Energy-Efficient Power Control in D2D Networks: A Distributed ADMM Approach With Dynamic Penalty Coefficient
Yuting Huang, Xiaozheng Gao, Minwei Shi, Neng Ye, Kai Yang
Abstract
Device-to-device (D2D) communication plays an important role in future networks due to the explosive growth in Internet of Things (IoT) devices. The surge of mobile devices necessitates higher data transmission rates and resource utilization efficiency. To address the requirements, developing effective power control algorithms is crucial to improve network performance and reduce energy consumption. In this work, we propose a distributed power control scheme based on the combination of the alternating direction method of multipliers (ADMM) algorithm and the successive convex approximation (SCA) algorithm. With the aim of maximizing energy efficiency, the original problem is decomposed into multiple convex subproblems through SCA, and the distributed optimization of ADMM is used to solve each subproblem, thus making the solution of the problem more efficient. In particular, a dynamic penalty coefficient strategy is also developed to improve the convergence performance of the distributed algorithm. The simulation results demonstrate that compared with centralized power control methods, the proposed method can achieve similar optimal values and effectively distribute the computational load to each device, which can support the optimization design and implementation of future D2D communication systems.