The University of Southampton
University of Southampton Institutional Repository

Synthetic aperture radar signal processing on the distributed array processor

Synthetic aperture radar signal processing on the distributed array processor
Synthetic aperture radar signal processing on the distributed array processor

This thesis investigates the use of a highly parallel SIMD machine for fast Synthetic Aperture Radar (SAR) signal processing. Through the development of new parallel techniques for the SAR processing problem, the Distributed Array Processor is shown to meet the computational demands of SAR while remaining flexible and cost effective. SAR azimuth compression is shown to be the significant computational component of the SAR processing chain whereas coping with range migration is the significant non-numerical component. High precision time-domain, and fast frequency-domain based SAR azimuth compression algorithms are presented together with techniques for efficiently handling the range migration. New parallel DSP algorithms, applicable to SAR, are developed which include a comprehensive FFT suite of mixed precision, interpolation and filtering routines. A complexity analysis of the algorithms is presented. The efficient FFT algorithms use a block floating point precision which has been tailored to match the DAP architecture. The statistical properties of this BFP approach are compared to the standard fixed and floating point precision algorithms. A broad study of DFT algorithms showed that the PFA/WFTA were the most suited to range-Doppler SAR azimuth processing on the DAP. However DAP implementation of the permutations associated with these algorithms was found to be non-trivial. The permutation problems were solved by constraining the dimensions to yield toroidally ordered data sets on which new parallel algorithms, based on Floyd's matrix transposition algorithm, were applied. A complexity analysis of the new algorithms is presented and compared to the conventional power-of-two based algorithms. It is shown that the restricted dimensions associated with the developed algorithms may be used in a SAR design which actually increases the number of looks selectable with no additional data movement overheads, when compared to the power-of-two based processors. A range compressed SAR simulator is formulated and programmed. A set of SEASAT and ERS-1 data bases is presented. Three DAP based SAR azimuth compression processors are designed : the single look time-domain processor (TDA), the single look frequency-domain processor (FDA), and the multiple look frequency-domain processor (MLA3). The relative performance of these designs is studied using the simulated data. Complexity formulae derived for the processors are verified experimentally thus allowing extrapolated times for a large data set to be computed.

University of Southampton
Soraghan, John James
Soraghan, John James

Soraghan, John James (1989) Synthetic aperture radar signal processing on the distributed array processor. University of Southampton, Doctoral Thesis.

Record type: Thesis (Doctoral)

Abstract

This thesis investigates the use of a highly parallel SIMD machine for fast Synthetic Aperture Radar (SAR) signal processing. Through the development of new parallel techniques for the SAR processing problem, the Distributed Array Processor is shown to meet the computational demands of SAR while remaining flexible and cost effective. SAR azimuth compression is shown to be the significant computational component of the SAR processing chain whereas coping with range migration is the significant non-numerical component. High precision time-domain, and fast frequency-domain based SAR azimuth compression algorithms are presented together with techniques for efficiently handling the range migration. New parallel DSP algorithms, applicable to SAR, are developed which include a comprehensive FFT suite of mixed precision, interpolation and filtering routines. A complexity analysis of the algorithms is presented. The efficient FFT algorithms use a block floating point precision which has been tailored to match the DAP architecture. The statistical properties of this BFP approach are compared to the standard fixed and floating point precision algorithms. A broad study of DFT algorithms showed that the PFA/WFTA were the most suited to range-Doppler SAR azimuth processing on the DAP. However DAP implementation of the permutations associated with these algorithms was found to be non-trivial. The permutation problems were solved by constraining the dimensions to yield toroidally ordered data sets on which new parallel algorithms, based on Floyd's matrix transposition algorithm, were applied. A complexity analysis of the new algorithms is presented and compared to the conventional power-of-two based algorithms. It is shown that the restricted dimensions associated with the developed algorithms may be used in a SAR design which actually increases the number of looks selectable with no additional data movement overheads, when compared to the power-of-two based processors. A range compressed SAR simulator is formulated and programmed. A set of SEASAT and ERS-1 data bases is presented. Three DAP based SAR azimuth compression processors are designed : the single look time-domain processor (TDA), the single look frequency-domain processor (FDA), and the multiple look frequency-domain processor (MLA3). The relative performance of these designs is studied using the simulated data. Complexity formulae derived for the processors are verified experimentally thus allowing extrapolated times for a large data set to be computed.

This record has no associated files available for download.

More information

Published date: 1989

Identifiers

Local EPrints ID: 461602
URI: http://eprints.soton.ac.uk/id/eprint/461602
PURE UUID: c8f20228-bc1a-4016-92e0-cc6e0ec30d08

Catalogue record

Date deposited: 04 Jul 2022 18:50
Last modified: 04 Jul 2022 18:50

Export record

Contributors

Author: John James Soraghan

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of http://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×