摘要
开放最短路径优先(OSPF)协议在电力通信网络中有广泛应用,随着电力网络规模日益扩大,以往过多依赖于网管经验的区域划分方法面临着效率方面的挑战。为了降低网络中冗余数据包传输和拓扑变更时路由重计算量,提高网络收敛速度和网络稳定性,需要对大规模OSPF网络进行更加合理和高效的区域划分。研究提出一种基于密度聚类的改进区域划分(MDBSC)算法,通过将网络跳数抽象为密度的度量,结合考虑实际网络节点地理位置因素,采用聚类的方法指导电力网络的区域划分,并通过实验验证了方案的可行性。仿真实验结果通过OSPF网络收敛时间和流量状态统计验证了方案的可行性。
Open shortest path first （ OSPF ） protocol is widely used in electric power communicationnetworks, and with the increasing scale of power network, the area division method, which relies too muchon the experience of network management in the past, faces the challenge of efficiency. In order to reducethe amount of routing re-calculation in the transmission of redundant data packets and topology change inthe network, and to improve the convergence speed and network stability, a more reasonable and efficientregional division of large-scale OSPF network is needed. An improved region division （ mdsc ） algorithmbased on density clustering is proposed, by abstracting the number of network hops as the densitymeasure and considering the geographical location of the actual network nodes, the algorithm usesclustering method to guide the regional division of power network, and verifies the feasibility of thescheme through experiments. The simulation results verify the feasibility of the scheme through OSPFnetwork convergence time and traffic state statistics.
出处
《微处理机》
2018年第1期13-17,共5页
Microprocessors
关键词
电力通信网络
OSPF协议
区域划分
跳数
聚类
Electric power communication network
OSPF
Regional division
Hops
Clustering
