Performance Evaluation of Low-complexity Distributed MU-MIMO Beamforming Techniques for Underwater Acoustic Communications

We construct two distributed MU-MIMO systems using more realistic and accurate statistical underwater acoustic (UWA) channel models. Our study evaluates the performance of various beamforming (BF) techniques in three system layouts: centralized (Cen), two-duster distributed system (Dis2), and four-duster distributed system (Dis4). For centralized BF, we analyze analog maximum ratio combining (analog MRC), digital maximum ratio combining (digital MRC), and zero-forcing (ZF). In the case of two-stage BF, we evaluate A-D (analog MRC for the first stage and digital MRC for the second stage), D-D (digital MRC for both stages), and A-ZF (analog MRC for the first stage and ZF for the second stage). Our findings highlight that analog MRC performs optimally in a centralized system, while digital MRC offers advantages in a distributed system. In distributed systems, we show that A-ZF emerges as a promising candidate in high signal-to-noise ratio (SNR) scenarios, offering low hardware and computational complexity. We place emphasis on examining the performance characteristics of various BF strategies across a range of system configurations, SNR levels, and considerations for hardware and computational complexity.