Spatiotemporal processing and time-reversal for underwater acoustic communications

Abstract

High-rate underwater acoustic communication can be achieved using transmitter/receiver arrays. Underwater acoustic channels can be characterized as rapidly time-varying systems that suffer severe Inter Symbol Interferences (ISI) caused by multi-path propagation. Multi-channel combining and equalization, as well as time-reversal techniques, have been used over these channels to reduce the effect of ISI. As an alternative, a spatiotemporal focusing technique had been proposed. This technique is similar to time-reversal but it explicitly takes into account elimination of ISI. To do so, the system relies on the knowledge of channel responses. In practice, however, only channel estimates are available. To assess the system performance for imperfectly estimated time-varying channels, a simulation analysis was conducted. Underwater acoustic channels were modeled using geometrical representations of a 3-path propagation model. Multi-path fading was incorporated using auto regressive models. Simulations were conducted with various estimator delay scenarios for both the spatiotemporal focusing and simple time-reversal. Results demonstrate performance dependence on the non-dimensional product of estimation delay and Doppler spread.

(cont.) In particular, it has been shown that when this product is low, the performance of spatiotemporal focusing remains superior to simple time- reversal.