基于Hose虚拟专用网络模型的网络资源优化与路由方法研究-通信与信息系统专业论文.docxVIP

ABSTRACTABSTRACT ABSTRACT ABSTRACT With the development of the Intemet and network security,virtual private network (VPN)technology has become more and more perfect and secure．The growth of user traffic，pluralistic development of business model，security requirements，as well as the emergence of intelligent optical network has further driven the development and application of the VPN．The hose VPN model outforms the conventional pipe model in many aspects，such as flexibility to send traffic and reduction in the size of access links through stastistical multiplexing．However,the researches for Hose-VPN mainly concentrate on smile hose scenario and most of the exiting algorithms for hose-VPN do not considerate the influence of hose asymmetry and hose throughout for resource reservation in VPN．To solve these problems，this paper mainly studies the two fields for hose-VPN as mentioned below： 1．In the static hose scenario，this paper proposes a heuristic algorithm，named Iterative Clustering Steiner Tree Algorithm(ICST)，to construct a resource shared tree for hose-VPN．The ICST algorithm has been simulated in several random topologies generated by Waxman model．The simulation results show ICST Mgofithm Can achieve better performance on resource reservation compared with traditionM Hierarchical Iterative Spanning Tree algorithm,especially in asymmetrical hose and unbalanced hose throughput scenarios． 2．Based on the dynamic hose model and user-controlled VPN，this paper proposes a user-controlled VPN interface system based on dynamic hose in custom edge(CE) and provider edge(PE)．In this user-controlled VPN system,the customer Can dyrm=1ically resize the size of its hose interface and update its label switch path(LSP) through the prediction of local traffic and network information to support the user-controlled function． This paper moreover proposes a signaling control mechanism for the user—controlled VPN interface system using extensional RSVP—TE protoc01．111e extension of RSVP—TE