基于深度强化学习的毫米波通信资源分配方法  

Resource Allocation Method for mmWave Communication Based on Deep Reinforcement Learning

作　　者：顾玲玲[1] 袁新颜[1] GU Lingling;YUAN Xinyan(School of Electronics and Information,Jiangsu Vocational College of Business,Nantong 226011,China)

机构地区：[1]江苏商贸职业学院电子与信息学院,江苏南通226011

出　　处：《无线电工程》2023年第7期1578-1586,共9页Radio Engineering

基　　金：国家自然科学基金青年基金项目(62102199)。

摘　　要：针对5G毫米波通信系统资源管理难度大的问题,提出一种基于深度强化学习的毫米波通信系统资源优化分配方法。建立了符合实际应用场景的5G毫米波通信系统模型,模型考虑了移动通信基站、蜂窝用户终端与设备-设备用户终端,同时包含网络服务质量的多个关键因素。以最大下行速率与误码率为约束条件,以最大化系统总吞吐量为目标,利用深度强化学习技术寻找系统的资源分配措施。仿真实验结果表明,该方法通过资源优化分配措施改善了毫米波通信系统的总吞吐量。Considering the difficulty of resource management in 5G millimeter wave communication system,a resource optimized allocation method based on deep reinforcement learning is proposed for millimeter wave communication system.A 5G millimeter wave communication system model complying with actual application scenario is constructed,in which mobile base station,cellular user equipment and device to device user equipment are considered,and multiple critical factors of network quality of service are contained.Then,taking the maximum downlink rate and bit error rate as constraint conditions,and maximized total throughput of system as target,the deep reinforcement learning technique is utilized to discover the resource allocation strategy for the system.Simulation experimental results show that the proposed method improves the total throughput of millimeter wave communication system by resource optimized allocation method.

关 键 词：移动通信 蜂窝网络 深度学习 强化学习 毫米波通信 网络吞吐量 

分 类 号：TN929.5[电子电信—通信与信息系统]