Research On Security And Application Of Edge Computing For Domestic Processors

As the development trend of the Internet of Everything continues to deepen,the average growth rate of data exceeds the average growth rate of existing network bandwidth greatly.The new era of big data has arrived.The bigdata machine has attracted public attention as an integrated system product of software and hardware.However,the current bigdata machine based on domestic processors(hereinafter referred to as domestic bigdata machine)still has shortcomings in terms of safety,autonomy and performance.In addition,45% of the data generated by the Internet of Things will be processed at the edge of the network.Therefore,the centralized processing model with the cloud computing data model as the core technology cannot process a large amount of data generated by terminal edge devices efficiently,and edge computing has emerged as the times require.However,the current edge computing architecture has the problems of scattered computing resources and limited single machine computing performance resources,which lead to the aggravation of task processing delay.From the perspective of edge users,in order to avoid the leakage of sensitive information of edge users and cause serious data privacy threats to users,identity authentication between edge users and edge servers is indispensable.The dissertation mainly focuses on the security and load balancing issues of the domestic bigdata machine in the edge computing environment.The following works are specifically carried out:(1)Aiming at the problem of massive data processing and storage in the edge computing environment of the domestic bigdata machine,a three-tier architecture based on edge users,domestic bigdata machine,and cloud data center is proposed.Expand by introducing an edge computing layer between edge user layer and the cloud layer to better provide low-latency user services.(2)Aiming at the identity authentication security problem of the domestic bigdata machine in the edge environment,an identity authentication scheme for the edge computing environment is proposed:IBE-EC(Identity-Based-Encryption Edge Computing).This authentication mechanism provides two identity authentication schemes:(a)between edge users and domestic bigdata machines;(b)between different domestic bigdata machines.It meets the security requirements of user identity authentication and data privacy protection in the edge computing environment,strengthens user information security,and enhances the security of edge computing.(3)In order to effectively support real-time data,various tasks need to be optimally scheduled to meet various SLA parameters,such as response time,service availability,delay,and cost,and to improve the performance of domestic bigdata machines.The dissertation proposes a load balancing algorithm for edge computing,namely LBA-EC(Load Balancing Algorithm Edge Computing).The algorithm is divided into two stages: task selection and scheduling between agent nodes of the bigdata machine and task selection and scheduling between containers.Finally,all tasks or services are effectively mapped to available resources to make full use of network edge resources,so as to reduce user delay and improve user service experience.(4)In the dissertation,the security dispatch system EDS for the domestic bigdata machine is realized in accordance with the above-mentioned edge computing-oriented identity authentication scheme and load balancing algorithm.The domestic bigdata machine was selected as the edge server to build an experimental environment,and the EDS security dispatch system was deployed on the bigdata machine.