Algorithms for Joint Channel Estimation and Data Recovery—Application to Equalization in Underwater Communications

One of the main obstacles to reliable underwater acoustic communications is the relatively complex and unstable behavior of the ocean channel. The channel equalization method, that can estimate and track this complex and rapidly varying ocean response, may lead to reliable data communications at high rates which utilize fully the available bandwidth. Unfortunately, standardized equalization techniques fail in this environment. In this paper we derive methods for joint ocean-channel estimation and data recovery, using optimal, Maximum Likelihood (ML) estimation criterion. The resulting ML problems may be complex; thus we will use iterative algorithms; e.g., the Expectation-Maximization (EM) algorithm. The different methods correspond to different assumptions about the ocean channel. The theoretical derivation of these methods as well as preliminary results on a simulated ocean data experiments are presented.