Reduced complexity iterative equalizers for M-ary 2D ISI channels

This paper considers detection of a signal corrupted by 2D intersymbol interference (ISI) and additive white Gaussian noise (AWGN). We compute equivalent convolution masks for 4-ary and binary modulation with Gaussian and Sinc square point spread functions (PSFs). To maintain the same spatial density, the binary mask is wider than the corresponding 4-ary mask. Based on the maximum likelihood (ML) bound, we find that for larger values of the PSF spread parameter sigma_b, the bit error rate (BER) of a 4-ary modulation can be better than that of the equivalent binary system. Reduced complexity algorithms for detection of 4-ary signals on 2D-ISI channels are also proposed.