Dynamic User-Centric Clustering Design for Combined Transmission in Downlink LiFi System

Light fidelity (LiFi) provides a promising solution for indoor high-speed wireless transmissions. However, the dense deployment of access points (APs) and the unity frequency reuse in LiFi system may induce the inter-cell interference (ICI) issue. In this paper, we consider the combined transmission (CT) which converts harmful interference to useful signals and thereby combats the ICI, and propose a dynamic user-centric clustering scheme for the downlink LiFi system. In our design, with the objective of maximizing the system throughput under proportional fairness constraints, the CT clustering is formulated as a mixed-integer non- linear programming (MINLP) problem. Then we set up an exact potential game (EPG) model, where the Nash equilibrium is found via the best response algorithm, to provide a suboptimal solution to the MINLP problem. Simulation results demonstrate that the LiFi system using our presented scheme exhibits a higher throughput and a better satisfaction degree than that employing the single-point transmission (SPT) scheme.