Neural-network-based end-to-end learning for adaptive optimization of two-dimensional signal generation in UVLC systems.

Visible light communication (VLC) benefits from the underwater blue-green window and holds immense potential for underwater wireless communication. In order to address the limitations of various equipment and harsh channel conditions in the underwater visible light communication (UVLC) system, the researchers proposed to use the method of autoencoder (AE) to tap the potential of the system. However, traditional AE schemes involve replacing the transmitting and receiving components of a communication system with a large multilayer perceptron (MLP) network, and they have significant drawbacks due to their reliance on a single network structure. In this paper, a novel 2D adaptive optimization autoencoder (2D-AOAE) framework is proposed to realize adaptive modulation and demodulation of two-dimensional signals. By implementing this scheme, we experimentally achieved a transmission rate of 2.85 Gbps over a 1.2-meter underwater VLC link. Compared to the traditional 32QAM UVLC system, the 2D-AOAE scheme demonstrated a 15.4% data rate increase. Moreover, the 2D-AOAE scheme exhibited a remarkable 73% improvement when compared to the UVLC system utilizing the traditional AE scheme. This significant enhancement highlights the superior performance and capabilities of the 2D-AOAE scheme in terms of transmission rate.

Transfer Learning Strategy in Neural Network Application for Underwater Visible Light Communication System.

Post-equalization using neural network (NN) is a promising technique that models and offsets the nonlinear distortion in visible light communication (VLC) channels, which is recognized as an essential component in the incoming 6G era. NN post-equalizer is good at modeling complex channel effects without previously knowing the law of physics during the transmission. However, the trained NN might be weak in generalization, and thus consumes considerable computation in retraining new models for different channel conditions. In this paper, we studied transfer learning strategy, growing DNN models from a well-trained 'stem model' instead of exhaustively training multiple models from randomly initialized states. It extracts the main feature of the channel first whose signal power balances the signal-to-noise ratio and the nonlinearity, and later focuses on the detailed difference in other channel conditions. Compared with the exhaustive training strategy, stem-originated DNN models achieve 64% of the working range with five times the training efficiency at most or more than 95% of the working range with 150% higher efficiency. This finding is beneficial to improving the feasibility of DNN application in real-world UVLC systems.