Sparse Channel Estimation and Tracking for Underwater Acoustic Communications

Abstract

This thesis focuses on estimation and tracking of underwater acoustic channels. A model of underwater acoustic channel suitable to study the channel estimation problem is constructed. The model is used together with a Bellhop acoustic wave propagation simulator to generate channels for various propagation environments, depths and distances, and we studied their sparseness and variation properties. Finally a number of estimation methods such as basis pursuit and approximate message passing are adapted to channel estimation problems and tested on a number of channels. Window-based and Kalman- filter based channel tracking is also studied. A detailed comparison of estimation and tracking methods is presented.