A pinger localization algorithm using sparse representation for autonomous underwater vehicles

The problem of underwater Pinger localization is a crucial part of the underwater pipeline monitoring and maintenance by using an AUV in oil and gas industries. In this problem, a Pinger repeatedly transmits pulses of audio signals to an array of hydrophones. Due to the delay pattern, the direction of a Pinger relative to the hydrophone array can be estimated. However, since underwater environment is multipath in nature, multiple copies of Pinger pulses arrive at the hydrophones. In this paper, we employed the sparse representation to find the best combinations of the multipath signals, and use the direction of the first arrival path to be the direction of a Pinger.