Doubly-Spread Channel Equalization based on Channel Shortening in Underwater Distributed Acoustic Sensing Communications

Abstract

Distributed acoustic sensing (DAS) using submarine optical cables is a new marine technology for large-scale underwater data transmission and data collection of autonomous underwater vehicles (AUVs). However, the shallow water surroundings and the mobility of AUVs cause severe multipath propagation and Doppler effect, resulting in sound signal distortion in the delay and Doppler domains when sound waves reach the submarine optical cable. In this paper, we propose an underwater acoustic equalization method based on channel shortening to mitigate these issues. We use frequency-domain decision feedback equalization (FD-DFE) for channel shortening to suppress multipaths and lower the Doppler spread of the channel, followed by time-domain decision feedback equalization based on recursive least square (RLS-DFE) to eliminate residual delay and Doppler spreads. This dual decision feedback equalization (DDFE) approach combines block- and symbol-wise equalization and demonstrates robustness through numerical simulations. In the reservoir experiment, the DDFE method increased the output signal-to-noise ratio by an average 4.5 dB compared to the traditional method for medium-rate communications at 3 kbps, and reduced the bit error rate by 60%.