A new robust adaptive algorithm for underwater acoustic channel equalization

TL;DR

This paper presents a novel adaptive equalization algorithm for underwater acoustic channels that combines multiple error norms with a logarithmic cost function, improving stability and convergence in challenging environments.

Contribution

The paper introduces a new robust adaptive equalizer using a logarithmic cost function that enhances stability and convergence in non-stationary underwater acoustic channels.

Findings

Achieves convergence performance comparable to least mean fourth (LMF) equalizer.

Significantly improves stability in impulsive noise environments.

Validated through realistic simulations of underwater acoustic channels.

Abstract

We introduce a novel family of adaptive robust equalizers for highly challenging underwater acoustic (UWA) channel equalization. Since the underwater environment is highly non-stationary and subjected to impulsive noise, we use adaptive filtering techniques based on a relative logarithmic cost function inspired by the competitive methods from the online learning literature. To improve the convergence performance of the conventional linear equalization methods, while mitigating the stability issues, we intrinsically combine different norms of the error in the cost function, using logarithmic functions. Hence, we achieve a comparable convergence performance to least mean fourth (LMF) equalizer, while significantly enhancing the stability performance in such an adverse communication medium. We demonstrate the performance of our algorithms through highly realistic experiments performed on accurately simulated underwater acoustic channels.