Optimal Discrete Constellation Inputs for Aggregated LiFi-WiFi Networks

In this paper, we investigate the performance of a practical aggregated LiFi-WiFi system with discrete constellation inputs from view. We derive achievable rate expressions for first time. Then, study maximization problem via optimizing distribution and power allocation jointly. Specifically, multilevel mercy-filling scheme is proposed by exploiting relationship between mutual information minimum mean-squared error (MMSE) inputs. Meanwhile, an inexact gradient descent method obtaining optimal probability distributions. To strike balance computational complexity transmission performance, further develop framework that maximizes lower bound where can be obtained in closed forms distributions solved efficiently Frank-Wolfe method. Extensive numerical results show optimized strategies are able to provide significant gains over state-of-the-art schemes terms rate.