Upload
others
View
7
Download
0
Embed Size (px)
Citation preview
7~
AD-763 9444
MTI (MOVING TARGET INDICTORS) TRACKINGOF A REENTRY VEHICLE IN STRONG ATTACHED'WAKE
-I •M. Arm, et al
Riverside Research Institute
SIN
.•_•_• •_Prepared for: .
• • \White Sa~nds Missile Range
• • 15 June 197,3
DISTRIBUTED BY:
Nab"onl Technical Infomfi• Service .AU. S. DEPARTMENT Of COMMERCE5285 Port Royal Road, Springfield Va. 22151
" I
RIVERSIDE RESEARCH INSTITUTE
80 West End Avenue / New York, New York 10023 /1(212) 8-O-4000
,5 june 1973
TECHNICAL MEMORANDUM TM-64/241-4-60
MTI TRACKING OF A REENTRY VEHICLE
IN STRONG ATTACHED WA'(E
By
M. Arm and I. Wesssman
Prtpc-,'d for
Commanding Genera'
WIk,-te Sonds Mi;sile RangeNew Mexico
e8002
Contract No. DAAD07-72-C-0128
-•; ~~~~NAP-,r'4 .. A "LIC:.
C-DCI ,,,, II,
,1.7
16u'ill
+_3
++ D-A__________________________L +h z ->'+-+- . . ._L --.:---:- T U L-
RIVERSIDE RESEARCH INSTITUTE
AUTHORI ZATI ON
This report describes research performed
at Riverside Research Institute and was pre-
pared by M. Arm and I. Weissman.
This research is supported by the Depart-
ment of the Army, White Sands Missile Range,
New Mexico, under Contract DA.AD07-72-C-0128.
Approved by:
N. B. MarileDirector.
Radar Division
.- 6 4 /, -4-60 1 -4----
II ... RIVERSIDE RESEARCH INSTITbIE II. INTROIJCTIOUz
This memorandum presents data, o1g-ained by the AMRAD
2; radar at the White Sands Missile Range, which demonstrateSi the utility of _MTI clutter cancellation techniques for im-
proving the tracking of reentry vehicles in the presence
of strong wake echoes.SIf usprsewake returns within thetrcigae
of aBBMD radar could degrade the quality of metric data or
even steal the RV track, with ccitsequent detrimental effects
on the designation, discrimination and intercept functions.
MTI (moving target indicator) waveforms and processing tech-
niques provide an fficietit Ieans of wake suppression by
cancellation of echo Doppley components differing from that
of the RVM Other methods for eliminating or reducing the ef&* -fects of wake exist, including use of wideband tracking pulses,
coherent burst waveforms, or tracking using a long up-chirp
S-pulse, and these approaches may be advantageous under vari-
ous conditions. However, M.1i has the advantages of not re-
quiring particularly large pulse bandwidths, of not having
a long waveform duration more prone to ambiguous rangealecho interference, and of not being wasteful of radar energy
resources (as may be the case with a long pulse).
s As de:cribed in Ref. 1, MTI processing can be used to
suppress either the body echo or the wake echo within the
same range resolution cell. An application in which the wake
echoes are canceled in data processing in order to permit
nmeasurement of RV cross sections in strong wake is discussed
in Ref. 2. The data presented here are intended to demonstrate
S1 7--64/241-4-60 -1-
RIVERSIDE RESEARCH INSTITUTE
the performance of a range tracker, using a real-time 3-pulse
canceler, which has recently been implemented in the AMRADradar. Section II, which follows, briefly describes the
principle and the expected performance using MTI processing.
Section III discusses the AMRAD data obtained during the
&. RANT-01 flight and analyzes the performance of the MTI tracker.
II. THEORY
The improvement of tracking performance in wake clutter
I• is accomplished by pulse cancellation techniques similar to
I those commonly employed in MTI radar. Several coherent pulses
are transmitted., as indicated in Fig. Ia. Upon reception,
the signals are processed by forming a weighted complex sum-
mation of the echoes corresponding to the individual pulses,
thereby synthesizing, in effect, a Doppler filter. Cancella-
tion can, for example, be achieved at a particular Doppler ve-
Slocity by forming the summation of echo voltages E:,E2 .... Em
for m equally-spaced pulses (of spacing TS) as follows:
E kE + k EC + kmE m "' (l\
where, for achieving a null at the Doppler velocity, Vnu 11'
the phase weighting T = 4VniT/A. is used (for radar wave-
4i length )), and the amplitude weights are chosen to satisfy
the condition k. k 0. The residual value of E then:j-i •
corresponds to the uncanceled components. The Doppler separa-
tion between peak -and null of the filter characteristics is
equal to W/(4TS) so that, at L-Band, spacings of between 8
and 12 ,Jsec would be typTical under ICBM conditions in order
to suppress the low-velocity wake components. As indicated
TM-64/241-4-60 -2-
-• L
RIVERSIDE RESEARCH INSTITUE A
IT
I ~ timeI
*JCANCELLATION WAVEFORM
7j 7
0 [~ r
Sdp
-~05--~....*4*[0s
- m C---ii_ _ _ _ _
0pO :1Z
B;.i ASIN: WAVEF"OP. A-SO, CHARACTERISTICS-
T-64/241-.4-60
RI VERSIDE RESEARCH INSTITUTE
in Fig. lb, the use of a simple two-pulse canceler (k 1,k 2= -1) results in a relatively sharp null in the Doppler
filter characteristic. For suppressing clutter echoes char-
acterized by a large Doppler spread E broader wake rejection -
characteristic may be required, such as that obtained by the 0th~ree-pu l.5e or four-pulse canceler. (For the cases shown,
a bnomalsequence of weighting coefficients was assigned;
in enealmore optimum filters can be synthesized, depene.-
.Ln ontheclutter echo spectra, by other cc. binations of
weihr- orby nonuniform pulse spacings.)
Figre2 gives the degree of wake RCS suppression (power
a1 wake r.m.s. Doppler spread, aV/V~. frm=2- - n4-pulse cancelers, and -for different a-mounts of misaligment be-
tween the null Dopple.,. and -the true mean wake Doppler. Xt is
see tht or n ~m.-.wake Doppler spread as large as 1-0 per-
cen o th bdyDoppler~ velocity, a 4plecanceler can
achevemor thn 3 dBof ,7ae sppresson orcases where
themea wke oplerhasber. ccratlypredicted. This
suppesson igue i no sevrel dci~aed venfor a meanwak~ Dople whch i unnow tothe extent; of 10 percent of
-th boy vlocty.Thee ol-clatonswer baedon the as-sizptonof Gu~sia wkeDoppler setu;actual spectra
havngdaifeentchrace istc esecall ot o te-ir ta-Us,car' substantially alter the~ser~ls-
Ncteý that, in practice, the actual azhievable wlake. sup-
pression. could be 'Limited, -by quanttizati-on coarseness it, analog-to-dig-ital. conversion (if a digital traýcker is iised), by
21nstabilities in the radar, or by spixrioais ampli-tude o~r phase
madialatiUon.
T14-64/241 -4-63 4
RIVERSIDE RESEARCH MSONTE
"t -PUI,5E CANCELER-INOMIA,. WEIGHTING COEFFICIENTSGAUSS'AN WAKE SPECTRUM
-MEAN WAKE DOPPLERo'V: r.m.s. WAKE DOPPLER SPREAD
60 vpa• : v9: SODY DOPPLER
1 50
I m:4ý/V 9 :0
fe I*--.
Ft. AH VBVAWK 6 P0ESON!•
II
-20
gv o
0O 0.05 0.10 015 020
I *FI0.2 ACHIEVABLE WiAKEr SUPPRESSION
4 r-s4-64/241-4-:60 5
27
PRIVEFISIDE RESEARCH INSTPIMIE
on the raw pulse echoes, both range and angle tracking can
Sbe improved. In most situations of interest in a terminal
defense geometry, the wake would introduce a negligible"amount of angle-tracking scintillation (glint) relative to A
the noise-determined monopu'-se error signals. However, the
vresence of uncanceled wake in the tracking range cell can
in fact limit t-he range tracking precision, as will be illus-
trated. When wake cancellation waveforms are employed, the
r.a.t.. 1'•nge tracking jitter for a steady (nonfluctuating)
reentry x..il cross section can be expressed as
K n ( /R;'S.D. (r) n g w :
b
where %, n and q. are the RCS values corresponding to
the body, integrated receiver noise, and wake clutter within
the body range resolution cell, respectively, B is the band-
width of the tracking pulse and K depends on the tracker Char-
acteristics (including the tracking loop dynamics and smoothing:
tracking PRF, leading edge threshold criteria, and detailed
tracking pulse shape). Tnis expression is only valid for suf-I ficiently large ratios of body to random-echo signals (qb /a
Rb/6w 3l1). Also, if signrificantly large body RCS fluctuations
are present the tracking quality can be affected. These latter
effects, as well as the effects of receiv-er noise, are not re-
duced by the cancellation process. (Techniques are available
for coping -with the effects of body RCS fluctuations.)
II . Br.SLTS-
The results presented here were obtained from data col-
lected on the RANT-01 payload (Athe-na 137) flown on 12 Jan--A
T?4-64/241-4-6 0 -6-
S-- -
RIVERSIDE RESEARCH INSTITUTE
uary 1973 at the White Sands Missile Range, which was the firstflight test for which the AMRAD MTI tracker was operational.* A
In order to assess the effectiveness of the MT.L technique twoindependent range trackers were used at AMPAD to simultaneously
track the payload during reentry. The principal tzacker em- i
ployed a single pulse (3 Psec compressed to 0.1 p-sec). The
MTI tracker, which operated on the outout of a cancellation cir-cuit, used 3 pulses, each 0.1 P-sec wide,with a fixed spacing
4of 10.5 ;Lsec. Binominal amplitude weighting coefficients
(1,-2, 1) with no phase weighting (i = 0)were employed for thistest. Thus the transfer function null was at zero Doppler and,since the AMRAD frequency is 1300 MHz, the peak occurred at a
Doppler velocity of 5.5 km/sec.
A block diagram of the three-pulse cancellation circuitermployed is shown in Fig. 3. (A conventional analog design ap-
proach which would adequately demonstrate the tracking techniquewas selected.) The first loop provides the amplitude weighting
while the second loop provides the proper delay (and phase) to
form a null at the desired wake Doppler frequency. Two powerdividers (PD1 and PD.) and one gate (Gl) perform the amplitudeweighting function with coefficients k1 - 1, k 2 = -2, k3 = I.
Cable (0 ) cut to the proper length is used to maintain proper"1 • phase between the gated and ungated signal paths. The second
pulse of the received triplet is gated on by G1 as shown in the
timirg diagram. Power divider PD2 drives the cancellation loopwhich introduces a 10,5 lisec delay and a loop gain of two. (Theexact delay is adjusted ay the proper iezngth of cable 0,-N The
7-' output of the cancellation circuit feeds the MTI range trackerwhi-rh is gated on with the third pulse. In practice, this cir- A
cuit achieved a cancellation ratio (R') of about 30 dB.
-h* e PANRT-01 target was flown under the sponsorship of the V- -tUS. Air Force, SAMSO. AMRAD data are presented in Ref. 3.
rXRP/A1A46 -7-
RIVERSIDE RESEARCH INSTITUTE
I I
MIDi
0li
al
z go-: - 0
c-i IN
SFIG., 3-PULSE ANALOG CANCELLATION CIRCUIT EMPLOYED AT AMRAD
-1-q-4-64/241-4-60 -8- A
RIVERSIDE RESEARCH INSTITUTE
Both trackers operated in a leading edge mode at a wiuve-
form repetition rate of 50 Hz; the trackers were identical in
all other respects, including the tracking loop time constant
(approximately 0.1 sec).
N6 attempt was made to incorporate an MTI waveform for wake
suppression in the AMRAD angle tracker, which uses a 1.2-ýLsec
uncompressed- pulse. For "he RANT-01 flight, the slant range at
maximum waking was approximately 125 km and the radar aspect
angle was 20 deg. Thus, the angle subtended by the wake within
the angle tracking range cell was at most 0.5 mrad, which is not f N
significantly greater than the single-pulse angle tracking jitte.L..
due to noise alone. (In fact, the wake had no discernible effect
upon the angle tracking quality for this test. 3)
Fig. 4 shows the range vs. time history and the range trank-
ing residuals obtained during data processing for both the MTI
and single-pulse trackers over the interval (313 to 32 sec) forSwhich comparisons are meaningful. The residuals correspond to
the tracking fluctuations about a 2-sec running polynomial fit to
the range vs. time history. Fig. 5 shows the reentry body and
wake RCS in the body range cell, the wake mean Doppler and Doppic :
spread in the body range cell at various times,* and the sup-
pressed wake echo RCS calculated on the basis of both the ideal
cancellation ratio (as determined by: the wake Doppler moments)
and the actual cancellation ratio (29 d3) measured for the cir-
cuit employed. For reference purposes, the equivalent RCS acorresponding to the receiver noise is also shown.
The effects of wake upon the single-pulse range trackingcan be observed in Fig. 4, particularly when strong attached
* The body RCS and wa3ce RCS and Doppler measurements were ob-tained by burst waveform Doppler processing.
TM-64/241-4-SO -9-
.61I
RIVERSIDE RESEARCH INSTIUTE]
Li NMI
so in
14 IK
1~L0A
zlnl~ OV
FIG.4AMRA RANE TRCKIN DATTM-64/41-4-0 -1)
_______________________ - -u---
-- - - ----------- -- �--.---�---�---�-----F UVERSIDE RESEARCH ESTITUIE
-f ALTITUDE (ka�)
40 35 30 25 20 IS
24 -4
1 � 0.02C) A I WA�CE DOPPLER M0MEt�TS;
�c. 6V/VPEAK
0.005 -- K, i' 15�a.tsr SQOY RANGE CELL
0.002W �PEAK � . -�
4 ________________________
p0.001 I I I
01��� RCS� V�b(O.lsecAVERAGE)
. tI...
-�
a0I �!01'�
-- WAKE RCS. o�(0.�-sec AVERAGE IN15 mtsr BODY RANGE CELL)
ISmeter 301W RANGE CELLK q�/R (ft u TNEORETICAL WA�CE
** NOISE LEVEL-Mn TRACKER @�/R' IR"AC7UAL CANCELLATONa29dB)(3 Q.1-�&sec PULSES)
I
- - - - - - -
xx '1NOISE �c �x
�eoh LEVELSIt4GLE PULSE TRACKER0. (3ps.c-.0�p&.ec)
'UI ...7�I___________ j4 I I* 513 314 315 �'16 317 318 319 320 321 322 1 �
*1 - TIME (see)
A91 F�S. 5 AMRAD DATA RELEVANT TO MTI TRACKIZR PERFORMANCE FOR
RA#�'O� FLIGHT i<I !I�M- 64/241-4-60 1�
-- - - -. - - - -�
RIVERSIL'E RESEARCH INSTITUTE
wake is present (317.5 to 319.5- sec). During the entire inter-
val the wake RCS exceeds the noise level and therefore contrib-
utes more to the -tracking jitte-r. tor the 14TI trackleT. the
* ~wake echo is suppressed to levels where it is negligible corn-
pared to the receiver noise, except in the interval 317.5 to
319.5 sec when-it somewh-at exceeds the noise (see Fig. 5).
during this latter interval. In any case, note the general irl-
provement in tracking performance obtained with the MIT tra-:ker.
More striking examples of the tracking improvement using
MItechiiiques have been obtained for more recent WSMR fli.31hts,
frwhich the alata are classified. Those interested in those AXrecent data are invited to contact M. Arm at i~he AMRAI) site forfurther details.
ACKNOWLEDGMENT
The authors acknowledge the Contzributions of D. Grieco and
~ I W. Edelson for helping to develop the M1TI technique for reentry
$measurement applications (Refs. 2 and 4), of K. Krabbe and 3. Harrell t 2
for implementing the circuitry, and of F. McIntyre for assistance
in the preparation of the RANT-Ol1 data.
REFERENCES
-f eissman, I,, "Low-A1*-itude Discrimination Waveforms (U%,,"in Proceedings of t.-: 1.97i le5w-Altitude Workshop, pulb-lished b~y River-side Resaarch Institute, 4., 5 M.arch 1971,ISSecret,/RD.
2. Edelson, W., "A Cancellation Technique for Processing RadarReentry M~easurements (u),, Riverside Research YnstLituteFI Report TM-45/186-4-10, 1 January 1973. $ecret/AD.
3. "Preliminaryv AMRAI) Data Report,, Ath-ena, Flight No. 137 I,H005OJ/RAIU-.01< Ri.verside Research Institute ReportPADR-2.5/241-4-l0, 9 February 1973, Uncl:-,ssified.
4. Weissman, 1. and Grieco, D., "A Low-Altitude Discrimina-
side Research institute, Technical Memorandum TM-33,/077-4-00, 30 June 1970, Secret.
jTM-64/241-4-60 -12-
- - -V-0