A.R. Paradis 6 September 1977

Lincoln LaboratoryMASSACHUSETTS INSTITUTE OF TECHNOLOGY

LEXINGTON, MASSACHUSETTS

Project ReportATC-77

L-Band Air-to-Air Multipath Measurements

A.R. Paradis

6 September 1977

Prepared for the Federal Aviation Administration,

Washington, D.C. 20591

This document is available to the public through the National Technical Information Service,

Springfield, VA 22161

This document is disseminated under the sponsorship of the Department of Transportation in the interest of information exchange. The United States Government assumes no liability for its contents or use thereof.

1. Report No.

FAA-RD-77-87

2. Government Accession No.

Technical Report Documentation Page

3. Recipient's Catalog No.

4. Title and Subtitle

L-Band Air-to-Air Multipath Measurements

7. Author( s)

A. R. Paradis

9. Performing Organi zotion Name and Address

Massachusetts Institute of TechnologyLincoln LaboratoryP.O. Box 73Lexington, MA 02173

12. Sponsoring Agency Name ond Address

Department of TransportationFederal Aviation AdministrationSystems Research and Development ServiceWashington, DC 20591

15. Supp lementary Notes

5. Report Date

6 September 1977

6. Performing Orgonization Code

8. Performing Organi zation Report No.

ATC-77

10. Work Unit No.

Proj. No. 034-241-012

11. Controct or Grant No.DOT-FA77-W Al-727

13. Type of Report ond Period Covered

Project Report

14. Sponsoring Agency Code

.~.

The work reported in this document was performed at Lincoln Laboratory, a center for research operatedby Massachusetts Institute of Technology under Air Force Contract F19628-76-C-0002.

16. Abstroct

A series of air-to-air earth-scattered L-Band multipath measurements are describedand experimental results presented. During these measurements RF pulses were transmitted between two instrumented general aviation aircraft flying coaltitude, diverging pathsover a variety of terrain and water surfaces. Multipath data was collected over grazingangles ranging from ~5° to ~75°. The objectives of these measurements were the:

1. Characterization of the multipath environment in which beacon basedairborne colliSion avoidance (BCAS) equipment would operate.

2. Investigation of the merits and limitations of various degrees of antenna diversity in the rejection of multipath.

.... --

17. Key Words

antenna diversitygrazing anglemultipathRayleigh distributionscattering

18. Distribution Stotement

Document is available to the public throughthe National Technical Information Service,Springfield, Virginia 22151.

19. Security Clouil. (of this report)

Unclassified

20. Security Clossif. (of this poge)

Unclassified

21. No. of P oges

100

Form DOT F 1700.7 (8 -72) Reproduct ion of compll'tcd pagf' authori zed

..

TABLE OF CONTENTS

Section

Part 1 - Overview and Summary

1. OVERVIEW

2. SUMMARY OF MULTIPATH DATA

3. CONCLUSIONS

Part 2 - Experiment Description and Results

4. DESCRIPTION OF THE EXPERIMENT

4.1 Flight Operations

4.2 Data Acquisition

Page

1

2

3

10

11

12

12

13

13

19

23

28

28

50

59

66

68

79

89

91

92

Preliminary Flights

System Description

4.2.1

4.2.2

4.3 Data Reduction

5. EXPERIMENTAL RESULTS

5.1 Ocean Surface (Sea State 1)


5.3 Frozen Lake Surface

5.4 Calm Lake Surface

5.5 Smooth Land Surfaces

5.6 Rough Land Surfaces

5.7 Banking Over Ocean Surface

Acknowledgment

References

iii

Figure

2.1

4.1

ILLUSTRATIONS

Mu1tipath-to-Signa1 Ratio Variation With Terrain.

Effect of Antenna Diversity on Ocean Scattered Mu1tipath

(Log-Video Photographs).

4.2 Effect of Antenna Diversity on Land Scattered Mu1tipath

Page

5

15

16(Log-Video Photographs).

4.3 Comparison of Land and Water Scattered Mu1tipath (Log-Video

Photographs With Long Time Scale). 17

4.4

4.5

4.6

5.1

5.2

~5.3

5.4

5.5

5.6

Airborne Measurement Facility (AMF). 20

Data Acquisition Crosslink Signals. 22

Mu1tipath Experiment Pulse Data Dump. 25

Ocean Scattered Mu1tipath (Log-Video Photographs). 29

Ocean (Sea State 1) Mu1tipath Time-Raster. 31

Mu1tipath Delay. 33

Ocean (Sea State 1) Mu1tipath Instantaneous Signal Strength. 35

Normalized Autocovariance (Ocean Sea State 1). 36

Signa1-to-Mu1tipath Ratio Cumulative Distribution (Ocean

Sea State 1).

5.7(a-b) Ocean (Sea State 1) Multipath Average Signal Strength.

37

39

5.7 (c) Ocean (Sea State 1) Multipath Average Signal Strength

(Top-To-Bottom Link). 40

5.8 Signa1-to-Mu1tipath Ratio, Antenna Gain, and Total Scattering

Loss (Ocean Sea State 1).

5.9 Signa1-to-Mu1tipath Ratio Distribution Variations (Ocean

Sea State 1).

42

44

5.10 Altitude Variations in Signa1-to-Mu1tipath Ratio (Ocean

Sea State 1).

v

ILLUSTRATIONS (CON'T)

5.13

5.14

Figure

5.11 Geometrical Dependence of Signal-to-Multipath Ratio (Ocean

Sea State 1).

5.12 Signal-to-Multipath Ratio Cumulative Distribution (Ocean

Sea State 2).

Ocean (Sea State 2) Multipath Average Signal Strength.

Signal-to-Multipath Ratio, Antenna Gain, and Total

Scattering Loss (Ocean Sea State 2).

5.15 Signal-to-Multipath Ratio Distribution Variations (Ocean

49

52

54

55

67

69

70

71

63

64

58

61

62

60

57

Normalized Autocovariance (Frozen Lake).

Frozen Lake (Low Grazing Angle) Multipath Instantaneous

Signal Strength.

Frozen Lake (Low Altitude) Multipath Average Signal

Strength.

Frozen Lake (High Altitude) Multipath Average Signal

Strength.

Calm Lake Multipath Instantaneous Signal Strength.

Desert Multipath Time-Raster.

Desert Multipath Average Signal Strength.

5.23

5.24

5.25

5.21

5.19

5.20

5.18

Sea State 2).

5.16 Altitude Variations in Signal-to-Multipath Ratio (Ocean

Sea State 2).

5.17 Geometrical Dependence of Signal-to-Multipath Ratio (Ocean

Sea State 2).

Frozen Lake (High Grazing Angle) Multipath Instantaneous

Signal Strength.

5.22

\

vi

Figure

5.26

",

ILLUSTRATIONS (CON'T)

Signal-to-Multipath Ratio, Antenna Gain, and Total

Scattering Loss (Desert).

5.27 Signal-to-Multipath Ratio Distribution Variations (Desert,

73

75High Grazing Angles).

5.28 Signal-to-Multipath Ratio Distribution Variations (Desert,

Low Grazing Angles). 76

5.29 Altitude Variations in Signal-to-Multipath Ratio (Desert). 77

5.30 Flat Plain Multipath Average Signal Strength. 78

5.31 Normalized Autocovariance (Smooth Land). 80

5.32 Geometrical Dependence of Signal-to-Multipath Ratio (Desert). 81

5.33 Geometrical Dependence of Signal-to-Multipath Ratio (Flat

5.34

5.35

5.36

5.37

5.38

5.39

Plain).

Rough Land Multipath Time-Raster.

Signal-to-Multipath Ratio Distribution Variation (Rough

Land).

Rough Land Multipath Average Signal Strength.

Snow Covered Rough Land Multipath Average Signal Strength.

Forested Mountain Multipath Average Signal Strength.

Multipath Sensitivity to Banking.

vii

82

83

85

86

87

88

90

Table No.

2.1

4.1

TABLES

Summary of Mu1tipath Data

Scattering Surfaces Investigated

viii

4

12

"', II

PART 1: OVERVIEW AND SUMMARY

1

1. OVERVIEW

An airborne collision avoidance system employing Air Traffic Control (ATC)

beacon transponders is under development by the Federal Aviation Administration.

This system (called BCAS for Beacon Collision Avoidance System) will be capable

of operating with the present Air Traffic Control Radar Beacon System (ATCRBS)

transponders as well as those of the future Discrete Address Beacon System (DABS).

The occurrence of terrain induced multipath on the air-to-air link is of

considerable interest in the design of the BCAS system. Previously published

multipath data was judged inadequate to support the BCAS design due to incorrect

geometry or incomplete documentation. For this reason, Lincoln Laboratory

performed a study of air-to-air multipath based on actual field measurements

using a pair of instrumented general aviation aircraft. Aircraft geometries,

RF frequency, scattering surfaces, and antenna configuration were all selected

to have maximum relevance to the BCAS design effort.

A summary of the key findings of this study is presented in Chapters 2

and 3. Details of the experimental data collection and results are provided

in Chapters 4 and 5.

2

2. SUMMARY OF MULTIPATH DATA

Detectable multipath echos were observed from all reflection surfaces over

which measurements were conducted. Observations were made as to the multipath

delay, waveform, short-term variability, power levels, dependence on geometry,

dependence on reflecting surface, and the performances of various antenna con-

figurations in rejecting multipath. Data collected over similar surfaces on

different days, separated by as much as a year, exhibited striking consistency

with regard to each multipath parameter.

The data presented refers to the direct and multipath receptions including

the effects of the aircraft antenna patterns. The primary results are sum-

marized in Table Z-l, Fig. Z-l, and in the following paragraphs.

Delay. When compared with the "signal" (namely the received pulse

which traveled the direct path from transmitter to receiver), the echo was

in every case delayed by an amount which agrees with the simple geometrical

formula

Delay ~ R(secant G - l)/c

where R is the air-to-air range, c is the speed of light, and grazing angle

G is

-1G = Tan «AI + AZ)/R)

in terms of the altitudes Al and AZ of the two aircraft above the level of

the reflecting surface.

Multipath Waveform. The waveform of the muitipath echo is determined

primarily by the roughness of the scattering surface. When observed at the

output of a log-video amplifier, the waveform may be described as consisting

3

TABLE 2.1

SUMMARY OF MULTIPATH DATA

Multipath Reflecting SurfaceProperty Ocean I Frozen Lake I Smooth Land (Desert Flat PIa in) Roul!h Land (sub. areas Mtns.)-

Multipath Waveform Low Noise Level ofResulting from Pulselike Long Duration ( ~ 35 ~s)0.5 ~s RF Pulse

Delay Time Closely Follows Theoretical Curve: R(secant(G)-l)/c **K

Short Term Variability -DdB -2dB -13dB(10% to 90%)

CorreIa t ion * <50 msec -1 sec <50 msecTime

Multipath to Signal As highRatio (Bottom-to- as -OdBBottom Antennas, less --5dB --15dB --20dBMedian Value, severeGrazing Angle - for rough30°) ocean

Signal to HultipathRatio Improvement Significant (-15 dB for grazing angles >20°) lessDue to slngli*~op improvement at lower grazing anglesAntenna Link

*Defined in more detail in the text.**R range

G = grazing anglec = speed of light

*** The improvement is greater for the top-to-top link.

(

10r-----------.-------------,-------------,

Bottom-to-Bottom Antenna ink

Ocean (Sea S ate I, Alt: 9500 ft.)

90

Ot-----------.-.~..::":_~~.:-::.:::.-:.-:.:::_'::-::'.:-..-~L-----J-------------1"... '.. "., ".

: '~I \ Ocean (Sea Sate 2, Alt: 12000 ft.), ",--,~

-10 I------;-T~~----+----""::-=\r:'~'-.._...:..Jt."""=_+ ~

'~.... ..... , .....- ..

\ ...", -'.•(All. : 10000 f~,)- 20 +---~~9Ik=__--__:t-....:.-=-......;:1Iir____:,,~---r--r-......;r-------

,-" ..(Suburban Boston, Alt.1\,

9500 ft.~',

"- 30 l--.-,r-!.~'.LL!.!..,!...!...!...!/.J.'.L~'.LLJ.~~...--.....--~.....--4--.,..-~-.......-~.....-----J

60

Grazing Angle (deg.)a.

o,------------,.----------r-------.------rTop-to-Bottom Antenna Lin

-10 r--------:,:-;..:-,----+----------+_--------~I ..

I ..

I '" .....,~'l r ....

.....J \ - ....- 20 r---#rt:~0"'l'....:.....;.~.....:::~---------+----------l

Grazing Angle (deg.)b.

Oi-----__-r---------,----------.Top-to-Top Antenna Link

- IOi----------jf---------+-----------J

9060Grazing Angle (deg.)

c.

30- 30 r----,r_.....:.,~_L.;~4..£!.

o

-20r---75'~~:_"";:__--+_--------+----------l

Fig. 2.1. Multipath-to-Signal Ratio Variation With Terrain.

5

of two components: a delayed slightly distorted pulselike replica of the

direct pulse signal followed by a low-level noiselike waveform lasting tens

of microseconds. In cases of flights over relatively smooth surfaces, such

as ocean or desert, the pulselike component of multipath dominates the other

component by many dB. For these surfaces the peak received multipath power

is the strongest relative to the direct signal level. In other cases, i.e,.

over more ordinary land regions which are relatively rough, the pu1se1ike

component is not present and the multipath waveform consists totally of a

noiselike level of long duration which is slightly stronger than the noise

level component associated with smooth surfaces, but still extremely small

with respect to the direct signal.

Short-Term Variability. The mu1tipath measurement was repeated at the

rate of 20 per second. Multipath echos in successive measurements were com

pared to assess the short-term variability over 10-second periods (which

is a short enough period that aircraft altitudes, orientations, and the

air-to-air range are reasonably constant). In almost all cases the mu1tipath

power varied greatly during the 10 second period, with the span between

10 percentile and 90 percentile being 10 to 15 dB. The power distribution

was analyzed and found to be in reasonable agreement with a "Rayleigh Model"

(in most cases). The power distribution of the Rayleigh Model is that

which results when amplitude A has a Rayleigh distribution and when received

multipath power in dB is 20 log A. In this model, the span between 10

percentile and 90 percentile is 13.4 dB.

6

...

..

Statistical autocovariance calculations were carried out, and these

showed that, in all cases agreeing with the Rayleigh Model, successive multi-

path samples are essentially uncorrelated. That is, the correlation time of

the multipath variability is less than 50 ms (the measurement repetition

period).

A few exceptions to this behavior were noted when flying over certain

very smooth surfaces, such as small inland lakes on a windfree day or when

frozen, or over larger bodies of water when frozen. In these cases the 10 to

90 percentile power variabilities were much lower -- on the order of 2 dB.

Autocovariance calculations showed the correlation time to be on the order of

1 sec in these cases.

Power Level. The power levels of measured multipath echos are summarized

in Fig. 2-1. The figure shows the median value of relative multipath power,

given as the multipath-to-signal ratio (MSR), plotted as a function of grazing

angle G. Evidently, under worst case conditions the median MSR can be as high

as about 0 dB. In these cases, about 50% of multipath echos exceed the direct

signal power, and about 10% of multipath echos exceed a level 5 dB above the

signal.

Dependence on Grazing Angle. Figure 2-1 shows directly the dependence

on grazing angle. Certain trends are evident. For example, as grazing angle

increases above 30°, multipath power levels drop relative to the signal. As

°grazing angle decreases below 30 ; and for a bottom-antenna-to-bottom-antenna

link, ocean reflections and desert reflections exhibit opposite trends -- the

former decreasing while the latter is increasing. Very low values of G, below

about 100

, are relatively unimportant in the context of BCAS.

7

Dependence on Altitude. The multipath data have been analyzed

to assess the dependence on altitude. Results show that MSR does have an

altitude dependence, and that this consists primarily of the dependence

on grazing angle discussed above, with very little additional dependence

on altitude. Thus, for example, if aircraft A and B are both at one altitude

and aircraft C and D are both at twice that altitude, then provided the range

between C and D is twice the range between A and B, the two pairs of aircraft

will experience approximately the same multipath-to-signal ratio. Because

of this property, the format of Fig. 2-1 serves as a convenient summary of

lUulitpath effects over a range of altitudes.

Dependence on Aircraft Antennas. Aircraft antenna patterns playa major

role in determining the relative level of received multipath since the effect of

antenna pattern variations must be considered at both ends of the link. The

data indicates that a bottom-to-bottom antenna link has gain variations that

amplify the ground scattered multipath signal strength while reducing the

direct signal strength, an expected result considering measured aircraft

antenna patterns. The use of a single.top mounted antenna in the link

results in.si~nificant (15 dB) reductions in received multipath at high grazing

angles (100 < G < 75°) with less multipath rejection observed at smaller

grazing angles. The top-to-top antenna link reduces multipath levels still

further for grazing angles above about 100. For very low grazing angles,

where multipath reduction would be especially beneficial, the additional MSR

reduction due to the second.top antenna is not very significant.

Dependence On Reflecting Surface. Figure 2-1 and Table 2-1 serve as

a summary of the dependence on reflecting surface. The echos from land

surfaces are. with few exceptions, found to be appreciably weaker than echos

8

from water. Mu1tipath over the ocean is seen to depend on sea state, with

stronger mu1tipath occurring on calm days. Among the surfaces summarized

in Fig. 2-1, the ocean in Sea State 1 generally stands out as being the

worst-case producing strong echos with approximately 0 dB median MSR over

a broad range of grazing angles.

9

1.

3. CONCLUSIONS

Multipath scattered from smooth surfaces, especially water

surfaces is a significant form of interference on the air

to-air channel.

2. Employment of top-mounted antennas appears to be warranted in

preventing strong multipath from interfering with BCAS operation.

a. A single top-mounted antenna in the link appears to provide

significant multipath rejection for grazing angles above

_100 which includes almost all geometries of interest

in BCAS.

b. At lower grazing angles, the use of a top mounted antenna

on each aircraft provides little additional multipath rejec

tion.

10

..

PART 2: EXPERIMENT DESCRIPTION AND RESULTS

11

4. DESCRIPTION OF THE EXPERIMENT

4.1 Flight Operations

In order to gather multipath data over a wide spectrum of scattering

surface types, flights were conducted over several oceanic and CONUS regions.

Table 4.1 lists the surface conditions and their locations which were investi

gated in the measurement program.

TABLE 4.1

SCATTERING SURFACES INVESTIGATED

..

SURFACE TYPE LOCATION

Ocean (sea state 1)

Ocean (sea state 2)

Frozen snow covered lake

Lake (sea state 0)

Desert

Flat Plain

Rough Land surface

Snow covered rough surface

Forest covered mountains

Massachusetts Bay

Massachusetts Bay

Lake Champlain, Vermont; and

Lake Cochituate, Massachusetts

Lake Cochituate, Massachusetts

Mojave Desert, California

Central Kansas

Massachusetts Suburbs

Massachusetts Suburbs

White Mountains, New Hampshire.-------- -l- --'

12

These surfaces are typical of many that would be encountered on an

air-to-air channel over which BeAS equipment would operate. Data gathered

from these areas show marked differences in both received multipath signal

strength and time structure.

Several data runs were flown at various altitudes over each surface. Most

data runs consisted of two aircraft flying a V shaped divergent flight path.

Both aircraft flew at the same altitude during each data run. As each pass

was being flown, the headings of the aircraft were varied randomly. These

random changes in heading tended to wash out the variations in multipath due

to the fine structure of the antenna patterns associated with the antennas used.

Several times during the flights large heading changes (> 100

) were made to

determine the effect of aspect angle on the received multipath. During one

flight, data was recorded as t~e two aircraft converged on each other.

4.2 Data Acquisition

4.2.1 Preliminary Flights

During the determination of a meaningful sampling scheme, several flights

were conducted to gather data on the time structure of multipath signals scat

tered in response to pulse transmissions. During these flights two equal alti

tude aircraft were flown parallel to one another at a range of two nautical miles

and at an altitude of 7500 feet. Pulses at a 63 dBm level and 1030 MHz fre

quency were transmitted from a bottom mounted L-Band blade antenna on one of the

13

aircraft. Direct and multipath signals were received via both top and

bottom mounted antennas on the receiving aircraft. Photographs were taken

of the analog signals at the outputs of log-video detectors.

The data from these flights indicated that the multlpath signals in

response to a pulse transmission can exist in one of two distinct forms

depending on the roughness of the surface. Fig. 4.1 and Fig. 4.2 show

log-video photographs taken during the flights. In all the photographs the

pulse on the left is the direct pulse transmitted from a bottom mounted

antenna. The captions indicate the location of the receiving antenna. The

signal to the right of the direct pulse in each photo is the multipath scat

tered from the surface. Fig. 4.1 shows that the multipath from a smooth

(Sea State 1) ocean surface had the appearance of a distinct and strong pulse

with a width somewhat greater than that of the direct pulse. Later data

verified that this was typical of signals scattered from smooth surfaces in

general. Fig. 4.2 shows similar photographs taken over a rough land surface.

Here the multipath signals were not pulselike but rather appeared as low

level noise of long duration. Again, such signals proved to be typical of

multipath scattered from a rough surfa c such as forest land or a residential

area.

Fig. 4.3 shows log-video photographs taken during later flights illu

strating the energy distribution of the multipath signal over long (-SO ~sec)

time intervals. During these flights an interrogation was sent from one

aircraft to a second aircraft which responded with a four pulse reply.

14

DI RECTPULSE

MULTI PATHSIGNAL

I~--::-·7-;·" _.-

RECEIVED ON UPPER ANTENNA

RECEIVED ON LOWER ANTE NNA

OCEAN (SEA STATE 1

ALTITUDE: 7500 ftRANGE: 2 nmiGRAZING ANGLEI

51°

Fig. 4.1. Effe.ct of Antenna Diversity on Ocean Scattered Multipath(Log-Video Photographs).

15

DIRECT MULTI PATHPULSE SIGNAL

..

-----.

ROUGH LANDALTITUDE: 7500ftRANGE: 2.5 nmiGRAZING ANGLE:

470

RECEIVED ON UPPER ANTENNA

I_.. _------

RECEI VED ON LOWER ANTENNA

Fig. 4.2. Effect of Antenna Diversity on Land Scattered Multipath(Log-Video Photographs).

16

WATER (SEA STATE l)

, BOTTOM TOPTRANSM ITTED ON, ANTENNA ANTENNA

-'-- ~

RECEIVED ON:

TOP ANTENNA

BOTTOM ANTENNA

ROUGH LAND

BOTTOM TOP

~I ~50fLsec

117.5 dB

T

•

INTERROGATION

8 BACKSCATTERREPLY 8 FOWARD

SCATTEREDMULTIPATH

INTERROGATION REPLY 8 FORWARD8 BACKSCATTER SCATTERED

MULTIPATH

LOG VIDEO PHOTOGRAPHS

Fig. 4.3. Comparison of Land and Water Scattered Multipath (Log-VideoPhotographs With Long Time Scale).

The first two reply pulses were transmitted from a bottom antenna while the

second two were transmitted from a top antenna. The bottom trace in each

photograph shows signals received on a bottom antenna while the top trace

shows signals received on the top antenna. The left most pulses in each

trace represent the two interrogation pulses. On each bottom trace, back

scattered multipath was observed directly following each interrogation

pulse. Note that the backscatter is stronger by approximately 6 dB for the

rough land surface than for the relatively smooth water surface as would be

expected. Scanning the pictures from left to right, the third and fourth

pulses (spaced 55 ~sec apart) are the first two direct pulses. Fig. 4.:3a

presents the multipath return from the water as a concentrated spike of

energy (occurring shortly after the direct pulse). very strong (comparable

in strength to the direct pulse) for the bottom-to-bottom antenna combination,

but much weaker for the other antenna links. Fig. 4.3b shows that the multi

path return over land was much weaker than the water reflected multipath and

much more dispersed in time. The bottom-to-bottom signals show that land

scattered multipath can persist for as much as 35 to 40 ~sec, although at a

very low relative signal strength. The last two pulses in each trace show

the direct pulses transmitted from a top antenna. The top-to-bottom and

bottom-to-top antenna combinations exhibited much weaker multipath than the

bottom-to-bottom link for both surfaces. Virtually no multipath was observed

on the top-to-top combination for both surfaces.

18

4.2.2 System Description

The aircraft used in the program were a twin engine Navajo Chieftain

and a single engine Beech Bonanza. The Navajo Chieftain was equipped with

a set of top and bottom mounted L-Band antennas, a mode D interrogator unit,

and the airborne subsystem of a digital processing/recording device known

as the Airborne Measurement Facility (AMF). Fig. 4.4 is a photograph of the

AMF (Ref. 1). This device was developed at Lincoln Laboratory for the pur

pose of gathering data on the RF environment at the ATC frequencies of 1030

MHz and 1090 MHz. The AMF records aircraft state data (heading, altitude, etc.)

and navigation (VOR-DME) information as well as pulse data. Data from several

data runs were recorded on a single high density instrumentation tape. The

Mode D interrogator unit used a relatively low power (52 dBm) output stage.

The interrogations were transmitted at a 20 Hz repetition rate.

The Beech Bonanza was equipped with a Crosslink Transponder Unit (CTU)

and a set of top and bottom mounted L-Band antennas. The CTU was specially

designed at Lincoln Laboratory for the multipath measurements. It used a

modified King 76/78 Transponder RF receiver section and a high power

(63 dBm) APX-76 transmitter. Defruiter logic was included so that the

transponder could track the mode D interrogations from the Navajo. If the

•

19

No

Fig. 4.4. ~rborne Measurement Facility (AMP).

logic detected the absence of an interrogation, the transponder would still

transmit a reply, but delayed in time by 30 ~sec from its normal response

time. The use of these contingency replies provided link reliability informa

tion while continuing to provide multipath data in the absence of a good

interrogation link. If an interrogation was not received the CTU continued

to "coast" the interrogation track and as long as interrogations were absent the

transponder emitted the delayed contingency replies but when an interrogation

reestablished the link, the CTU timing was resynchronized with the interro

gation timing.

The signals used and their relative timing are illustrated in Fig. 4.5.

Interrogation timing was derived from the AMF clock. Every 50 msec a mode D

interrogation pulse pair was transmitted from the Navajo. An external

trigger pulse was generated 150 ~sec after the first pulse of the Mode D

interrogation. The external trigger pulse caused interrogation time informa

tion to be recorded and enabled the recording circuitry to accept data for

a 500 ~sec window following the external trigger. The 150 ~sec delay was

required to prevent recording backscattered multipath generated by the

interrogation as was observed in Fig. 4.3. During the recording window

pulse data recording was initiated when a signal was received which was above

the minimum triggering level (MTL) of either top or bottom receivers in the

AMF. Once initiated, a series of data samples was taken in rapid sample

fashion with a sample spacing of 0.25 ~sec until the signal fell below the

MTL. For each data sample, the AMF recorded the time of arrival and signal

21

'"

AMF CROSSLINK INTERROGATION TIMING

20 Hz INTERROGATION REP RATEI. 50 ms ---------......-011

_Ilfl fUl__MODE D INTERROGATION PULSE PAIR

-11 n,------l ~O.5 fJ-sI-25 fJ-s --lII

: EXTERNAL TRIGGER PULSE

:.. 150 fJ-s .. I: ~

CROSSLIN K TRAN SPONDER TI MING

NORMAL REPLY:

r RECEIVED INTERROGATION

~25 J1-s ----l

IIl REPLYJ· 168,32 fJ-s ------l..~~"~L-__

TRANSMITTED FROM: BOTTOM ANT TOP ANT

CONTINGENCY REPLY:2fJ-s SEARCH PEROIO

!'--50ms-lfJ-s =-1' "IFROM LAST INTER

CONTINGENCY REPLY TRIGGER

~II

CONTINGENCY REPLY :--1983~55JLS-+60JLsII

TRANSMITTED FROM: BOTTOM ANT TOP ANT

Fig. 4.5. Data Acquisition Crosslink Signals.

22

strength measured on both top and bottom antennas. Also. the analog log

video signals at the output of the detector were available for observation

or photographing.

When the CTU on the Bonanza received a mode D interrogation in the

proper time frame it waited for 168 ~sec before responding. This delay

corresponded to the 150 ~sec external trigger delay previously mentioned.

After the delay the four pulse reply w~s transmitted at 1030 MHz. As

mentioned earlier the first two pulses were transmitted from the bottom

antenna and the last two were transmitted from the top antenna.

Before collecting data. a short recording was made of calibrated

pulse information. During a data flight the Mode D interrogations were

continuously transmitted throughout the duration of each data run.

4.3 Data Reduction

The first step in the reduction of the multipath data was preliminary

tape processing at the AMF ground playback facility. A Data General Nova

minicomputer accepted data from the high density instrumentation tape and

generated three types of output. The first type was an integrated data

dump (ID dump) which presented the time of day. aircraft state. and naviga-

tion information associated with each second of recorded data for the entire

instrumentation tape. This output made it possible to locate any given data

block on the large instrumentation tape. The actual flight path of the Navajo

can be reconstructed from the navigation data. The recorded AMF switch settings

were also presented in the ID dump.

23

The second type of output was the AMF pulse data dump. This provided

a look at the unprocessed data on an individual sample basis. Fig. 4.6

illustrates a section of such an output. The top line provides interrogation

time information used in range calculations. The TOA column represents the

pulse time of arrival. The receiver columns show the relative strength of

the sampled pulses as received via top and bottom antennas. These values

are linearly related to absolute power levels in dBm. The 6T column repre

sents the time between samples in l/8th ~sec clock increments. The last

column provides sample identification numbers. The direct pulses can be

identified by the proper spacing between the four pulses (see Fig. 4.5). The

multipath pulses can be identified by examining the 6T column and searching

for a pair of samples with each sample delayed by the same amount following

a direct pulse. For example, pulses 374 and 383 were separated by 55 ~sec

(440 clock counts) indicating that they were the first two direct pulses.

Pulses 377 and 386 both occurred 129 clock counts after the direct pulses.

Thus, these pulses were multipath signals. The time correlation between

multipath pulses was necessary in order to filter out the multipath signals

from uplink interference. Since the AMF operated in a rapid sample mode,

a wide pulse had several data samples associated with it as illustrated for

example by pulses i.e., samples 374, 375 and 376 which were all associated

with the first direct pulse.

24

..

Interrogation l-l-lu..1

time reference RECE IVER RECEIVERBOTTOM TOP .f;ample

TOA ANTENNA ANTENNA 6T Identification0 5«:,4.J~' I;U';"'i,:j r --0 (I ti I) C:-,·".JL ,,,."23 {5'" o {"I 0 19!' 5415 454198'514 PULSE • 3740 I) (1 C,': 11. ',,12~':,DBB 5'" OD8T ,:., 0 199 2 4'5419951'5 PULSE • 37'50

"I) (1 ':,7'·! --:: .: 12:-'~ ~7' o 'i-l 0 109 2 4541B851'5 PULSE • '376

0 I.~ 1 (1 ,::.C: -J .1:em· r or 0 154 129 4'54199532 PULSE • 377

0 J •.1 I) ,:c,I·12 I~ I ~~':> 2':1 o 2.) 0 109 2 4'54199532 PULSE • 378[l n 0 0 5: -1..1':' ';11 1,0 1':' OM I" 0 IO~' 2 454198'532 PUL!>E • 3790 (1 'J .) a::,~ -;t..:..:: .) 1'3.: ... "'18 10:. (1 8T I':, 0 75 4 454199533 PULSE • 3880 1 (\ I] ,)C:,~..1. .; 1:'1)/0) 13 o 1-, 0 64 2 454 1116'533 PULSE • 3810 I lJ I) ':·'51·..1.,,:' .:' I '3;--I~ II:, [l 17 0 €-4 18 4'54199'535 PULSE • 39Z0 1 I)

"C:-C".-1.J2 ,:",,,3 {%

o r' 0 199 297 4'54199569 PULSE • 3930 .) I lJ r::,,: ~.l~:

:~ :~,~~ 088 ~~,) D

BT.;, II 199 2 4'54189570 PULSE .. 394

0 l) 0 .) ~.:,.l L o l).J 0 'q 2 454199578 PULSE • 38'5

''--'" 0 IJ 0 lJ ~,-: .1.J':: 0:,17% {2~o {"

0 17~ . 129 4'54198596 PULSE • 3960 ,j .) 0 -:.1)-1-'': ';1j'"~'8.. 2:: o 1~ 0 Ii':" 2 454198'596 PULSE • 387lJ lJ (1 l) ~':.1.J2 ,;IH1-I BB 21 OM8T I~ 1'I 64 6 454199588 PULSE • 3980 I 0 ,) ':,':-.J-C I~ 18t1'; 11 o i lJ ';4 2 4'54188588 PULSE • 389{1 I (' lJ S~-t-l::' 0:.21.13 {C"~ Of;; 1'I l~~ 33;- 454188638 PULSE • 3980 11 1 I) '5'5-l-C 1:,21-1~, 0 ~'_1 [l :3 [l 19!1 2 45-1189630 PULSE • 3910 lJ l' ,j 5"~·1 ..k' t.';'l..J7 rl 51 lJ ~T ~'5 lJ 164 2 4'54189';30 PULSE • 3920 0 lj l' Sr'" -1 ..L ';25~":· r: I) o {":' lJ 1~9 452 454189';86 PULSE • 3930 I) I (1 ':o~,~,!2 ~;2~O l D

YB5-;-' o ·3 lJ l~~ 2 454188';87 PULSE • 394

0 0 0 I) ':-''j.,l.....? ',20;0, '51 oo,.T ;-; (I IO!l 2 4'54188687 PULSE • 3958 t1 I 0 ':·'5-1-12 ';-1590 20 0 19 0 33. 'M' 4I5....t16 PULSE • 3960 0 I II '5'544,,' ',-I!,~5 « • 5 0 87 !I 41541 I88t1l PULSE • 3'7

Fig. 4.6. Multipath Experiment Pulse Data Dump.

25

Since the time order of the direct pulses indicated the location of

the transmitting antenna, it was possible to separate the signal strengths

associated with the four possible antenna combinations. Thus, in Fig. 4.6

MBT designates the relative strength of a multipath signal received on the

top antenna as a result of a pulse transmitted from a bottom antenna.

As well as providing a quick look at the unprocessed data, the pulse

data dump allowed a visual inspection of the calibrated sections of

recorded pulse information. Thus, the pulse data dump was extremely useful

in verifying the validity of the data prior to analysis. Also, any given

section of data which showed interesting behavior could be accessed via this

output mode for detailed study.

The last type of output from the Nova computer was nine track IBM data

tape. Each instrumentation tape was used to generate a number of nine track

tapes corresponding to each data run. These tapes were then processed on

an IBM 370 computer.

The analysis programs executed on the IBM computer extracted crosslink

and multipath information and generated several kinds of plots and printouts

which served to spotlight the multipath parameters of interest. Computer

generated plots included the following:

1. Time-raster plot which showed signal tracks indicating the receipt

of direct and multipath signals, the variation in multipath delay

with range, and the multipath signal time dispersion.

26

2. Instantaneous signal power plot which showed the time variation

of successive direct and multipath signal strengths over short

(usually 10 second) time intervals.

*3. Average signal strength plot which showed direct and multipath

signal strengths averaged over ten second intervals for the duration

of an entire data run.

*4. Signal-to-multipath ratio plot which indicated the variation in the

statistical distribution of the signal-to-multipath ratio as the

grazing angle varied during the course of a data run.

Software was available to generate plot types 2 to 4 for each of the four

possible antenna combinations. The plots will be described in greater detail

as they appear in the data presentation.

In addition to the plots, the analysis program generated a printout

which presented several types of statistical summaries of data averaged over

ten second intervals for the duration of a data run. Probability distribu-

tions were generated for both the multipath signal strength and the signal-to-

multipath ratio for each antenna combination. Mean and variance values of

several multipath parameters were also computed for each antenna combination.

The normalized autocovariance of the multipath signal received on the bottom-

to-bottom antenna combination was computed for time lags up to two seconds.

*These figures are actually plotted versus time which increases linearly asthe data evolves. Since the absolute time scales provide no insight into aphysical understanding of the data the time scales have been suppressed.

27

5. EXPERIMENTAL RESULTS


The strongest multipath signals observed were reflected from smooth water

surfaces and were received via a bottom-to-bottom antenna link. As mentioned

earlier in the discussion of the multipath photographs, the multipath signals

generated in response to RF pulse transmissions at L-Band were very pulselike

in nature. The photographs in Fig. 5.1 show more structural detail than the

previous photographs. Again the pulse on the left in each photograph is the

direct pulse, the signal on the right is multipath. The leading edge of the

multipath signals rose sharply following the leading edge of the direct pulse.

Over a range of grazing angles from 15.20

to 57.40

the peak strength of the

multipath signals varied over a range from 20 dB below the direct signal to

as much as 8 dB above the direct signal strength (on the bottom-to-bottom

antenna link). Across the top of the multipath signals the pulse was some

times relatively flat probably indicating a uniformly smooth local reflecting

surface for the duration of the incident direct pulse as in Fig. 5.la. At

other times the peak of the mulitpath signal was ragged as in Fig. 5.1b, taken

a short time later, indicating a slightly rougher local scattering surface.

Following the trailing edge of the direct signal the peak of the multipath

signal dropped sharply. However, as indicated in Fig. 5.lb the trailing

edge drop-off for an apparently rougher scattering area was not as sharp as

in Fig. 5.la.

28

..

DIRECTPULSE

MULTIPATHSIGNAL

II

UNIFORMLY SMOOTH SCATTERING AREA

ROUGH SCATTERING AREA

OCEAN (SEA STATE 1)ALTITUDE: 5500ftRANGE: 3 nmiGRAZING ANGLE: 31°BOTTOM TO BOTTOM

ANTENNAS

Fig. 5.1. Ocean Scattered Multipath (Log-Video Photographs).

29

Fig. 5.2 is a time-raster plot showing mu1tipath tracks from signals

scattered from a calm ocean surface. The vertical scale shows the times of

the external trigger pulses which enabled the recording circuitry for a 500

~sec period. Since the external trigger times and the interrogation times

were related by·a fixed time offse~ the vertical axis also provides a scale

of the interrogation times from the beginning of the flight, at the bottom

of the figure, to the end of the flight at the top. The horizontal axis

shows time in ~sec as measured across the 500 ~sec recording window associated

with each external trigger. Each dot on the plot represents the occurrence

of a pulse sample received by the AMF on the bottom antenna. Thus, for a

particular external trigger time on the vertical scale, the dots viewed while

scanning horizontally from left to right show the time distribution of pulses

received during the associated recording window. The data evolves sequentially

in time from left to right and bottom to top. The dots form four continuous

pairs of pulse tracks representing four sets of direct pulses and associated

mu1tipath signals. As indicated in the figure the first two direct pulse

tracks were due to bottom antenna transmissions while the last two were from

top antenna transmissions. The tracks diverge from the vertical axis indicating

the divergence of the aircraft flight paths. The variation of multipath delay

with range is clearly indicated by the convergence of the direct and mu1tipath

trails as the aircraft range increases. Also, the width of the mu1tipath

trails illustrates the time dispersion of the mu1tipath signals. (A comparison

of Fig. 5.2 for the calm ocean with Fig. 5.34 for a rough land surface shows

the variation in mu1tipath dispersion with surface roughness).

30

i I I ' i I I '

_ r . I ~./. ;. /. ~ J. ! I 2. !.:_ --1-: -I:' -. ;:i--"iT~~~~:-:f--··_ll-_-

.-. -~.-'- · .. ··1~·MJ -:.~- r": '--:l--~ I

Ocean (sea state 1)Alt: 9500 ft.Receiving Ant: Bottom

- t- I I.. " j., f.., --.

• - .. : Ij. ~ ~ ~ -••__pirett Pulse 'I -... "., . J .,.

.. -

..:

........

ransm tted rom:} •• J . I !Bottr"J ,,' ,,',,'

Ir T P,,-,~•"" ',. ,I.. If' .. ,/''~~~ . ;~I .. ~!r

~.....H

I-lQJCDCD.....I-lH

>'.....'"-....

C)

QJ

'"oN............

''j~ ..'j '.j' I~·'" ..~'. :•.a_ •

! .,.! ,.1: .. ': .. -: .'I J, f: I ~ , ',:' ..., .. " I-....!: - ~ ...... 1 . ~.:- '. . .__.~T'"H-r - . ~~~-Re-'c-+r~~iv-e-d-~~'U;~-i'-pa-+-h--:---+~~ -[1-----

j I i

P:JlSE DEl"" !\ForE!' TPIG'.EP "J~[':)

Fig. 5.2. Ocean (Sea State 1) Multipath Time-Raster.

31

Since the receivers detected any signal of sufficient strength at a

frequency of 1030 MHz there is a random background of pulses due to uplink

interrogations and sidelobe suppression pulses from local ground stations.

In Fig. 5.2 the multipath trails were not much broader than the associated

direct signal trails evidently indicating a relatively small scattering

area. The mu1tipath trails for the first two sets of pulse tracks, corres-

ponding to a bottom-to-bottom antenna combination, were much stronger than

for the last two sets of pulse tracks, corresponding to a top-to-bottom antenna

combination. Also, at the beginning of the flight the last two sets of

pulse tracks indicate that ~he mu1tipath pulse trails were not well established

until the aircraft had separated by about two miles indicating the reduced

mu1tipath rejection capability of the top antenna at lower grazing angles.

Still referring to Fig. 5.2, between trigger times from 420 seconds to

480 seconds there was a degradation of the interrogation link due to inter-

ference other than multipath. Contingency replies, as indicated by the delayed

tracks, continued to provide mu1tipath data during this period.

For the data shown in Fig. 5.2 the multipath delay was measured and the

..

results are shown in Fig. 5.3. Also, shown is a theoretical curve that

represents the delay of a simple specular mu1tipath for equal altitude aircraft:

where

D R (secant (G) - 1) / c (5.1)

D = multipath delay, R = range, c

Grazing angle G is given by:

G = tan-1 ( 2A)R

speed of light, and G grazing angle.

(5.2)

32

en:l.-

20

15

+

2

OCEAN (SEA STATE 1)ALTITUDE: 12000 ft

THEORY

3 4 5 6 7 8 9 10 11 12 13 14RANGE (nmj)

Fig. 5.3. Multipath Delay.

33

where A = altitude above ground level. As seen in Fig. 5.3 formula 5.1 agrees

with the measurements quite accurately. This accurate predictability of

multipath delay has been observed in all the data regardless of aircraft geometry

or scattering surface.

One feature of the multipath scattered from the ocean which in fact was

true of all but the very smoothest surfaces was the rapid variation in the

multipath strength from sample to sample. Fig. 5.4 shows the instantaneous

direct and multipath signal strengths for successive replies for a ten second

period at different grazing angles. The signals in the figure were received

via a bottom-to-bottom antenna combination. In both cases successive multi

path values changed by as much as 25 dB over a 50 msec time interval. Such

variability in the multipath seems to imply that a significant portion of the

multipath is not due to purely specular multipath. In conducting this

experiment interest was focused on the total multipath signal received. No

attempt was made to separate the coherent and incoherent components of the

multipath signal.

Fig. 5.5 shows normalized autocovariance plots corresponding to the data

in Fig. 5.4. Both cu~ves confirm in a quantitative'way that there is very little

correlation between successive multipath samples. Such lack of correlation

further indicates that even a relatively calm sea gives rise to a large

incoherent component in the multipath at large grazing angles.

Several features were observed in the data which are illustrated in the

signal-to-multipath ratio (SMR) cumulative distribution curves in Fig. 5.6.

The curves have a shape similar to the exponential distribution curve cor

responding to Rayleigh amplitude statistics associated with diffuse multipath

(see Ref. 2). The Rayleigh property was commonly observed in the data at all

34


0.0

-10.0

-20.0

-30.0

~-40.0E

CD~ -50.0a.~ -60.0

-70.0

-80.0

, I I ; I I ~...L-._-! , I, , I ,

'7TH i : I I I i j I

i I ! I !I I . ,

''''++4-,. i

I ! I

-II I ,I I I , ,, ,

!!>lt~. •',•.1._ , ,~

, , ,~- .. .. .. .

,\ ,'.' , ... , i , ,~..

I , , ,,,

,,, , , ,

-90.0

-100.0

TIME n 5seamlnor div)

GRAZING ANGLE: 380

RANGE: 4 nmi

0.0

-10.0

-20.0

-30.0

~ -40.0Em~ -50.0a.~ -60.0

-70.0

-80.0

-90.0

-100.0

I

, , .. "-, .. .'. • I., , ,.... ..~ .. w-

• .-1, r-. •• •

, , •.. ,

.-

TIME (0. 5 sec/minor dlvl

GRAZING ANGLE: 190

RANGE: 9nmi · -Direct• -Multipath

Fig. 5.4. Ocean (Sea State 1) Multipath InstantaneousSignal Strength.

35

1.0GRAZING ANGLE =3SO

0.5

OL.--------"---------l....------'----------=--.Jo 0,5 1.0 1.5 2,0TIME (sec)

1.0GRAZING ANGLE =19.2°

0,5

- -o '--- ----I1 ----1...1 --J1'----_-_-_""""---I

o 0,5 1,0 1.5 2,0TI ME (sec)

OCEAN (SEA STATE 1)BOTTOM-TO-BOTTOM

ANTENNA COMB,

Fig. 5.S. Normalized Autocovariance (Ocean Sea State 1).

36


40302010SMR (dB)

(0)

o-10

BOTTOM-TO-BOTTOM

0.5

0,0 L- ---.J.-...::::::::::: L- --L...~=_____ _____l. ___U...~~__l._...:I.........:J

-20

1.0

40302010SMR (dB)

( b )

o-10

BOTTOM-TO-BOTTOM

0.5

0.0 L- ----L._~=----_....L_~=----_-----.L __l...._______L~ll....:l~J.......l.......lo.....lo_ll

-20

Fig. 5.6. Signal-to-Multipath Ratio Cumulative Distribution (OceanSea State 1).

37

ranges, altitudes, antenna combinations and all surfaces with the exception

of the very smoothest surfaces. For the bottom-to-bottom link curves, the

probability of SMR < 6 dB was 0.86 at the higher grazing angle and 0.76 at

the lower grazing angle. At both grazing angles the top-to-bottom curves were

shifted to the right from the bottom-to-bottom curves indicating an improve

ment in SMR due to the shielding of the top transmitting antenna from the

active scattering surface. However, the degree of improvement was reduced

from 20 dB at the higher grazing angle to 9 dB at the lower grazing angle.

At lower grazing angles the two distributions continued to converge to a

single limiting curve with a median SMR of -10 dB. Two mechanisms could

contribute to the reduced SMR improvement at the lower grazing angle. The

bottom-to-bottom curve in Fig. 5.6b has shifted slightly to the right of its

position in Fig. 5.6a indicating slight reductions in both antenna gain in

the multipath direction and scattered multipath strength. The primary reduc

tion in SMR improvement was due to the shift of the top-to-bottom curve to the

left at the lower grazing angle since the top antenna was less shielded from

the scattering area.

Distribution curves for the bottom-to-top and top-to-top antenna com

binations followed the same trends as the top-to-bottom curve. The bottom-to

top antenna link behavior was slightly less sensitive to grazing angle than

the top-to-bottom link. The aSYmmetry in the top-to-bottom and bottom-to-top

links was probably due to slight differences in aircraft size and antenna

locations.

The strong multipath reflected from the ocean is rather dramatically

illustrated in Fig. 5.7 which shows the average direct and multipath signal

38

, ! !i I

!

.. (0)........ ..

••~~....-. . ......'" .....1-\" ... "';';---'". ....

-80,0

OCEAN(SEA STATE 1)BOTTOM TO BOTTOM LINK

RANGE (nmi! .75 2.5 4.0 6.75 9.0 11.25 13.0GRAZING ANGLE 76 51 38 25 19 15 13

(deg)

-10.0

ffi -60.0~oQ.

-70,0

~ -50.0'0

-40.0

-30.0

-20.0

-90,0TIME (200 secJdlvl

ALTITUDE: 9500ft

RANGE (nmi! 0.3 1.5 2.0 3.7 5.0 8,7 13.0 18.0GRAZING ANGLE 78 45 37 22 17 10 7 5

(deg)0,0

-10,0/---f--- --', -·--·------~i---

i-20,0 f---of---+---+---.,-----j

-30,01----h-----+------I---+I---I

I (b)

-80.0

-90.0

-70.0

~ -40.0 /---.""';.,1-:::.,,;.,,,-'-'-'.,::--::..+-----+---+----1~ ~ "\...•..a: -50.0 I---~'-'-+--J''-:-' - .::;.+:---+--_._--w : ~.'~ ¥ K::::'.....~-60.0 t----------1e-----t---+-""''li-~.\o,-&--.....-.-:_.

\ ..-.--- _. f----.--~--+_----.,----

I "j .....

--- t··---- i----':---' !.~~/l, -

TIME (288 sec/divlALTITUDE: 4500 fl

. -Direct• -Multipath

Fig. 5.7(a-b). Ocean (Sea State l) Multipath Average Signal Strength.

39

•

"."

Direct Si,€r"..ll

Oc~J.;. (:3.::3. st..;,-~c 1)A1 ~: 9500['.Ant. Co::tb. :To;, to Eo ... t:::-:

13.013

11. ::'515

9.019

:•. 038

2.551

.7576

.' '. (C)'.' ....

.....'.........

• .' .,.. ..

II

• . n h ..a.o' ....I • "It'" I I ... --I I

• I •I·

n' , I.I

I '.r..'

-:.'. ~

-c:~.C

-3~ .~

-".t

p -~e.e

0

'"[R -u .•(

I-I

" -7C ••

-H.e

-g~ .•

R.1"~'-' (llmil:Grazin~.~ngle (deg)'

Time (200 sec/div)

Fig. 5.7(G). Ocean (Sea State 1) Mu1tipath Average Signal Strength.

40

strength as a function of time into a data run. For grazing angles from

o 015 to 57 the multipath signals were very strong. The reflections were

strong enough so that for grazing angles between 250 and 400

the differential

antenna gains consistently increased the average multipath signal strength

to a level greater than or equal to the average direct signal strength.

Fig. S.lc showing average signal strengths for a top-to-bottom antenna

link illustrates a very significant multipath feature regarding the effective-

ness of antenna diversity. o 0For grazing from 14 to 40 the average multipath

was nearly constant. The phenomenon appears to be due to two competing

mechanisms. If the top antenna had not been shielded from the scattering sur-

face the multipath would initially increase and then decrease with range as

in Fig. 5.7a. The shielding of the top antenna reduces the received multipath

level but by an amount that decreases with increasing range. The variations

in multipath strength and top antenna shielding with range evidently offset

each other producing the relatively constant level of average received multi-

path. This phenomenon was observed on antenna links employing one or two top

mounted antennas. Also, this phenomenon was observed in all the data associ-

ated with smooth surfaces i.e., desert, flat plain, and lake surfaces.

The drop in multipath signal strength at the higher and lower grazing

angles was apparently due to two mechanisms as indicated in Fig. 5.8. This

figure shows the measured signal-to-multipath ratio at several grazing angles

for the data in Fig. S.7a. Also, shown are the total differential antenna

gain and the reflection loss. The differential antenna gain is the total

41

OCEAN (SEA STATE 1)

BOTTOM-TO-BOTTOM ANTENNAS

SMR: SIGNAL-TO-MULTIPATH RATIO (MEDIAN)L: TOTAL SCATTERING LOSSG: TOTAL ANTENNA GAIN IN SPECULAR DIRECTION

80706020-10 L--__--l....-__-----.l ....L- -----L ..l....-.__---l...::l8--_-----I

10

30m~ 25...Jw~ 20....J

<i 15zC)

(/) 10...J~

..... 5zwa:w 0IJ..LL

0-5

~ig. 5.8. Signal-to-Multipath Ratio, Antenna Gain, and Total ScatteringLoss (Ocean Sea State 1).

42

The loss curve represents the total

gain in the specular multipath direction over that of the crosslink direction

taking into account the gains of both transmitting and receiving antennas.

*The antenna data was taken from Ref. 3

reflection loss (absorption and scattering loss) computed from the measured

signal strengths and the antenna gain variations.

From Fig. 5.8 it is seen that for these grazing angles, the variations

in signal-to-multipath ratio were primarily dependent on antenna gain varia-

tions rather than on variations in the scattering loss.

For this data the reflection loss increased at the lower grazing angles.

This is most likely because of the surface roughness. The "Rayleigh criterion"

states that a surface is considered smooth if the height of the surface

irregularities, h, satisfies:

Ah < 8 sin y

where A is the signal wavelength and y is the grazing angle. The waveheights

encountered exceeded the Rayleigh smoothness factor even at the lower grazing

angles at which data was recorded. Hence, the surface was electrically rough

and scattered less energy in the direction of the receiving antenna. This

qualitative behavior was anticipated by theoretical analyses of rough surface

scattering (see for example Ref. 2). At lower grazing angles, shadowing also

becomes an important factor in reducing the energy scattered in the specular

direction.

Fig. 5.9 shows plots of signal-to-multipath ratio (SMR) distribution

variations versus time, range and grazing angle. Each figure corresponds to

one of the four possible antenna combinations. In evaluating the performance

*Crude antenna patterns were measured for the actual aircraft used in the rn1l1ti-path measurements and were found to agree with the more detailed patterns inRef. 3 to first order.

43


RANGE (nmi)

GRAZING ANGLE(deg)

80,0

70,0

0,8 2,5 4,0 6,75 9,0 11,3 13,0

76 51 38 25 19 15 13

I

0,8 2,5 4,0 6,759.0 11.3 13.076 51 38 25 19 15 13

80.0

70.0 ~~-- -,~-t-~~-+~~----lr--'~----1

600 1---~+~~--+~~~I--~~+_~~--1 60.0 I-~~+,~~-+~~--+~~--+---

40.0 f--._-";';&-'h,---~+_~~_+~~_+~~__I

50,0 I----+---+-~-+--+--_l

"'1·-"10,0 I-_~+-~~-+--c'---';;~_~---'-~I-_~_

----

.:..

.".\::~. ~'~'--';_:I-~~+-~~-t--~vtc;-,----'-~t---

~~---+----+-'''''-;''''f--

I-~~-+-~~-+-~~-t~~--t---50,0

40,0

30.0CD:g

20.0a:::~VJ 10.0

0,0

-10.0

20,0

......... .0,0 ~---~t-----t-~~-+----":-"':"•....:&.f'-'~

-1 0,0 .....-'~+-~~-+---~_+~~-II_~--1

-2 a ,0 L-__...:...-__--'-__---'-__----J'--_----J

CD 30 0 1---~+.;r.-'--'--+"::.;..=-,;.,-,_-,.--+~~~I-~--1~ ~~~:.'. .~..~ 200 - --'~+-~--:.....r-~~+':.-. ....--~+_~~-iVJ

T1M~ 1200 sec1dlvlTOP TO BOTTOM LINK TOP TO TOP LINK,

40,0 I-~--"".±-..,---+----+~~~+----j

80,0 ,..---.......--..,....---...,....--...,...--....,

70,0

60,01_--+---+----t-----+----I

50.0 ----I~--I_--+_---+_-__l

iii . '.-'-;.~

; 30.0 I-~--+-'-.."-_,...,..~,-.~.'~.'.-'.. ~.,-'::::-.--+~--+-------j

~ 20,0 f----+--..-"',....:..,.:..,.".:,-~·-,"·4"""""'~-,-l-~--l-... ' ....'". .

10,0 1-----Jr-------1----J~..._-.::..:"'-+--_j'.1-..," .

-10.0 J---t---+---+-----+-------I

-20,0 "'-_---l__-..l..__...:..__-"-_---I

T1Ml (200 sec1divl

.' ..',..;.;.. .II' . ..' " ~

....

!,

" -'

0,0 1------+

11

f....;...; .......-100

·20,0 L-__-'--__..:.....__....l-__....L__-'

800

700

600

500

40,0

CD:g 30,0

a:::~ 20,0VJ

10.0 -,

. BOTTOM TO BOTTOM LINK BOTTOM TO TOP LINK

- Sensitivity IhresholHProblSMR< XI' 0.9. ProbISMR-::. Xl' O.~: 'ProoISMR S'xl· 0.1

Fig. 5.9. Signa1-to-Mu1tipath Ratio Distribution Variations (OceanSea State 1).

44

of the top antennas in rejecting multipath, the angle of flight divergence

(nominally 300 for the data under consideration) plays an important role in

determining the degree of top antenna shielding from the scattering region by

the aircraft structure. A 300 divergence provided a nearly minimum wing

shielding of the top antennas. In constructing these figures the cumulative

distribution of the signal-to-multipath ratio was accumulated over successive

ten second intervals for each antenna combination and certain points of each

distribution were plotted. In each figure the "+" corresponds to the SMR

value below which 90% of the SMR values measured during a ten second interval

fell. The "." corresponds to the median SMR value while the "*" corresponds

to the 10% point on the SMR distribution. The "_" represents the limiting

measurement threshold imposed by the minimum triggering level reflecting the

instrumentation sensitivity. Points above the MTL symbol do not represent

valid data.

The effects of antenna combination on the SMR are evident from the plots

in Fig. 5.9. In comparing the median SMR curve for the bottom-to-bottom

antenna link.with the median SMR curves associated with antenna links

which include at least one top antenna there was a substantial improvement at

grazing angles greater than 19°. At 38° the top-to-bottom link caused a 20 dB

improvement in mediam SMR link while the top-to-top antenna link provided a

32 dB improvement in median SMR. As the grazing angle descreased there was a

reduction in the ability of the top antennas to discriminate against multi

path signals. This is evident in Fig. 5.9 from a comparison of SMR improve

ments at various grazing angles. For instance, at 380 the top-to-bottom link

afforded a 20 dB improvement while at 190 there was a 10 dB improvement over

the bottom-to-bottom case.

45

The overall effect of the top antennas was to increase the minimum SMR

value and to cause it to occur at a lower grazing angle, i.e., at a greater

range for a given altitude. While the top antennas always produced an increase

in SMR over the bottom-to-bottom case, the degree of improvement at lower grazing

angles depended strongly on the roughness of the surface since the top antennas

were more exposed to the active scattering area. The data in Fig. 5.9 show

that the minimum SMR improvement was provided by the top-to-bottom antenna

combination which shifted the bottom-to-bottom minimum median SMR of -2 dB at

350 to 8 dB at 150. This amounted to raising the minimum average SMR by 10 dB

and causing it to occur at a range six nautical miles greater than for the

bottom-to-bottom antenna link.

Another feature of the data is that for all antenna links the 10% to

90% spread in SMR distributions was somewhat constant for grazing angles

from 170

to 760 having values from 10 dB to 15 dB depending on the antenna

combination. For grazing angles lower than 170 the spread tended to increase

as the grazing angle decreased for all antenna combinations. This was

possibly due to sampling irregularities at low multipath signal levels.

Fig. 5.10 shows a comparison of median SMR curves for two different

altitudes. Two features are evident in the curves. First, both curves

varied somewhat inversely to the antenna gain plot as is also indicated in

Fig. 5.8 for the higher altitude. Also, for higher grazing angles (above

450

) h hwere t e antenna gain was rapidly dropping below its peak and for

lower grazing angles (below 250) where the scattering loss seemed to be

increasing, the SMR for the lower altitude was several dB less than for the

46

OCEAN (SEA STATE 1)

BOTTOM-TO-BOTTOM ANTENNAS-CD"U 25-0~ 20<{a:::::I: 15~a:~ 10::::>:E0 5~

-' 0<{z(!)

(f) -50 10 20 30 40

GRAZING ANGLE (deg)

ALT: 9500ft

50 60

Fi~. 5.10. Altitude Variations in Signa1-to-Mu1tipath Ratio (OceanSea State 1).

47

higher altitude. This might be attributed to the increased size of the

active scattering area at the higher altitude. Since the scattering area

was presumably larger at the higher altitude, the multipath energy was

probably more dispersed than for the lower altitude. Hence, the higher

altitude might be expected to have a slightly higher SMR than the lower

altitude.a 0

For intermediate grazing angles from 25 to 45 this effect was

swamped out by the combined effects of strong antenna gain and any reduc-

tion in scattering loss indicating that variations in SMR due to altitude

variations were second order effects. Thus, mllitipath has a definite

dependence ~n grazing angle, as would be expected. The additional dependence

on altitude is minor with no significant change in the worst-case multipath,

and with a small change in the grazing angle at which the worst-case multipath

occurs.

The SMR variations in Fig. 5.11 summarize the Sea State 1 ocean multipath

data. The main features are that a bottom-to-bottom antenna link over the

calm ocean surface experienced a severe multipath environment over a large

variation in grazing angle and over an altitude variation of at least 5000 ft.

The use of top antennas in the link provided substantial multipath rejec-

tion at high grazing angles but the multipath discrimination capability of

top antennas decreased with decreasing grazing angle. The minimum SMR for

a link employing a top antenna was 8 dB for the top-to-bottom link. The top

antennas caused the minimum SMR to occur at a greater range than for the

bottom-to-bottom case. The amount of the range shift in the SMR minimum

point varied directly with altitude.

48

OCEAN(SEA STATE I)

BOTTOM-TO-BOTTOMANTENNAS

ZING ANGLE: 30·o1l.::o:-----~5:------....L..10----·---:-l,51.---------J20

10tOOO~__2..,..0__---,-J5__JTT0TT7'TTT~IO~d_B--,

--.....-

201510

RANGE (nmi)

WORSTSMR- 8 dB

~-15°G.A. TOP-TO-80TTOM ANTENNAS

Fig. 5.11. Geometrical Dependence of Signal-to-Multipath Ratio (OceanSea State 1).

49


Photographs taken of multipath returns from a sea state 2 ocean indicated

that the signals for this slightly rougher surface were still well defined

pulses. While the increased surface roughness tended to reduce the average

level of multipath signals by several dB from the Sea State 1 level. occ.a-

sionally multipath pulses were observed which equalled or exceeded the direct

signal in strength. The envelopes of the received multipath signals generally

were more ragged than for the smoother Sea State 1 ocean as in Fig. s.le.

Also. as in Fig. s.lb the bases of the main portion of the multipath signals

were somewhat broader than that illustrated in Fig. s.la. The peak multipath

levels tapered down more slowly indicating the increased time dispersion

indicative of a slightly larger scattering area associated with a rougher

surface.

A raster-plot of multipath data recorded at an altitude of 1200 ft.

exhibited the same temporal characteristics illustrated by the multipath

trails in Fig. 5.2. The multipath trails were quite thin indicating a

relatively small scattering area but were slightly broader than the direct

pulse trails. Instantaneous signal strength plots showed rapid variations

over a 20 dB range in the strength of successive multipath samples in

Fig. 5.4. The autocovariance computed from the sea State 2 data indicated

that successive multipath samples were slightly less correlated than was

the sea state 1 case in Fig. s.sb. There was very little variation in the

multipath correlation over the range of grazing angles for which data was

o arecorded (14 to 75 ).

50

Fig. 5.12 shows several signal-to-multipath ratio cumulative distribu-

tion curves. As in Fig. 5.6 a decrease in grazing angle was accompanied by

a slight (~2 dB) improvement in the median SMR for the bottom-to-bottom

antenna link indicating a slight drop in the total antenna gain in the

multipath direction and possibly a slight increase in the scattering loss

at the lower grazing angle. There was only a very slight reduction in the

median SMR for the top-to-bottom link at the lower grazing angle unlike

the curves in Fig. 5.6. The primary reason for the slight shift in top-to-

obottom distribution is that at a grazing angle of 38 the top-to-bottom link

SMR distribution was at a lower SMR range than was the case for Fig. 5.6a. This

degradation in SMR was caused by increased multipath reception via two pos-

sible mechanisms. oWhen the aircraft were at a grazing angle of 38 • their

flight paths were diverging at an angle of -350 allowing a relatively high

exposure of the top antennas to the scattering surface due to the absence

of wing shielding thus increasing the level of received multipath. In addi-

tion to the increased top antenna exposure to the scattering surface. the

larger scattering area associated with the rougher sea tended to expose the

top antennas to multipath contributions from the outer portions of the

scattering area at higher grazing angles than would be the case for a smaller

scattering area associated with a smoother surface. The latter mechanism was

probably a second order effect.

In comparison with the Sea State 1 data in Fig. 5.6, the bottom-to-

bottom SMR distributions at both grazing angles were shifter to higher SMR

values (by -7 dB at 380

; -4.5 dB at 190

) indicating the effect of increased

51


'0SMR (dB)

o-10o

-20

,.Or---------------==::::=:::::=====:::::::==-~"'\\:\\l

0,5

r---------------------=:::;pl~~t"'"T"'\"""\'\T""In1.0

0.5

OL- ---L-__~......I.II!~_~___:_l ______:~-..i..l....l......),..~..l......),..~-l.......l.~

-20 -10 0 10 20SMR (dB)

Fig. 5.12. Signal-to-Multipath Ratio Cumulative Distribution (OceanSea State 2).

52

surface roughness on received multipath strength. Thus, the improvement in

the bottom-to-bottom SMR distribution due to the rougher surface and the

degradation in the top-to-bottom distribution apparently caused a significant

decrease in the SMR improvement at the higher grazing angle. In spite of the

premature reduction in the effectiveness of top antennas in rejecting multi

path at closer ranges the top antennas still provided substantial multipath

rejection. The SMR improvement at the lower grazing angle of Fig. 12b

was not significantly different from that shown in Fig. 5.6b since flight

conditions were such that top antennas were exposed to approximately the

same scattering area in both cases.

At the lower grazing angle the bottom-to-bottom link curve shifted to

higher SMR values as rough surface scattering loss evidently increased as in

the Sea State 1 data. The bottom-to-bottom and top-to-bottom curves tended

to converge toward a single limiting curve with a median SMR approaching

-17 dB at lower grazing angles as the top antennas viewed an increasing

portion of the scattering area "seen" by the bottom antennas.

The average signal strength plots in Fig. 5.13 bear out the median SMR

variations in Fig. 5.12. A comparison of Fig. 5.l3b with the Sea State 1

data in Fig. 5.7a clearly indicates the reduction in received multipath

due to increased scattering loss. The variation in average multipath strength

in Fig. 5.l3a indicates another rough surface effect in that the relative

constancy of the average multipath signal strength range observed over smooth

surfaces (as in Fig. 5.7c) is absent for the top antenna links.

The antenna gain and scattering loss curves in Fig. 5.14 again indicate

that over most grazing angles the variations in SMR were due to antenna gain

53


2.0 3.5 6.0 9.0 12.5 13.8 14.3 15.063 49 33 24 18 16 15 15

-20.0

RANGE (nmi)GRAZING ANGLE

(deg)-10.0 .-----l---l..-\_I--..I..-t-

I........-+---'----,

---+-----1------+------+-------I :I

-30.0 ------+---+---+---+----------j

E -40.0lD:sa::w -50.03:0ll.

-60.0

-70.0

-80.0

-'90.0

.: ........

.....:.

.... ", ,','., ....1' .. .. ......

.- --*.... I ....,' • .... J ....-........

~ . "~.- ........- .", ,.

TIME ( 240 sec/dlv)

TOP TO BOTTOM LINK

-10.0

-20.0 ._-

-30,0

E -40.0CD:g

a::-50.0w

3:0ll.

-60.0~

-70.0

-800

-90.0

I

............".

.. ,- ....,~ .1.-0'.. I -..... '.:'.' - : ..:.... .. .

...... .".." "" I ....

TIME ( 240 sec/dlv)

BOTTOM TO BOTTOM LINK. -Directo-Multlpath

Fig. 5.13. Ocean (Sea State 2) Multipath Average Signal Strength.

54

OCEAN (SEA STATE 2)BOTTOM-TO-BOTTOM ANTENNAS

SMR: SIGNAL-TO-MULTIPATH RATIOL: TOTAL SCATTERING LOSSG: TOTAL ANTENNA GAIN IN SPECULAR DIRECTION

ALT ITUDE: 12000 ft

80

SMR

70

--_--L

30 40 50 60GRAZING ANGLE (deg)

20

- 30CD"C

...J25

w> 20w...J

...J 15«z(!)

en 10...J« 5~zw 00::WLL.LL. -5-0

-1010

Fig. 5.14. Signa1-to-Mu1tipath Ratio, Antenna Gain, and TotalScattering Loss (Ocean Sea State 2).

55

variations rather than scattering loss variations. While the total loss

curve did not vary greatly over most grazing angles (i.e., 200

to 700

) its

average value was raised by ~7 dB over that in the Sea State 1 data in

Fig. 5.8 consequently improving the SMR curve. Again, the increased

scattering loss at the lower grazing angles indicates the rough surface

scattering of less energy in the direction of the receiving aircraft.

The signal-to-multipath ratio curves of Fig. 5.15 illustrate the SMR

improvement due to different levels of antenna diversity as the grazing

angle varied. A comparison of Figs. 5.l5a and 5.l5c with Figs.5.9a and

5.9c indica~es that for grazing angles higher than 400

the reduction in

SMR improvement was probably due mostly to a decrease in the multipath level

raising the SMR curves for the bottom-to-bottom link above the level shown

in Fig. 5.9c. This was during a period when the aircraft flight paths

diverged at the same angle for both data sets. The sharp drops in the SMR

curves in Figs. 5.l5a and 5.l5b at~400 indicate that the SMR improvements

was apparently further reduced by a change in aircraft heading which increased

the exposure of the top antenna on the transmitting aircraft. All top

antenna links had a minimum median SMR > 10 dB.

The median SMR curves in Fig. 5.16 show the same trends at all altitudes.

At high grazing angles the SMR increased due to antenna nulls in the multi

path direction in addition to scattering loss. At low grazing angles

the SMR increased presumably due to increased scattering loss rather than

a reduction in antenna gain. The minimum SMR of the lower altitudes was

56

OCEAN (SEA STATE 2)ALTITUDE: 12000 fl

60,01----1----+-----+---+------1

1.5 2.0 3,5 6.5 9.0 12.4 14.2 14.5 15.069 63 48 31 24 18 16 15 15

. . ..

0,0 l------J---f---+---+----l

10,01---+----+----+----+---1

-10,0 f--- ~.--+_--+----j----

- 20.0 '--_-.JL....-_----'__-.l.i ....J

80.0

70,0

60,0

50.01D~ 40.0n::~ 30,0(f)

20.0

"' ...... -..,,' ,.,""":.. . .. :'. I, '.,: . -.:....: ..:

"..eo ,'. • /--.. ....• II

'..' ~--......".

RANGE (nmil 1.5 2.0 3,5 6,5 9,0 12.4 14,2 14,5 15,0GRAZING 69 63 48 31 24 18 16 15 15ANGLE (dj.l:e:lLol_.....,..__--r__--,r--__'r""'"_--,

80,0 i70,0/----+------ c----- _._. .L ._ ---

50,0m~ 40,0n::~

30,0(f)

20.0

10.0

00

-100

-20.0

TIME rl40 secldivl

TOP TO BOTTOM LINK TOP TO TOP LINK

-"":'.. .' : .. ~.' ,...• ·· ·1'1 ..... ":"....... ...,. .J.'

--..... ...---:"..:r:::-~::,.-n"-

.... • to •

..~t--......: - t' :"-.~ 7"-

80.0

70.0

60.0

50,0

40.01D~

30.0n::~(f) 20,0

10,0

0,0

-10.0

-20,0

TIME 1240 secldivl

'-,,' .....:. ".. '. .... .t,.. ••••••••••.. i; .'

K ..·..·..·..-1-'.' -.' :.. . ....... ..

.,.-.-~,- ~---. -.' ~. ~--,.. " .' ':" .....

0L.-_---l__---J.__---I.__.....l.__....J

-10.0 t-----t-----+-----+----+----J

800

700

600

500

400

ill~ 300n::~ 200lfJ

100

BOTTOM TO BOTTOM LINK BOTTOM TO TOP LINK

- Sensitivity threshold:tProbISMR< xl ' O. Q

. ProbISMR,:: xl ' 0.5: 'P rObISMR 'S-XI' o. I

Fig. 5.15. Signal-to-Multipath Ratio Distribution Variations (OceanSea State 2).

57

OCEAN (SEA STATE 2)

BOTTOM-TO-BOTTOM ANTENNASr---------------------------40

..- 35CD'l:J.......0 30~a:z 25~a.:-~ 20~

:E

0 15........J<!Z 10C)

U5

5

00 10 20 30 40 50 60 70 80

GRAZI NG ANGLE ( deg)

Fig. 5.16. Altitude Variations in Signa1-to-Mu1tipath Ratio (OceanSea State 2).

reduced by ~. dB compared to the higher altitude curve possibly since multi

path signals were detected from a larger scattering area than at the high

altitude. Other than small downward shift in minimum SMR the curves exhibited

no significant altitude dependence.

The graphs in Fig. 5.17 indicate the SMR improvement with a top antenna

link over the airspace for which data was obtained. While the bottom-to

bottom link was susceptible to relatively strong multipath interference over

a wide volume of airspace, links which used at least a single top mounted

antenna did not experience a high level of multipath intensity over any

portion of the airspace investigated.

5.3 Frozen Lake Surface

While very strong multipath was scattered from the Sea State 1 ocean as

described above, the surface was still rough enough to render the multipath

samples taken 20 msec apart essentially uncorrelated as shown in Fig. 5.5.

Data from flights over two frozen lakes indicated that not only was the

multipath very strong but it was much more correlated than for rougher

surfaces.

Fig. 5.18 shows the instantaneous multipath measured over a small lake

(Lake Cochituate) at a grazing angle of 330• When the scattering area was

primarily over the frozen lake the multipath level at times exceeded the

direct signal level by as much as 9 dB on the bottom-to-bottom antenna link.

As indicated in Fig. 5.l8a the use of a top transmitting antenna provided

significant reduction in the multipath.

59

OCEAN (SEA STATE 2)

-I-WWu..-

10000

WORST SMR=2 dB

33° G. A, BOTTOM-TO-BOTTOM ANTENNAS0 ---1- """""- _

o 5 10 15 20wCl~

I-

~<3: 10000

20

WORST SMR = 11 dB

5000 f----..:..,~;....r..:;.~__T'_--------__,

SMR = 9 dB

TOP-TO-BOTTOM ANTENNASOwt.....- ---'- --'--- ---'-- -----'

o 5 10 15

RANGE (nmj)

Fig. 5.17. Geometrical Dependence of Signal-to-Multipath Ratio (OceanSea State 2).

60

FROZEN LAKEALTITUDE: 9500 ftGRAZING ANGLE 33°

0,0

-10,0

·20,0

-30,0

-40,0

1D.::g -50,0a..:!'<{ -60,0

-70.0

-80.0

-90.0

-100,0

I I I I 1 I--r ..1 ! I I -r-, I

I I 1 +-+--+ II

I,

-4 1

I I

, . .. _.... ·1'- .. .... -.. .. --+ ,~

.1 . ~ • ....I • ' I.

.... • • I"~ • ... ••... [' 'H~ I Ir,; d Il- • • • ••• • n .- '11 • • .'..II

r' • • • .ra•

l-I-1-. • -- • ..'~ .l- ...• •

TlM~ (1 sec/minor diy)

TOP TO BOTTOM LINK

0.0

-10,0

-20,0

-30,0

~ -40,0

a..~ -50.0

-60,0

-70.0

-80,0

-90,0

-100,0

I I I I 4--rT ! : I!

I1 I T. __I

-+_.1

I

W-, , ~ Ir • . lJ al, I'. ._. .- -I.: .....·1" . i, . - . .

!• •. loll a I'- I' I" • II I~ ~ ro;; ,• I~ lOy • , , . • • , ~ I~,,, . I • • • 1, • •'II .,..

• ••

, .. .. • • • I-

TIME! I sec/minor dlvl

BOTTOM TO BOTTOM LINK · -Direct, -Mu Itipath

Fig. 5.18. Frozen Lake (High Grazing Angle) Multipath InstantaneousSignal Strength.

61

FROZEN LAKEBOTTOM-TO-BOTTOM ANT, COMa

',0 r------------------------.

GRAZING ANGLE =330

0,5

2,01.51,0

TIME (sec)

0.5OL..- .L-- L-- L..- -----J

o

Fig. 5.19. Normalized Autocovariance (Frozen Lake).

FROZEN LAKEALTITUDE: 3500 ftGRAZING ANGLE: 11°

0.0 ....,....,..IT....-TI"--rTl-r"TIIT-r-rT....."

-100 HH-++++-HH-++++-HH-t++-1 -1 00 ~+-H++-I-++t-1-++H+-Hrt-\

200 H-+-++-+-,H-+-++-H-+-++,-H--r-1 -200I-l-++-H++-,I-t++-H++-H+-H

-30,0 H-+++-H-+-+-H-+-+-+-H-1r--+--+-H -30 ,0 H-++-I...-+-++-t-+-++-t-+-++-H-+--H

• ,. 1\ ••

-80,0 HH-++++-HH-f+-:!.-to-t~::1-~+o••;-:;HH

-400 H-+++--oH-++HI-+++-H-+-++-HE~ - 500 H-+-+-+--I--I-H-++-+-+--I-H-f-+-+-+-i~ -60,0 I~..-- .. -0 ••

<! , II '.1-'" ', ..

.-80 ,0 H-+-+-+-IH-++-H-+-+-'.l-f-C+-IIi~°Y+-H--I

II ••

-40,0Eal

- 50,0:3a..:::;; -60,0<!

·70,0o l"o' •fl.: - ••

I... . o ..

-90,0 H-++-HH-++-H-+-+-+-H-+-+-t-1......- 100.0 ......-J..................L...JL-l.-l..-'-..L-.L....l---L..........L-.L..-I--L-l.....

_90,01-+..++-H-++-f-+-++--=H-+-+-f-+-+-t-I-1 00,0 Ll...J....l-.LJ~~..I-l.....l...J....l-.L...IL...L.....I.....l-.l.....I....J

TIME 10.5 sec/m inor divl

TOP TO BOTTOM LINK TO P TC TO P LIN K

·90,0 H-+-+-H'-++-H++-H++H-+-+-l-1000 L..J........-'-..L...JL..J......I......J-.l.....II.....I-...l-..L-I-J......l..-'--"--L-J

TIME 10.5 sec/minor divl

'n, I'..... ,I.. I':

0.0

-100

-200

-30,0

-40,0Eal

-50,0:3

- • I ... Q

;. :::;; -60,0 l.o'•. 0 0 <!

J-70,0

-80.0

H-+--+--+-+-+--H-+-+-+-+-+-H-+· ---

0.0

-100

,-". -200

-300

-40,0Eal

-500:3Q

:::;; -600 I't' .. r" .<!

-70,0

·800

-90,0

-100.0

f3'JTT(:M TO BOTTOM liNK . -Direct• -Mullipalh

BOTTOM TO TOP LI ~J K

Fig. 5.20. Frozen Lake (Low Grazing Angle) Multipath InstantaneousSignal Strength.

63

•

17.0 11.6 5.6 3,3 0,8 1,24 6 12 19 55 44

FROZEN LAKEALTITUDE: 3500 II

17.0 116 5.6 3,3 0,8 1.2 RANGE (nmi)4 6 12 19 55 44 GRAZING ANGLE (degl

. 0,0

-10,01-------1r-----t-----t----t----;

RANGE (nmi)GRAZING ANGLE (deg)

0,0

-10.0 I---f---+--+---t-----j

.... l

-30,0r-----t-----t----t---t-:-----1

- 20.01-------1f-----t-----t----t----1

.. .

...'.

_70,01---+---+--....&....:·-/·'-----\------1..• • I,

ECD:g -40.0r---t---r----r----r---"'-:10:

.~ -50,0oa.. -60,0 I---+----t--~.,..,.;:-.--..-'-1-----1..".-70,0 1---~-t---l---o---l----l-------1

- 200 I----+--R-OU-G+H--FR-O-+Z-EN----t-----;

L/IND LAKE-300 L---+-------':4~±=::'~-=--__l_--1

E~ -400 1---~--+---+---_+_-__:__t_-~..:_:1..

0:~ -50.0 \---+---+---....,j.:..,-..c:-:..-:-.-:--__t_-----Ioa.. -60,0 I-----+---+---.:{.~"~"'.............-I-"""....--l.....

-800 \---+---.+••-:••:Y---+----+----j

-900 L-__.L..-__.L..-__l....-__l....-_---'

-80.0 I---+----.t:o..••:i"• ...----i---t-,::",.-;-.~---j

-90.O.L..---'------'-----'----

TIME (144 seddlvlTOP TO BOTTOM LINK TOP TO TOP LINK

-3001r-----+----t----t---+-----j

0,0

-100

-200

-300E(I]:g -4000:w 500~0CL

.) ,)0

-70,0

-800

-900

.'. . ;~ '.I, .• II ••••••

," o''...••t'

o,O,.-------,r---,---,---....,...-----,

-10,O'I-----1f----+-----t----t------i

-20.011---t-----+---+---+----;

E(I]

:g -400'1----t---_+_---+---+-~-----1

a:: ." '. .t~ -50.0'r---+----+----+--,r-·-i-------j::> ',' • t,o ~

a.. -60,0 1---f-----;-f:-.-:-~.:..._:_f--_=____of-_;r"'9• • ...... ',, ".I

- 70,01---\---+-- +---'-'-+"'-'----/

.-8001----+--+••-='•• ,!..·--f---t---j

-90.0l....-_.......__....L...__...L...--L.-----'

T1Ml 1144 seddlvl

BOTTOM TO BOTTOM LINK . -Direct• -Mullipalh

BOTTOM TO TOP LI NK

Fig. 5.21. Frozen Lake (Low Altitude) Multipath Average Signal Strength.

64

• •

Fig. 5.19 shows the correlation observed between the multipath samples

shown in Fig. 5.18. Multipath samples taken 1.25 seconds apart were some-

what correlated. A comparison with the Sea State 1 data in Fig. 5.5 indicates

the degree of increased correlation as surface roughness decreased from

waves 1.5 feet high peak-to-peak to a nearly flat surface.

The lower grazing angle data in Fig. 5.20 was collected at an altitude

of 3500 feet over Lake Champlain. Again, a high degree of correlation was

observed between successive multipath samples received over the bottom-to-

bottom antenna link. The poor performance of the top antenna in rejecting

the multipath illustrates the degree to which the top anten~as were exposed

oto the scattering area at a grazing angle of 11. The drop in received multi-

path during the last 5 seconds of Figs. 5.20a and 5.20b was most likely due

to a slight change in heading of the transmitting aircraft which caused the

top antenna to be more shielded, by the aircraft structure, from the active

scattering surface.

The effect of aircraft heading on antenna shielding is shown in Fig. 5.21.

The aircraft were initially converging at_63° until a range of 5.6 nmi

when a heading change reduced the angle of convergence to _~Oo. At

a range of roughly 1.2 nmi the aircraft heading changed such that they

were diverging at 19°. Also, during the first portion of the flight

up to -300 seconds the aircraft were over land accounting for the very low

multipath observed initially. As soon as the scattering area was primarily

over the frozen lake surface the multipath level rose sharply. The low

grazing angle and large converging aspect angle caused a significant degrada-

65

tion in the performance of the top antenna li~k8 in rejecting multipath.

The first heading change caused the top transmitting antenna to become slightly

more shielded resulting in a noticable reduction in the multipath for the

top-to-bottom and top-to-top antenna links. As the aircraft approached and

departed from .the closest point the high grazing angles reduced the multipath

received on all four antenna links. After diverging slowly to 3 miles the

aircraft flew a nearly parallel flight. The increased multipath reduction

for the small diverging angle is very evident in the significant increase in

SMR for the top antenna links in Fig. 5.21. This was most likely due to

increased wing shielding of the top antennas.

Data waS also collected at an altitude· of 1200 feet over a frozen (~ake

Champlain). Again, strong multipath was observed on the bottom-to-bottom

link as shown in Fig. 5.22b, but since the grazing angles were higher

(minimum 110) the top antenna links performed much better in reducing the

received multipath level as indicated in Fig. 5.22a. The top-to-bottom and

bottom-to-top links had minimum average SMR values of 7 dB and 8 dB respec-

tively. While the top-to-top link had a minimum average SMR value of 17 dB.

The top antennas were not particularly well shielded by the wings during

this flight as a result of aircraft aspect angle since they were diverging

oat an angle of -54 .

5.4 Calm.Lake Surface

During a portion of a flight which was intended to be primarily over a

rough land surface a small lake was straddled which had a very smooth sur-

face. As the location of the active scattering surface moved from perdominantly

66

• I' •.

FROZEN LAKEALTITUDE: 12000ft

RANGE (nmil 0.5 4.0 8,1 11.0 15.0 17.0GRAZING ANGLE 83 45 26 20 15 13

(deg)O,o .....--+--'--t--......_+---'--.,.....---,

-10.01---+---+---+---+-----1

-20.01---+----+----+------+-----11

-30,01----+---+----+-----+----11

E~ -40,0 1---+-..,---+----+----+-----1

a:~ -50,0 I---+--~+----+----t-----IIoa.

-60,0 1---I---I-----Ot----+-----11

..... .,' I

-70;0 I---t--a:......-.........#-.~·,.....-f!··---t---I

-80.01---1r---+---+---+------l

1-90.0 '--__"'--__........__-'-__.......__~

TIME (170 serJd Iv)

TOP TO BOTTOM LINK

0,0 .-------,,----,-----r-----r----,

-10.01----t----t---t---+------i

-20.01---\---+----+----+----1

-50.01---\---+----+----+----1E~ -40.01----t-.,----+----+----+-----j

a:~ - 50,0 I----t---..;.-..-+----+----+-----j~ ..,,

-60,01---+--&-..!..·~~·..c.·.,··~···~·I_'·:...·~"::'-'-':,..'"'-+---1• .' r,". ..•.,' .

-70,01-----+---+----f------+----j

-80.01-----+---+----f------+----j

-90.0 L.--__L.--__L.--__.L...-__.L.-_--J

TIME 1170 serJdivl

BOTTOM TO BOTTOM LINK . -Direct• -Multipath

Fig. 5.22. Frozen Lake (High Altitude) Multipath Average Signal Strength.

67

land to predominantly water and back again very sharp changes took place

in the nature of the multipath as shown in Fig. 5.23. When the active

scattering surface was over the water, not only did the average multipath

level rise by ~15 dB, but the samples became much more correlated exhibiting

slightly weaker correlation than for the frozen lake shown in Fig. 5.19, but

still significant for 1.25 seconds. Since the grazing angle was relatively

high (330) the top antennas were very effective in reducing the received

multipath. The lowest median SMR observed on a link using a top antenna

was observed on the top-to-bottom link and was 17 dB, a 12 dB improvement

in the average SMR value over the bottom-to-bottom antenna link. Since the

data in Fig. 5.23 was taken from the same lake as the data in Fig. 5.18, it

is interesting to observe that in the frozen lake multipath of Fig. 5.18b

there was a -10 dB increase in the multipath level over the calm lake multi

path in Fig. 5.23 due most likely to the smoother surface afforded by the

frozen lake.

5.5 Smooth Land Surfaces

Data collected over the Mojave Desert in California and the flat plain

of central Kansas exhibited very similar properties. For an altitude of

14,500 feet (12,000 feet above the surface) the desert time-raster plot is

shown in Fig. 5.24a while at 7,500 ft the plot in Fig. 5.24b was observed.

At the high altitude the multipath tracks were similar to those observed for

the Sea State 1 ocean surface in Fig. 5.2 except that they were much weaker

68

(' (

~.3

-Hl.J

-Z~.~

-30.0

-.4'.11

~O."

1Ii

'" ~

\.0 ,. -6e .•

I

D

• -?e.O~

-a•. e

-te .•

-1••••

I I ! i I i I ! i! I,

II j

_··~--t--i--·I I

! ! I-'"T""'.-

• I : I I I I ! I j! I I I II I I

i I I ! i iI

i : I I :i ; I :I I I !

-t-~--~- ! I -: 1 I

, i I I I I I . .I : I ! i I ; I I I :

I I I I II i II I

i I TI I ! I II I

/: Rou,,~ ! ""- j I1and I 7I

: ! Ij

I ;I

1--1. _ ......L-••~~ . -.-.-~

I.. •~."='. .. '•.. - -_. . _.

: I I !~ i'Ii

It- -I •~rl i_I! I ~

J"1'& ! • r I • .t

,.".1...-r--.~ l~.l: iA. y..-, \1 I, 'i J r.I•• I

~l~F..~ fl' \ ~ cj • P I I~~ r,1

II I .. -.. I • ~;. 7- - ~ ... T IIjw1i.1. I. I I -

I •• I"..• I II I I ~. 1 I I I II

II

1I . .1 i

I I I i[

Very Calm LakeAlt: 9500 ft.Range: 4.5 nmiGrazing Angle: 33.3°Ant. Comb.: Bottom-to-Bottom

(Low level multipath scatteredfrom rough land surface).

. '" Direct* '" Multipath

Time (2 sec/minor div)

Fig. 5.23. Calm Lake Multipath Instantaneous Signal Strength.

I

~ , " !.". J ~ ' ••"

1-' "

... '

'. ;. i r·, I ' . ,•••• ~...... ~- ':1':- :: .. " •

'fi. • r. .. " '; ....j .'; ~

lJ .' l! .... " .' .''ij :"'1; "l ~t .. ;;:;;:: ,. ., ; .! _....

!! f~ •.'.~ .::: .'. /. ? ...... 0:••0" .'. -:;':.' .....:: .;-.' ';0 • ".'

" / ,,-I '. '/ '. I '. ,.. ;..

-1°·, I.',. ., , , . ., ..o 60 120 180 240 300 360 420 480 540 600

PULSE DELAY AFTER TRIGGER (usee)

'"ALTITUDE: 12000 ft

: ! I I

I !I

T'~-r- / I'I r,.. II:/'

-~-~r- •'1 J rIo

',. 'r'1 iI

. ,J 'rl '--L- -+-

1/ /..

i .! .. 'j' / ' . .'I j .

--1 .j.. -/ / /', ... I

·:--t ':( ! .:. If' .' ! .., .. . "; ••• #

..., ,.] !.. .

.. .-J " j 'j: . ; . , ...... -.1/ V. V V . .

0

·'.1;{.f /. 0 .0.. '"to ,.. I I

o

'Height above surface

]?ig. 5.24.

60 120 180 240 300 360 420 480 540 600PULSE DELAY AFTER TRIGGER (usee)

*ALTITUDE: 5000 fl

Desert Multipath Time-Raster.

70

•

DESERT SURFACEBOTTOM TO BOTTOM LINK

RANGE (nmi) 0.5 1,5 6,0 7,5 8.5 10,5 12.0 14,5 15.5GRAZING ANGLE 73.1 47.7 15,3 12.4 11.0 8,9 7B 6,5 6,0

(deg)

-10,0 ...---t----f------~--+----

-20,0 ----f---t---+---+----j

E -30,0lD~

a: -40,0w~

-50,00ll.

-60,0

-70,0

-80,0

-90.0

,...'

... .• • II • I'

I' " • II

I,IJ . .. .............

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

TIME 1192 secldivl

*ALTITUDE: 12000 II

RANGE (nmil 0.5 1,5 6,0 7.5 8,5 10.5 12.0 14,5 15,5GRAZING MG.E73.1 47.715,3 12.4 11.0 8,9 7B 6,5 6,0

(deg)0.0 r--....L-----I.._1-....J1_..J...-_+_!--'-----L_...L...--,

i I I-10,0 - -------t---+---+----,.-----'----

. I I-20.0 ---..--"'-r-----+----+-----1f----

" .

E -30.0lD~

a: -40.0w~

-50,00ll.

-60,0

-70,0

-80,0

-90,0

....

...... .. ',I. I,'"

. . ........ • •• iI-' .

, ...~ •• £& ... .::. /III • ~ .~ .. II ....

...

TIME (216 secldM

·Height above surface ...ALTITUDE: 5000 fI . -Direct

·-Multlpath

Fig. 5.25. Desert Multipath Average Signal Strength.

71

at the high grazing angles near the beginning of the flight. Also, the tracks

for the top-to-bottom link did not become significant until the aircraft were

much further apart. At the lower altitude the multipath tracks were no

longer narrow tracks as in the higher altitude case but during the early por

tion of the flight were quite broad persisting for as much as 20 ~sec. This

indicated that although the multipath detected was very low level as sub

sequent data will show, signals were detected which were scattered from a

larger area.· While such tracks were indicative of multipath scattered from

a rough surface, the rapidity with which the tracks became narrow as the 'air

craft diverged indicates that the surface was relatively smooth.

The average signal strength plots in Fig. 5.25 indicate that at higher

grazing angles (> 160

) the desert scattered multipath though not negligible

was still relatively weak. However, at the lower grazing angles the multi

path became relatively strong with respect to the direct signal, i.e.,

within 10 dB.

The SMR and scattering loss curves in Fig. 5.26 show that the total

scattering loss was much higher than for the water surfaces in Figs. 5.8 and

5.14. The increased reflection loss was due most likely to increased

absorption loss by the sandy desert surface since the smooth surface would

not scatter much incident energy away from the specular direction toward the

receiving aircraft. Furthermore, aircraft aspect angle cannot account for

the reduced multipath strength since the aircraft were diverging at the

same angle as in the previous flights. A higher absorption loss is con

sistent with theoretical considerations (Ref. ~).

72

..

•

DESERTBOTT 0 M- TO-BOTTOM ANTE NN AS

SMR: SIGNAL-TO-MULTIPATH RATIO (MEDIAN)L: TOTAL SCATTERING LOSSG: TOTAL ANTENNA GAIN IN SPECULAR 01 RECTION

ALTITUDE: 12000 ft

.-----r~--__ L

80

SMR

7030 40 50 60

GRAZING ANGLE (deg)

-CD 50"0-...J

40w>W-1 30-1<tZ 20(!).'-" -(f)

...J 10<t-..... 0zwa::w -10~

~

0 -2010 20

Fig. 5.26. Signal-to-Multipath Ratio, Antenna Gain, and TotalScattering Loss (Desert).

73

As in the previous data the SMR decreased with decreasing grazing

angle at the higher grazing angles reflecting the increase in antenna gain

in the multipath direction. However, unlike the previous data the SMR and

scattering loss decreased at lower grazing angles. This was probably due

to the smooth surface which scattered more energy in the direction of the

receiving aircraft at the lower grazing angles than the rougher surfaces

previously considered.

Figs. 5.27 and 5.28 summarize the measured SMR data for all antenna

combinations over a wide range in grazing angle. In Fig. 5.27c the median

SMR was relatively high (i.e., > 10 dB) for grazing angles greater than

_160 while Fig. 5.28c indicates predominantly low SMR values at lower grazing

angles for the bottom-to-bottom antenna link. While the top antenna link

curves in Fig. 5.27 all indicate very high multipath rejection, the top

antenna link curves in Fig. 5.28 show median SMR values below the 10 dB

level for grazing angles lower than 90.

The SMR curves in Fig. 5.29 show relatively little dependence on

altitude. These curves differ from previous curves in that the SMR con

tinued to decrease at lower grazing angles due to the smoothness of the

scattering surface as noted earlier.

The data collected over the flat plain of central Kansas exhibited

scattering properties very similar to the data from the desert except that

the level of multipath was slightly higher as shown by comparing the average

signal strength curves in Figs. 5.25a and 5.30. There was probably less

absorption loss over Kansas plain due to the moisture in the foliage present.

Altitude dependence was similar to the desert case shown in Fig. 5.29.

74

•

" .

RANGE (nmll 1.25GRAZ ING ANGLE 72

(deg)

2,5 4,0 6,0 7:3 8,5 11.0 120 14,558 45 33 29 25 20 18 15

DESERT SURFACEALTITUDE: 12000 fl

1.25 2,5 4,0 6,0 7.3 8,5 11.0 120 14,572 58 45 33 29 25 20 18 15

20,0 --r I • ao"~,

10,0~

0,0 ------- t---e------ e---+-----I ,

-10,0 - -.---j------ -----+_ -

-20,0 t....-_....ll ---L__..l..'_--I

80,0 i !

70,0 ----r-- !--- --1- - ,

60,0 I~' ~~+i-----50,0~~ ~---+----l

......~ 40,0 • ,

~ 30,0 t---t-~'~~"-,'',i'-.'....~,i"'::;.~_~!:j'==--o:--~--~(/) .,"~, r'::I::~:-:~:. -..__-_._-

20.0 /----+__-t' "1', .... ~

10,0 -----_t__ ----1--$;:."~~~-,:: ... t~~r :l=- ~ ..~..·20,0 I ' i

80,0

70,0

60,0

50,0

iii 40,0~

II: 30,0

"VI

--- ~------- .

J.-..--+~--.---+;---~I '

......I---f",",,=--+---+--·---t--_·---

.......~---

--

TOP TO BOTTOM 1I NK

80,0 1 I i

;::~t-H_:50,01-+-i---·-f---.j---i-·--1

TIME 1192 secJd ivI

TOP TO TOP LINK

80,0r---,--------------,

70,01--__-+ -_-"- ----'-__ ---_ .. --

60,01-__}-__+-__...L.. ....:.-__--l

~ 30,0VI

BOTTOM TO BOTTOM LINK

TlMl (192 secJdivl

flOTTOM TO TOP LINK

- Sensitivity threShold:tProbISMR< xl • 0.9, ProblSMR < xl • 0,5; 'p roblSMR -c-xI• 0, I

Fig. 5.27. Signal-to-Multipath Ratio Distribution Variations (Desert,High Grazing Angles).

75

..

DESERT SURFACEALTITUDE: 5000 It

RANGE (nmi) 0,5 1,5 6,0 7,5 8.5 10,5 12,0 14.5 15.5GRAZING ANGLE 73 48 15 12 11 9 8 7 6

(deg)

.~ I.~'"

• I

.\ I

i I I

.. ir~---:-- --;-'.'

0,5 1.5 6.0 7.5 8.5 10,5 12,0 14.5 15.573 48 15 12 11 9 8 7 6

1......:'-::._.......... ~'.' ''l~"'.. '.."::..~r•••~ _~'. ~~_.:,-:'

I-------I--".~ . . . . . - -:;""-i ...... .. ,~.•_ • i : .. :-'1 1 ',' ..

I----+I-----r- . ...... .. ....

.'-~t----+~l-~----10.0"·--+- -- --·1-------1 ,.:.. - ..

I I : .-20 ,0L-----" .....l.. -----.l

80,0

70.0

60.0

50,0

iii 40,0Ea:

30.0~VI

20,0

10,0

0,0

'.

!I

................20,0

80.0

... ..' "' ..' ,.' ".. ,' ~

10,0 f---"~---+--.L,~_...;.~.'-.. 4.-c...~.;-.......:._.•- ~II'i;-;.,.--......... -t

0,0~I I • i" .;.. .-10.0 1-,-. I [,!

I-20,0 L-__:....-__~_----J__......;:...-_--J

TIME 1216 seddlvl

TOP TO BOTTOM LINKTOP TO TOP LI NK

0.0 -----·-1---10.0 !

I

r'. '.

i.

I

: : : : .. " .. "..". " - ~.....-.. . . .... .

. .;-'-

..../ ~ ....• ; ';';""-: -V

'.'..:."-,,.- ~..'...:::.-..~-:-_. --~,....;---.. -. "'.-"""", ..

.'.,"--.

.. '

. '\

80.0

70,0

60,0

50,0

iii 40,0Ea: 30,0~VI

20,0

10,0

80.0 I I I -1.-70,0---.~.-.r-' -,--.T- ..--60,0 >---~•.-.-+-!-'~+----+I----1----1

"'-. I50,0 ~--'...cyl---+---+----+-----t

-100-

_20.0l-_---L__....;... L-_-.l

TlMl 1216 seddivlHOHOM TO BOTTOM LINK BOTTOM TO TOP LINK

. Semitivity threshold;tProbISMR< xl • 0.9

. ProbISMR, xl" O.~; ·proD1SMR.::XI· 0.1

~ig. 5.28. Signal~to-MultipathRatio Distribution Variations (Desert,Low Grazing Angles).

76

( (

DESERTBOTTOM-TO-BOTTOM ANTENNAS

•

70

60

50

CD 40'0-

..... a::: 30.....~CJ)

20

10

00 10 20 30 40 50 60 90

GRAZING ANGLE ( deg )

Fig. 5.29. Altitude Variations in Signal-to-Multipath Ratio (Desert).

Flat PlainA1t: 10,000 ft.Ant. Comb: Bottom-to-Bottom

. _ Average Direct

* - Average Multipath

14.3 18.014 11

8.0 11.523 17

5.731

I -, -, T

.......

.... '

..-... .• .

.' • .'• ·- •• oooo .......... "

• • '. . . .• .,. . .. ...-......

• • ••••• ............ •• ....• ......• ...... ."..•

Range (nmi): 0.3 2.5Grazing Angle (Deg): 85 54

•••

-4e.'

'-JP -!;e.1

000II[

~ -6e.1,D

•.. -"N.II

-...-91.1

Time (200 sec/div)

- .Fig. 5.30. Flat Plain Multipath Average Signal Strength.

(, ( •

The multipath signals scattered from the surface of both the desert

and the flat plain were well defined pulses as in Fig. 5.la with the multi

path pulse generally several dB lower than the direct pulse.

The correlation properties of data from both terrains are summarized

in Fig. 5.31. Most of the multipath pulse seq~ences were weakly correlated

as shown by the lower curve. Occasionally, bursts of multipath pulses were

observed which showed very strong correlation as indicated by the upper

curve. Such instances were most likely due to locally very smooth areas

over which the active scattering area passed.

Figs. 5.32 and 5.33 summarize the variations in SMR as explicit func

tions of aircraft geometry and antenna combination for the desert and flat

plain surfaces respectively.

5.6 Rough Land Surfaces

In this section multipath data for very rough scattering surfaces is

presented. Flights were conducted over the suburban Boston area and the

White Mountains of New Hampshire. In all of these flights the multipath

observed was not pulselike but rather appeared as weak noise level of long

duration which eventually decayed as pictured in Figs. 4.2 and 4.3b. The

dispersed nature of the multipath is clearly brought out in Fig. 5.34 where

the multipath tracks for the bottom-to-bottom link were much broader than

for smooth wat~r or land surface as in Figs. 5.2 or 5.24a. Successive

multipath samples varied rapidly over a 20 dB range as in the first ten sec

onds of Fig. 5.23. The samples were essentially uncorrelated exhibiting

less correlation than shown in Fig. 5.5b.

79

CORRELATION BEHAVIOR

- - USUAL• - OCCASSIONAL (VERY SMOOTH)

DESERT

ALT: '2000 f1RANGE: 5,3 nmiGRAZING ANGLE: 36.7°

2.01,0

TIME (sec)

0.5

OL.- ----lL.- ---L .....J.... --l

o

2.0

0,5

00o

Fig. 5.31. Normalized Autocovariance (Smooth Land).

•

DESERT

20

20

WOR§TSMR :: adB

WORSTSMR =6 dB

15

15

TOP-TO-BOTTOM ANTENNAS

BOTTOM-TO-BOTTOM ANTENNAS

10

10RANGE (nmi)

5o WI!.::......- -'- ---'- ......I...- ----"

o

5000 1-------....L-~....L..----LL.l..4..--.....LlILt'""-------'

2010000 t===I=====1::==:]!ZlJ2~~

5000

-....wWlL.- ",7° G, A,W 00 0 5:::>....-I-.-J

~<t

10000

~ig. 5.32. Geometrical Dependence of Signa1-to-Multipath Ratio (Desert).

81

r=;=r==;p:===========,------"'/I FLAT PLAI N

20

20

15

1510

TOP-TO-BOTTOM AN"rENNAS

10RANGE (nmil

5

5

12,60 G.A. BOTTOM-TO-BOTTOM ANTENNAS

OIoC.- --I. ---l... ....J...- -'

o

10000 t==2JO~lC5J10::========~~~~

5000

5000

LIJo:::J~

~«10000 1=--~2~0""'-------='1'=""5----~~"""""'"'"'~""'i'

-

Fig. 5.33 .. Geometrical Dependence of Signal-to-Multipath Ratio(Flat Plain).

82

I I Iii ! i- ! ,I +._~)-_.) ,1__1 _

---:-':l~' !' • ..;. - - r-j" -;c.-. ,/~7 ;. ; --• i ' 10" I,. I

.. I 'i i" -.. I' I . "l I,' !-.=-~____+_~~f------"-+--+--

, I .1 /1 / ~I . i' _ \ _. II I' . l ../ : /. ../ '

t----~i--_+--·-I-~IW+'-~_+-'J'i__+_~"'-+_-~f_-_+--~

I .. . / .;. I I .1 /i ~.r/. ~, 1/

•

• -. :,¥i4. ~/' • ",,'" .,-'.' .~ /..' . .'. ..' ,.,

./ .~ ~/ .;, ..~r lr ...'

,,- ~ ..' ",. I. ' / ...-~-A1J(/: ·./i '1: . ,..

,s,o i~' ::--~ --t'• ~ ~;..J ••1' ~":•.J.. ·.1 .• :_--.•.

I '.l' .:., i/ .., . r••I.~ ' .. ". I~,' •. . I E:'1'" to .".- I' , ../. "-.7: ..:', . . I .I 'j ~)4 .1l~ "~I , - •

., .(1., L ·ltIt 10 I I

-" ..~--:- ;..~.:- .

• I

~..-". - ~

~ .• :. •• 1----1--

Rough LandAlt: 9500 ft.

t. 12'. 11•• 24'. 31•• :M'. "2'. 4". 5.... 'M.

Fig. 5.34. Rough Land Multipath Time-Raster.

83

Fig. 5.35 shows SMR distribution variations for one of the suburban

flights. Fig. 5.36 shows an average signal strength plot for the same flight.

Except for the indicated lake and ocean regions. the multipath was essentially

negligible. (At the high grazing angle in the lake region the top-to-bottom

antenna link provided a 10 dB SMR improvement while at the lower grazing angle

of the ocean section there was only a 5 dB improvement). The data in Fig, 5.37

was gathered over the same area covered with snow. While the multipath from

the snow covered land was about 7 dB stronger than in Fig. 5.'36 it was still

very weak with respect to the direct signal.

In Figs. 5.36 and 5.37 the multipath level changed very rapidly when a

water surface was encountered. The lake straddled (Lake Cochituate) was

quite small. The stronger multipath signals received under these conditions

were discussed above in Sections 5.3 and 5.4 in connection with Figs. 5.18

and 5.23 (which correspond respectively to the lake regions in FigsJ 5.36

and 5.37).

The average signal strength plot in Fig. 5.38 indicates that negligible

multipath was scattered from a forest covered mountain area. The data in

this section implies that multipath scattered from rough land surfaces is

not a significant form of interference.

84

•

, I

--..-.-r-

~J< Lake ~cea~L' ~.....,.~. ---

L ---:------.... --. . ~"r..... • :1'_':-• • I • I ,-...~...

• I.... ...'. -.... . _.. , • '2lt' 'I' 9'&_. • .\ t,-I • , '-

I I I - -I"-a '-III I I , ". ',-•... .., , +, 1 . , I,•. 'I , •.. ....-

('

Range (nmi):Grazing Angle (deg.):

at .•

7'.'

M .•

W.•

40."

3e."

00 SVI I':

R 21••

tI

• 1'.'

•••

-1'.'

-M••

1.564

4.057

6.326

8.221

(

11. 315

14.412

17.010

19.09

Rough land (suburban Boston)Altitude: 9500 ft.'Bottom-to-Bottom Antennas

- Sensitivity Threshold+ Prob [SMR < xl 0.9

Prob [SMR 7 xl 0.5* Prob [SMR ~ xl 0.1

( •

Time (240 sec/div)

Fig. 5.35. Signal-to-Multipath Ratio Distribution Variation(Rough Land).

Rough LandA1t: 9500 ft.Ant. Comb: Bottom-to-Bottom

.....Average Direct*..... Average Mu1tipath

11.3 14.4 17.0 19.08.26.31.5 4.0g): ~4 5I 2

16 '11 If 12 10 9

~

I I I

I I I -,i

I I

i II,

---- I I I !

II I ,

I r!

I I

Ii

Lake,f

Ocean. I I~I

•J..".

..' ..

'" " ...- ......... .

• ..'" .....

• .... -.~ •

.............- .,.• -, • .~

II,

fta/a"." • ..'.-- • • .... • r! '.. , ,.... ;. ....."..,...--~

-4'.'

It -5'.'eII£

• -H.'ID

•" -78.'J

-....-81 ••

-3'.'

Range (nmi):Grazing Angle (de

r.

Time (240 sec/div)

~i~. 5.36. Rough Land Multipath Average Signal Strength.

,

(

Range (nmi):

Grazing Angle (deg):

1.2

69

3.5 6.2

42 27

8.1

21

(

10.0 11.4 13.2 14.0

17 15 13 12

(

i: '.,

-~~~ ...,

-Z~.~

-3~."

-"".t'ex>'-J

p -5•• 00

'"ER -6f.8

(

DI

" -78 ••J

-8'.'

-M.'

1 1 I I I II

II Ii i t

I ! i; I _.+------+-...------r-- II

I I IL

~~:~-h,.. Lake

I~

I

. ,l 1(' )1..I .I . ..•.. . .....• .. ....... ...

• .......... . ... ...... ...• • • •• ....u. I • ••• I I • • I • I •.,I -,

I' - • ,- I

• In,I'II' ., • I I

" I "I I II I

"I , I

I

I

Time (156 sec/div)

Snow Covered Rough LandAlt: 9500 ft.Ant. Comb: Bottom-to-Bottom

. _ Average Direct*-Average Multipath

Fig. 5.37. Snow Covered Rough Land Multipath Average Signal Strength.

Range (nmi):

2 7 3 3 10 0 14 2, "I 1 I

4 4I

6 7r I

.:~ .0.

Tree covered gentle mountainsA1t: 8500 ft. (average)Ant. Comb: Bottom-to-Bottom

. _Average Direct* -Average Kultipath

III! ':~J"....:.....II.

_' I • I" IJI I 1,"1 '. ,. LI~"II .. II

l'fl .. I,. .I I _I._______--fI -.N_

i---- -------.----- -----+-I-------,---------t----.-- -

I • I

-----+-:---!---+--I---t--.. '~"':'" I -J,II----+--!----I

! -1"-........ I-------r-i------11--·':7""•••-:-•••-.-+!----I-----i

! I ., I

-2('.t

-J~.e

-9'.'

-....

-4'.'

00p -S'" J....--.

00 'J~

£II -&e.'

t1I" -?e •• •

Time (230 sec/div)

Fig. 5.38. Forested Mountain Multipath Average Signal Strength.

5.7 Banking Over Ocean Surface

In order to investigate the sensitivity of the various antenna combina

tions to banking maneuvers, a flight was conducted over an ocean surface in

a Sea State 2 condition. The aircraft followed a roughly parallel flight

path with a nominal separation of three nautical miles at an altitude of

5000 feet. During the flight the receiving aircraft continuously banked

alternately 300 away from the transmitting aircraft (indicated by _300

in

Fig. 5.39 then 300

toward the transmitting aircraft (indicated by +300).

Fig. 5.39 shows average signal strength curves for each antenna combination

for a portion of the flight.

Multipath received on the bottom antenna experienced small variations

as the aircraft banked while multipath received on the top antenna experienced

large variations in signal power. Direct signals received on the top antenna

experienced smaller signal strength variations than those received on the

bottom antenna. Direct and multipath signals were out of phase on the top

to-bottom link while they were in phase on the bottom-to-top link. Even

though the aircraft were at a high grazing angle it is evident that banking

degraded the SMR improvement of the top-to-bottom link when banking toward

each other. The bottom-to-top link also suffered a slight degradation in

SMR improvement from the level flight condition since banking away from

each other evidently tended to shield the top receiving antenna from the

direct signals while allowing slightly more exposure to the multipath signals.

The top-to-top link continued to provide high SMR values at all bank angles.

89

'...I-30 0 +30 0

-30 0 +30 0

I I tlCMO ...." I I nc.. .......0,0 0.0

-10. 0 -10.0

-20.0 -20.0

-30.0 -30.0E...

_ -4>.0. :g -4>.0", IX ...

.15 ..~ . '

: -50.0 0 -50.0l1.

~ ' ...., , ,lI. ~.O ,,' .' ... ~O, .

•••••u ••• , " ".. . '. . , , •'I.··' ..... ........ .,'"

o •

-70.0 " -70.0 • , . 10' .. ..• • ...

• "-11.0 -11.0 ..' ..-'MI. 0 -90.0

0 !42 283 425 566 708 0 142 213 425 566 , 701TIME (secl TIME (secl

TOP-BOTT<N TOP-TOP

"CMI~tf - '1=- ....."0.0 0.0

-10.0-10.0

-20.0 -20.0

-30.0 -30.0

ECD

i -41.0S! -41.0IX

.... ..

I ":sa

-50.0 .. .. ..' ,', a:: -50.0"

.... ........ ~ .' .• .. .. ' .. '.• I' .. Go " . '" .

~.O,

~.O .', , .'

:' '. , '.-70.0 -70.0 .. ,

"

-Sl.O -11).0

..... 00

-9ll. 0142 283 425 566 lIN 0 142 283 425 566 701

TIME lsecl TIME !secl• - direct AVERAGE SIGNAl STRENGTHI - multlpath BOTT<N-BOTT<N BANKING OVER OCEAN (SEA STATE 21 BOTT<N-TOP

AlT: 5000 tt

Fig. 5,39. Multipath Sensitivity to Banking.

90

Acknowledgment

A great many people who were involved in conducting the Air-to-Air

Measurement Program have my appreciation for their cooperation. The success

ful completion of the operational aspects of the program was due in large part

to the efforts of Everett A. Crocker and Joseph V. Gagnon. The insights and

encouragement from William H. Harman and Jerry D. Welch throughout the project

have also been greatly appreciated.

91

REFERENCES

[1] G.V. Colby, "The Airborne Measurement Facility (AMF) System Description,""Project Report ATC-60, Lincoln Laboratory, M.LT. (25 March 1976),FAA-RD-75-233.

[2] P. Beckmann and A. Spizzichino, "The Scattering of Electromagnetic Wavesfrom Rough Surfaces," Pergamon Press Inc. NY 1963.

[3] K.J. Keeping and J.C. Sureau, "Scale Model Pattern Measurements of Aircraft L-Band Beacon Antennas," Project Report ATC-47, Lincoln Laboratory,M.I.T. (4 April 1975), FAA-RD-75-23.

92

..

A.R. Paradis 6 September 1977

Documents

Transcript of A.R. Paradis 6 September 1977