Investigation on Probabilistic Shaping for Symbol-Level Uplink Non-orthogonal Multiple Access Visible Light Communication Systems

In this paper, we investigate the performance of probabilistic shaping for uplink non-orthogonal multiple access (NOMA) visible light communication (VLC) systems. The optimal decision threshold for hard decision decoding is analyzed to improve symbol error rate (SER) performance. Furthermore, we propose a constellation-adaptive approach to find the optimal tradeoff between altering the constellation scale and changing the probability distribution. Finally, we evaluate the achievable sum rate in two-user uplink NOMA VLC systems. The simulation results show that the proposed approach can both decrease the SER and increase the achievable sum rate.