Enhancing RIS-aided D2D communication for IoT applications

Reconfigurable intelligent Surface (RIS) is a potential 5G/6G technology that provides energy efficient solutions to the increasing demand of communication network resources. The number of Internet of Things (IoT) nodes in a network is also on the rise as automation is being implemented in different fields. In this paper, the data rates for RIS-aided Device to Device (D2D) communication between two IoT nodes is maximized by defining a non-convex optimization problem with Quality of Service ($\mathrm{Q} 0 \mathrm{~S}$) constraints. Due to the non-convex nature of the problem, an alternate Reinforcement Learning based approach is implemented instead of a traditional mathematical approach as it is computationally exhaustive. The results obtained on the implementation of the proposed approach gives enhanced performance in the communication link.