Deep neural network-based underwater OFDM receiver

Due to the characteristics of the underwater acoustic (UWA) channel, the process at the receiver is complicated to match the channel. To simplify receiver design and match UWA channel better, this study proposes a deep neural network-based orthogonal frequency division multiplexing receiver for UWA communication. Different from existing receivers needing a neural network and several other processing parts, the proposed receiver only uses a single neural network to implement the whole signal processing. Moreover, it is a general receiver which is suitable for other modulation schemes. Simulation results show that the proposed receiver offers better bit error rate performance over traditional ones.