Multi-Channel Multi-Variate Equalizer Design

Research paper by Ravikiran Rajagopal, Lee Potter

Indexed on: 01 Jan '03Published on: 01 Jan '03Published in: Multidimensional Systems and Signal Processing


The polarimetric calibration of synthetic aperture radar (SAR) imagery requires the equalization of a multiple-input, multiple-output (MIMO) distortion system. For wide-band, wide-angle SAR systems, the distortion is frequency and angle dependent and can be accurately modelled as a two-dimensional finite impulse response (FIR) filter bank. This paper presents a design algorithm, using a Gröbner basis, to compute a FIR filter bank that exactly inverts the multi-channel distortion. The results presented for polynomial inversion of MIMO FIR systems hold for sampled data signals of arbitrary dimension.