Folie 1

1
Ambiguity Suppression by Azimuth Phase Coding in
Multichannel SAR Systems
F. Bordoni, M. Younis, G. Krieger
DLR - Institut für Hochfrequenztechnik und
Radarsysteme
IGARSS 2011, 24-29 July, Vancouver, Canada
2
Outline

• Introduction
• APC (Azimuth Phase Coding) technique
• APC in multichannel SAR (Synthetic Aperture
  Radar) systems
• Figure of merit
• Numerical analysis
• APC performance versus system parameters
• Example two multichannel systems for high
  resolution wide swath imaging
• Conclusions

3
Introduction
Current spaceborne SAR systems limitation
trade-off spatial resolution v.s. swath width
Research in two main directions
Processing methods for removing the
ambiguities APC - low implementation complexity
- effectiveness for point and distributed
ambiguities

• New, more flexible SAR systems
• - Multichannel systems
• Digital Beamforming (DBF) on receive
• Multichannel processing

APC is conceived for conventional SAR systems ?
APC in multichannel systems based on DBF on
receive?
4
Review of the APC Technique
APC is a technique for range ambiguity
suppression, conceived for conventional (1 Tx and
1 Rx) SAR systems Dall, Kusk 2004
Dal04 J. Dall, A. Kusk, Azimuth Phase Coding
for Range Ambiguity Suppression in SAR, IGARSS
2004.
APC is based on three main steps 1) Azimuth,
i.e. pulse to pulse, phase modulation on Tx
APC modulation phase
Tx pulse number
2) Azimuth phase demodulation on Rx
APC demodulation phase
_at_ round-trip delay
APC residual phase
azimuth sample number, order of range ambiguity,
APC shift-factor
3) Azimuth filtering over the processing
bandwidth
5
APC residual phase ? Doppler shift
Time domain linear phase
Frequency domain Doppler shift
order of range ambiguity (0 useful signal)

• M2 ? maximum Doppler shift of the 1st order
  range ambiguity
• ? Larger oversampling ? Larger
  ambiguity suppression

6
Application to Multichannel Systems
Multichannel SAR system 1 transmitter, N
receivers
N
2
1
N Rx az. signals sampled at PRF
APC residual phase
MULTICHANNEL PROCESSING
? The behavior of the APC changes when applied to
a multichannel system
7
APC Reconstructed Multichannel Signal
The APC residual phase has no more a linear trend
versus the azimuth sample (pulse) number ? no
shift of the Spectrum
(uniform PRF)
PRF
,
,
? The residual phase a stair shape (lt?gt Doppler
shift)
? The ambiguity spectrum
PRF matched to the antenna length and No. of
apertures gt regular sampling in azimuth results
8
Figure of Merit
Measurement of the ambiguity suppression induced
by APC
APC Gain
Computed on the SAR signal after multichannel
processing
PSD (Power Spectral Density) range ambiguity of
1st order if APC is not applied
useful signal after multichannel reconstruction
(neglect. elev.)
processed bandwidth
PSD range ambiguity of 1st order if APC is applied

• Note the Gapc depends on the azimuth pattern
  shape

9
APC Performance Analysis
Reference Multichannel Planar Systems
Parameter System System System System
Parameter 1 2 3(Ref.) 4
Orbit height km 520 520 520 520
Carrier frequency GHz 9.600 9.600 9.600 9.600
Rx antenna total length m 3 6 12 24
Tx antenna length m (and Rx subapert. length) 3 3 3 3
No. of az. Rx channels 1 2 4 8
PRF Hz (uniform) 5068 2534 1267 633.5
PRFeff Hz 5068 5068 5068 5068
The systems have the same azimuth patterns
Processing bandwidth 2316 Hz Bp 4168 Hz

• Behavior of APC versus the number of Rx channels,
  N

Investigation

• Effect of the Doppler oversampling
• The effect of the pattern shape is not evident

10
Numerical Results Gapc
APC Gain v.s. oversampling factor
For the considered systems, for M2

• 0.1dB Gapc 3.13dB
• for a given N, the Gapc increases with the
  oversampling factor, ?
• the Gapc decreases for increasing number of
  channels, N
• the sensitivity of Gapc to ? decreases with
  increasing N

11
Numerical Results PSD v.s. N
Normalized PSD 1st range ambiguity after
multichannel reconstruction
without APC
with APC
N 8
N 2
N 1

• larger N, the upper profile PSD with or without
  APC are similar and Gapc reduces

12
HRWS SAR Multichannel Systems
HRWS (High-Resolution Wide-Swath) SAR
System promoted by the German Aerospace Centre
(DLR) conceived to obtain high resolution and
wide swaths (1 m resolution,
70 km swath width in stripmap mode)
Different Rx azimuth patterns multichannel
reconstruction
Parameter Planar Reflector
Orbit height km 520 745
Carrier frequency GHz 9.600 9.650
Tx/Rx antenna total length m 8.75
Paraboloid diameter (elev., az.) m 10, 12
Total number of feeds (elev., az.) 60, 10
No. of az. Rx channels 7 10
PRF Hz 1750 2792
Processed bandwidth Hz 6252 5946
Oversampling factor 1.960 4.696
Planar system currently adopted
design Reflector system alternative design
option, studied in DLR
13
Peculiarities HRWS Systems
Reflector system
Planar system
The pattern of each Rx channel covers
Bp Multichannel processing Multi-Aperture
Reconstr. The patters do not change along the
swath
The pattern of each Rx channel covers 1/N of
Bp Multichannel processing Spectral
decomposition The patters change along the swath

• Evidence of the dependence of the APC
  performance on the pattern shape

14
Numerical Results Planar HRWS System
Normalized PSD 1st range ambiguity used to
compute the Gapc (after multichannel
reconstruction)
with APC
without APC

• 
  For M2, Gapc 0.69 dB
• The high number of channels (7) and the small
  oversampling (1.96) associated low Gapc

15
Numerical Results Reflector HRWS System
Normalized PSD 1st range ambiguity used to
compute the Gapc (before multichannel
reconstruction, single Rx channel)
without APC
with APC

• For M2, 3.2 dB Gapc 8.6 dB over the swath,
  depending on the azimuth pattern
• The azimuth pattern strongly affects the APC
  performance
• The reflector based system, characterized by a
  higher oversampling factor (4), takes better
  advantage from the application of APC

16
Conclusions

• In multichannel systems, the APC effect is no
  more a frequency shift of the range ambiguity.
• Also in multichannel systems, the APC allows for
  improved ambiguity suppression.
• The azimuth pattern strongly affects the APC
  performance.
• For a given azimuth pattern, the suppression is
  directly proportional to the oversampling factor
  and inversely proportional to the number of
  receive channels.
• In a conventional SAR system with g 2, the
  achievable suppression of each ambiguity of odd
  order is about 3 dB. In multichannel systems
  based on planar antenna architectures, the
  suppression is generally poorer.
• Reflector based systems reach better
  performance, because of the higher oversampling.
• In the planar and reflector based HRWS systems
  the APC suppression is about 0.7 dB and between 3
  and 8 dB, respectively.