Abstract
This study proposes a novel receiver structure for underwater vertical acoustic communication in which the bias in the correlation-based estimation for the timing offset is learned and then estimated by a deep neural network (DNN) to an accuracy that renders subsequent use of equalizers unnecessary. For a duration of 7 s, 15 timing offsets of the linear frequency modulation (LFM) signals obtained by the correlation were fed into the DNN. The model was based on the Pierson–Moskowitz (PM) random surface height model with a moderate wind speed and was further verified under various wind speeds and experimental waveforms. This receiver, embedded with the DNN model, demonstrated lower complexity and better performance than the adaptive equalizer-based receiver. The 5000 m depth deep-sea experimental data show the superiority of the proposed combination of DNN-based synchronization and the time-invariant equalizer.
Highlights
Vertical underwater acoustic (UWA) communications are critical in deep-sea activities such as scientific exploration with human-occupied vehicles
We used the vectors n oL The deep neural network (DNN) consists of one input layer, L, hidden layers, and one output layer
The model total number of the linear frequency modulation (LFM) signals was N Frm + 1 = 15, which was equal to the sizes for both the input was trained in the simulation data generated in accordance with Figure 1, with a constant wind speed and output of the DNN model
Summary
Vertical underwater acoustic (UWA) communications are critical in deep-sea activities such as scientific exploration with human-occupied vehicles. The deep neural network (DNN)-based OFDM receiver proposed in [13,14] was shown to be more robust than conventional methods; this model was extended to deal with the time-varying underwater acoustic channel and was verified by simulations [7]. A novel receiver structure for underwater vertical acoustic communication is proposed, where the correlation-based estimation bias of the timing offset is learned and estimated by the DNN to an accuracy that renders the real-time adjustment of the subsequent equalizer unnecessary. The model was based on the PM random surface height model with a moderate wind speed and was verified to be applicable for various wind speeds and experimental waveforms This receiver, embedded with the DNN model, demonstrated lower complexity and better performance than the adaptive equalizer-based receiver.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
