MIMO TFT-OFDM for Underwater Acoustic Communication with Bidirectional Soft Decision Feedback Equalization

Abstract

Conventional underwater acoustic multiple input multiple output orthogonal frequency division multiplexing (MIMO-OFDM) system adopts orthogonal pilots, which leads to low spectral efficiency. In this paper, we propose an iterative time-frequency training OFDM (TFT-OFDM) scheme for underwater acoustic MIMO systems to improve spectral efficiency. Each TFT-OFDM symbol contains the time-domain training sequence and frequency-domain orthogonal pilots. The time-domain training sequence is used to estimate the path delays without eliminating the influence of OFDM symbols. By using the estimated path delays and the sparse characteristics of the underwater acoustic channel, the path amplitudes can be obtained with fewer pilots. In addition, the bidirectional soft decision feedback equalizer (BiSDFE) is introduced into the underwater acoustic MIMO system. BiSDFE eliminates intercarrier interference from both directions by combining the results of forward and backward decision feedback equalization to mitigate error propagation. Simulation results demonstrate that the proposed MIMO TFT-OFDM system achieves better bit error rate (BER) performance and higher spectral efficiency than the MIMO ZP-OFDM system.