Traffic Offloading for RIS-Aided Systems: Number Configuration and How Much Traffic Should be Offloaded?

Traffic offloading has been considered as an effective solution to address the problem of network congestion during peak hours. However, with the advent of the emerging huge number of wireless users and applications, a new problem arises: how to enhance the traffic offloading performance without utilizing more frequency bands. In this letter, we exploit the reconfigurable intelligent surface (RIS) for cellular and access point networks to provide users that support dual connectivity with a stable and high-quality connection. Specifically, we aim to explore the traffic offloading trade-off by minimizing the total power consumption. Considering the complexity of the problem, we design an inner approximation-based algorithm with successive convex approximation. Furthermore, we analyze how much traffic should be offloaded and optimize the number of reflecting elements, which can serve as the valuable guidance for effectively implementing RIS in practical applications. Simulation results verify that the proposed algorithm outperforms the existing offloading algorithms in terms of saving power consumption.