Pragmatic Performance Optimization of a Multichannel DFE System for a Wideband 100 kbps 1-km Subsea Acoustic Modem

We recently demonstrated a real-time wideband 100 kbps 1 km subsea acoustic modem in a horizontal channel off Falmouth, MA, USA, where we conducted the experiment in collaboration with the Woods Hole Oceanographic Institute. To build a real-time receiver, we must allocate limited computing resources judiciously, such that the performance of the wideband acoustic system is optimized. In this paper, we present a pragmatic performance optimization of the decision-feedback equalization (DFE) system we used, based on the number of input channels, feedforward taps, and feedback taps. This allows us to obtain the best performance using best-in-class recursive least squares (RLS) adaptive filters that require significant computational resources. This optimization gives good guidance to optimize the performance of a real-time system where computing resources are limited.