Modified Multichannel Reconstruction Method of SAR With Highly Nonuniform Spatial Sampling

Abstract

The traditional filter bank algorithm (FBA) for a multichannel synthetic aperture radar (SAR) system can provide an unambiguous recovery. However, the FBA may result in a seriously signal-to-noise ratio (SNR) loss or might even fail to reconstruct the spectrum if there is highly nonuniform sampling in azimuth. It is the general case because pulse repetition frequency (PRF) cannot be selected flexibly for spaceborne imaging system. In this paper, a new algorithm for processing highly nonuniform sampled SAR data to improve the SNR is presented. The proposed method is based on the operation of ambiguity-index screening (AIS) and the minimum mean square error (MMSE) criterion. It can be utilized to reconstruct the spectrum even in the presence of highly nonuniform sampling or coinciding sampling. The approach can significantly improve the SNR especially for the highly nonuniform sampling, and the azimuth ambiguity-to-signal ratio obtained by the proposed method is superior to the traditional one for a wide range of PRF. The proposed algorithm is validated by the simulated data and the real two-channel radar raw data acquired by C-band airborne SAR system with a bandwidth of 210 MHz, which is designed by the Department of Space Microwave Remote Sensing System, Institute of Electronics, Chinese Academy of Sciences.