Artificial Intelligence for Smart Resource Management in Multi-User Mobile Heterogeneous RF-Light Networks
Zi-Yang Wu, Muhammad Ismail, Erchin Serpedin, Jiao Wang
Abstract
Recent trends in 5G and beyond wireless networks have encouraged the migration from the already congested radio frequency (RF) spectrum to higher frequency bands. In this context, the ubiquitous presence of lighting systems supports wide scale deployment of wireless communication links via light. However, the susceptibility of light to user mobility hinders its wide adoption. Hence, the coexistence of RF and light-based wireless communications has the potential to offer seamless heterogeneous network (HetNet) coverage through intelligent vertical handover policies in the presence of users' mobility. In this article, we first present new insights on the implementation of realistic indoor mobile optical channels and the impact of crowd mobility on relevant channel statistics. Then, an artificial intelligence (AI)-based framework for efficient resource management in mobile multi-user RF-light HetNets is proposed using a deep learning-empowered optical link predictor and a multiagent reinforcement learning-based link assignment strategy. The proposed AI-based framework helps to lay down the foundations of smart resource management in mobile multi-user HetNets.