In the context of fifth generation (5G) technology, Device-to-Device (D2D) communication plays a pivotal role, requiring swift and intelligent decision-making in mode selection and device discovery. This study addresses the challenge of rapid mode selection and device discovery within 5G communication networks, focusing specifically on enhancing spectral and energy efficiency for Internet of Things (IoT) applications. A novel self-centered game theory-based algorithm is introduced to optimize spectral efficiency and support intelligent mode selection. Additionally, the utilization of the support vector machine (SVM) expedites mode selection decisions. For D2D discovery, the Frank-Wolfe method is adopted, significantly improving the differentiation between D2D and Cellular users based on signal strength and interference, thereby enhancing spectral efficiency. The proposed approach maximizes spectral efficiency while adhering to strict power and interference constraints, intelligently partitioning bandwidth into two subparts using game theoretic principles to amplify spectral efficiency. Furthermore, the emphasis on energy efficiency is underscored through iterative calculations to achieve maximum energy-efficient spectral allocation. Numerical analyses validate the efficacy of the proposed technique, revealing substantial improvements in accurately predicted labels. As the number of devices increases, precision and recall rates experience noteworthy enhancements, ultimately leading to superior bandwidth utilization. This research presents a significant contribution to the field of 5G communication, particularly concerning energy efficiency, which is paramount for IoT applications. By accelerating D2D connectivity and optimizing energy and spectrum resources, it advances the goals of energy-efficient D2D communication within 5G-IoT networks.
Read full abstract