Design And Implementation Of Cloud-Edge-Collaborative Service Function Graph Mapping Method And Prototype System

With the development of Internet of things(IoT)and 5G technology,the number of terminals and business types in the IoT are gradually increasing,and users’ requirements for quality of service(QoS)are quite different.Servers in the traditional cloud center network are far away from users and cannot meet the delay requirements of some services.Mobile edge computing(MEC)technology can support the deployment of network services on the edge to reduce service latency.Through the introduction of MEC technology,the cloud-edge-collaborative IoT(CECIoT)architecture is formed,which can integrate the computing power of cloud and the low latency characteristics of MEC technology,and effectively realize the differentiated management of different types of business.Network virtualization function(NFV)technology is introduced to solve the problems of traditional network resource management,such as resource consolidation and low network scalability.In CECIoT architecture based on NFV technology,virtual nerwork function(VNF)can form a service function chain(SFC)and be deployed in the network to achieve unified management of cloud and edge resources.However,the existing SFC deployment scheme has the following problems to be solved:1.the traditional serial SFC construction and processing method leads to high service latency;2.1:1 VNF instance backup consumes a lot of resources;3.most deployment schemes optimize the performance of a single service QoS index,which makes it difficult to balance the QoS and the utility of network resources.Therefore,a reliable service function graph(SFG)construction and mapping mechanism is proposed in this paper to solve the difficulty of balancing services among multiple SFCs in CECIoT architecture.Firstly,a delay and reliability based SFG generation mechanism is proposed to realize parallel arrangement between VNFs and graphical combination of multiple SFCs.Then,an SFG deployment algorithm with QoS and cost balance is proposed to realize the comprehensive optimization of delay and resource consumption by applying differentiated mapping strategies to different SFC services.Finally,a traffic-change-aware SFG scaling algorithm is designed to delete,expand or migrate network services on related nodes and links according to the traffic changes of each SFC,so as to improve the success rate of network services.Simulation results show that the proposed algorithm can effectively improve the reliability and reduce the average service delay and comprehensive resource consumption.Furthermore,in order to verify and apply the proposed algorithm,this paper designs and implements a prototype system of cloud-edge collaborative service function graph mapping.Firstly,the requirements analysis of the system was completed,and the prototype system was divided into four modules:CECIoT network configuration module,user SFC request configuration module,SFG mapping algorithm configuration module,and algorithm execution and result display module.Then the system architecture and workflow are designed in outline,and the workflow of each functional module is designed in detail.After the database design and module development and implementation,the function and performance of the system are tested.The test results show that the developed prototype system can efficiently complete the construction and mapping of SFG in CECIoT,and provide system support for the network function deployment in this scenario.