2004 IEEE Aerospace Conference Proceedings

Precision Dynamic SAR Testbed for Tactical Missiles12

Brian I. Smith Wayne Gamer Randal Cannon
United States Army Aviation and Missile Command
ATTN: AMSAM-RD-MG-RF
Redstone Arsenal, Alabama 35898
256-876-1330
Brian.Smith@rdec.redtone.army.mil

Abstract-The United States Army is currently investigating advances have been made in real time generation of SAR
the feasibilityiutility of a high resolution Synthetic Aperture images and image resolution, cost of S A R for a tactical
Radar (SAR) mode for the detection and subsequent missile system is still a significant concern. The typical life
engagement of targets by tactical missiles. A precision expectancy for an Anti-Tank Guided Missile (ATGM) is
dynamic S A R testbed has been designed, constructed and typically measured in seconds, or at most, minutes; the
installed by the United States Army Aviation and Missile Army simply can’t afford a $100,000.00 seeker on its
Command‘s (AMCOM) Aviation and Missile Research, ATGMs.
Development and Engineering Center (AMRDEC) at
Redstone Arsenal, Alabama. The purpose of the testbed is 2. Background
to allow the development and evaluation of image formation The US Army is currently developing the next generation of
algorithms, motion compensation techniques, and autofocus ATGMs. This next generation of weapon systems must he
algorithms in a controlled and repeatable environment. able to operate at extended ranges, in adverse weather, and
battlefield obscurants. Active millimeter wave (MMW)
TABLE
OF CONTENTS radars are a proven choice for the detection and engagement
of moving targets. However, MMW seekers normally can
1. INTRODUCTION....................................... 1 not utilize real beam operation against stationary targets at
2. BACKGROUND ........................................ 1 extended ranges due to ground clutter masking the desired
3. PROBLEMFORMULATION .......................... 2 target. Doppler beam-sharpening (DBS) can be an effective
4. TESTBED REALIZATION ............................... 3 solution for improving the signal-to-clutter ratio of targets at
5. INITIAL TESTRESULTS .................................... 4 longer ranges, especially when the target volume in the
6. CONCLUSION ......................................... 4 radar footprint is low and the location of the target is known
within reasonable accuracy. However, in situations where
these caveats aren’t satisfied, DBS has serious limitations.
1. INTRODUCTION
Today’s battlefields often lie in urban environments with
The testbed was originally conceived as a tool to support the military targets interspersed among civilian, non-
development of a S A R mode for the AMRDEC Multisensor, combatants. In these situations it is imperative that missiles
Multimode Suite ( A M M S ) [l]. The AMMS is an be as accurate as possible. At the same time, threat anti-
instrumentation system under development to collect target aircraft systems are becoming more effective at longer
and background signature data and is designed to be flown ranges. However, as engagement ranges increase so does
in a specially configured UH-1 aircraft. The testbed will be the target location error. Automatic target recognition
used to validate the operation of the AMMS S A R mode techniques have been employed in infrared (IR) seekers for
prior to integration into the UH-1. One of the early many years. Unfortunately, IR has the disadvantage of
applications of the testbed will be to verify the proper use of being relatively short range, and suffering significant
inertial measurement unit (MU) data in the development of degradation in adverse weather. S A R techniques have not
motion compensation algorithms. been an option for ATGMs in the past due to the associated
data processing requirements and cust. Recent advances in
In addition to supporting the AMMS system, the testbed will microprocessors and the reduction in size and cost of MMW
be used to perform trade studies between radar hardware, components resulting from advances in the
image formation algorithms, IMU quality, etc. [2,3,4]and telecommunications industry have now made S A R a viable
seeker cost for tactical missile seekers. While significant candidate for seeker applications. In turn, SAR techniques

’ U.S. G o v m c n t work not pratcctcd by U.S. copyright.

IEEEAC paper U1086,Version 6, Updated Dcccmbw 11,2003

55-6/04/$17.00 02004 IEEE

0-7803-81 2220
have the potential of providing radar ATR algorithm On dynamic platforms such as a helicopter, STI data at
developers new metrics providing significant improvements longer ranges, can be collected for either SAR or DBS
in performance. processing. For shorter range stationary targets high-range
resolution (HRR) data can be collected. The STI mode is
While S A R may offer a unique solution for relatively long designed to provide 6 inch resolution in spotlight S A R or
range ATR, it cannot provide terminal homing and aimpoint HRR operation. Due to the extensive effort required to
selection. Monopulse techniques can he used for these develop SAR and motion compensation algorithms, it would
functions, hut IR sensors provide much better performance. be extremely time consuming and costly to test software
IR performance is dependant on image quality and the iterations on an aircraft. The S A R testhed will provide a
sophistication of the algorithms. For truly precise targeting, tool on which the radar can he mounted and tested
or when engaging targets for which the IR seeker has not repeatedly in a timely and cost effective fashion.
been “trained”, semi-active laser (SAL) seekers provide a
solution. Unfortunately, SAL seekers require the target to 3. PROBLEM
FORMULATION
he lased throughout a majority of the missile, thus exposing
the lasing platform. Clearly, each sensor has both The fundamental requirements used in the design of the
advantages and disadvantages. These factors and the current testbed were:
state of technology have led seeker developers to investigate
multimode seekers utilizing various combinations of i. Generate 6”xV resolution S A R images.
sensors. ii. Support a 125 Ib radar system.
iii. Operate when the radar is connected to power
The AMMS is being developed as a tool to allow the and data cables.
quantitative evaluation of single mode and multimode iv. The ahsolute position of the antenna must be
seekers. The AMMS is a modular design allowing multiple known to within U8 = 1.1 mm in (x,y,z)over
sensors to be integrated and tested using a common the coarse of the rail.
interface, and in some cases, a common data collection v. The velocity of the antenna must he known to
system. The suite will initially include a K,band radar, a within 0.1 d s .
mid-wave i n h r e d (MWIR) sensor and a SAL sensor. The
entire sensor suite is designed to fit into a four-axis The length of the testbed was constrained by the size of the
stabilized platform which can he mounted on either a fixed- tower available at the desired test site. The tower
wing aircraft or a helicopter as shown in Figure 1. dimensions limited the rail to 36 feet in length. Allowing
for radar acceleration to the desired speed and a similar
amount of distance for stopping, the rail was designed to
have 21 feet of effective synthetic aperture. The maximum
distance at which SAR images can he generated, denoted as
R,is given by the following equation [5]

where
1-Radar wavelength (m)
S, - Azimuth resolution (m)
,S
;,, - Radar beam squint angle (rad)
La, - Synthetic Array Length (m)
K - Broadening factor due to windowing
Figure 1 - AMMS Unit mounted in the “Fathoy”
The testbed was designed using the following parameters
The Guidance and Operations Computer (GOC) is the
heart of the suite providing a common, well documented 1 = 0.0086 m
intefiace for up to four sensors. The AMMS radar, depicted L, = 6.4 m
in Figure 2 features dual circular transmit and receive
polarization, simultaneous elevation and azimuth monopulse
6.= 0.1524m
S,,,, = 1.57 rad
via a four channel receiver, Y60MHz bandwidth, 4W solid-
K = 1.33
state transmitter, and a seven inch diameter 2YdB gain
antenna. The radar utilizes pulsed-Doppler waveforms for
From eq (1) the maximum distance that SAR images can be
moving target indicator (MTI) and STI operation.
generated is 170 m for 6 inch resolution. The mechanical
and electrical design of the testhed was performed from the
2
2221
specifications given in this section and numerous The mechanical support structure is designed to rigidly
consultations with the design team. support the along and cross track rails. The mechanical
structure turned out to be relatively massive to support the
4. TESTBED
REALIZATION accelerations of the radar package with minimum deflection
and motion of the rails. The 5 main support stands are made
A photograph of the testbed under going integration testing from 3”xY rectangular steel tubing. The stands are 6’2”
in the laboratory is shown in Figure 3. The along track and wide at their bases and 4’ tall. Two Vx4” steel I-Beams are
cross track directions of motion are denoted on the fastened on top of the 5 stands to support the radar package
photograph for reference. The main elements of the testbed for along track and cross track motion. The tops of the I-
are the 2 precision actuator rails (along track and across Beams are not machined surfaces; a 5/8”x 4” aluminum
track), mechanical support structwe, 2 precision motors, and guide rail spacer plate is mounted on top of the 2 I-Beams to
the motor controllers. The actuator rails are designed to insure a flat surface. Machined stainless steel guide rails
provide sub-millimeter position accuracy and are typically (l’x3i8”) are mounted on top of the guide rail spacers.
used in the electronics industry in pick-and-place machines. Alignment of the guide rails was a critical step in the
The along track rail is 34’-3” in length while the across assembly of the testhed; a theodolite was used to aide in the
track rail is 48”. An aluminum receiver in attached to the alignment process.
drive belt on the along track rail to which the across track
plate and subassembly are mounted. The receivers on the A 4400 watt, 200 volt, 3 phase AC servo motor is used to
rails are driven through a pulley gear on each of their ends. drive the along track rail and a 750 watt, 200 volt, 3 phase
AC servo motor is used for the cross track motion. The
cross track motor is attached to the small actuator rail
through a 1O:l reduction gear. Servo amplifiers and
controllers are used to control the motion of both actuator
rails through an interface to the main control computer. The
controllers are under software control and can quickly be
program to give essentially any desired trajectory. The

, . . Timing & Contro’l Board

Signal Distribution

. . . . .
. .
, .
. .. . ..

Figure 2 - Layout of the Ka-Band AMMS Radar

3
2222
controllers are programmed through a graphical user performed investigating image formation algorithms, motion
interface (CUI) which makes setting up new profiles very compensation techniques, and autofocus algorithms.
easy.
Acknowledgements
A problem experienced during initial trails was twisting of
the along drive track belt during high speed rail motion.
The authors would like to thank Mr. Roger Brown of the
The along track drive belt caused a torque of the along track
Boeing Company for performing the mechanical and
rail that quickly caused slack in the drive belt. The feedback
electrical design of the testhed and also for doing the
of this slack into the motor controller caused undesirable
physical construction. Mr. Brown's performance has been
perturbations in the radar's acceleration and position. An
exceptional.
additional bracket was added to the drive belt pulley system
to relieve the unbalanced stress on the along track rail
eliminating the tendency of the drive belt to become loose
during testing. References

After the rail was assembled and integration testing [I] B. Smith, D. Garner, Initial Spolight SAR Algorithms for
completed in a hi-bay, the testbed was moved and installed the AMRDEC Multimode Sensor Suite, TR-RD-MG-02-15,
on a 50 foot tower on Test Area 3 (TA-3). Redstone US AMCOM, April 2002.
Arsenal, Alabama. Original plans called for the testhed to
he disassembled, trucked to the tower and then re-assembled [2] G. Stimson, Introduction to Airborne Radar, El
on top of the tower. It was felt that moving and lifting the Segundo, CA: Hughes Aircraft Company, 1983.
testbed onto the tower as one unit would likely result in
damage to the unit. However, disassembly and then re- [3] D.Wehner, High Resolution Radar, Boston, M A Artech
assembling the testbed would have been very time House, 1995.
consuming. A local crane service that is routinely used to
move large/delicate hardware-in-the-loop equipment was [4] M. Soumekh, Synthetic Aperture Radar Signal
contacted and they proposed a method of moving and Processing with M T L A B Algorithms, New York John
installing the testbed intact. The crane service did an Wiley & Sons 1999.
outstanding job with no damage to the testbed. The testhed
simply had to be leveled and the machined guide rails [5] W. Garrara, R. Goodman, R. Majewski, Spotlight
realigned. Figure 4 shows the rail being lifted onto the Synthetic Aperture Radar Signal Processing Algorithms,
tower and its final installation. Nonvood, M A : Artech House, 1995.

TESTRESULTS
5. INITIAL

A series of tests were conducted using a Ka-band radar at

TA-3. The test matrix consisted of straight, parabolic, and
sinusoidal radar profiles on the testbed. Three target sets
were used in the initial testing: 1) comer reflector array 2)
main battle tank and 3) an artillery piece. The purpose of Biography
the comer reflector array was to quantitatively assess the Brian Smith is an electronics engineer ^^__".__._l__l.
~

performance of the testhed in generating high quality with the Applied Sensor, Guidance
images. Along track velocities up to 3.0 d s were used in and Electronics Directorate, Missile
the test process, higher along track velocities cause Research, Development and
unacceptable strain on the system. The initial SAR images Engineering Center, US AMCOM. He
generated using the test bed are very good. has worked in modeling and
simulation of Air Defense Systems,
6. CONCLUSIONS Anti-Radiation Missiles, SANISAR
systems, data compression, and battle$eld communications
A precision testbed for the development and evaluation of
systems. He has conducted extensive research in localized
tactical SAR systems has been successfully designed and computerized tomography f o r medical imaging and defense
constructed by the U.S. Army. The ability to program related areas. He holds a Ph.D. in Electrical and Computer
desired flight profiles into the testbed has been
Engineeringfrom The University of Alabama in Huntsville.
demonstrated. The testbed is in the process of being
transferred to a 40 foot tower at Redstone Arsenal, Alabama
where extensive SAR development testing will be
4

2223
 Wayne Garner holds a Bachelor of
Science Degree in Engineeringfrom the
Universit?,ofAlabama in Huntsville. He
is currently employed in the RF
Technology Area of the U. S. Army
Aviation and Missile Research,
Development, and Engineering Center
(AMRDEC). He has 21 years experience
in the field of radar seekers, focusing on multi-mode seekers
for the past seven years. He has worked on the Pershing,
Tacit Rainbow, MLRS TGW. Longbow, and BAT PI
programs. He has just completed an assignment as Chief
Engineer for the Common Missile Science & Technology
Objective seeker development program and currently serves
as the AMRDEC Primaly Support Director for the Non-
Line-Of-Sight Launch System Task Force. He has
experience in the areas of MMW and multi-mode seeker
system design and development, test and experimentation,
and ECMECCM

Randal Cannon is an electronics engineer

with Applied Sensors. Guidance, and
Electronics Directorate, Missile Research,
Development and Engineering Center, US
AMCOM. He has worked in development,
modeling, and simulation of Infrared
seekers for air defense and ground target
applications. Over the past 16 years, Mr, Cannon
contributions have included development of IR seekers for
missile flight technical demonstrations, development of
countermeasure and clutter rejection algorithms f o r both
ground and air-target defense systems, and the development
of high fidelity seeker models for integrated flight
simulations. Most recently, he conducted development of
hi-mode seekers f o r a Science and Technology Objective
(STO) program called Common Missile. He has a
Bachelors degree in Electrical Engineering from the
University of Memphis.

5
2224
 -
Figure 3 Photograph of the SAR Rail Te.rtbed

6
2225
Figure 4 SAR Rail Testbed being Lifted onto 50.17 Tower and Installed.

2226