Research on Modulation Format Recognition for Underwater Visible Light Communication based on BiGRU

【Objective】Underwater Visible Light Communication (UVLC), as a cutting-edge technology replacing traditional acoustic communication, has garnered widespread attention due to its high data rates, broad bandwidth, low latency, and enhanced security. However, it also faces challenges such as the complexity of underwater channel environments, signal power attenuation, and the imperfections of optoelectronic devices leading to nonlinear effects. The signal degradation caused by these low Signal-to-Noise Ratio (SNR) and nonlinear effects can make it difficult for the receiving end to correctly identify and demodulate the original signal. However, accurately identifying the signal modulation format lays the foundation for subsequent use of other algorithms, such as nonlinear compensation and frequency offset compensation, to enhance communication performance. Hence, this study aims to address the performance limitations in Modulation Format Recognition (MFR) within UVLC systems, with a special focus on improving recognition accuracy and system robustness in complex underwater environments.【Methods】To address these issues, we propose an innovative MFR algorithm that combines Bidirectional Gated Recurrent Units (BiGRU) with coordinate transformation. This algorithm leverages the advantages of BiGRU in sequential data processing and the efficiency of coordinate transformation to effectively extract the signal features, significantly enhancing the modulation format recognition accuracy in underwater environments.【Results】Experimental results demonstrate that under various transmission voltage conditions, the algorithm achieves recognition accuracy rates exceeding 96% for ten different Quadrature Amplitude Modulation (QAM) and Amplitude Phase Shift Keying (APSK) signal modulation formats, including 2QAM, 4QAM, 8QAM, 8APSK, 16QAM, 16APSK, 32QAM, 32APSK, 64QAM and 64APSK. The training speed has doubled, and robustness in low SNR and nonlinear distortion conditions is significantly improved.【Conclusion】The algorithm proposed in this study notably enhances the MFR performance in complex underwater environments for UVLC systems, holding significant application value and technological innovation. It lays the groundwork for the future development of high-speed underwater communication technologies.