Research On Fog Computing Resource Scheduling Algorithm

With the increasing number of the Internet of Things users,the amount of data transmission is increasing rapidly,resulting in heavy burden on cloud servers.However,fog computing has the advantages of low latency,real-time performance and supporting mobility.It provides a new way to alleviate the pressure of cloud servers.At present,resource scheduling is one of the key points in the research of fog computing,and it is the key factor affecting the performance of fog computing services.Especially when large-scale service requests appear,if the problem of resource scheduling is not effectively solved,service delay will be increased,resource utilization rate and user satisfaction will be reduced.In addition,there are many types of devices and applications in the Internet of Things,the increasing diversity of users service requests and the complexity of big data processing have brought challenges to the resource scheduling in fog computing.How to make the scheduling algorithm adapt to more service types has become the focus of research on resource scheduling problem in fog computing.Therefore,this paper chooses the resource scheduling problem in fog computing as a research topic.This paper mainly studies fog computing resource scheduling algorithms from two aspects.On the one hand,it designs a user-oriented fog computing resource scheduling algorithm.On the other hand,it designs a fog computing resource scheduling algorithm based on task priority and cost constraints.The details are as follows.(1)A fog computing resource scheduling algorithm based on improved spectral clustering algorithm is proposed.In order to improve the users satisfaction of fog computing in the Internet of Things,this paper proposes a user-oriented improved spectral clustering algorithm(ISCM).The ISCM algorithm uses a spectral clustering algorithm to reduce the dimensions of the matrix and solve the eigenvectors.Based on the improved k-means algorithm,the initial clustering center selection algorithm is added,and when the initial clustering center is selected,irrelevant data points are eliminated,thereby reducing the computational amount of the algorithm.The ISCM algorithm solves the problem that the clustering result is sensitive to initial values,and implements re-clustering so that the resulting clustering result is more stable.Finally,a fog computing resource scheduling scheme is obtained according to the clustering results.The experimental results show that the resource scheduling scheme based on improved spectral clustering algorithm is superior to the traditional spectral clustering algorithm in terms of reliability of clustering results and algorithm response time.(2)A resource scheduling algorithm based on task priority and cost constraints(PSCC)in fog computing is proposed.By using the Dijkstra algorithm and the AOE network theory,PSCC algorithm solves the shortest time,shortest path and critical tasks from the beginning of execution to the completion of all tasks.PSCC algorithm considers task priority and system cost,quality of service and so on,and maps the tasks to lower cost resources.According to the score matching rules,the algorithm matches the task demands with the attributes of resources by weighted matching,and get the resource allocation result that meets the task requirements.The PSCC algorithm not only reduces the service cost of fog computing,but also improves the accuracy of resource allocation.The experimental results show that PSCC algorithm can meet the cost-constrained resource scheduling and reduce the task completion time in the heterogeneous environment of fog computing.