Research On The Algorithms For Resource-efficient Service Function Chain Deployment

With the increase of network users and the development of business diversification,the telecommunication industry needs to store and transmit a large number of data.The traditional communication network based on dedicated hardware has many problems,such as the rigidity of network,the difficulty of sharing resources and the high cost.Network function Virtualization(NFV)has been proposed to solve the problems faced by traditional networks.As one of the key research directions of resource allocation in NFV,Service Function Chain(SFC)deployment affects the performance of network in NFV environment.There are some problems in the current research on SFC deployment,such as low efficiency of resource utilization,incomplete consideration of performance indicators,topology limitations and so on.This thesis aims at the high resource efficiency in SFC deployment,and thus the corresponding deployment algorithms are proposed.The main contents and innovations are as follows:(1)For the offline SFC deployment in NFV,the corresponding mathematical model is established.Based on the mathematical model,aiming at the efficient utilization of resources and the reduction of end-to-end delay,an SFC deployment algorithm SFCDO based on BFS is proposed.The SFCDO uses BFS algorithm to perceive the topology information of physical network and preferentially selects the path with the number of hops close to the length of SFC to realize the deployment.In addition,simulation experiments and performance comparison between the proposed algorithm and G-SA algorithm are carried out.Experimental results show that SFCDO algorithm can effectively reduce the bandwidth resource consumption and end-to-end delay of deployment path.Compared with G-SA algorithm,SFCDO algorithm reduces bandwidth resource consumption and end-to-end delay by 49% and 40%,respectively.(2)To solve the problem of online SFC deployment in NFV,a mathematical model for online SFC deployment is established.Based on the mathematical model,an online SFC deployment algorithm OSFCD is designed to efficiently utilize the physical network resources and achieve network load balancing.In order to improve the performance of online SFC deployment,the algorithm optimizes the path of SFC deployment and fully considers the load ratio of nodes and links.In addition,the proposed algorithm is compared with the other three existing algorithms.The experimental results show that the OSFCD algorithm can effectively reduce the bandwidth resource consumption,the maximum node load ratio and the maximum link load ratio in the network.In addition,the success ratio,end-to-end delay and other metrics of online SFC deployment are also optimized.