Enhanced Performance of a Phosphorescent White LED CAP 64QAM VLC System Utilizing Deep Neural Network (DNN) Post Equalization

Guoqiang Li,Fangchen Hu,Yiheng Zhao,Nan Chi
DOI: https://doi.org/10.1109/iccchina.2019.8855926
2019-01-01
Abstract:In this paper, a phosphorescent white LED CAP 64QAM VLC system utilizing deep neural network (DNN) and LMS linear equalization (LE) has been experimentally demonstrated. We successfully achieve a data rate of 2.4Gb/s at a 1.1-m indoor free space transmission with the bit error rate (BER) below the 7% forward error correction (FEC) limit of 3.8 × 10 -3 . Compared to the LE and volterra NLE+LE, DNN+LE brings a better performance to mitigate nonlinear distortion. The results show that DNN+LE will be a promising solution for indoor high-speed VLC system.
What problem does this paper attempt to address?