Blind turbo equalization with EM algorithm for DPSK systems

Abstract

Iterative joint equalization and decoding is called turbo equalization. If the channel is unknown, the channel estimation is an additional signal processing task. In this study, a blind turbo equalization method is developed for digital communication systems using DPSK modulation, which is a promising candidate for future optical wavelength division multiplexed transmission systems. The reason for choosing DPSK is to combat the phase ambiguity. The expectation-maximization algorithm is used to estimate the unknown channel and the noise variance. Super-trellis structure is made up of differential encoder and channel. This structure is exploited for doing differential decoding and equalization together at the receiver. Performance of the proposed method is tested via computer simulations.