Time reversal precoding in the presence of non-gaussian underwater acoustic noise

Abstract

Time Reversal (TR) precoding is well-established for mitigating the effects of multipath in an underwater acoustic (UWA) link. However, it has traditionally been studied with additive white Gaussian noise (AWGN) receivers, and, has been found to be optimal for SNR maximization of correlation receiver. Recent research has presented several scenarios where conventional AWGN assumption does not hold in UWA channel; rather, channel noise is better described by heavy-tailed non-Gaussian stochastic processes. This motivates an interest to study the optimality of TR precoding in the presence of non-Gaussian noise as well. In this paper, we use three very popular non-Gaussian statistics, Generalized Gaussian (GG). Gaussian mixture (GM), and Cauchy-Gaussian mixture (CGM), to approximate UWA noise and evaluate the performance of a traditional TR precoder based communication system in different underwater channel conditions.