Scheme Design And Applicaion Research Of Lightweight Edge Computing Platform

In recent years,with the deepening of the "Internet+" strategic pattern and the maturity of 5G technology,the Internet of Things will meet a period of rapid development.The explosive growth of core network data traffic presents a huge challenge to the current cloud computing architecture in terms of response latency and communication burden.Edge computing provides computing and storage at the edge of the user's network,providing flexible services with low latency and high bandwidth efficiency.It becomes a feasible supplementary solution under the cloud architecture and represents a new trend of the Internet of Things solution.Edge computing is designed to meet key requirements such as scalability,multi-tenancy,security,privacy,and flexibility.In these areas,rapidly evolving lightweight virtualization technologies and software-defined networking technologies attempt to solve these problems such as improving the utilization of computing resources and implementing flexible networking.On the other hand,single-board computers are favored in the field of Internet of Things with their dual advantages of low cost and low power consumption.More and more studies have shown that a single-board computer can be used to build a cost-effective edge computing network.For the above network and technology development,this paper mainly studies the lightweight edge computing platform solution,the specific content is as follows:(1)With the in-depth study of edge computing research progress and edge computing platform solutions,combined with the requirements of edge computing platforms in the Internet of Things environment,this paper proposed a lightweight edge computing platform architecture.The lightweight of the architecture is embodied in lightweight edge nodes,lightweight network switching,and lightweight virtualization.Based on the platform architecture,the design of the lightweight edge computing platform is further proposed.The technical details of each functional module in the scheme are designed.The low-cost,low-power,small-volume single-board equipment is used to form the edge computing node.Lightweight virtualization technology and software-defined network technology are used for resource abstraction and control and orchestration,realizing a lightweight edge computing platform from networking to monitoring management to service provisioning.(2)Based on the lightweight edge computing platform design,we have completed the technical implementation of each functional module in the scheme.We implement an Edge Intelligent Server that combines network forwarding and computing storage functions on a low-cost,low-power single-board computer;edge distributed cluster is built according to master-slave distributed architecture using Edge Intelligent Server;service delivery of platform is realized by combining Docker and Kubermetes technology And container orchestration;using Open vSwitch and SDN technology to realize network switching and network control of the platform,the controller dynamically configures the flow table of the network switching node in the networking to realize the dynamic construction of the platform low-latency network.(3)We have built a test environment based on the lightweight edge computing platform solution for platform performance analysis,including network forwarding delay,request processing capability,power consumption,migration efficiency,and container orchestration perform?ance.The experimental results show that the Edge Intelligent Server has a single-hop forwarding delay of 0.4ms,which can reach the level of commercial SDN switches.The distributed node can process requests with similar response times to commercial servers,which is cost-effective.Edge Intelligent Server has obvious power consumption performance and migration efficiency advantages;the platform's container orchestration method can effectively reduce the service request response time.Then,based on the platform,the wireless probing user behavior analysis,the mobile video service QoE test system,and the lightweight edge video analysis system application are respectively implemented to verify the feasibility of the lightweight edge computing platform providing different IoT services.Finally,the performance of the video application on the platform is evaluated,and the platform scheme is verified to improve the QoE of the delay-sensitive service.