Adaptive Diversity Combining Technology with Deep Neural Network for High-Speed and Reliable Underwater Visible Light Communication System

Wangwei Shen,Hui Chen,Zhongya Li,Junhui Hu,Sizhe Xing,Chao Shen,Ziwei Li,Junwen Zhang,Nan Chi
DOI: https://doi.org/10.1109/ogc52961.2021.9654410
2021-01-01
Abstract:We proposed and experimentally demonstrated a high-speed and reliable underwater visible light communication (UVLC) system using single-input multi-output (SIMO)-based adaptive diversity combining technology (ADCT) with deep neural network (DNN) for performance enhancements. The Maximum Ratio Combine (MRC) algorithm is used for ADCT with SIMO receiver-array to combat the underwater turbulence, and the multi-layer perception (MLP)-based DNN is employed to compensate the nonlinearity impairments. To the best of our knowledge, this is the first study to combine SIMO system, MRC algorithm and MLP in the UVLC system, achieving a maximum 2.4Gbps data transmission in balanced received-power case and a 0.25 Gbit/s data-rate increase while maintaining system stability in unbalanced received-power case under the BER threshold of $1x10^{-2}$.
What problem does this paper attempt to address?