Downlink Interference Management in Cell-Free VLC Network

Abstract

Visible light communication (VLC) network is a candidate complement to its radio frequency counterpart with the benefits of wide free spectrum and low energy consumption. Due to the dense distribution of access points (APs) and users in VLC network, interference management is crucial to ensure the quality of service. Meanwhile, VLC link relates not only to the transmission power and the orientations of APs, but also to the orientations of users. Hence, the coverage of APs can be irregular and aligned with the motion of users. Cell-free network model, which is composed of irregular cells, is suitable for the VLC network. Therefore, in consideration of the directionality of VLC links, we investigate the joint optimization of AP grouping, AP-user association and AP power allocation for interference management (IM) in the cell-free VLC network, which is formulated as a network utility maximization (NUM) problem. The NUM problem is a mixed non-convex and combinatorial problem, and a graph-based IM algorithm is proposed to solve it. Specifically, a heuristic AP grouping algorithm is proposed based on an interference graph, which is defined according to the locations of APs and users as well as their orientations, to preselect several candidate AP groupings from the huge number of feasible AP groupings. Given each select AP grouping, the AP-user association and AP power allocation is optimized via the successive convex approximation method. Then, the proposed IM algorithm will choose the AP grouping from those candidates, as well as the corresponding AP-user association and AP power allocation, that attain the highest network utility. Simulations demonstrate that the proposed IM algorithm could achieve a network utility close to the maximum one obtained by searching through the entire solution space, with comparatively low complexity, and outperforms the conventional IM schemes.