Yao Li,Haoyu Zhang,Chaoxu Chen,Z. H. Lu,Jianyang Shi,Nan Chi
标识
DOI:10.1109/oecc56963.2023.10209983
摘要
In this work, we proposed a mixed precision training-based transformer (MPT-Transformer) model to process signal post-equalization in underwater visible light communication channel (UVLC). The application of MPT-transformer model in 1.2m underwater experimental platform using 64 quadrature amplitude modulation-carrierless amplitude (64QAM-CAP) modulation achieves 3.08Gbps while the bit error rate (BER) below the 7% forward error correction (FEC) limit of $3.8\times 10^{-3}$, which is 120Mbps faster than traditional least mean squires (LMS) method.