弹性光网络中基于DRL的RMSA算法  

RMSA algorithm based on DRL in elastic optical networks

作　　者：侯临风 何荣希[1] 吴梓敬 HOU Linfeng;HE Rongxi;WU Zijing(College of Information Science and Technology,Dalian Maritime University,Dalian Liaoning 116026,China)

机构地区：[1]大连海事大学信息科学技术学院,辽宁大连116026

出　　处：《光通信技术》2024年第3期57-63,共7页Optical Communication Technology

基　　金：国家自然科学基金项目(61371091、61801074、62371085)资助;大连市科技创新基金项目(2019J11CY015)资助。

摘　　要：为了更好地解决弹性光网络(EON)的路由、调制格式与频谱分配(RMSA)问题,进一步降低网络阻塞率,提出一种基于深度强化学习(DRL)的RMSA算法。该算法在奖励设计中将考虑影响RMSA决策的资源占用度和频谱邻接度这2个指标,以鼓励智能体优先选择资源占用度低、频谱邻接度高的路径来建立光路,并对比该算法与其它算法在不同网络中的性能。仿真结果表明:与几种典型的DRL算法相比,所提算法的网络阻塞率更低。In order to better solve the routing,modulation format and spectrum allocation(RMSA)problems of elastic optical networks(EON),and further reduce the network blocking rate,an RMSA algorithm based on deep reinforcement learning(DRL)is proposed.This algorithm will consider two indicators,resource occupancy and spectral adjacency,which affect RMSA decision making in reward design,to encourage agents to prioritize selecting paths with low resource occupancy and high spectral adjacency to establish optical paths,and compare the performance of this algorithm with other algorithms in different networks.The simulation results show that compared with several typical DRL algorithms,the proposed algorithm has a lower network blocking rate.

关 键 词：弹性光网络 路由、调制格式与频谱分配 网络阻塞率 深度强化学习 奖励设计 

分 类 号：TN929.11[电子电信—通信与信息系统]