Joint Optimization Of Scalable VNF Placement And Flow Scheduling In Service Function Chain

Network Function Virtualization(NFV)technology is considered to be a revolutionary technology in the future network because it significantly reduces the OPEX and CAPEX of operators,improves the flexibility and effectiveness of network deployment,and shortens the deployment time.As an important part of NFV architecture,Virtual Network Function(VNF)greatly improves the dynamic and flexibility of network,but it also brings great challenges to the management and deployment of VNF.Since VNFs need to be arranged and combined into different service function chains(SFC)to provide services for network users,there are traffic changes and dependencies between VNFs,which makes the deployment of service function chains need joint consideration and optimization in multiple stages.This paper focuses on the joint optimization of VNF scaling,VNF placement and flow scheduling in the process of service function chain deployment.The optimization goal is to minimize the deployment resource consumption and optimize the performance indicators of network services represented by the service function chain.In view of the above research content,this paper has done the following work:(1)Aiming at the three stages of SFC placement,this paper uses the queue update equation to model the packet processing flow of SFC,and then puts the model into the scope of Lyapunov method.Then the resource consumption function related to network traffic is established,and the Lyapunov drift plus penalty method is used to analyze the network model,which can ensure the scalability of SFC and reduce the cost of SFC placement.(2)Aiming at the joint optimization model,this paper proposes a Lyapunov based joint optimization algorithm for SFC placement,and designs the mechanism of VNF scaling and flow scheduling based on the best match priority,which can ensure the stability of the system and achieve the approximate global optimal SFC placement scheme.