Effective Communications over Doubly-Selective Acoustic Channels using Iterative Signal Cancelation

Abstract

A low-frequency doubly-spread acoustic channel is used to illustrate the effectiveness of iterative (turbo-) equalization for rapidly changing communications channels. The channel is modeled as a 1 km shallow ocean channel and its physical parameters are supported by recent measurements conducted by Dalhousie University. The proposed signal processing mechanism for equalization and decoding borrows from turbo equalization, but uses repetition coding and frequency-domain signal processing. Both have been shown to be advantagous for signaling on radio-frequency channels. Several simulation experiments verify that iterative interchannel interference cancelation can, in principle, effectively remove such interference, even for channels with very short coherence times. A low-complexity signal processing structure for effective communications over doubly-spread channels is presented and its performance is quantified in a series of experiements and related to the theoretically obtainable maximum performance, providing a well-rooted baseline benchmark for other methods, or the future inclusion of channel estimation and tracking algorithms.