面向多层次、多指标的光网络规划与评估方法  

Multi-level and multi-index planning and evaluation method for optical network

作　　者：曹晓宏 徐思雅[1] 周桂平 王英杰 于波涛 CAO Xiaohong;XU Siya;ZHOU Guiping;WANG Yingjie;YU Botao(China State Key Laboratory of Networking and Switching Technology,Beijing University ofPosts and Telecommunications,Beijing100876,China;State Grid Liaoning Electric Power Co.,Ltd.,Shenyang110006,China;Beijing Guodiantong Network Technology Co.,Ltd.,Beijing100070,China)

机构地区：[1]北京邮电大学网络与交换技术国家重点实验室,北京100876 [2]国网辽宁省电力有限公司,辽宁沈阳110006 [3]北京国电通网络技术有限公司,北京100070

出　　处：《电信科学》2019年第9期58-68,共11页Telecommunications Science

基　　金：国家重点研发计划基金资助项目(No.2016YFB0901200)

摘　　要：首先深入分析了影响业务可靠性的关键指标,并建立了面向可靠性和建设成本的网络规划模型,然后设计了自适应免疫算法以求解网络规划方案。进而,建立了面向多层次、多指标的网络规划方案评估模型,并使用神经网络算法获得指标权重,实现对网络规划方案的全面、有效评估。通过仿真可知,提出的面向多层次、多指标的光网络规划与评估方法能够提高光缆资源利用率和站点成环率、降低网络建设成本以及保障网络可靠性,有助于提高运维人员的工作效率和运维水平。The key indicators affecting business reliability were analyzed in depth,and a network planning model for reliability and construction cost was established.Then an adaptive immune algorithm was designed to solve the network planning model.Then,a multi-level and multi-index oriented evaluation model for network planning method was established,and the weights of the indexes were obtained by using neural network algorithm,so as to achieve a comprehensive and effective evaluation of the network planning method.The simulation results show that the multi-level and multi-index oriented optical network planning and evaluation method proposed can improve the utilization of optical cable resources and the site ring rate,reduce the network construction cost,ensure the network reliability,and help to improve the efficiency and quality of network operation and maintenance.

关 键 词：光网络 网络规划 评估模型 自适应免疫算法 神经网络算法 

分 类 号：TN929.1[电子电信—通信与信息系统]