Performance Analysis of Multichannel Lattice Equalization in Coherent Underwater Communications

This work examines the numerical fixed-point performance of a new multichannel lattice RLS filtering algorithm using data from two underwater acoustic communication experiments. The algorithm may be an appealing choice for underwater equalization due to its robust numerical behavior and linear scaling of the computational complexity with filter order. Simple modifications to widely-used methods for carrier/timing synchronization and symbol slicing in transversal equalizers are proposed. Experimental results show that the algorithm is as accurate as the similarly array-based QR-RLS, tolerating word lengths as low as 16-20 bits with minor degradation relative to floating-point benchmarks. These features, coupled with a very modular and regular structure, are highly desirable in energy- efficient hardware or embedded implementations.