On Generating Monte Carlo Simulations of Underwater Acoustic Communication Systems with Application to Transmit Beamforming

Abstract

Underwater Acoustic (UWA) communication systems still rely heavily on at-sea trials. This work presents an operational framework that significantly reduces the need for practical experiments. The key idea is to generate channel impulse responses (CIRs) drawn from probability density functions constructed based on trusted information and to employ Monte Carlo simulations to develop new UWA communication systems. Hence, the proposed operational framework depends only on cheaper-to-acquire physical measurements to produce CIRs. It comprises a model-based CIR replay tool and a stochastic-based UWA channel simulator. The former can be any model-based CIR replay tool, and the latter is proposed in this work and validated using data from four different practical experiments. We also carried out experiments for a transmit beamforming with signals digitally modulated in binary phase-shift keying, which were transmitted by an array and by a single source with equivalent power. For the array, the ideal transmit direction comes from the lowest bit error rate (BER) obtained with computer simulations. This paper compares the performance of the transmit array to the single source transmission and the results of a practical experimental transmission with a Monte Carlo simulation employing the proposed technique. We show that both achieved close results regarding BER and mean squared error. The conclusion is that the proposed operational framework, once adjusted to the specific transmission site, can be used to design new UWA communication systems, eliminating the burden of at-sea trials for tests of new transceivers. Finally, we conducted real-life transmit beamforming experiments to verify the BER gain obtained in practice using the steering angle obtained from simulations.