| With the rapid development of Big Data,Internet of Things and other technologies,the number of terminal devices and new applications is rising rapidly,and the era of Internet of Everything has begun.The traditional cloud computing model has problems in real-time,network bandwidth,energy consumption and other aspects because all the business data need to be transmitted to the cloud data center through the core network for processing and computing.In order to make up for the shortage of Cloud Computing model,the academia put forward the Edge Computing model.Edge Computing uses the computing and storage resources of edge node close to the edge of the network to provide faster response and improve the quality of service.Management node in edge computing system does resource allocation and task scheduling according to the service capacity of edge node.However,the current research ignores the dynamic of edge node service capability.The number of resources used in the decision as service capability cannot reflect the precise description of the edge node service capability,and there is the problem that the status information perception is not timely.Therefore,the optimal decision cannot be guaranteed and the user experience will be degraded.In order to realize the rapid perception of edge node service capability,this thesis studies the service capability assessment of edge node and the status information perception of edge node,and its contributions are mainly reflected in the following three aspects:(1)Aiming at the problem that existing service capability characterization methods cannot accurately describe the edge node service capability,this thesis proposes a service capability evaluation model of edge node based on service quality.Firstly,the service demands of edge node in the edge computing system are analyzed,and the service capabilities of edge node are determined to include computing capabilities and storage capabilities,and the service capabilities are described from the two dimensions of resource quantity and service quality.For the realization of the service quality,after the first from the various factors that influences the quality of computing and storage,and quality index is determined based on the index selection algorithm of Spearman correlation set,and then the weight of each index is determined based on the combination of empowerment,finally the service quality is described based on the fuzzy comprehensive evaluation,so as to realize the evaluation of the edge node service capability.(2)Aiming at the problems of poor timeliness and extra network overhead in the existing node state awareness schemes,this thesis proposes a quick awareness mechanism of edge node state information based on Web Socket.First proposed a multi-domain collaborative edge node management mechanism,on the basis of using the Web Socket protocol implements the design of management information to complete the status information embedded in the transmission,and put forward a kind of state information based on triggered update mechanism to ensure the timeliness of state information update,finally designs the corresponding state information sending and receiving process,Thus,the state information of edge node can be quickly perceived.(3)In order to verify the above theoretical research,this thesis designs and implements the edge node service capability perception prototype system,and verifies the effectiveness of the service capability evaluation model and status information perception mechanism proposed in this thesis in a real environment,so as to realize fast perception.To sum up,in order to provide decision support for resource allocation in the edge computing environment and improve the service quality of the edge computing system,this thesis conducts an in-depth study on the fast perception of the edge node service capability.Firstly,the edge node service capability evaluation model based on service quality is proposed.Secondly,the status information transmission mechanism of edge node based on Web Socket is proposed.Finally,a prototype edge node capability status sensing system is designed and implemented.Through a series of simulation experiments and system experiments,the feasibility and effectiveness of the research results in this thesis are verified. |