A Robust ADMM-Based Optimization Algorithm For Underwater Acoustic Channel Estimation

Tian Tian,Agastya Raj,Bruno Missi Xavier,Ying Zhang,Feiyun Wu,Kunde Yang

2023-08-25

Abstract:Accurate estimation of the Underwater acoustic (UWA) is a key part of underwater communications, especially for coherent systems. The severe multipath effects and large delay spreads make the estimation problem large-scale. The non-stationary, non-Gaussian, and impulsive nature of ocean ambient noise poses further obstacles to the design of estimation algorithms. Under the framework of compressed sensing (CS), this work addresses the issue of robust channel estimation when measurements are contaminated by impulsive noise. A first-order algorithm based on alternating direction method of multipliers (ADMM) is proposed. Numerical simulations of time-varying channel estimation are performed to show its improved performance in highly impulsive noise environments.

Signal Processing

What problem does this paper attempt to address?

This paper aims to solve the key problems in underwater acoustic channel estimation, especially in coherent systems. Due to the severe multipath effect and large delay spread, the estimation problem becomes very complex and large - scale. In addition, the non - stationarity, non - Gaussianity and impulsive characteristics of the ocean environmental noise bring further obstacles to the design of the estimation algorithm. Under the framework of Compressed Sensing (CS), this paper pays special attention to the robust channel estimation problem when the measurement data is contaminated by impulsive noise.

To deal with these problems, the author proposes a first - order optimization algorithm based on the Alternating Direction Method of Multipliers (ADMM). This algorithm reformulates the optimization problem by introducing an auxiliary variable \( z \) as follows:
\[ \min_{x, z} \tau \| z \|_1 + \| x \|_1 \]
\[ \text{subject to } Ax + z = y \]

Here, \( y \) is the observation vector, \( x \) is the vector of unknown channel parameters to be estimated, \( A \) is the matrix constructed by training signals/symbols, \( z \) is the auxiliary variable, and \( \tau \) is the regularization parameter. Through this method, the algorithm can handle the channel estimation problem in a highly impulsive noise environment more effectively.

The numerical simulation results show that the algorithm performs excellently in a highly impulsive noise environment and has high estimation accuracy and stability.