0% found this document useful (0 votes)
161 views6 pages

Amsar - A European Success Story in Aesa Radar: Abstract - DGA, BWB and D&ES Initiated A

The AMSAR project was a European collaboration between defense organizations in France, Germany, and the UK to develop and demonstrate an active electronically scanned array (AESA) airborne radar system. The project involved building a 1000-element X-band AESA demonstrator radar that was integrated onto a test aircraft. Flight tests were conducted from 2008 to collect radar performance data against real targets under different digital beamforming techniques. The successful AMSAR program proved the capabilities of AESA radar technology for airborne applications in Europe.

Uploaded by

Ahmed Hussain
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
161 views6 pages

Amsar - A European Success Story in Aesa Radar: Abstract - DGA, BWB and D&ES Initiated A

The AMSAR project was a European collaboration between defense organizations in France, Germany, and the UK to develop and demonstrate an active electronically scanned array (AESA) airborne radar system. The project involved building a 1000-element X-band AESA demonstrator radar that was integrated onto a test aircraft. Flight tests were conducted from 2008 to collect radar performance data against real targets under different digital beamforming techniques. The successful AMSAR program proved the capabilities of AESA radar technology for airborne applications in Europe.

Uploaded by

Ahmed Hussain
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

AMSAR - A EUROPEAN SUCCESS STORY IN AESA RADAR

Jean-Luc Milin Pierre-Yves Triboulloy


DGA/DET/CELAR Thalès Airborne Systems,
Bruz, France, Elancourt, France,
jean-luc.milin@dga.defense.gouv.fr pierre.triboulloy@fr.thalesgroup.com

Stephen Moore Mike Royden


DSTL, Porton Down SELEX,
Salisbury, United Kingdom, Edinburgh, United Kingdom,
samoore1@dstl.gov.uk mike.royden@selex-sas.com

Wolfram Bürger Joachim Gerster


FGAN/FHR, EADS,
Wachtberg, Germany, Ulm, Germany,
buerger@fgan.de joachim.gerster@eads.com

Abstract— DGA, BWB and D&ES initiated a The first purpose was to mature E-scan technology and
radar programme, called AMSAR (Airborne demonstrate all operationally significant attributes and
functions of an AESA airborne radar system focusing on air-
Multi-role Solid-state Active-array Radar) to to-air performance, including Electronic Counter-Counter
demonstrate the enormous potential of Active Measure (ECCM), but also on Ground Moving Target
Electronically Scanned Array (AESA) radar. Indication (GMTI).
This paper describes all the stages of the
AMSAR project and the results obtained. The Work was contracted to GTDAR, a company owned by Thales
AMSAR demonstrator is still the only European, (France), SELEX Sensors and Airborne Systems Ltd (UK) and
forward-looking multichannel active array radar EADS (Germany). GTDAR is under contract to the French
Authorities, who act on behalf of the French, German and UK
with the capability of recording the outputs of Ministries of Defence.
more than 4 quadrants.
The AMSAR demonstrator is an X-band active electronically
Keywords-AESA, Airborne Radar, Fighter, ABF, STAP scanned adaptive radar with a circular 1000 element antenna.
Each element is connected to a high-power low-noise transmit
1. INTRODUCTION: THE AMSAR PROJECT / receive module (TRM) mounted on a vertical plank to ease
mechanical integration. Each TRM has a phase shifter with
DGA, BWB and D&ES initiated a radar programme, called attenuation control that allows precise phase and amplitude
AMSAR (Airborne Multi-role Solid-state Active array Radar) control. Thus the demonstrator is capable of inertia-less
to demonstrate the enormous potential of Active Electronically electronic scanning out to wide angles. See Refs 1,2.
Scanned Array (AESA) radar. AMSAR was designed and Furthermore, the array architecture provides 8 digitised
built in co-operation with Thales (France), SELEX (UK) and channels to support digital adaptive beamforming. Adaptation
EADS (Germany). Throughout the program, industries results is possible across a large trade space, e.g. AMSAR can adapt
were assessed by CELAR, CEV, FGAN, WTD61, DSTL and transmit / receive pattern, waveforms, update rates, mode
QinetiQ. interleaving, etc.

The AMSAR programme started in 1993 between UK and


France, with Germany joining in 1995. AMSAR’s objective
was a flying airborne AESA technology demonstrator with
real time operation, including Adaptive Beam Forming (ABF)
techniques. The target platform for the technology was a fast
jet aircraft.

1. Uhlmann, M.; Tanner, J.S.; Albarel, G., "Design characteristics of the 2. Arnold, E. , "A radiating element for an active airborne antenna"
AMSAR airborne phased array antenna " IEE Colloquium On Electronic Antennas and Propagation Society International Symposium,
Beam Steering - 98/481 , vol., no.pp.3/1-3/5, 28 Oct 1998 1999. IEEE , vol.1, no.pp.134-137 vol.1, Aug 1999
Integration, adjustment and evaluation of the AESA Antenna In receive the RF Interface generates the Sum, DiffAz
were carried out between early 2005 and spring 2006 at EADS (difference pattern in azimuth) and DiffEl (difference pattern
Defence Electronics in Ulm. In 2006 / 2007 the radar system in elevation) channel from the Quadrant Subarrays.
was integrated in EADS Ulm and then intensively evaluated
on the ground in France – both on the Bruz CELAR facility The RF Manifold distributes the RF from / to the RF Interface
BEDYRA (hardware-in-the-loop test bench) against simulated to / from the TRMs. The TRMs (see figure 2) weight the
targets and jammers, and also against real targets and jammers signal in amplitude and phase to achieve the desired aperture
at the CEV flight test centre at Cazaux. The AMSAR system distribution to produce the requested far field pattern.
was then installed in the QinetiQ BAC 1-11 test bed aircraft.
A total of 22 flights were subsequently performed in 2008
over land and sea in France, Germany and the UK. The radar
performances for air-to-air and air-to-ground modes were
assessed against real targets like Falcon 20, Tornado, Alpha
jets and Ground Targets.

A huge amount of real data was collected at the outputs of the


antenna sub-arrays in presence of not only real targets, but
also real jamming signals. The data collected during ground
and flight trials permitted the performance of numerous
different Digital Beam Forming (DBF) techniques to be
established, including:

• Adaptive Beam Forming (ABF) for sidelobe and


mainlobe anti-jamming,
• Jammer mapping by means of super-resolution,
• Deterministic nulling using weighting at TRM level,
• Ground Moving Target Indication (GMTI) and air-to-air
clutter rejection by Space Time Adaptive Processing The radiating element and the WAIM (Wide Angle Impedance
(STAP), Matching) sheet ensure that the array can scan to wide angles
• Simultaneous multiple beamforming on receive. with low reflections. The BSCU (Beam Steering Control Unit)
and DPSU (Distributed Power Supply Unit) supply the TRMs
3. THE AMSAR ACTIVE ARRAY MULTICHANNEL ANTENNA
with control and power. The antenna structure is responsible
The well-known major advantages of an AESA with respect to for mechanical positioning accuracy, stiffness and liquid
a mechanical scanned array are: cooling of the TRMs.

• Rapid electronic inertia-less scanning, The sub-array architecture second level, consisting of a Side
• High average transmit power possible due to high Lobe Blanking (SLB) guard channel, the 8 digitised sub-arrays
efficiency of array architecture, and the monopulse four quadrants is shown on figure 3.
• E-scan array architecture with multiple receive channels
facilitates a number of advanced modes,
• AESA can be adapted to number of other functions,
• AESA share a highly reliable radar design:

The principal
architecture of the
AMSAR antenna is
shown in figure 1. The
RF InterFace (RF-IF) is
the interface between
Receiver / Exciter and
the RF Manifold. In
transmit it amplifies the
Exciter signal to
provide the required
input power for each
Transmit and Receive Figure 1. Block diagram of the AMSAR Figure 3. 8 Sub-array Architecture
Module (TRM).
Before evaluation, the antenna must be adjusted. Amplitude Figure 9 shows
and phase are adjusted for each Radiating Element, to the Standard Rx
compensate for manufacturing tolerances of the sub-units Beam. (Red trace
(RF combiners, manifold, cables, etc.). In this single element shows cross-
adjustment process each radiating element is measured for polarisation
some TRM states. The amplitude / phase deviations are rejection> 35 dB.
corrected to get a
uniform amplitude /
phase distribution on
the antenna aperture.
Figure 4 shows the
farfield pattern of the Figure 9. Standard Beam Rx
AESA without single Figure 10 shows the standard beam scanned to 70°.
element adjustment.

Directivity is only
Figure 4. Farfield pattern before single 4dB less than the
element adjustment
boresight pattern.
Amplitude and phase
deviations on individual This validates the
elements result in high design of the
side lobes of the main radiating element
beam. Fig 5 shows the and the WAIM
improved farfield sheet.
Figure 5. Farfield pattern after single
element adjustment pattern after single
element adjustment. Figure 10. Standard Beam Rx @ 70°

After single element adjustment, the challenging performance The far sidelobe
requirements set at the beginning of the project were fulfilled - levels of an AESA
in particular, the average far sidelobe level. are a determined by
the residual random
During evaluation the errors in amplitude
cardinal pattern requirements and phase after
of the antenna were adjustment.
measured in a Near Field
Test Range (see figure 6) and A very low sidelobe
a Far Field Test Range (see beamshape
figure 7 and 8). (Hamming - 40dB) is
shown in figure 11. Figure 11. Hamming - 40dB Beam
Measurements included Figure 6. The AMSAR antenna Such low sidelobes
directivity, beamwidth, peak on Near Field Test Range are the result of a
side lobe level, average far very accurate single element adjustment.
sidelobe level and beam pointing accuracy. The liquid cooling
system made it easy to study temperature effects. SAR mode
performance is
shown in figure 12.
This pattern was
measured with the
frequency offset
500 MHz from the
frequency used for
single element
adjustment. There
Figure 7. AMSAR antenna in is only a slight Figure 12. Standard Beam @ f + 500MHz
the Far Field Test Range increase of the
sidelobes, showing that a broad SAR chirp is possible without
Figure 8. AMSAR antenna in the losing the quality of the farfield pattern.
roll frame used on the Far Field
Test Range
Finally, the AMSAR antenna was measured in free space, with 5. THE CEV FLIGHT TEST CENTRE CAMPAIGN
and without a fighter radome. After its integration and
acceptance in EADS Ulm
4. THE AMSAR SYSTEM Germany, the AMSAR system
Figure 13 shows went to Centre d’Essais en Vol
the AMSAR radar (CEV) flight test centre in
system block Cazaux near Bordeaux in
diagram. France. Fitted on a ground
An exciter trolley (see figure 14), the
generates the system was exercised with
signal to be combinations of real targets
Figure 14. The AMSAR radar
transmitted to the such as Falcon 20, Hunter,
system during ground trials in
RF-IF in the Mirage 2000 and Alpha Jet (see Cazaux
antenna. The figure 15). Some runs where
output of the 4 recorded with the Falcon 20
conventional Figure 13. The AMSAR radar system block and/or the Hunter emitting
channels: sum, DiffAz, DiffEl and guard channel for SLB, jamming signals in presence of
formed in the antenna RF-IF, are connected to the 3 channel a target. Different
receiver (DiffAz and Guard channel are multiplexed). This 3 configurations were tested with
conventional channel receiver performs real time signal sidelobe jamming, mainlobe
digitisation. The outputs of the 8 sub-arrays are also digitised jamming, or both. This
in real time, by routing the 8 outputs from the antenna RF-IF campaign was mainly dedicated
to an 8 channel Sub-Array receiver. to radar parameter adjustment Figure 15. The AMSAR radar
and data recording with real system during ground trials in
The Antenna Assembly was designed to be installed in the targets and jammers. Cazaux facing different targets
nosebay of the QinetiQ BAC 1-11. It is supported by a heat
exchanger mounted in the forward cargo bay, and a primary 6. THE BEDYRA TEST CAMPAIGN
power supply in the cabin. Also in the cabin is: After the CEV test campaign, the
AMSAR demonstrator moved to
• The Antenna Control Computer, for command and CELAR in Bruz near Rennes in
control of the antenna. France for intensive
• A digital recorder which captures signals on the radar characterisation of the system.
bus and the high bandwidth signals from the two This used the hardware-in-the-
receivers (12 channels total, capacity 3 TB) loop dynamic test bed BEDYRA
• The Trials Monitor Computer which comprises (Banc d’Evaluation DYnamique Figure 16. The AMSAR
radar system on the 3-axis
Engineering Display, Signal Processor, Data pour Radars et Autodirecteurs table of BEDYRA CELAR
Processor and Radar Display. The Engineering électromagnétiques). BEDYRA is
Display provides system control and the HMI. a French MoD facility dedicated to
The Signal Processor uses data from the 3-channel evaluation of real radars and missile
receiver to provide classical radar processing for seekers.
target detection, and the Data Processor is The material under test is mounted
responsible for computation of range, velocity and on a 3 axis table (see figure 16),
bearing, as well as track formation. Finally, the Radar facing a wall of horns which
Display shows the antenna beampointing, target reproduce an RF environment which
range, velocity and tracks. simulates different types of targets
and jammers, including real ones
The AMSAR radar system is able to detect and track multiple plugged into a horn. (see figure 17).
targets simultaneously, using these air-to-air modes: The BEDYRA Campaign was mainly Figure 17. BEDYRA
used to finalise radar parameter horn wall
Velocity search only adjustment and assess the radar performances in terms of
SO1 (HPRF detection and confirmation) detection range and track formation range against various
SO2 (HPRF detection and MPRF confirmation). targets. It was also used to perform scenarios impossible in
Search and Track flight such as crossing targets. A huge amount of data was
ST1 (HPRF search, MPRF track), recorded with multiple targets, mainlobe jamming, sidelobe
ST2 (MPRF search and track) jamming, both simultaneously, false targets, etc.
ST3 (ST1 and ST2 interleaved). - but without any clutter…
• Adaptive Jammer Rejection by spatial Adaptive Beam
7. THE FLIGHT TRIALS TEST CAMPAIGN Forming using the 8 sub-array channels,
Flight trials are needed to assess the radar performance in a • Probability of Detection improvement for Air-to-Air
real environment and to record data with sea or ground clutter. mode by means of ground clutter rejection using STAP,
The AMSAR radar system was consequently installed on the • Minimum Detectable Velocity improvement for Air-to-
QinetiQ BAC 1-11 aircraft (see figure 18). Ground modes by means of STAP GMTI,
A total of 22 • Rapid reacquisition of lost under-track targets by means
flights have been of simultaneous multiple Beam Forming on Receive and a
performed. The broadened Transmission.
first ones, over
UK, from 1) Digital sum channel construction
Boscombe test The first challenge was to prove that it was possible to
Figure 18. QinetiQ BAC 1-11
centre, were accurately calibrate the 8 subarrays, i.e. compensate for
mostly dedicated differences in phase and amplitude between the 8 channels, to
to radar adjustment in the BAC 1-11 environment followed by reconstruct a sum beam. AMSAR achieved this to the extent
air-to-air mode performance assessment. Multiple targets were that the sum beam reconstructed off line using the 8 subarray
detected and tracked at the expected ranges. outputs was even better than the conventional sum beam. The
holes between adjacent sidelobes in the reconstructed sum
The next flights, over Germany, from the Manching WTD 61 beam were deeper than in the conventional one, proving that
test centre, were dedicated to air-to-ground modes using a the aperture illumination was closer to the ideal.
STAP / GMTI waveform. The outputs of the 8 sub-arrays
were recorded for further offline signal processing on a ground 2) Digital Adaptive Beamforming
moving target equipped with GPS and reflectors (see figure The aim of digital adaptive beamforming is to:
19) was driving in circles.
• Form nulls (i.e. points of zero return) in jamming
Finally, the last flights were direction,
over French ground and sea • Cancel multiple jammers (numbers dependent on degrees
clutter, from Cazaux CEV. of freedom),
Dedicated mostly to ABF
• Adapt the antenna pattern to the environment in real-time.
and air-to-air STAP, the
AMSAR system faced high
Different ABF
ground clutter (with the
Algorithms under
target going through the
different constraints
clutter notch); or up to two
were tested on real
simultaneous jammers Figure 19. Ground moving target data recorded during
trying to mask an Alpha jet. for GE flights
the trials. With ground
The outputs of the 8 sub-
trials data recorded in
arrays were again recorded in various configurations including
BEDYRA it proved
multiple sidelobe jamming, mainlobe jamming and
possible to reject the
simultaneous sidelobe and mainlobe jamming with real clutter
jammer down to the Figure 20. ABF for anti-jamming
and real masked target. The hundreds of Terabits of data
noise power level floor
recorded have then been analysed using different types of
in all cases.
multichannel antenna processing.
The target masked by the jamming signal was always detected
8. THE COLLECTED DATA ANALYSIS after ABF processing. Adapted radiating patterns were plotted
and the degradation caused by countering sidelobe jamming
The AMSAR program proved the feasibility of an airborne
was so small that having ABF on, even when no jamming is
AESA radar, and showed that all the benefits of such a system
present, would not be unreasonable. For mainlobe jamming,
were achievable in real time. In addition, the AMSAR
the angular limit of acceptable adapted pattern degradation
program aimed to support the evaluation of multichannel
was established. Up to this limit the masked target was always
antenna processing techniques. All the data recorded during
detected after ABF processing. Results obtained were in line
trials has been used for this purpose. Among the techniques with textbook predictions.
evaluated were:
With data recorded on flight trials, results obtained were
• Jammer mapping using angular super-resolution, nearly as good as during ground trials. In this case, the
• Non-adaptive Jammer Rejection using spatial processing influence of clutter prevented jammer rejection from reaching
applied to the total degrees of freedom of the antenna by the noise floor.
means of deterministic nulling,
3) Space-Time Adaptive Processing for air-to-air modes 5) Jammer mapping
The aim of STAP for Data collected in flight and in BEDYRA with crossing
air-to-air modes is an jammers have been used to assess the capability of AMSAR to
improvement in perform jammer mapping. Jammer mapping uses data from a
probability of detection total scan of the entire space surrounding the radar. In regions
by means of Ground where signals are high, spectral processing is applied for super-
Clutter rejection. Figure resolution. Although the real limit of the angular separation of
21 shows the influence 2 close jammers has not been assessed, results obtained were
of the antenna beam Figure 21. Antenna radiating pattern better than expected and show that such a mode is very
pattern on the influence on the Range-Doppler map promising.
range/Doppler map.
Clutter intercepted by the 6) Deterministic nulling
main beam prevents When the position of the jammer is known with precision a
detection of targets flying priori, for example using jammer mapping, one can in theory
with the same velocity as form a null in the jammer direction using the total degrees of
the aircraft. Furthermore, freedom of the antenna - i.e. the number of TRMs. This
sidelobes of the antenna technique is known as deterministic nulling. This non-adaptive
also create a «cactus» that concept can be used on antennas that are not equipped with
degrades the target ABF. Radiating patterns with deterministic nulling were
detection in range and Figure 22. The STAP ground clutter- recorded in the BEDYRA anechoic chamber proving the
velocities close to the
filtering concept feasibility of this concept.
main beam region. 7) Simultaneous multiple Beams Forming on receive.
As presented on figure 22, STAP counters this effect and the Even for an AESA radar, losing a target under track usually
result is a range Doppler map with virtually no cactus and with means spending time going back into search mode for
a main beam clutter area that spreads much less in velocity. reacquisition and confirmation. To avoid these time losses, it is
The actual performance of STAP processing was assessed possible to form, on receive, multiple beams around the latest
using data collected during special flight trials from Cazaux. known target position. The beam giving the detection indicates
These trials used a lookdown configuration, with high clutter where the target has gone. Reinitializing the Kalman filters
and a target going through the clutter notch. Good results were with this value avoids having to go back to search mode.
obtained, proving the practical benefit of STAP. AMSAR was evaluated in the BEDYRA anechoic chamber,
4) GMTI: Space-Time Adaptive processing for air-to- forming two simultaneous Receive beams in azimuth, and a
broadened Transmit beam. The expected performances were
ground modes
reached.
The aim of STAP here
is to ensure the
detection of slow 9. CONCLUSION: THE FUTURE OF AMSAR
moving ground The technological progress made on AESA architecture and
targets. What is Transmit and Receive Modules (TRMs) in the early years of
expected is a the AMSAR program permitted the companies involved to
reduction in the confidently invest in a number of operational products. Two of
Minimum Detection
these products are the RBE2AA (Radar à Balayage
Velocity (MDV) for Figure 23. The STAP ground clutter- Electronique 2 plans à Antenne Active) for the Rafale fighter
various scan angles. filtering result on slow ground moving
Ground clutter and Captor-E for the Eurofighter. Furthermore, this program
targets
filtering by STAP allowed the testing of a large family of multichannel antenna
permits detection of targets that processing techniques. Their performances now being known
were masked by the clutter without in a real environment, they will be soon implemented in
any adaptive processing, as shown operational modes. The data recorded during these trials has
on Figure 23. The resulting created a large library of data covering a diverse set of
detection can be superposed on scenarios that can be used for algorithm development for
SAR maps as shown on Figure 24. many years. Finally, by a combination of antenna testing in
To assess the performance of STAP BEDYRA and antenna simulation the ‘graceful degradation’
in this application, data was assumption has been confirmed by showing minimal effect on
collected using lookdown performance even with a random selection of 10% of TRMs
configuration during dedicated inhibited.
flight trials from Manching with a Figure 24. GMTI
GPS equipped, ground-moving detection on a SAR map
At the time of writing, the AMSAR demonstrator is still the
target. Results were very good only European, forward-looking multichannel active array
proving that slow moving targets can be detected in real radar with the capability of recording the outputs of more than
environment. The specified MDV was surpassed. 4 quadrants in flight.

You might also like