by the pilot to be 3, 5,10, 15, 18, or 20 nautical miles. Vie selected range is

* given in the upper right corner of the display.

The surveillance range gate is identified by an oval rowof dots ranging 16.6 miles directly ahead to 8 miles inthe rear.

This boundary represents the maximum range from which anintruder flying 600 knots could reach Own aircraft in 15seconds plus a 10,000 foot detection zone when Own air-craft is flying at any speed from 600 knots down to 250knots.

The o~clock positions are always displayed at 4 nauticalmiles except when the 3 nautical mile range is selected.Four nautical miles is chosen to give the pilot a refer-ence at the range where he should be able to visuallylocate intruders.

An indication of Own heading is given by the compass arcat the top of the display. The arrow points to Own air-

- craft's heading. When Own aircraft turns, the compasswill turn accordingly. Own aircraft's flight level isdisplayed in the upper right corner of the display. InFigure 2.7-1, the flight level is 005 or 500 feet and itsheading is approximately 210 degrees from magnetic North.Also displayed in the upper right corner is the altitudeband within which targets are displayed. In this example,it is +2000 feet of Own's altitude. This parameter isadjustable by the pilot.

The lower left corner of the display contains the time.This is used for correlation with recorded flight data.

The location of each target symbol indicates the measuredrange and bearing of the target relative to Own aircraft.The shape and color of each symbol indicates its threatlevel. The number above the shape gives the target'saltitude relative to Own aircraft in hundreds of feetwhere positive is above Own aircraft. An arrow beside therelative altitude indicates the direction of change ofaltitude (up indicates increasing). A target with norelative altitude displayed indicates that the target hasa non-altitude reporting transponder. A vector is

* attached to each target that is a Traffic Advisory orResolution Advisory. The head of the vector indicates theestimated position of the target in 25 seconds based oncurrent range and bearing rates.

(a) An open white diamond indicates a non-threateningtarget.

13

(b) A solid white diamond indicates a proximity warning.A target will generate a proximity warning when itcomes within four nautical miles in slant range and1200 feet in relative altitude of Own aircraft.

(c) A solid yellow circle indicates a tric advisorytarget. When a target first becomes a trafficadvisory, an aural alert is given - the word"Traffic" is repeated two times. This indicates thatthe target is projected to pass closest point ofapproach at less than one nautical mile and within1000 feet relative altitude within 30 seconds.

(d) A solid red square indicates a target that has becomea resolution advisory. The recommended manuever isdisplayed to the pilot.

The Resolution Advisory is shown in the upper left cornarof the display as seen in Figure 2.7-2. This area of thedisplay pictorially shows the resolution advisories. Whenthe resolution advisory is first displayed, an aural alertis given - the advisory is repeated three times. If atany time the advisory is changed, the new advisory isrepeated three times as well, alerting the pilot of thechange.

Figure 2.7-2 shows two targets. One is a resolution advi-sory that is at 4 nautical miles at 30 degrees relative bearingand 500 feet above Own aircraft and descending. The recommendedresolution advisory against this threat is "Turn Right" as shownby the arrow in the upper left corner. The other target is non-threatening, at 4 nautical miles, -30 degrees relative bearing,500 feet in altitude below Own aircraft, and ascending.

2.8 ALARM REDUCTION

TCAS III has several levels of filtering or target rejec-tion prior to initiating tracks on potential threats. Once atarget is in tralck, it is subjected to threat detection logictests to assess its potential for collision or near miss.

An important objective of the TCAS III system is to reducethe number of unnecessary alartts while ensuring a high probabil-ity of declaring a threat when it exists. To perform the complexdata processing required on each threatening target, the numberof targets being processed must be minimized.

The most general filter is the limitation of the threat

volume by range and interrogation power so replies from aircraftbeyond regions of interest are rejected. Next, the interrogationsector of each transmission is limited to an effective 22.5degree beamwidth. Then, the receive sidelobe suppression (RSLS)circuitry rejects fruit and replies beyond +13 degrees for ATCRBSand +32 degrees for Mode S. After a reply is decoded, it may be• discarded because it differs in altitude by more than the

14

A

* FLOO5' 2;' /±2000121

,10NM

co 08-351-29V)

% 'A

FIGURE 2.7-2. COCKPIT DISPLAY WITH RESOLUTION ADVISORY

threshold (7000 ft). These general or bulk filters in effectceduce the data processing load.

Trackad targets are prevented from causing unnecessaryalarms by testing range rate a.ad accurate three-dimensional missdistance against minimum thresholds. Since the determination ofa target as non-threatening is the complement of determining itis a threat, this alarm reduction is the negative result of theactions taken to declare a threat.

The TAU criteria, range divided by range rate as shown jin

Figure 2.8-1, is the best pdirameter for assessing when a targetis a potential threat. It reliably declares threats on everythreatening target. However, it indicates many targets arethreats when they will pass at a safe distance. Each threat thenis tested further for altitude and horizontal separation toeliminate unnecessary alarms.

15

~ -

FB-VG-1768A

TAU - P (DELAY AND MANEUVER TIME)

r- INTRUDER

TCAS 2031) 20 10 A 10

CPA

ALTITUDE - VERT:CAL MISS > THRESHOLD [eA+200] feet ALTITUDE ERROR e A

-22'"tI NTRLUDER

T...AS l 1 0 20 30 4 0 _TCAS•

4 G 30 20 10

CPA

BEARI!G !ATE - NORIZONTAL (MISS) > THRESHOLD 3o ERRORS..IN POSITION

ppt > (Cd+lO000 PREDICTION

(a) ALARM , ({ =Ed

TCAS

HORIZONTAL MISS DISTANCE d(b) NO ALARM

TCAS

FIGURE 2.8-1. MISS DISTANCE FILTERING

16

• .'v,- .• - .- - - "1 =0Y-•"\" • :' 2' "-

The vertical miss distance is tested next to see whethe:there is adequate vertical separation at CPA time. where time isdetermined by the TAU criteria. Graphically, this test is shownin Figure 2.8-1. The vertical separation must be greater thanthe expected altitude error plus a minimum of 200 feet.

The final test of a potential threat is on horizontal missdistance. The end result of the horizontal miss distance test isto see whether there is safe passage if both aircraft continue onstraight courses. Safe passage is miss distance being greaterthan the three sigma error in computing miss distance plus 1000feet. As shown in Figure 2.8-1, in (a) an alarm would be given,but in (b) the accuracy of estimating bearing rate is improvedand the distance reduced. Even though the time to CPA isreduced, no alarm is needed.

2.9 THREAT DETECTION AND RESOLUTION

The TCAS III Threat Detection and Resolution logic hasbeen developed by the MITRE Corporation. It takes advantage ofpast Collision Avoidance work and adds new concepts to takeadvantage of the bearing measurement.

As stated in the discussion of miss distance filtering,the preliminary detection of threats is based upon the predictedtime to closest approach, derived by dividing measured range bythe range rate estimate. This time test is augmented with analtituC] test and a horizontal plane projection to avoid alarmsfor safe passage. A new concept used in the TCAS III logicreplaces the fixed time thresholds with a variable alarm time.An alarm is deferred as long as possible, since most apparentconflicts are safely resolved either by Air Traffic Control orpilot visual procedures. Also, the effects of measurement errorsfrom bearing and altitude rate decrease at smaller ranges. Analert may provide from 10 to 35 seconds of warning time, and isI. finally given when further delay would cause a standard escapemaneuver to provide insufficient separation at the point ofclosest approach. Figure 2.9-1 snows this concept for verticaland horizontal maneuvers for their potential separation.

Figure 2.9-1 also illustrates the process of evaluatingseparation by modeling potential maneuvers. Using one-secondtime intervals, the intruder is projected ahead on an unaccele-rated path using the current three-dimensional velocity estimate.Also using the one-second time intervals, Own aircraft's track isprojected ahead on a number of different flight paths. Each path"simulates a standard response in compliance with a different oneof the available advisories. For each of these paths, the pointof closest approach may be reacned at a different time. Whenthis point is found, the three-dimensional separation and itsvertical and horizontal components are stored in a matrix. Thisis depicted by the middle section of Figure 2.9-1.

17

FS.VG.1769

MODELING THREATS

ALARM CAN BE DELAYED ALARM IS GIVEN

OW -CA ThRESHLD -trTHRESHOLD

INTRUDER'

I --3 ERROR i-IN PREDICTION I30 ERROR

IN PREDICTIONMANEUVER THRESHOLD CPA EQUALS

CPA MAEEUVERTHRESHOLD.

PREDICTED MISS DISTANCE MATRIX

CLIMB

LEVELOFF

MAINTAINR~ATE

DESCEND VERTICAL COMPONENT OF 3-D MISS DISTANCE

HORIZONTAL COMPONENT OF 3-D MISS DISTANCE

TURN I3- MiSS DISTANCE (VERTICAL WEIGHTED 5:1)LEFT

CONTINUESTRAIGHT

TURNRIGHT

RESOLUTION ADVISORY SELECTIONVALIDITY ADVISORY EVALUATION

TESTS TESTS

e_ TIE BREAKERCL 0 O00-G TEST

DESTL •SELECTED

ADVISORY

POTENTIAL TRRESOLUTION CLTL 0 - 0 0 -ADVISORIES CLTR 0- 0 0

DESTL 0

DESm 0----- i

FIGURE 2.9-1. 3-D RESOLUTION ADVISORIES

18

Choice of the individual Resolution Advisory is made byconsidering a list of test criteria. Evaluation of these cri-teria is the means of reducing the list of potential advisoriesuntil one choice remains. This methoi is used because often morethan one advisory would provide satisfactory separation, andother considerations take priority over simply maximizing separa-tion. Some examples of the principles motivating these criteriaare:

(a) The pilot cannot comply with two contradictorysories simultaneously, such as "climb" and "descend".

(b) The pilot should not be asked to change the escapemaneuver between the vertical and the horizontalplane unless absolutely necessary, and the logicmust accomodate typical measurement noise withoutselecting such changes.

(c) It is preferable to initially display an advisozy ina single plane; both a horizontal and a verticaladvisory may be given together if the conflict shoulddeteriorate.

(d) An advisory that is compatible with a pre-existingmaneuver is preferable to an advisory that wouldreverse it.

Prior to issuing an advisory, a complementary advisory is

coordinated with th3 intruding aircraft if it is TCAS equipped.

2.10 ENGINEERING TEST UNIT HARDWARE

Figure 2.10-1 is a photograph of one of the TCAS III Engi-neering Model eight element, directional, electronically steeredantennas. It is 14.4 inches in diameter to the outer bolt circleand is 7/8 of an inch high. The too and bottom antennas areidentical. Each antenna array also contains a center elementwhich is used by the Mode S transponder. The Engineering Modelantenna is curved to fit the 74-inch radius of the 727 fuselage.Although This curvature distorts the pattern slightly from thosewhich would be measured on a flat ground plane, stored monopulse

* * correction curves in the TCAS III processor compensate for thoerror. Figure 2.10-2 shows the top antenna mounted on the FAA'sBoeing 727. The bottom antenna is mounted similarly under thefuselage.

--- The associated antenna electronics equipment required tosteer and shape the patterns is located in a separate package,which is mounLed inside of the fuselage near each antenna. Thiscontrol box interfaces with the TCAS Signal Processor via a 6-bitdigital beam steering command and RF connections to the threeantenna ports: sum, difference, and omnidirectional. The sum anddifference beams, formed by the Beam Forming network, are usedfor monopulse tracking. The center antenna element is used with

19

FB.VGl1 774

FIGURE 2.10-1- ENGINEERING MODEL DIRECTIONAL ANTENNA

-F B-VG-1775

FIGURE2.102 TCAS III ANTENNA MOUNTED ON FAA'is BOEING 727

20

the TRU-2B for Mode S transponder functions and air-to-air coor-dination.

Figure 2.10-3 shows the principal functional blocks orsubsystems of the TCAS III Engineering Unit (ETEU). Both the topand bottom mounted antennas are directional. The antenna sub-system also includes all the beam forming and beam steeringcircuits necessary for directional operation.

The Interrogator RF subsystem includes all the ETEU trans-mission and reception circuitry. An existing airline qualitytransponder was modified to serve this function in the Engineer-ing Model. The Mode S transponder is a slightly modified unitfrom the FAA's DABS Engineering Evaluation program. The Interro-gator Processor controls the transmission waveforms, controls thedirectional antennas, and performs reply signal processing.

The Computer is comprised of four Intel 8086 micropro-cessors, associated memory banks and power supplies, which sharethe main TCAS III processing load. It also includes severalspecial purpose circuit boards used in interfacing the ETEU toother aircraft subsystems and performing other special functions.The Computer Programs which it exercises are loaded through theMonitor and Control equipment. These programs include all thetracking algorithms and the logic necessary to determine theexistence of a threatening aircraft.

The Monitor and Control subsystem includes instrumentationand display facilities that allow control and ease of changingparameters during flight tests. The display includes a cockpitmounted graphic display of traffic advisory and resolutionadvisory outputs of the system to the flight crew.

The Own Ship Data interface includes attitude and trueNorth-referenced position data from the on-board InertialNavigation System. In the TCAa III, it is necessary to makecorrections for aircraft attitide to achieve the required systembearing measurement accuracy.

Figure 2.10-4 is a photograph of the TCAS III EngineeringTest Unit installed in the FAA's Boeing 727. Rack number 1 onthe far left contains the essential TCAS functional equipmentwhile the other three racks contain support and test facilities"that would not be provided in an operational application. Thetop shelf of the first rack mounts the Interrogator (TRA-65A),the Mode S transponder (TRU-2B), and the System Control andPerformance Monitor. The second shelf contains the InterrogatorProcessor Unit and the third shelf the Data Processor Unit whichhas all the logic needed for threat assessment.

The second rack incorporates peripheral equipment to loadthe computer (disk drives) and - Micioprocessor DevelopmentSystem to be used to maintain oi make modifications to the systemoperating programs.; The third rack contains the Traffic Advisory

21

I AD

INTERNES ANTE INA IF PROCESSOR CMU[

SlPPRESSIO@ MiS CU SIPS 3ATA

U.I

FIGURE 2.10-3. BLOCK DIAGRAM OF TCAS III ENGINEERING MODEL

ii

FIGURE 2.10-4. TCAS III ENGINEERING MODEL INSTALLATION ONFAA'S BOEING 727

22

Display on the top shelf and the Data Extraction and RecordingUnit magnetic tape recorder which is used to log data during theflight tests on the second shelf. The fourth rack-contains theIn-Flight Performance Monitor (IFPM), a small personal computerthat has an alphanumeric status display and provides operatorinterface to the TCAS system.

3.0 TCAS III CONTRACT MODIFICATIONS

While under contract to the FAA, a modification to thecontract was added to target several areas that requiredadditional work. This section describes the problem areas andthe solutions.

3.1 COCKPIT DISPLAY

The Contractor shall investigate techniques to display theTCAS video output and develop the necessary interface.

After much investigation, an off-the-shelf ElectronicFlight Instruments System (EFIS) display was purchased fromBendix Avionics Division for installation in N-40. The EFISdisplay is a raster-stroker system that provides superiorresolution and brightness. This display system interfaces withthe Engineering Test Unit by installing an Intel single boardcomputer (iSBC) in the multibus data processor rack. The iSBCaccesses the track file and formats the appropriate data fordisplay and aural advisories. The iSBC forwards the display datato the EFIS symbol generator using a high speed ARINC 429 bus.This display contains all the information found on the TFPMgraphic display with the addition of color to indicate threatlevel and arrows indicating direction of altitude change forevery altitude reporting target.

3.2 WHISPER/SHOUT TECHNIQUES

The Contractor shall analyze high density data obtainedluring flight testing to determine the effectiveness of the4hisper/shout levels chosen during simulation. New levels shallbe investigated and implemented if warranted.

The whisper/shout performance was analyzed using data4• recorded during the February 1985 flights in the Los Angeles

Basin. A summary of the whisper/shout reply statistics for theFebruary 5th test flight shows that the whisper/shout levels arebalanced. The February 3rd A.M. and P.M. flights show almostidentical statistics.

Referring to Figure 3.2-1, it is seen that targets at themost dense range (3-4 nautical miles) are almost evenly balancedamong the four whisper/shout levels. The whisper/shout inter-rogation levels of 1 to 4 elicited the following percentagereplies:

23

jn% K -'K 21ýx xrizsu- z xx --x -z-x -wW -K i- v-u- v -

25io WHISPER/SHOUT LEVEL I

w"- 290-a.

w

cc

5 1 15 20

SLANT RANGE (NH)

WHISPER/SHOUT LEVEL 2

aJ

c i s@

15 Is is 20

SLANT RANGE (NM)

WHISPER/SHOUT LEVEL 3W

2_6

-J(LUi

m 29

:3 - -

1

,8 15 20

SLANT RANGE (NM)

2S60,

WHISPER/SHOUT LEVEL 4U)

lose

M

cmJ

(-h7

SLANT RANGE (NM)

FIGURE 3.2-1. THE FOUR-LEVEE, WHISPER/SHOUT USED BY ETEU#l SHOWSEXCELLENT SEPARATION OF1~ TARGETS IN THE L.A. BASIN

24

Level 1 14 %Level 2 24 %Level 3 32 %Level 4 30 %

If the distribution peaks were estimated from the shape ofthe curves, the following ranges were observed:

Level 1 1.5 nmiLevel 2 3.5 nmiLevel 3 5.0 nmiLevel 4 6.5 nmi

Figure 3.2-2 shows the whisper/shout level 4 distributionfor three power levels transmitted as a function of the rangegate range in search and the predicted range in track.

From this summary analysis, the whisper/shout levelsappear to be chosen correctly. The data validates the fourwhisper/shout levels selected by analysis of the 1985 L.A. Basintraffic model early in the contract.

3.3 AUTOMATIC ANTENNA PERFORMANCE MONITORING

The Contractor shall investigate automatic methods ofchecking antenna performance -rom an external source and shallimplement hardware and/or software to accomplish this.

An antenna pattern check routine was added to the Calibra-tion program to record target reply amplitudes on magnetic tape.The use of an external transponder was required. Data for thesum and difterence patterns were gathered separately. A postprocessing program was written to read the data from magnetictape and plot it.

3.4 EFFECTIVE ANTENNA BEAM WIDTH

The Contractor shall analyze a cross-section of trans-ponders interrogated during flight tests to determine effectivebeam width. Modifications to optimize operation shall be made ifwarranted.

An attempt was made to analyze data from several flightsinto Newark, New Jersey. Data tapes from the ARTS-3 radar duringthe flights were available to provide an absolute reference. Thecomparison of data between the two sources was found to be insuf-ficient to determine the effective beam width.

3.5 ALTITUDE DECODING ERRORS

The Contractor shall investigate the problem associatedwith altitude decoding errors and shall incorporate hardwareand/or software modifications to minimize the problem. At aminimum, hardware modifications to delay the "in-beam" decision

25

WHISPER/SHOUT LEVEL 4ATTEHURTIONs 6-7 D2

0.

z

8 5 18 15 26

SLANT RANGE (NH)

I WHISPERISHOUT LEYIL 4t..RTTEHUArrOH I 3-6 DI

- 2084-/a.U1

518 _ _ _ _ _ _2__ _

SLANT RANGE (NM)

2s- WHISPER/SHOUT LEVEL 4(_ RTTEHUATIOHt 8-2 D1LI

a.

5 le 15 20

AANE LEVELi I LEE 2 LEE 3 LEVE•L 4 LEE 4A, LEVE 41 LEVEL. 4C0- 1 1427 24, 16, 1 29 04 40

'- 2 1 103 s86 709 583 4 11 11 1 6l

S2- 3 854 048 1197 837 14.4 30 243•:- 4 60 so 12 1507 1348 930 45 373

4- :ý t161 591 1 190 1 488 M9 83 312

1 28600

A- 7 39 3w 1138 2 129 1211 ",

7- 8 10 199 712 2017 120 . 72141- 1 4 511 443 1,-~d 32, '73 5A9- 10 1 27 242 122' 0 810 373

'0--I 2 21 207 9"4 0 42•4 340.1-12 0 20 152 806 0 393 4 1312-.13 0 a 74 648 0 0 646

'€i 13-14 I 9 5b 3c)1 14-,63 0 0 20 393 0 0 "3r9

15-1 0 1 25 197 0 0 197l&-17 0 0 O 16 00 146

<: 1 7- 18 0 0 0 0 0 0 0" "s- 1.- 0 0 0 o 0 0 0f'" 19-20O 0 0 0 0 0 0 a

FIGURE 3.2-2. THE SHOUT LEVEL #4 STATISTICS ARE EXTENDEDBY COMPUTING THE NUMBER OF REPLIES FROM THE DIFFERENT

ATTENUATION LEVELS

26

until after bracket decode has been accomplished, and to includeconfidence bits for each altitude code pulse, shall be investi-gated. Software modification to implement a voting scheme usingthe confidence bits at track initiation shall be explored.

This area was the second highest priority task of the con-tract modification. Both hardware and software changes wereinvestigated. The hardware changes required modifications to theInterrogator Processor. The first change was to use the sum beamvideo for decoding. When the equipment was built, the video fromthe HE + jA) + (Z - jA)] monopulse circuitry was used for decod-ing rather than the sum beam as planned. The hardware was modi-fied to use the sum beam only for bracket detection and altitudedecoding. This reduced the interference from fruit, reduced thepunch-through and improved the RSLS in-beam circuitry. To incor-porate this change, the Analog/Digital card was modified bychanging the wire wrap connections. Also the Minimum ThresholdLevel (MTL), Dynamic Minimum Threshold Level (DMTL), ReceiverSide-Lobe Suppression (RSLS), and other thresholds were recal-ibrated.

The second hardware change was also made to the Interro-gator Processor. This modification was to remove the altitudedecode pulses from RSLS in-beam filtering. A new board (A4) wasdesigned, so when installed, altitude code bits are not elimin-ated by the RSLS threshold. This new board permits confidencebit flags to be logged when pulses are in-beam. When the A4 cardis removed, operation reverts to its original state, such thatall pulses are gated by the RSLS in-beam criteria.

In the process of incorporating the confidence-bit-flagchange in the IP, two circuit problems were identified. Oneresulted from the clocking of the quantized video (QV) and thedifferential pulse positive slope (PS) prior to qualifying the PSwith QV. The 125 nanosecond quantization on each allows up to250 nanosecond variation between the two pulses. In addition,the PS position changes approximately 100 nanoseconds as thepulse cmplitude approaches MTL. Thus the PS relative to the QVchanges approximately 350 nanoseconds depending upon clock timingand pulse amplitude.

The second circuit problem was more subtle. Prior to thelog amplifiers, the pulse is differentiated to obtain leading andtrailing edge pulses. It was also found that high frequencynoise generated in the log amplifier can result in extra positiveslope declarations.

The combination of noise in the circuit and quantizationerrors necessitated an investigation into possible hardwarechanges that would increase the percentage of decodes. Fourhardware configurations were analyzed. After testing, it wasdetermined that the analog detection of PS and QV provided themost improvement in the altitude decode performance. This changewas installed in the Interrogator Processor of both units.

27

Another problem with the Interrogator Processor was dis-covered during flight testing in the Los Angeles Basin. Incertain situations, real altitude code pulses were eliminated,causing the non-linear vertical tracker to predict an erroneousaltitude rate. The problem was traced to the suppression drivecircuit in the TRA-65A. The solution was to add a 470 ohmresistor at the amplifier output of the circuit.

Various software modifications were also investigated toimprove the system's altitude decoding capability. The first wasan altitude bit voting scheme. After range and bearing correla-tion, the altitude code of each defruited reply was compared tothe three previous replies from the same target snd adjusted ifit seemed that therc wcre bits in error. The method for deter-mining bits in error was to compare the number of l's and O's foreach bit over the current reply and the three previous ones. Thecurrent reply bit was "voted" to be either a "1" or a "0"depending on the number of occurrences of l'-s and O's from theprevious three replies. The "voted" altitude provided the inputto the vertical tracker. The flight testing of this algorithmand the analysis of its performance indicated that anotheralgorithm needed to be implemented.

.ihe next algorithm was to implement an alpha-beta trackerfor altitude correlation. After range and bearing correlation, adefruited reply is then correlated in altitude to the predictedvalue output of the alpha-beta tracker. If it is within +200feet, the decoded altitude itself (not the output of the alpha-beta tracker) is sent directly into the non-linear verticaltracker for altitude rate estimation. This algorithm along withthe hardware modifications described above resulted in a largeimprovement in the altitude decoding capability of the system.

3.6 ADDITIONAL AURAL ADVISORY CAPABILITY

The Contract or shall investigate schemes to incorporate alow cost digital voice simulator dnd shall program it to provideaural advisories if warranted.

The addibional aural advisory capability was added alongwith the upgraded cockpit display. The aural alert and voicesynthesizer unit is controlled by the same single board computer(iSBC 86/35) that controls the display. Aural advisories areforwarded to the aural alert unit using the iSBC parallel outputport. The iSBC identifies each discrete phrase to be spoken tothe pilot. Also, the aural advisories were added to the pilot'sintercom.

The vocabulary of the aural alert unit consists of thefollowing phrases: Traffic, Climb, Limit Vertical Speed,Descend, Turn Left, Turn Right, Limit Turn, RA Clear, RA Invalid,and Crossover. The unit provides aural alerts when a targetbecomes a traffic advisory, when a target becomes a resolution

28

W T ý -- A -ýA,_-Xn l • • •' "i U• - • • • E'

advisory, when a resolution advisory changes, when the advisoryis finished and when TCAS is unable to resolve the conflict.

3.7 TARGET SPLITTING

The Contractor shall investigate more thoroughly theproblem of target splitting (multiple tracks on the same targetin range and/or azimuth) and shall incorporate hardware and/orsoftware modifications to minimize the problem.

Target splitting was considered the highest priority itemin the contract modification. Investigation into the problemshowed that multiple tracks were also generated due to faultyaltitude decoding. Improvements in altitude decoding, asdescribed above in section 3.5, provided a reduction in thenumber of multiple tracks as well.

Even with the improved altitude decoding, there were stilla number of target splits. So, a new software module, GeneralCorrelation (GC) was added to the operational p:ogram. The pur-pose of this module is to identify multipath, punch-through, andsplit targets. A multipath target is one that correlates inbearing and altitude with another target, has greater range, thesame sign of range rate and a smaller magnitude of range ratethan the target. A target that is declared to be multipath isflagged as such in the track file and is not actively updated.The target split test consists of correlation in range, bearing,and altitude. A target that is declared a split is deleted fromthe track file. Punch-through targets occur when sidelobes ofthe sum beam "punch-through" (hate higher amplitude than) thedifference beam at the same azimuth angle. A punch-through tar-get will correlate in range arid altitude and will have a lowersignal level than the actual target. The punch-through targetwill be flagged and retained in the track file until it is drop-ped. Both multipath and punch-through targets are not activelyupdahed in track, do not go through the chreat detection andresolution logic, and are not displayed. The installation ofthis module provided a significant reduction in the number ofmultiple tracks.

Another software modification implemented was range gatingon the bottom antenna. After flights in high density areas suchas Atlanta and Los Angeles, it was noticed that a large number ofnuisance replies were detected by the bottom antenna. So, arange gate of four nautical miles was placed on the bottom anten-na for search on v. Active track updates on the bottom antennahowever are stilt performed to full coverage. To compensate forreduced coverage on the bottom antenna, the top antenna now pro-cesses replies down to - 5 degrees elevation instead of 0 degreesas was previously set. When Oaqn aircraft is flying above 10,000feet and at high speeds, the top directional antenna providessufficient coverage to long ranges. Below 10,000 feet where themaximum relative velocity is 500 knots, the four nautical milerange provides 30 seconds protection. This range gate reduced a

29

large amount of fruit returns and enabled the system to operatereliably in a high density environment. In addition, multipathreturns whicn are usually detected on the bottom antenna due tosurface reflections, were also reduced.

An additional source of target splits was found to becaused by Mode S targets. Orxginally, Mode S targets were cor-related by Mode S address only. It was discovered, however, thatdue to punch-through and/or multipath effects, target replieswere being correlated by the ID even though the range and/uzbearing was in error. These inputs to the trackers causederroneous rate and position estimates and very poor Mode S track-ing performance. The solution was to subject all Mode S repliesto the same range, bearing, and altitude correlation algorithmsthat apply to all ATCRBS replies. The Mode S address test isstill performed for correlation of Mode S replies. This softwaremodification resulted in a significant imprcvement in Mode Stracking performance.

3.8 HORIZONTAL THREAT DETECTION PROCRAM

The Contractor shall analyze various schemes to improvethe efficiency of the horizontal threat detection logic andmodifications to accomplish this.

Once the horizontal threat detection and resolution logicwas operational in the TCAS, analysis of the efficiency of thealgorithms was begun. The most important discovery was that thetime involved with modelinA the flight paths of the intruder wasrather large, especially when there was more than one threat.This prompted MITRE to propose an alternate method for modelingaircraft maneuvers for a revised version of the CAS logic.

3.9 HORIZONTAL THREAT DETECTION AND RESOLUTION LOGICVALIDATION

The Contractor shall test the horizontal threat detectionand resolution logic by simulation (using both real target dataobtained in flight tests and simulated targets) to determine itsperformance for a wide range of scenarios, and shall recommendimprovements.

At the request of the MITRE Corporation, certain trackfile and CAS logic variables were logged on tape to facilitatepcst processinq of flight data for CAS validation. The FAATechnical Center conducted a large number of flight tests thatcontained a wide variety of scenarios. From the data logged ontape, an estimate of bearing and bearing rate accuracy wascompiled. This accuracy translates directly into miss distanceaccuracy which is essential to the CAS logic algorithms. A largeamount of post processing of flight test data was analyzed todetermine search and track performance in addition to v-lidationof the threat detection and resolution algorithms. More detailson the flight data analysis can be found in section 4.

30

4.0 FLIGHT TEST DATA

The FAA is in the process of testing TCAS III to validateits ability to reduce unnecessary alarms, to provide relativemotion on traffic advisories, and to give appropriate horizontalresolution advisories. The first Engineering Test Unit wasinstalled on the FAA's Boeing 727, N-40, in 1983. So far, themajority of -he operational testing has been done using thisunit. The major test objective of assuring that the systemangular accuracy is sufficient to pcedict miss distance has beencompleted. More testing is scheduled to evaluate Version II ofthe TCAS III logic and the air-to-air coordination.

To measure the antenna angular accuracy, two aircraft wereflown in both linear encounters and circular orbits. They wereflown from the FAA Technical Center over their instrumented testrange. A Convair 580 with a regular ATCRBS transponder on boardwas used as the target aircraft. The TCAS III installed on N-40acquired, tracked and logged range time-referenced data on tapefor the many encounters and orbits flown. The FAA Technical Cen-ter precision tracking radars locked on to the range instrumenta-tion x-band transponders mounted on the test aircraft and pro-vided an accurate position of one aircraft relative to the other

*! as a function of time. After each flight, data from the twoseparate sources were compared. Because the radar is five to ten

+ times more accurate than the TCAS III, its position estimates areconsidered to be perfect and all errors are attributed to theTCAS IIT.

To test operational performance in typical and highdensity environments and to build a data bank from which per-formance estimates can be made, the TCAS III has been flown intoand over many areas such as Los Angeles, Boston, New York,Atlantic City, Philadelphia, Washington, and Atlanta. Inaddition, encounters with a target aircraft have also been run inthese airspaces to evaluate the effect of traffic density on theTCAS III's ability to track and issue appropriate resolutionadvisories.

4.1 ANGULAR ACCURACY TESTS

To test the angular accuracy of the TCAS III, the targetaircraft was flown in 360 degree orbits about the EngineeringTest Unit #1 at elevation angles from -20 degrees to +20 degrees.As previously discussed, the radar data is assumed to be perfectand all the angular error is attributed to the TCAS III. Figure4.1-1 shows a summary graph of six such orbits.

Bias errors shown at the top of the figure do not impactcalculations for a horizontal maneuver because true relative

*• bearing does not appeat in the formula for computing miss dis-tance. The effect of such bias errors is to give the pilot asmall error in the direction visually search-ý for a target.Bearing rate is used to calculate horizontal miss distance.

31

I

~' ~ i~PP r TO. x~tcR.i XjflK w m v llW Ik I? tIf

8 DIAS ERROR

6 EL L EL20.

-150 -105 5' 5 0 5

21 EL EL10'

ix0

3.0 EL EL :10

S1.0 T

00 -150 -100 -50 0 50 100 150

4IGUR 4.-1 RANGULARRRU C SIGMAG)PRORA

3.

2.52

Bearing rate errors are caused by random errors or bias errorsthat appear as random errors due to changes in bias as the anglechanges slightly. In the assessment of angular coverage per-formance, the combined random errors and changes in bias areobserved because errors are computed as the difference betweenradar measurements (assumed to be correct) and the TCAS III anglemeasurements. It is seen that these one-sigma random errors arefrom 1 to 2 degrees near the nose, and in a worst case approach,3.5 degrees at the tail. It should be noted that the radar datawas updated ten times each second and that Own aircraft data wasupdated only once per second during this test rather than fourtimes per second.

Angular errors over a wide range of elevation angles havebeen tested several times. In addition to both TCAS and targetaircraft flying, such testing includes placing the TCAS III air-craft on the ground near the center of the airport and flying thetest aircraft at positive elevation angles. With own aircraft onthe ground and a five mile TCAS to target aircraft orbit range,the random error is reduced to one-half the dynamic result thatwas shown in Figure 4.1-1. The dynamic errors plotted in thefigure (assumed to be worst case errors) are still well withinthe toler..-ce needed to compute horizontal miss distance forresolution advisories.

4.2 TCAS ENCOUNTER FLIGHTS, FEASIBILITY OF HORIZONTALMANEUVERS

Encounters between the TCAS III and a target aircraft havebeen run to evaluate the feasibility of horizontal resolutionadvisories. In these cases, the encounters were run on the FAATechnical Center instrumented test range with one radar trackingthe TCAS III aircraft and the other tracking the target aircraft.The data measured by the TCAS equipment was compared to the datameasured by the radar as described in Figure 4.2-1. The encoun-ters were run at different speeds and at different angles todetermine errors in miss distance derived on board the TCAS IIIequipment in real time. In addition, both vertical and horizon-tal fake-outs were run. The fake-outs are encounters where thetarget flys a stiraight and level trajectory, then makes a suddenmaneuver in either the vertical or horizontal plane. The datapresented in this section are taken from 3 flight tests:December 19, 1983; January 23, 1984; and October 25, 1984.

Figure 4.2-2 is a typical encounter. Although the trackwas started at a much longer range, only the last 40 secondsbefore CPA (closest point of approach) are shown on the plots.In this particular encounter, the tazget aircraft is on a course90 degrees from the TCAS III. The speeds were approximately 200knots and t1- altitude separation at intercept was 300 feet.

The first plot is an x-y presentation of the target track.It indicates the target's position relative to the TCAS aircraft.The next set of plots show the bearing accuracy. Each asterisk

33

INTRUDER

TARGET TCAS MEA3UREMENTS

TCAS h

TRACKER/

CNRTH t

/ /t

dR

TCAS T

LATITUDELONGITUDE

(10z)NORTH ROLPITCHiio•IIG

LATITUDE/LONGITUDE a- COORDINATE

TO CARTESIAN TRANSFORMATION

TRANSFORMATION x STABILIZED

Y 8 x YTCAS CENTERED

RAW BEARING ERROR

COMPUTE w B= 5y/ BR 1

a-0l a-#. a-, a-,,FILTER FILTER -AZ/ERoFILTER FE

COMPUTE HORIZONTAL COMPUTE HORIZONTAL

VjI ISS DISTANCE HORIZONTAL MANEUVER ESCAPE MISS DIS7ANCESII •ENVELOPE

o m TIME TO CPA

SM.: { Xi.Yb)IVI X2 +j2)

FIGURE 4.2-1. EVALUATION OF TCAS III HORIZONTAL MISSDISTANCE ESTIMATE

34

FI-VG- 1420

TARGET TRACK

J.0 TRACK NUMBER: 37,.5 ENCOUNTER DATE: 12/19/834.0 TIME t 12:24 :00

IsENCOUNTER ANGLE: 90 DEGREES30 "OWN: SPEED - 200 KNOTSS 25 ALTITUDE q 4500 FEET=20 TARGET: SPEED = 200 KNOTS015 t ALTITUDE a 4800 FEET

10 SEPARATION: .300 FEET VERT05

I, ,4 -s -e 1 2 3 4 5RELATIVE Y (NWo)

BEARING ACCURACY

S 1o 0 10

C a.~100

1 . NUMBER OF POINTS 4170 " EAN - -6.12 DEG70., ,STANDARD DEVIRTION 0 8.69 DEG

2 -10

.40 -35 -30 -25 -20 -15 -10 -5 0 -40 .35 -30 -25 -20 -13 -10 -5 0TIME TO CPA (SEC) TIME TO CPA (SEC)

BEARING RATE ACCURACYtJ 3

2 20

S1 5S•O2'Si ~1.0

0. s

0 :i0.0.

RADAR 00, 05 3 ~ I-1P,, ••-1 05 EH .6 EE

-1 0 HUM.PER or POINTS 41

- MEA N - e.0 46 DEG/SEC

1-3 - .-40 -35 -30 -25 -20 -15 -10 -5 0 -4. 35 -30 -25 -0O .1 0 Is 0

TIME TO CPA (SEC) TILE TO CPA (5EC)

FIGURE 4.2-2. 340-KNOT, 90-DEGREE ENCOUNTER

(SHEET I OF 2)

35

FO-VG-1421

1_ 10000I-

v 7500

z 5000

(-$1- 2500 RADAR

0 -2500-w 4

S-5000C30

tn -7500

-100001-40 -35 -30 -25 -20 -15 -10 -5 0

TIME TO CPA (SEC)

10000

S7500

/5000

., P2500

u 0

ca • •"DISTANCE

(A -5000, ENVELOPEEn

::E -7500-

-10000- 0 -35 -30 -25 -20 -15 -10 -5 0

TIME TO CPA (SEC)

FIGURE 4.2-2. 340-KNOT, 90-DEGREE ENCOUNTER(SHEET 2 OF 2)

36

on the TCAS-derived north bearing plot indicates a single bearingmeasurement made by the TCAS III. The straight line is the bear-ing derived from the radar measurements. The plot next to thenorth bearing gives the error between the TCAS-derived and radar-derived measurements. The third set of plots shows the bearingrate accuracy of the TCAS relative to the radars. The importanceof these encounter flights is shown in the final set of graphswhich compare miss distance derived from TCAS and the radars. Onthe miss distance error plot, heavy lines are drawn to representthe distance own aircraft can move by executing a coordinatedmaneuver within 6 seconds of the graphed time to CPA. For exam-ple, it is assumed that a traffic advisory has already been given-o the pilot is alerted to the fact that traffic exists in thedirection indicated. Then, if a horizontal resolution advisoryis given at 25 seconds befdre CPA, the pilot could achieve asmuch as a 6000-foot or one nautical mile lateral separationshould the situation require it.

Some additional examples of different types of encountersare shown on the following pages. Figure 4.2-3 shows an 810knot, 0 degree closing encounter. Figure 4.2-4 shows a 415 knotclosing, 15 degree encounter. The next three encounters arehorizontal fake-outs where the aircraft are separated 300 feet inaltitude and the encounter begins with aircraft offset horizon-tally as well. As the encounter progresses, the target crossesdirectly over TCAS in Figure 4.2-5, crosses in front of TCAS inFigure 4.2-6, and crosses behind TCAS in Figure 4.2-7. These sixexamples are just a few of the many encounters run at the FAATechnical Center.

The conclusion after these encounters is that warninginformation given to the pilot prior to a resolution will alerthim should an advisory become necessary. However, with thethree-dimensional separation modeling, an alarm may be safelydelayed to the last possible moment, so that in most cases, thepotential conflict will be resolved by ATC procedures or apilot's normal action. If the conflict persists and a resolutionadvisory is given, the pilot will respond quickly with emergencytype actions. Since the modeling is repeated every second, theresolution advisory is removed as soon as the threat is resolved.

4.3 LOS ANGELES BASIN FLIGHTS

TCAS III was flown in the Los Angeles Basin in February1985 to assess its performance in a high density aircraft envi-ronment. Over the span of five days, the TCAS III EngineeringTest Unit logged over 15 hours of flight time and flew in 60encounters with an FAA target aircraft. The Los Angeles Basinflight tests indicate that TCAS III performs well in all types ofenvironments. It detects targets at the maximum surveillancerange and tracks them reliably even in high density areas. Itassesses threatening situations and determines the approoriatemaneuver to avoid a possible collision. In addition, it provides

37

FB-VG-1434

TARGET TRACK

10o TRACK NUMBER : 14C/ ENCOUNTER DATE :1/23/8/ TIME: 11 : 28: 00

65 ENCOUNTER ANGLE:O DEGREESOWN ALTITUDE:15,000 FEET

A • 'TARGET ALTITUDE: 15,600 FEETSEPARATION: +600 FEET VERT

2

-10 -6 -6 -4 -2 2 4 6 S 10RELATIVE Y ( I)

BEARING ACCURACY

-10 0 10

Sso

~~~~~~ *O1rH- g3 r

so- Eo

-40 .35 -30 -25 -20 -15 -10 -5 -40 -35 -30 -25 .20 -15 -10 -5 0TIUE TO CPA (SEE) TIME TO CPA (SEC)

BEARING RATE ACCURACY60 6 MEA 1,0 D_j -l

I'

0-~ .0

-0.S0 *.... ..*

S-1 .0.5"

SNUMBER OF POINTS 31-2 -AAR1.0 MEAN * -aZ DEGSEE

SSTANDARD DEVIATION C 0.327 DEG/SEC

-40 -35 -30 -25 -20 -IS -10 -5 0 -40 -35 -30 -25 -20 -15 -10 -S 0TIME TO CPA (SEe) TIME TO CPA (SEC)

FIGURE 4.2-3. 810-KNOT, HEAD-ON ENCOUNTER(SHEET 1 OF 2)

38

q. - .1 4t4 _A - ,: % N

FB-VG-1435

-. 15000I--LL

'W 10000-< 5000_,..i RADAR

S5000

tn qcn* 4

0

>. -5000

b -i0000

150

-5 0 -35 -30 -25 -20 -15 -10 -5 0TIME TO CPA (SEC)

10000

I- 7500i _ 5000 5EC ROLLOVER TO 45*

m 5000S2500t

w

0

'n -25000,-, / • \ ESCAPE

-7500-

-°o -35 -30 -25 -20 -15 -10 -5 0TIME TO CPA (SEC)

FIGURE 4.2-3. 810-KNOT, HEAD-ON ENCOUNTER

(SHEET 2 OF 2)

39

FB -VG-.440

TARGET TRACK5 0* TRACK NUMBER: ZFC4.504 0 ENCOUNTER DATE: 12/19/833 50 TIME: 13:24:00

ENCOUNTER ANGLE: 15 DEGREES3 0 OWN: SPEED c 200 KNOTSALTITUDE c 4500 FEET

2 0 TARGET: SPEED r. ZOO KNOTS1.5 IALTITUDE a 4800 FEET10 SEPARATION: .300 FEET VERTo05 2000 FEET HORIZ00-5 -4 -3 -2 -1 1 2 3 4 SRELATIVE Y (NM!)

BEARING ACCURACY

11" .- 40

22

s•o RADAR -4~- M HUIMPER OF POIHT$ - 3?6 MEAN - -1.08 DEc

8. STANDARD DEVIATION 1.34 DEC6O____ ___ ____ ___ ____ ___ ____ ___ ____ ___ -10

-40 -35 -33 -25 -20 -15 -10 .5 0 -40 -35 -30 -25 .20 -15 -10 -5 0TIME TO CPA (SEC) TIME TO CPA (SEC)

BEARING RATE ACCURACY

Liw 2.0

-cz, 1.5

2 1.0

S1 • 0.5 .S0 0."

StrtUMBER or POINTS * 372 u -1i~I 0 MENv 06 DEC SECCCRADAR \' ýTQ.I~IDAR DEVIATIOlI - 0 .740 DEG'SEC

* -4 -2.0-40 -35 -30 -25 -20 -15 10 5 0 -40 -35 -30 .25 -20 .1 0 -TIME TO CPA (SEC) TIME TO CPA (SEC)

FIGURE 4.2-4. 415-KNOT, 15-DEGREE ENCOUNTER(SHEET I OF 2)

40

FB-VG-1441

10000'uL

7500

j 5000-

C3 2500-

-7500

E -7500-

10000- t

-40 -35 -30 -25 -20 -15 -10 -5 0

*TIME TO CPA (SEC)

10005000 6 SEC ROLLOVER TO 450

cr- 2500

u 0

S-2500 ESCAPEm DISTANCE

cn' -5000 ENVELOPE

-7500

-10000--40 -35 -30 -25 -20 -13 -10 -5 0

TIME TO CPA (SEC)

FIGURE 4.2-4. 415-KNOT, 15-DEGREE ENCOUNTER~

(SHEET 2 OF' 2)

41

TARGET rRACK

TRACK NUMBER:. 8AS4. ENCOUNTER DATE: 10/25/844." TIME. 11.18:00

z S3.1;

SX 3.0 TARGETW 2.s 3300'/19OKTS

Sz.0 \ 3 °/ S E Ca 2.5 TCASSw 3600'/200KTSc:t 1---. ,NM -----A

.. 5-4 3 -z -" "

RELATIVE Y (NMI)

BEARING ACCURACY

0, to TROCK NUMIrRc OR0 * SCENARIO DATEi 1o'2.'94

XG 54 RAARSCCHARRO 11/•C: l11161Z60wLit: HUlMBER OF POINTS - 41

0•: S T93NDARRD bCVIATION 1 1.919 :)CC.... 140 aA 21,

wo

cJ a" -

21e #1 -toVsi its 1115 lie its 1211 129:; 39 131;i 1*4 I ll lIi e I Its; 12l 125 0131 136

RELATIVE TIME (SEr.) RELATIVE- TIM9 (SEC)

BEARING RATE ACCURACY

• I0 TRACIX HUMBER, OR•1.5 SCENARIO DaTE 10/*S 1.94

u. XJ SCEARRIO TliPIar 11,,1,,20

a - X .0 NUMBCR or POINTS - 41

to- . .. .• .l• .. .. I . . . . . uJ0- .7 EI

t2 s

1W

InRDA a~r'-

42

L) -40

fl s7 toll 1eg lie 22 129 225 239 125 Is* 2 og Ile g2ig1 120 125 130 t35

RELAT3VE TIME (SEC) RELATIVE TIME (SEC)

Lb~SH E 1 .O F 2) HOIR;B

Z 42

isseo RADAR

z

U7

0

r -N-Il- I l I I

Il

• u-10 0

95 100 18 l 115 120 125 130 13SRELPTIVE TIME (SEC)

Me

LL. 6 SEC ROLLOVER TO 456

w, 40 Aa

°• ~ ~ ~~~~~~~ ~~~~ FIUE425SOIOTA AEUTRrTCAPSESBLWTA

c~-2588

1008

• ~(SHEET 2 OF 2':

43

TARGET TRACK

TRACK NUMBER: 19F4.9 ENCOUNTER DATE:, 10/25/8434.0 TIME: 11.53:00

S3.6

3. TARGET

2. 0 3'e\ /SEC/l < - -. -- TCAS

,C-,- --.- -310W/200KTS

• •; K-,--INM K-" l NM -A•

-4 -2 -2 I - 3 4 -- 2,7NMRELATIVE Y (NMI)

BEARING ACCURACY

- ..- 16 F TRACK NUMBER. 19r

RADAR SCERIO DATE.g 1012S'C4

,•~~ 4'''' MEAN H" - -3.7'1 DECZ49 . . 21. STANDARD DEVIATIO - 1.3r DEC

o-, n- Z30 (m * O O T 1•

c.................................................................... ...

a i Z21 t - 6.

0

S166 i ts 116 II 1 !2 126 135 135 16 1to Islie I 11 I 125 130 135RELATIVE TIME (SEC) RELATIVE TIME (SEC)

BEARING RATE ACCURACY

A. ' TRACK tIUR~R: I9S.

I.S. SCENARIO D rt, 10/26/84

J0 ISCENARIO T IMEa 11:61:26I, - NUMBER OF POINTS 4 48

Co" 2' MCAN - 6.979 DE$;'SC

2 , C3- I ANDRD DEYVIRTION - .29l ,*S SC DECa ... .-... .. . .. .

•J-3 RADAR a-o "

- -2.0i 1 leo 11; Its its 1zo 1zs 136 135 is* log 166 1 t1 126 o 25 126 135

RELATIVE TIME (SEC) RP.LATIVE TIME (SEC)

FIGURE 4.2-6. HORIZONTAL FAKEOUT, TARGET CROSSES IN

FRONT OF TCAS (SHEET I Or 2)

44

Iseee0 RADAR

1)z°) aZN

W..seee

-it ., .

w0

U)

I,--1599,.

95 198 105 118 115 120 12S 130 135

RELATIVE TIME (SEC)

18809

Li.. 6 SEC ROLLOVER TO 45-

-2500 -

ww 0 44

z -

/ "•..•ESCAPE

U) *" • ENVELOPE

o -18000LL

ci *

S9s 100 106 119 115 120 1'2S 130 13S

S~ RELATIVE TIME (SED)

FIGURE 4.2-6. HORIZONTAL FAKEOUT, TARGET CROt•-ES INFRONT OF TCAS (SHEET 2 OF 2)

45

TARGET TRACK

4 .$ TP.ACK NUMBER:, A6ENCOUNTER DATE: 10/25/84

"" :.I;TIME: 12:56:00

X 3.0

WJ 2.5 TARGET2 .8 3 300 '/ 190 KTS

p2..~ °~~3/SEC:

w TCASuJ1.0 __1_. •3600' /ZOOKTS

.I;K-.-!.•.7NM

-, -4 -3 -z -i z 3 4 5RELATIVE Y (NMI)

BEARING ACCUPACY

c 10 {TRACK I•430SER, A68 " •!4AR•OTC' 18/25084

I- 270 SCC(HOR7. 71M')t 12.4.4,40 to MWIL. vk" POIm7s - 41

ZL RADAR J 4 C -CC.3 DEcr.• . 00 .. - 2| StAHNOE5 DE'V$TIOH .llD€

C. . . 1. DC

,-z 250 _•

SaW 242 *aw

u 10

?S SO 85 98 95 ISO 1eg lie Sig 75 SO IRS is 96 ISO lag its £11RELATIVE TIME (SEC) RELATIVE TIME (SEC)

BEARING RATE ACCURACY

2. •Z 4TRACK HUMBER. A6

*~,0 1.5 SCEHARIO0 DATE, 1S2016,43 cc SCEHARIO TIEt: 152;4:45

a-1- N ~ e IUMBER OF POINTS 41C U), 2 MnEA - 8.096 DEC/SEC

to wo 0 STANDARD DEVIATION 6 .409 DEC,=E

> W ,

wp- -2- Z .

cc-3 ar

(1) 76 s 5; 9i 9 , 165 119 Its 76 SO 8 to 90 s is0 1065 11 Its

C-) RELATIVE TIP., RELATIVE TIME (SEC)

FIGURE 4.2-7. HORIZONTAL FAKEOUT, TARGET CROSSES BEHIND TCAS

(SHEET 1 OF 2)

46

L

I-

U.

ci

W tsees

0 •50884U)0

0W..seee

>

0

C3.

-zoeeE75 86 85 90 95 166 165 lie 115

RELATIVE TIME (SEC)

"1FOUS

LL e 6SCRLOVER TO 450

aX2S889w

a

U),-5809

r- 756B969 5le 0 i t

RELATIVE TINE (SEC)

FIGURE 4.2-7. HORIZONTAL FAKEOUT, TARGET CROSSES BEHIND TCAS(SHEET 2 OF 2)

47

M 1~

accurate targa relative motion to the pilot on the traffic

advisory dis_,ay.

4.3.1 DENSITY MEASUREMENTS

One of the major reasons the TCAS III equipment was flownin the Los Angeles Basin was to monitor its operation in a highdensity environment. Density measurements were calculated over a4-nautical mile circle during each of the 60 encounters. Thenumber of targets under track each second were averaged over 60seconds during each encounter. Then this average was divided bythe ared (50.3 nmi) to give the aircraft density per squarenautical mile. The highest density during the 60 encounters was0.155 aircraft per square nautical mile (or 7.8 aircraft over a 4nmi circle) and the average over all the encounters was 0.08aircraft per square nautical mile or 4.1 aircraft over the 4 nmicircle. The maximum number of targets under track within a 4nautical mile circle in any second during the 60 encounters was13, equivalent to 0.26 aircraft per square nautical mile.

Figure 4.3.1-1 shows two slant range versus time plots.Each is a plot of all target reports received during a 5-minutetime interval in the Los Angeles Basin flight tests. The topplot shows a period of average aircraft density while the bottomone is of a higher density environment. Both plots includereports from both altitude reporting and non-altitude reportingtransponders. Ezch asterisk on the plot indicates a receivedtarget report at a particular time and slant range. Note thatthe asterisks are not evenly spaced in time. This is due to thethreat-driven surveillance feature of TCAS III in that threate-ning aircraft are interrogated more often than non-threateningaircraft.

4.3.2 HIGH DENSITY ENCOUNTERS

Aircrafr encounters between the TCAS III Engineering Modelon the Boeing /?7 and an FAA Convair 580 were flown in the LosAngeles Basin area. Representative encounter geometries areshown in Figure 4.3.2-1. Of the 60 total encounters, 9 werehead-ons, 6 were 20-degree crossovers, 4 were 45-degree cross-overs, 9 were tail-chases, and the remaining 32 were fake-outs.The fake-outs are those encounters where the target aircraft

*• maneuvered in either the vertical or horizontal direction in anattempt to confCse the threat detection and resolution logic.Many ot these encounters have already been flown at the FAA'sTechnical Center in New Jersey, but never in such a high densityenvironment.

Figure 4.3.2-2 shows a typical encounter from the LosAngeles flight tests. This particular encounter is a 20-degreecrossover with a 200-foot vertical separation. The plot showss Ant range versus time for a 5-minute interval around theencounter. fhe track of the encounter aircraft is marked on theplot. Also included in the figure are three snapshots of the

48

AVERAGE AIRCRAFT DENSITY18i I SCENARIO DATEt 02/'0/'9

SCENARIO TIME: t1j351get6 6

14

12 N

a .

a 60 I23

- t 4AG *A

RCTIV E TleIME 20 258 308

~ijj HIGH AIRCRAFT DENSITY

SCENARIO DATE: 82'83'86SCENARIO TIME: 18:28:00

I16

'.14

ikt

z 6

*9z 4 00 a

0 2

co I e ISO2' s 8

u. RELATIVE TIME (SEC)

FIGURE 4.3.1-1. TYPICAL RECEIVED TARGET REPORTS, LOS ANGELESFLIGHT TEST

r4

HEAD-ON TARGET

TCAS 3 00 300'2, 'TS

l000200KTS

200 CROSS-OVER

W 2000" TCAS

Q" 80007200KTS

$200"/200KTS

450 CROSS-OVER

TOP VIEW PROFILE

TARGET - TARGET00[0KT- --o 0

4OO7200KTS / g0007200KTS

/ It 500 FPM/ ~~TCAS •

o5007/180KTS - I NM I \TCAS

ISO00"lOKTS _ _ -

A 3000 "

TAIL CHASE

TARGET•' 1800'1110K, S

"300, TCAS

1500'I240KTS

VERTICAL FAKEOUTTCAS

10.000"7200KTS

I 2100 FPM8000.

TARGET * N

7500"/200KTS 1 1 NA

HORIZONTAL FAKEOUT

STtAS0007200KTS

SI

fTP- 11 4 NM

4. -, NW I I

> TARGET j7700*1200KTS 3 Nm

FIGURE 4.3.2-1. AIRCRAFT ENCOUNTER GEOMETRIES

50

FB-VG-1759.1

13 SCENARIO'DATE. 02103v@S

V ENCOUNTER TRACK SCENARIO TIME: 10:14:0016

lt14 4,

*a• * 4 ••

z 12 *4

w 10 W* a 4443

a:4 * *ý * 4* * 44 A*** **4 t*4A$-t0

(1 (10:15:944

XR E 4 . 2 * YPICA E U "I SPLAYS,4444-t

6 ~64*

* 4'

*444 * -

8 108ISO 288 250 300

RELPTIVE TIME (SEC)ACQUISITION -oL RESOLUTION ADVISORY

(10 14:55) TRAFFIC ADVISORY (10:16:06)14 (10:15 :49)

* ACQUISITION

FIGURE 4.3.2-2. TYPICAL ENCOUNTER DISPLAYS,LOS ANGELES FLIGHT TESTS (SHEET 1 OF 2)

51

FB-VG-1759.2

TRAFFIC ADVISORY

4'-

RESOLUTION ADVISORY

FIGURE 4.3.2-2. TYPICAL ENCOUNTER DISPLAYS,LOS ANGELES FLIGHT TESTS (SHEET 2 OF 2)

52

traffic advisory and relative motion display during the sameencounter. The display shown in this figure is not of thecockpit display as described in section 2.7, but of the displaylocated on rack number 3 along with the rest of the engineeringmodel equipment. This display is of lower resolution than thecockpit display and has slightly different positioning of ther-esolution advisory and Own aircraft information. The firstsnapshot is at acquisition. The open diamond symbol shows theposition of the target relative to Own aircraft. The target wasacquired at 12 nautical miles in slant range with a relativebearing of 40 degrees at 10:14:55. The +2 next to the symbolindicates the relative altitude of the target. In this case, thetarget is +200 feet above Own aircraft. Target symbols without arelative altitude indicate aircraft whose transponders do notcontain Mode C altitude encoding. It became a traffic advisoryat 10:15:49 at 5 nautical miles slant range and 24 degreesrelative bearing (snapshot 2). Note that a traffic advisory isdisplayed as a filled-in circle with a 25-second predictionvector attached. The third snapshot i• at 10:16:06 when aresolution advisory was generated. The threat is at 3 nauticalmiles in slant range and 23 degrees in relative bearing. The"Turn Left" advisory is shown to the right of the display. Thisadvisory was also displayed to the pilot on the cockpit display.The average aircraft density during this encounter was 0.066 air-craft over 4 nautical miles.

Some additional encounters during the Los Angeles testing"are included in the following pages. Figure 4.3.2-3 shows ahead-on encounter. The first plot in the figure shows slantrange versus time for a 5-minute interval around the encounter.The track of the encounter aircraft is marked on the plot.Acquisition of the target was at 16 nautical miles at 9:53:16,approximately 129 seconds before closest point of approach. It"became a traffic advisory at 9:54:46 at 5.1 nautical miles slantrange and 359 degrees relative bearing. A resolution advisory of"Climb" was generated at 9:54:57 when the target was at 3.7 nau-tical miles at 356 degrees. The closest point of approach (CPA)was at 9:55:29 when the target was just 830 feet in slant rangeand 200 feet below the TCAS aircraft. Average density during the

I encount'r was 0.078 or 3.9 aircraft over 4 nautical miles. Thesecond plot in the figure shows relative bearing versus time.The third plot is the target's altitude with own TCAS's altitudeoverlaid on the plot. Each asterisk represents a receiveC targetreport. The last plot in the figure shows the track of the

.4 target relative to Own TCAS. In this example, the resolutionadvisory of "Climb" given to the pilot was not followed.

Figure 4.3.2-4 shows a 45-degree encounter. TCAS acquiredthe target at 15:22:31 at 11.6 nautical miles slant range and 45degrees relative bearing, 114 seconds before CPA. The targetbecame a traffic advisory at 15:23:32 at 4.7 nautical miles and21 degrees. It became a resolution advisory at 15:23:56 at 2nautical miles at 15 degrees. The displayed advisory was "Don'tDescend Faster than 1000 FPM" followed by "Don't Descend Faster

53

SLANT RANGE VS TIME PLOT

SCENARIO DATE: 082'09g2 SSCENARIO 'IM* e: 9: 2:88

16 RA-GE/AHO/Rl.T LIMITED REPLIES

14 *1 ENCOUNTER TRACK

12/ z/z

8I 8

a 6j

4 -ft

s s ies so f 280 260 399

leI RELATIVE TIME (SEC)ACQUISITION TRAFFIC ADVISORY I CPA9:53:16 9:54:46 .9:55:29

RESOLUTION ADVISORY9:54:57

RELATIVE BEARING VS TIME

TRACK NUMBERs 342SCENARIO DATE: 82/8s/OSSCENARIO TIME: 89ts2:,0

w

CDz

w

mcrLd

.ia

Ise -5

0-- -Zee *'... ..

!' ~-208,

aso lee Ise 288 258 388RELAT]VE TIME (SEC)

FIGURE 4.3.2-3. HEAD-ON ENCOUNTER(SHEET 1 OF 2)

54

ALTITUDE VS TIME

S•9090 TRACK HUM DERt 343SC ENARIO D ATE: :2/96/086SSCENARIO TIMEx 99:52 88

8568

OWN ALTITUDE

I-.4

Aw

0

S7688

6see

6886 --6 58 too ISO 289 256 308

RELATIVE TIME (SEC)

TRACK NUMLE,* 343SCENARIO DATE: 82'86'8SSTART TINE: e98:S:08END TIME , 89:57:80DIS Y RANC£* 19.9 naj

xll

@±co'

FIGURE 4.3.2-3. HEAD-ON ENCOUNTER(SHEET 2 OF 2)

55

SLANT RANGE VS TIME PLOT1SCrARI.O DAT• 02/92/gr/

SCENARIO TIMEs 168210@8RANfG'ANG/ALT LIMITED REPLIES

14 " ENCOUNTER TRACK * .

z 12 . '4 *I. "

U./ 8 'J .4,

z 6 f.. 4 zg

a W

4 04

eso toe * tas 250 308I IL PA RELATIVE TIMETRAFFIC ADVISORY 15:24:12 (SEC)

ACQUISITION _ 15:23:32 LRESOLUTION ADVISORY15:22:31 15:23:56

RELATIVE BEARING VS TIMEIS TRACK HUMBER: 617

SCENARIO DATE: 02/e3/85SCENARIO TIME: 15:2•1:88

leg

wo

LDz,-, I8 •

aLiJ

0-- -se 4Ld

-~10

I.4

9, 0 5s lee 158 288 260 388* RELATIVE T]ME (SEC)

FIGURE 4.3.2-4. 45-DEGREE ENCOUNTER(SHEET 1 OF 2)

56

ALTITUDE VS TIMElease

TRACK HUMtEI: 697SSCENARIO DATE A 2/13/85[ SCENARIO TIMEs 15:21S9i

9589

go eeesL ON ALTITUDE

a* •9 -__ •

7909

9o too 169 299 26G 398

REL"TIVE TIME (SEC)

TRACK HUMBER: 637SCENARIO DRTE: 02/83/85STRRT TIME; 15:21:99

/ • END TIML 1 15261 a

.... DISPLAY RAHGEs 14.9 nui

X

cr,(A

FIGURE 4.3.2-4. 45-DECPFE ENCOUNTER(SHEET 2 OF 2)

57

wA E 7 ý )

than 2000 FPM". The average density was 0.155 or 7.8 aircraftover a 4 nautical mile circle.

Figure 4.3.2-5 is an example of a vertical fakeout. Atacquisition (14:42:32, 122 seconds before CPA), the target was at15.5 nautical mil(z slant range, 346 degrees relative bearing,and 1500 feet above Own altitude. During the encounter, thetarget remained at a constant altitude while Own climbed 1000 fpmand leveled off 600 feet below the target. A traffic advisory

'* was generated at 14:43:52 at 5.3 nautical miles, 1 degree and1000 feet above Own altitude. The target became a resolutionadvisory at 14:44:10 at 3 nautical miles, 600 feet above Own air-craft. A "Don't Climb" advisory was displayed.

The last example shown is a horizontal fakeout in Figure4.3-2-6. The encounter began with the trajectories of the t,oaircraft spaced 3 nautical miles in the horfzontal plane. Thetarget aircraft maneuvered until it was within 1/4 mile. Acqui-sition was at 10:00:45 at 12.9 nautical uiles slant range and 20degrees relative bearing, 112 seconds before CPA. The targetremained 200 feet below Own aircraft during the entire encounter.It became a traffic advisory at 10:02:00 at 4.7 nautical milesand 13 degrees. A resolution advisory of "Turn Left" was gener-ated at 10:02:21 when the target was at 2 nautical miles slantrange and 19 degrees relative bearing.

These are just a few examples of the many encounters runin the Los Angeles Basin. The rest of the high densityencounters exhibited similar performance. Table 4.3.2-1 give.ssome summary statistics for all the L.A. encounters. As shown,the target aircraft was acquired with more than sufficient timeto track and assess its collision potential before closest pointof approach. The average slant range at acquisition is slightlymisleading since the TCAS III maximum search range varies as afunction of bearing. In the case of a tailchase encounter whenthe target overtakes the TCAS aircraft, the acquisition rangewill be small since the closing rate is so low.

4.3.3 TARGETS OF OPPORTUNITY

In addition to the preplenned encounters ±n tLre LosAngeles Basin, there were many other aircraft that were trackedby the TCAS. Some of these targets of opportunity generatedtraffic advisories from the TCAS and some even became resolutionadvisories. The importance of these targets is the opportunityto analyze the TCAS performance with different transponders. Allof the planned encounters wet3 run against a specific aircraftand transponder. So, the statistics generated from th.ese encoun-ters really show performance with a specific transponQer. Thetargets of opportunity ;uovide information about a larger cross-section of aircraft transponders.

On the five days ýhlt tne TCAS aircraft flew ir the L.A.

Basin, 34 Mode C targets of opportunity generated traffic

58

SLANT RANGE VS TIME PLOT

SCENARIO DATr: 6z,2'82'8SCENARIO TIME: 14141:38 .•'•q'•'••'•''•'•16 -RANG;E/flHO/ Li LIM lTE ) REPLIES • e

" • e, 14 "%, ENCOUNTER TRACK

I.//%N

"If"" ~ - --1- - *''U) 4;

2

8 ISO 186 208 256 368

ACQUISITION TRAFFIC ADVISORYj I CPA RELAT]VE TIMME (SEC)14:42:32 14:43:52 14:44:34

RESOLUTION ADVISORY14:44:10

RELATIVE BEARING VS TIME

TRACK NUMBERe 966SCENARIO DATE: 92/82/86 *SCENARIO TIMEs 14:4l:38

(D$ e

crcr

Cd

>m- -lg

.--I

,: "-Ise

--1- 1 OF 388

59II -

A7-

ALTITUDE VS TIME

TRACK NUMBE•: 966SCENARIO DATE: t2/82/85SCENARIO TIME: 14:41:30

lee..

908 -

OWN ALTITUDE7e88

e se t1oI ý.jo 26e 380

RELATIVE TIME (SEC)

TRACK NUMBER, 966SCENARIO DATE: 8082'/85/" •: • START TIMEs 14:41:08

/ ". • END TIME 1 14346:39

DISPLAY RANGE: 15.8 nmj

•'•i

)<

FIGURE 4.3.2-5. VER"ICAL FAKEOUT(SHEET 2 OF 2)

60

SLANT RANGE VS TIME PLOT

SCENARIO DATE, 92-8S'G8SCENARIO TIME: 891596EE

16 !RANGE'RNG/ALT LIMITED REPLIES

4ENCOUNTER TRACK

12 z7

aL to~ ..,

44

0so Ise, 1se 206 258 3098

RELATIVE TIME (SEC) I CPATRAFFIC ADVISORY 10:02:37

10:02:00 RESOLUTION ADVISORY

10:00:45 10:02:21

RELATIVE BEARING VS TIME

ttooTRACK NUMBER: 388SCENARIO DATE: 82/OS858

I S O S C E N A R I O T I M E , 9 1 6 9, 5 9, 9 9 *• . a . -

w0*

EDz

2 4S

so

a It

w

wI

I -5

-Jw

Ise

so le5 I5e 299 258 309C RELATIVE TIME (SEC)

FIGURE 4.3.2-6. HORIZONTAL FAKEOUT(SHEET I OF 2)

61

ALTITUDE VS TIME

9800 TRACK NUMBER: 38D

SCENARIO DATE: 02/3S'85SCENARIO TIMEs 89259680

8588

OWN ALTITUDEb-v 880g 0_

0 l

..Jcr 7988

6580

0 so too ISO 2I9 250 309

RELATIVE TIME (SEC)

TRACK MUMZERI 38DSCENARIO DATE: 82.'86/8SSTART TIME: 09:99800END TIME 1 l : G mie4,DISPLAY RANGEI 15.6 nml

c.'J

'I)

I-

FIGURE 4.3.2-6. HORIZONTAL FAKEOUT(SHEET 2 OF 2)

62

La 0u ::.L j -o co CD Q k

c)=( DO3- , L~. L')!

'm J .D Co lC CD) C.Ij 0

:I- V) * .

V) C,(3 c e

1-4.E-

U).

1-0 C' (' 'S - (' '

C1) cm4C,...

'S C.

z< U-f oI

z<CD)

C 0 -. (n , CD LOt

z~ 0L L) -4 - '- CS

V) C

'< I ý- (.,

4/

U) < ..

<- 1-

-. 4'-' ~ ~ 0 0 'S - 4' 4

I~j - <

I V)L) < o C))

Li ~ C\- - U. ... 4) U

(3(63

advisories. Of these 34 targets, 7 targets also becameresolution advisories. There were also quite a number of non-Mode C targets that generated traffic advisories. However, thesetargets are of less interest since their altitudes were unknownand they would never become resolution advisories.

Figure 4.3.3-1 is an example of one target of opportunitythat generated a resolution advisory from the collision avoidancelogic. The first plot is a slant range versus time plot for thefive-minute interval around the encounter. The target wasacquired at 10:29:52 at 13.5 nautical miles slant range, 8degrees relative bearing, and 1100 feet above Own altitude. Itbecame a traffic advisory at 10:30:57 at 5.2 nautical miles, 12degrees relative bearing, and 1000 feet above Own altitude. At10:31:19, the target generated , resolution advisory at 2.4nautical miles slant range, 16 degrees relative bearing, and 700feet above Own altitude. The displayed resolution advisory was"Descend". The closest point of approach occurred at 10:31:37when the target was just 0.7 nautical miles in slant range and+300 feet above Own aircraft. Average density was 0.1 or 5 air-craft over a 4 nautical mile circle. The second plot in thefigure shows the measured relative bearing versus time of thetarget during the entire track. The third plot indicates thetarget's absolute altitude with Own's altitude overlaid on theplot. The last plot shows the target's track relative to Ownaircraft. Each asterisk indicates a received target report.

A second example is shown in Figure 4.3.3-2. This parti-cular target was acquired at 10:24:16, 136 seconds before closestpoint of approach, at 12.1 nautical miles, 43 degrees relativebearing, and 1400 feet below Own altitude. A traffic advisorywas generated at 10:25:53 at 4.3 nautical miles as the targetwas just 100 feet above Own altitude. At 10:26:22, the targetbecame a resolution advisory at 1.2 nautical miles, 600 feetabove Own altitude. An advisory of "Don't Climb" was displayed.CPA occurred at 0.4 nautical miles at 10:26:32. Average densityduring the encounter was 0.063, or 3.2 aircraft over a 4 nauticalmile circle.

Both of these target encounters are examples of unplannedencounters. The first example occurred while the TCAS aircraftwas just flying around the Los Angeles Basin. The second exampleoccurred in the Seal Beach area as the TCAS aircraft was runninghigh density encounters with an FAA target aircraft. In both ofthese examples plus other cases where advisories were generatedby targets of opportunity, TCAS III acquired the target wellahead of CPA, kept a steady trnck on the target and posted areascnable advisory against the target.

S4.4 PERFORMANCE STATISTICS

The performance of TCAS III can be summarized as shown inFigure 4.4-1. Two types of tracks were considered in thisanalysis: encounters and targets of opportunity. The encounters

64

SLANT RANGE VS TIME PLOTIs

SCENRRO DATE: 12-'041'9SCENARIO TIMEs 829•: eRANGE/A/G/ALT LIMITED REPLIES

14 ENCOUNTER TRACK

z00

a

a .* /

cc (j. •. - , .-. 4. •_ So

8iý P 8 26s 390

CPA RELATIVE TIME (SEC)TRAFFIC ADVISORY 10:31"37

ACQUISITION 10:30:57 -RESOLUTION ADVISORY10:29:52 10:31:19

RELATIVE BEARING VS TIMETRACK NUMBER: 410

SCENARIO DATE: 02/04/8SSCENARIO TIME: 10:298o8

1O8

0 .•

.10

z 0

FIUE ..- 1 AGE FOPOTN*Y#

a 'elwm 8 I

w

I-

-leec'j

-IGO8 5 188 ISO 28e 268 380

RELATIVE TIM2 (SEC)

FIGURE 4.3.3-1. TARGET OF OPPORTUNITY,#1(SHEET 1 OF 2)

65

ALTITUDE VS TIME

TRACK NUMJER: 416SCEHARIO DATE: 020/4/8SSCENARIO TIME: 18:29:88

~~out

4500

Sm

so0 to-S 0 6 9END OWME ALTITUD

D I 8 180 2883i• RELATIVE TIME (SEC)

:cN

S~TRACK HUMIER: 418

~SCENARIO DATE: 82'84'85~START TINF: 10:29:08

DISPLAY RANGE: 18.9 flnt

C~jmCD"

I-

FIGURE 4.3.3-1. TARGET OF OPPORTUNITY, #1(SHEET 2 OF 2)

66

SLANT RANGE VS TIME PLOT

SCENARIO DATES 02,,05e&5SCENARIO TIME: 16:23:36

16 RANGE/ANG/gLT LIMITED REPLIES

14

ENCOUNTER TRACKr- 12

Z 6ac4WO

ACQUISITION TRAFFIC ADVISORY L CPA RELATIVE T IME (SEC)

RESOLUTION ADVISORY10:26:22

RELATIVE BEARING VS TIMEISO

TRACK NUMBE~R# 4PASCENARIO D AT E: a82/85/85SCEHNRIO TIME, 180:23:30

lee

L

9 59-loop

CD -s8

a

-LJ

a Jo ie ISO 200 250 300

RELATIVE TIME (SEC>

FIGURE 4.3.3-2. TARGET OF OPPORTUNITY, #2(SHEET 1 OF 2)

67

ALTITUDE VS TIME

999s TRACK HUMIER: 42ASCENARIO DATE: 82/'S'83SCENARIO TIMEs 1s823:3e

- * 44*am

I-- LSeeoc -,

w0 9 -mw*e

SOWN ALTITUDE

6see65eee

6e e is o•1• 288 258 388

RELATIVE TIME (SEC)

TRACK HUMNERr 4BASCENARIO DATE: 02/es/85START TIME: 10:23:38£END TIME : I9t28t3O

DISPLAY RANCEs 10.0 nal

S~FIGURE 4.3.3-2. TARGET OF OPPORTUNITY, #2

(SHEET 2 OF 2)

68

FAA TECHNICAL CENTERSUCCESSFUL TRACK BLIP/SCAN

NUMBER TRACKS (%) CONTINUITY 1%) RATIO (%)OCTOBER 25, 1984

ENCOUNTERS ONLY 13 100 98.0 96.1

TARGETS OF OPPORTUNITY 6 100 100 92.3

NOVEMBER 13, 1984

ENCOUNTERS ONLY 4 100 100 92.0

TARGETS OF OPPORTUNITY 2 100 100 90.0

NOVEMBER 28, 1984

ENCOUNTERS ONLY 6 100 100 89.3

DECEMBER 20, 1984

TARGETS OF OPPORTUNITY 7 100 lGL 93.1

TOTAL 38 100 99.2 93.1

LOS ANGELES BASINSUCCESSFUL TRACK BLIP/SCAN

NUMBER TRACKS (%) CONTINUITY (%) RATIO (%)

FEBRUARY 3, 1985 AM

ENCOUNTERS ONLY 12 100 100 94,2TARGETS OF OPPORTUNITY 5 100 100 89.2

FEBRUARY 3, 1985 PM

ENCOUNTERS ONLY 15 100 100 92.6

TARGETS OF OPPORTUNITY 6 100 97.3 90.0

FEBRUARY 5, 1985

ENCOUNTERS ONLY 24 100 100 94.9TARGET OF OPPORTUNITY 5 100 100 93.6

TOTAL 67 100 99.7 93.3

SUCCESSFUL TRACKS = PERCENTAGE FOR WHICH A TRACK EXISTED 25 SECONDSPRIOR TO CPA

TRACK CONTINUITY = FOR THE 50. SECOND PERIOD PRIOR TO CPAPERCENTAGEOF TIME DURING WHICH A TRACK EXISTED (INCLUDINGCOASTS)

BLIP/SCAN RATIO = LIMITED TO THE 50- SECOND PERIOD PRIOR TO CPA ANDLIMITED TO THE TIME IN TRACK, THIS IS PERCENTAGEOF THE INTERROGATIONS FOR WHICH TARGET REPORTS

LL WERE GENERATED

FIGURE 4.4-1. TCAS III PERFORMANCE STATISTICS

69

were the pre-defined trajectories flown between the ETEU and anFAA target aircraft. Targets which came within 2 nautical milesand +2500 feet in relative altitude of the ETEU at closest pointof approach (CPA) were determined to be targets of opportunity.The computed statistics: successful tracks, track continuity,and blip/scan ratio are as defined in "Active BCAS: Design andValidation of the Surveillance Subsystem,'" Project ReportATC-103, W. H. Harman, R. R. LaFrey, J. D. Welch, M. L. Wood,

A Lincoln Laboratory, MIT, Lexington, MA, 17 December 1980. Suc-cessful tracks indicates the percentage of tracks that existed 25seconds before CPA. Track continuity is the percentage of timeduring which a track existed (including coasts) for the 50-secondperiod before CPA. The blip/scan ratio shows the percentage ofinterrogations for which target reports were generated, limitedto the 50-second period before CPA and further limiteG to thetime under track.

The upper portion of Figure 4.4-1 shows the performancestatistics for 38 ATCRBS tracks, 23 encounters and 15 targets ofopportunity at the FAA Technical Center in New Jersey. The num-ber of successful tracks is 100%, the track continuity is 99.2%and the blip/scan ratio is 93.1%. The performance of the ETEU inLos Angeles in a much higher density environment is shown in thelower p~rtion of the figure. Here, the number of successfultracks is 100%, the track continuity is 99.7% and the blip/scan ratio is 93.3% for 67 ATCRBS tracks. These statistics areslightly better than those compiled at the FAA Technical Center.There was no degradation in performance due to aircraft density.

5.0 CONCLUSION

The major goal of proving the feasibility of horizontalmaneuvers for collision avoidance has been accomplished. Ahighly directional antenna that gives one to two degrees bearingaccuracy is the major reason this is possible. The use of accu-rate bearing measurements adds the extra dimension needed forhorizontal maneuver capability for collision avoidance. In addi-tion, bearing is also used to filter target data and alloo'successful operation in high density environments.

The concept and design of YCAS III has been validatedduring this contract through the development and test of the twoEngineering models. The extensive flight testing proved to beinvaluable in the analysis and evaluation of processing algo-rithms. In addition, many operational features of the TCAS IIIequipment have been refined due to observations by the testpilots and FAA Technical Center personnel. it has been pioventhat the TCAS III's accurate bearing measurement capability per-mits the issuance of horizontal in addition to vertical resolu-tion advisories and reddces the system's susceptibility to inter-ference, allowing full operation in high traffic densityenvironments.

70

I ----U- -----

APPENDIX A

MANUALS

BCD-TR-076 TCAS III Tnstruction Manual Jul 1986

BCD-TR-099 TCAS IIT Computec Progr-am Documenta- Jun i1986tion, Volumes I-:3

BCD-rR-143 TCAS Ill Operator's Manual Jul 198b

BCD-TR-144 TCAS III Cockpit Display Manual. Jul 1986

REPORTS

BCD-TR-053 Task-I Factory Performance Test Plan Jan 1982for Scalar Mode Computer Prooram

BCD-TR-058 Task-I Factory Performance Test Apr 1982Report, Volumes I-II

13CD-TR-063 TCAS II Engineering Model Design Apr 1982Specification (Model A and Model B)

BCD-TR-069 TCAS II Model A TEU Installation Nr- 1982Plan

BCD-TR-070 Addendum to Task-I Factorv Sep 1982Performance Test Report

BCD-TR-071 TCAS Ii Model A TEU Factory Nov 1982Accept•,,ce Test Plan

BCD-Th -073 TCAS II Model A TEU Factory Accept- Nov 1982Test Results Functional Thsts

BCD-TR-073 Addendum to TCAS II Model A TEU Feb 1983Factory Acceptance Test ResultsFunctional Tests

RD--TR-074 TCAS II Model A and Model B Field May 1983Acceptance Test Plan

BCD-TR-084 TCAS TI Model B TEU Factory Accept- 2un 1963ance Test Plan

BCD-TR-G90 Enhan:ed ''CAS II Antenna Character- Oct 1983isticF Percr':

A-1

REPORTS (Continued)

BCD-TR-092 Simulation of Enhanced TCAS II Feb 1984Signal Processor Hardware Design,Volumes 1-2

BCD-TR-098 TCAS TIT System Summaty Apr 1985

TECHNICAL NOTTS

BCD-TN-81-013 Systems Considerations Mar 1981FB-467-001 P. Kerrian

BCD-TN-81-012A Preliminary Estimate of the Effect of Mar 1981FB-467-002A Full BCAS on ATCRBS Round Reliability

A. I. Sinsky, P. Kernan

BCD-TN-81-014 Mode Selection Mar 1981FB-467-003 M. B. Covington

BCD-TN-81-015 Full BCAS Antenna Transfer Function Mar 1981FB-467-004 Algorithms

A. I. Sinsky, M. V. Wilhelm

BCD-TN-81-016 Miss Distancp Estimation from Two Mar 1981FB-467-005 Range, Bearing Measurements

T. L. Gabriele

BCD-TN-81-023A Target Motion Model Jun 1981FB-467-006A A. I. Sinsky

BCD-TN-81-024 Minimum Variance Estimation May 1981FB-467-007 A. I. Sinsky, E. A. Lew

BCD-TN-81-025 Parallel Path Turn-in Escape Time May 1981FB-467-008 T. L. Gabriele

BCD-TN-81-026 Notes from FAA-ER-250-1 May 1q81FB-467-009 J. W. Martin

BCD-TN-81-028 Self Reply Blockaoe Effects on Pas- May 1981FB-467-010 sive BCAS in an ATCRBS Environment

A. D. McComas

BCD-TN-81-032 Garble Analysis Jun 1981FB-467-011 M. A. Martin

BCD-TN-81-033 Alpha-Beta Tracking Errors foi Jun 1981FB-467-012 Orbiting Targets

A. I. Sinsky

A-?

STECHNICAL NOTES (Continued)

BCD-TN-81-036 Effects o' Unequal Update Intervals Jul 1981

FB-467-013 on Alpha-Beta Tracker PerformanceD. Silber, A. J. Spuria

BCD-TN-81-042 Full BCAS Covariance Analysis Aug 1981X FB-467-015 E. A. Lew, C. J. Smearman

B -TN-81-044 Interference to DABS Squitter Sep 1981FB-467-016 Detection in L.A. Basin

P. Kernan, A. I. Sinsky

SBCD-TN-81-054 TCAS Monopulse Equivalent Beamwidths Oct 1981

FB-467-017 C. J. Smearman

BCD-TN-81-055 Analysis of the Effects of Angle Nov 1981FB-467-018 Errors on Miss Distance Estimate

D. Silber

BCD-TN-81-056 Detection of DABS Squitters Nov 1981FB-467-019 D. C. Hecht

BCD-TN-81-057 Filter Requirements for TCAS Nov 1981FB-467-020 Interrogator Transmitter

A. I. Sinsky

BCD-TN-81-058 Monopulse Processor Analysis Nov 1981FB-467-021 A. I. Sinsky, E. A. Lew

BCD-TN-81-061A Preliminary Evaluation of MITRE Feb 1982FB-467-022A Hit or Miss Algorithms

S. J. Guarino

BCD-TN-82-01i Interference Analysis for Collision Feb 1982FB-467-023 Avoidance Systems in the Los

Angeles Basin from 1985 to 1995P. Kernan

BCD-IN-82-017 Souinted Beam Monopulse Processor Apr 1982FB-467-024 A. I. Sinsky

BCD-TN-82--220 Coordinate Transformation Algorithms May 1982FB-467-025 A. I. Sinsky, R. J. Tier

BCD-TN-82-021 Uncompensated Bearing Errcr May 1982FB-467-026 A. I. Sinsky, R. J. Tier

BCD-TN-82-035 System Angle Accuracy Prediction Jul 1982FB-467-027 A. I. Sinsky, R. J. Tier

BCD-TN-83-004 TCAS Electronic Beam Steering Unit Apr 1983"FB-467-028 Troubleshooting Procedure

A. I. Sinsky, R. J. Tier

A-3

___

TECHNICAL NOTES (Continued)

BCD-TN-83-008 TCAS Antenna Patterns Jul 1983

FB-467-029 A. I. Sinsky, R. J. Tier, K. Sirageldin

BCD-TN-84-004 Miss Distance Prediction Algorithm Feb 1984FB-467-030 A. I. Sinsky

BCD-TN-84-011 Logic Error Detection in PL/M Source Jul 1984FB-489-031 Code

J. C. Sipos

BCD-TN-85-005 CRC Generation for EPROM Based Jun 1985FB-489-032 Computer Systems

J. C. Sipos

DESIGN REVIEWS

Full BCAS Design Management Plan and Feb 1981Monthly Management Report

Full BCAS Preliminary Design Review, Apr 1981

Volume 1-2

Full BCAS Design Review May 1981

Full BCAS Design Review Jun 1981

Full BCAS Critical Design Review Sep 1981for the Engineering Model Task I

TCAS II Bimonthly Review Dec 1981

TCAS II Bimonthly Review Mar 1982

TCAS II Bimonthly Review Jul 1982TCAS II Bimonthly Review Oct 1982

TCAS II Bimonthly Review Dec 1982

Enhanced TCAS II Critical Design Apr 1983Review for Model B TEU

Enhanced TCAS II Bimonthly Review Sep 1983

Enhanced TCAS TI! Bimonthly Review Dec 1983

Enhanced TCAS Ii Bimionth-ly Review Feb 1984Revised Mar 1984

Enhanced TCAS II Bimonthly Review Jun 1984System Review / Computer Programming

A-4

DESIGN REVIEWS (Continued)

Enhanced TCAS II Bimonthly Review Nov 1984

MONTHLY MANAGEMENT REPORTS

Monthly Management Report Jan 1981

Monthly Management Report Feb 1981

Monthly Management Report Mar 1981

Monthly Management Report Apr 1981

Monthly Management Report May 1981

BCD-TP-039 Full BCAS Monthly Management Report Jun 1981

BCD-TR-042 Full BCAS Monthly Management Report Jul 1981

BCD-TR-045 Full BCAS Monthly Management Report Aug 1981

BCD-rR-048 TCAS II Monthly Management Report Sep 1981

BCD-TR-049 TCAS II Monthly Management Report Oct 1981

BCD-TR-050 TCAS II Monthly Management Report Nov 1981

BCD-TR-051 TCAS II Task I Monthly Management Dec 1981Report

BCD-TR-052 TCAS II Task II Design Management Dec 1981Report and Monthly Management Report

BCD-TR-054 TCAS II Task I Monthly Management Jan 1982Report

BCD-TR-055 TCAS II Task II Monthly Management Jan 1982Report

BCD-TR-056 TCAS II Task I Monthly Management Feb 1982Report

BCD-TR-057 TCAS IT Task II Monthly Management Feb 1982Report

BCD-TR-059 TCAS II Task I Monthly Management Mar 1982Report

BCD-TR-061 TCAS II Task I Monthly Management Apr 1982Report

A-5

MONTHLY MANAGEMENT REPORTS (Continued)

BCD-TR-062 TCAS II Task 1I Monthly Management Apr 1982Report

BCD-TR-064 TCAS HI Task !I Monthly Management May 1982Report

BCD-TR-065 TCAS Il Task II Monthly Management Jun 1982Report

BCD-TR-066 TCAS II Task II Monthly Management Jul 1982Report

BCD-TR-067 TCAS II Task II Monthly Management Aug 1982Report

BCD-TR-068 TCAS II Task II Monthly Management Sep 1982Report

BCD-TR-072 TCAS II Task II Monthly Management Oct 1982R'oort

BCD-TR-075 TCAS II Task IT Monthly Management Nov 1982Report

BCD-TR-077 TCAS II Task II Monthly Management Dec 1982Report

BCD-TR-078 TCAS II Task II Monthly Management Jan 1983Report

BCD-TR-079 TCAS II Task II Monthly Management Feb 1983Report

BCD-TR-080 TCAS II Task II Monthly Management Mar 1983Report

BCD-TR-082 TCAS II Task II Monthly Management Apr 1983Report

BCD-TR-083 WCAS II Task II Monthly Management May 1983Report

BCD-TR-085 TCAS II Task II Monthly Management Jun 1983Report

BCD-TR-086 TCAS II Task II Monthly Managenient Jul 1983Report

BCD-TR-087 TCAS II Task II Monthly Management Aug 1983Report

A-6

MONTHLY MANAGEMENT REPORTS (Continued)

BCD-TR-088 TCAS II Task II Monthly Management Sep 1983Report

BCD-TR-089 TCAS II Task II Monthly Management Oct 1983Report

BCD-TR-091 TCAS II Task II Monthly Manaqement Nov/Dec 1983Report

BCD-TR-094 TCAS II Task II Monthly Management Jan 1984Report

BCD-TR-095 TCAS II Task II Monthly Management Feb 1984Report

BCD-TR-097 TCAS II Task II Monthly Management Mar/Apr 1984Report

BCD-TR-100 TCAS II Task II Monthly Management May/Jun 1984Report

BCD-TR-I15 TCAS II Task II Monthly Management Jul 1984Report

BCD-TR-117 TCAS II Task II Monthly Management Aug/Sep 1984

IBCD-TR-123 TCAS II Task !I Monthly Management Oct 1984/Report Jan 1985

BCD-TR-123 TCAS II Task I! Monthly Management Feb/Apr 1985Report

BCD-TR-135 -CAS II Task II Monthly Management May/Jul 1985

*U.S. Governient Printing Iffice 1997 - 22-C,49/60392

A-7