Estimation and equalization for shallow water communication channel using geometric encoding of channel multipath

Abstract

The shallow water acoustic channel is well-known to exhibit rapid fluctuations in its impulse response, which is challenging to estimate and compensate for in real time. In particular, the moving ocean surface and static sea bottom reflect the acoustic signal in unpredictable ways to create rapidly time-varying multipath arrivals. The talk will focus on the nexus of shallow water acoustic propagation paths and how multipath can be discovered and exploited to improve shallow water acoustic communications. In particular, the talk will review sampling strategies for the time-varying shallow water acoustic channel and propose novel geometric encoding techniques that exploit non-uniform channel sampling strategies. We will also demonstrate how geometric encoding can be harnessed for real-time channel estimation as well as semi-blind channel equalization. Results based on experimental field data in recent work as well simulation results will be presented.