基于深度强化学习的多租户PON在线带宽资源分配算法  

Online bandwidth resources allocation algorithm for multi-tenancy PON based on deep reinforcement learning

作　　者：季晨阳 毕美华[1] 周钊 陈天宁 林嘉芊 徐志威 JI Chenyang;BI Meihua;ZHOU Zhao;CHEN Tianning;LIN Jiaqian;XU Zhiwei(School of Communication Engineering,Hangzhou Dianzi University,Hangzhou 310018,China;Information and Communication Branch of State Grid Hunan Electric Power Co.,Ltd.,Changsha 410007,China)

机构地区：[1]杭州电子科技大学通信工程学院,杭州310018 [2]国家电网湖南省电力有限公司信息通信分公司,长沙410007

出　　处：《光通信技术》2021年第9期36-39,共4页Optical Communication Technology

基　　金：国家自然科学基金(No.61501157)资助;浙江省自然科学基金(No.LY20F050004)资助;浙江省教育厅一般科研项目(No.Y201942093、No.Y201942104、No.Y202044258)资助。

摘　　要：为了解决运营商共享网络资源带来的资源分配问题,提出了一种基于深度强化学习(DRL)的在线带宽资源分配算法。该算法将多租户无源光网络(PON)系统映射到DRL模型中,DRL代理通过与环境交互,为各个待处理的带宽请求和当前剩余带宽做决策,并不断更新策略参数直至模型收敛,从而完成算法优化。搭建了仿真系统,对该算法进行了可行性验证,仿真结果表明所提的算法可以有效提高带宽资源利用率。In order to solve the problem of resources allocation caused by operators sharing network resources, this paper proposes an online bandwidth resources allocation algorithm based on deep reinforcement learning(DRL). The algorithm maps multi-tenancy passive optical network(PON) system to DRL model. The DRL agent interacts with the environment to make decisions for each pending bandwidth request and the current remaining bandwidth, and constantly updates the policy parameters until the model converges, so as to complete the algorithm optimization. A simulation system is built to verify the feasibility of the algorithm. The simulation results show that the proposed algorithm can effectively improve the utilization of bandwidth resources.

关 键 词：多租户 网络共享 带宽资源分配 深度强化学习 

分 类 号：TN915.6[电子电信—通信与信息系统]