Passive characterization of underwater sound channel using ray-based blind deconvolution algorithm

Abstract

Ray-based blind deconvolution (RBD) algorithm is a blind channel estimation method, which can estimate both the channel impulse response and the original signal broadcast from an unknown source. It is based on high-frequency ray assumption and requires only the elementary information of array geometry and sound speed at the array position. Its capability of estimating the channel in a passive manner can be used to infer the sound channel characteristics from a source of opportunity such as passing surface ship whose positioning information is available from geo-referencing systems like the Automatic Identification System. Recently, we applied the RBD algorithm to the shipping noise data recorded on vertical line arrays at multiple locations in shallow water and confirmed its applicability for channel estimation and performing environmental inversion. This presentation shows the channel impulse responses estimated from a variety of source-receiver geometry and discusses the characteristics of sound propagation as well as the environmental property such as bottom loss which can be observed using the RBD algorithm.