基于谱聚类的异构蜂窝超密集网络高能效资源分配算法  

Spectral clustering-based energy-efficient resource allocation algorithm in heterogeneous cellular ultra-dense network

作　　者：王雪[1] 刘京 孙佳妮 张继真[2] 钱志鸿[1] WANG Xue;LIU Jing;SUN Jiani;ZHANG Jizhen;QIAN Zhihong(College of Communication Engineering,Jilin University,Changchun 130012,China;Changchun Institute of Optics,Fine Mechanics and Physics,Chinese Academy of Sciences,Changchun 130033,China)

机构地区：[1]吉林大学通信工程学院,吉林长春130012 [2]中国科学院长春光学精密机械与物理研究所,吉林长春130033

出　　处：《通信学报》2021年第7期162-175,共14页Journal on Communications

基　　金：国家自然科学基金资助项目(No.61771219);吉林省科技发展计划基金资助项目项目(No.20190303137SF,No.20200401084GX,No.20190201187JC)。

摘　　要：为了解决5G移动通信超密集场景下功耗较大、频谱紧张、能效不高等问题,针对两层异构蜂窝非正交多址接入网络,提出了一种基于能效最大的资源分配算法。在超密集场景下行通信链路中,通过分步求解频率资源分配和功率分配方案将NP-hard优化问题转化为确定性的约束寻优问题,提出了基于谱聚类用户分组算法和改进的k-means基站聚类分簇算法,得到不同用户组的频率资源分配方案。基于Dinkelbach方法将能效优化的分式问题转化为可求解的连续凸优化问题,并通过拉格朗日乘子迭代算法实现功率分配。从基站分簇、用户分组、资源块分配与功率分配方面共同优化系统能效,最大限度地削弱基站簇间干扰与簇内干扰。仿真结果表明,所提算法在能效和计算效率相较对比算法均有明显优化。In order to solve problems of high power consumption,spectrum shortage and low energy efficiency in the ul-tra-intensive 5G mobile communication scenario,a resource allocation algorithm based on the maximum energy effi-ciency for the two-layer heterogeneous cellular non-orthogonal multiple access network was proposed.The original NP-hard optimization problem on the downlink communication link of ultra-dense scene was divided into two subprob-lem,such as frequency resource allocation and power allocation,which became a deterministic constraint optimization problem.The frequency resource allocation scheme of different user groups was obtained by using base station clustering based on the improved k-means algorithm and users grouping based on spectral clustering algorithm.The fraction of energy efficiency optimization was transformed into a solvable continuous convex optimization problem and power dis-tribution was realized by Dinkelbach method,and the Lagrange multiplier iterative algorithm,respectively.Jointly op-timize system energy efficiency in terms of base station clustering,user grouping,resource block allocation and power allocation,which minimized the inter-cluster interference and intra-cluster interference of the base station efficiently.The simulation results show that the proposed algorithm is better on energy efficiency and computational efficiency compared with existing algorithms.

关 键 词：异构网络 资源分配 功率分配 能量效率 

分 类 号：TN92[电子电信—通信与信息系统]