The expectation maximization (EM) algorithm is an iterative algorithm which can be used to obtain maximum-likelihood (ML) estimates in situations where a direct attack on the problem fails. In this research the PI will study the application of the EM algorithm to the problem of optimal decoding of trellis coded modulation signals in the presence of white Gaussian noise and random parameters. Specifically, the work will focus on two channels: the random phase channel and the random timing error channel. For each of these channels, the optimum receiver processing will be described, as well as its performance using existing trellis codes; further the implications of the optimal processing on the coding problem will be addressed.
