联合功率控制和信道分配的蜂窝网络能效优化算法
Energy efficiency optimization under joint transmittion power and channel allocation for cellular network
徐会彬1
作者信息
- 1. 湖州师范学院信息工程学院,浙江 湖州 313000
- 折叠
摘要
为了提高混合设备到设备(D2D)蜂窝网络中D2D干扰导致系统能效下降的问题,提出了联合功率控制和信道分配的能效优化(EEPC)算法,进而提升系统能效.以 D2D 用户和蜂窝用户最小速率为约束条件,建立最大化能效的优化问题;利用块坐标下降法将优化问题转化为信道分配和功率控制两个子问题,再分别利用Q学习算法、Dinkelbach算法和优化最小(MM)算法求解.并对Q学习算法中贪婪搜索因子进行改进,采用动态的搜索因子,平衡探索与利用间的关系.性能分析表明,提出的 EEPC 算法提升了系统能效.
Abstract
In order to solve problem of low-energy efficiency caused by device-to-device(D2D)interference in hy-brid D2D cellular network,energy efficiency improvement based on joint power control and channel allocation(EEPC)algorithm was proposed to improve the energy efficiency.The optimization problem of maximizing energy efficiency was established with the constraint of ensuring the minimum rate of D2D users and cellular users.By using block coordinate descent method,the optimization problem was transformed into two sub-problems of channel allo-cation and power control,which were solved by Q-learning algorithm,Dinkelbach algorithm and majoriza-tion-minimization(MM)respectively.The greedy search factor in Q-learning algorithm was improved,and a dynamic search factor was used to balance the relationship between exploration and exploitation.Performance analysis shows that the proposed EEPC algorithm improves the energy efficiency of the system.
关键词
混合D2D蜂窝网络/能效/信道分配/Q学习/Dinkelbach算法Key words
hybrid D2D cellular network/energy efficiency/channel alloction/Q-learning/Dinkelbach algorithm引用本文复制引用
基金项目
浙江省软科学研究计划项目(2021C35129)
湖州市自然科学基金资助项目(2021YZ20)
出版年
2024
NETL
NSTL
万方数据