Logo image
Back
Underwater Acoustic Channel Estimation Using Sparse Representations and Machine Learning
Conference proceeding

Underwater Acoustic Channel Estimation Using Sparse Representations and Machine Learning

Preeti Meena, Ananya Sen Gupta, Brejesh Lall and Monika Aggarwal
International Symposium on Ocean Electronics (Print), pp.1-6
12/10/2025
DOI: 10.1109/SYMPOL68153.2025.11395920

View Online

Abstract

Estimating the underwater acoustic (UWA) channel is challenging due to its non-stationary and time-varying nature under dynamic oceanic conditions. However, the channel's impulse response often exhibits a sparse structure, which can be effectively exploited to improve estimation accuracy. The delay-Doppler domain provides a suitable framework to capture this sparsity. Thus, this paper presents a machine learning-based framework for underwater acoustic channel estimation by exploiting channel sparsity in the delay-Doppler domain. The presented framework investigates the application of various machine learning techniques, including K-means, Gaussian Mixture Models (GMM), density-based spatial clustering of applications with noise (DBSCAN), and spectral clustering to accurately estimate sparse channel impulse response. Experimental results demonstrate the effectiveness of each learning technique and highlight their relative performance under varying channel conditions.
 Machine Learning Marine Technology Noise Accuracy Channel estimation Channel impulse response Delay-Doppler domain Estimation Gaussian mixture model Sparse approximation sparse channel estimation Underwater acoustic channel estimation Underwater acoustics

Details

Metrics

1 Record Views
Logo image