Model-Driven Based Deep Unfolding Equalizer for Underwater Acoustic OFDM Communications

It is challenging to design an equalizer for the complex time-frequency doubly-selective channel. In this paper, we employ deep unfolding approach establish underwater acoustic (UWA) orthogonal frequency division multiplexing (OFDM) system, namely UDNet. Each layer of UDNet designed according classical minimum mean square error (MMSE) equalizer. Moreover, consider QPSK equalization as a four-classification task and adopt Kullback-Leibler (KL) achieve smaller symbol rate (SER) with one-hot coding instead MMSE criterion. addition, introduce sliding structure based on banded approximation channel matrix reduce network size aid perform well different-length signals without changing structure. Furthermore, apply measured at-sea UWA offshore background noise evaluate proposed Experimental results show that performs better low computational complexity. Concretely, SER nearly order magnitude lower than MMSE.