QoS-based Service Function Chains Orchestration And Dynamic Optimization Method

With the rapid expansion and development of the Internet,the current network's architecture and service configuration have become "rigid" to a certain extent,which makes it difficult to adapt to the agile development and deployment of future services.Network Functions Virtualization(NFV)technology is considered to be an important emerging technology to solve the challenges which current network faces.NFV is dedicated to separating the network functions from physical-dedicated hardware,and then virtualizes them into the form of software running on the single,common physical platform,so as to reduce or even completely replace the dedicated hardware devices deployed in existing networks.Therefore,in order to realize the agile development and flexible deployment of services in NFV networks,how to ensure reasonable and efficient orchestration and dynamic optimization of Service Function Chains(SFC)are key issues in NFV's research.The orchestration of SFC means that a number of Virtual Network Functions(VNF)should be dynamically placed,sequentially linked and allocated resource in the network according to the specific requirements and necessary constraints,so that they can provide users with required network services.In a given network topology,a reasonable SFC orchestration scheme can effectively improve the resource utilization of the network,reduce the service's end-to-end delay and guarantee the quality of service(QoS).At the same time,the structure of physical network and environment often undergo unpredictable changes due to scaling,failure and throughout burst and so on.In order to improve the adaptability of the SFC orchestration scheme to the changeable network's structure and environment,it is usually necessary to change and optimize the existing SFC orchestration results dynamically to avoid the similar "rigid"phenomenon in SFC orchestration,which may result in the growth of end-to-end delay and waste of network resource and ultimately affect the QoS.Therefore,in order to give full play to the advantages of NFV network,it is necessary to design a reasonable and effective QoS-based SFC orchestration and dynamic optimization method.This paper will discusses the problems of SFC orchestration and dynamic optimization in the NFV networks.The main research contents include:1)A service function chain orchestration method based on delay-aware is proposed.This paper deploys the network service reasonably on the physical network by applying the improved genetic algorithm.It can minimize the end-to-end delay while taking into account the network resource overhead to ensure the network quality of service.The algorithm aims to decrease users' delay-aware,and firstly improves the genetic algorithm in orchestration scenario.And then it converts the optimal solution obtained by genetic algorithm to the resolution of SFC orchestration.Finally it uses the shortest path algorithm to link the VNF nodes orderly to achieve traffic guidance and resource allocation.2)A service functions chain dynamic optimization method based on delay-threshold is proposed.This paper tries to achieve the scaling operation for the whole SFC by judging the reasonable scaling timing,so as to assure the delay maintains within the threshold to meet the user's requirements for quality of service.The method firstly puts forward a static-threshold-based scaling determination mechanism,and then describes the dynamic optimization of SFC in the vertical and horizontal dimensions.Finally it raises the method to achieve dynamic resource configuration and the optimization of function nodes,which can minimize resource overhead as far as possible.In summary,this paper proposes and implements an efficient and reasonable QoS-based service function chains orchestration and dynamic optimization method.The method aims to guarantee the quality of service,and tries to meet the end-to-end delay while maximizing resource utilization and reducing network resource overhead at the same time.This is of great significance for the management of NFV networks and the study of SFC issues.