Mode Selection and Q-Learning Based Resource Allocation for D2D Communication Networks
Ali Majid Hasan Alibraheemi, Tengku Faiz Tengku Mohmed Noor Izam, MHD Nour Hindia, Kaharudin Dimyati · 2024
Integrating Device-to-Device (D2D) technology has emerged as an essential aspect of the prospective 6G network. This technology offers a configurable operational mode, minimal energy consumption, reduced latency, and substantial capacity. Nevertheless, the implementation of D2D communications is accompanied by several technological obstacles and challenges. This research examines the issue of joined mode selection and spectrum allocation in relay-aided D2D communications within a cellular network. The primary aim of this study is to enhance the average data rate of the system while simultaneously ensuring the quality of service (QoS) for both D2D and cellular users (CUs). To address these concerns, the proposed method introduces a greedy mode selection strategy for the purpose of determining the most appropriate mode for each D2D pair. A new channel allocation technique is introduced, which utilizes an adaptive Q-learning approach to choose the optimal channel for D2D pairs that share the same channels of CUs. The simulation findings provide evidence that the suggested method exhibits superior performance compared to existing algorithms with regard to total system throughput.