Logo image
Back
Braided multipath channel representations and locally constrained sparse optimization in shallow water acoustic communications
Abstract   Peer reviewed

Braided multipath channel representations and locally constrained sparse optimization in shallow water acoustic communications

Ananya Sen Gupta
The Journal of the Acoustical Society of America, Vol.159(4_Supplement), pp.A212-A212
08/01/2026
DOI: 10.1121/10.0045551

View Online

Abstract

The shallow water acoustic channel is well-known to exhibit rapidly fluctuating delay spread due to time-varying multipath arrivals. The channel support maybe sparse but the degree of sparsity can vary substantially as a function of changing unpredictable oceanic conditions, e.g., surface wave focusing events. This makes shallow water acoustic channel tracking an interesting sparse optimization challenge. Popular convex optimization methods to track this time-varying channel include but, are not limited to, classical least square error minimization techniques, as well as sparse optimization methods such as basis pursuit and mixed norm solver techniques. The talk, with recent interpretations on prior work with visions of how to move forward, will highlight trade-offs between some of these well-known techniques in terms of their channel estimation accuracy, as well as their agility to follow unpredictable changes in the channel delay spread. We will also provide an overview of different channel representations using braid manifolds, and highlight how different braid representations can capture different physical and statistical elements of the shallow water acoustic channel multipath. In particular, we will discuss, based on Bellhop simulations and SPACE08 experimental field data, how braided optimization can employ locally constrained gradient descent approaches for tracking and predicting multipath eigenray bundles.

Details

Metrics

1 Record Views
Logo image