Clustering Algorithm Based on Task Dependence in Vehicle-Mounted Edge Networks

Mobile Edge Computing can shift computing tasks from cloud servers to mobile edge servers for processing so that data and applications located closer to users. However, traditional clustering algorithm was not adopted to deal with the case where similar tasks exist in the edge computing networks. In addition, the repeated computation of similar tasks will lead to high loading in the edge server. In this paper, we propose a task clustering algorithm based on task dependency in Vehicle-Mounted Edge Networks. And we study task adaptation algorithm based on task dependency in Vehicle-Mounted Edge Networks. We first introduce the task clustering algorithm to car edge network. Then, we propose a task arrival model in Vehicle-Mounted Edge Networks. By constructing similarity matrix to cluster tasks, different tasks with similar characteristics will be clustered based on the maximum value of task dependency. Next, tasks will be assigned to edge servers according to data perception scheduling. Finally, the simulation results show the proposed algorithm improves the overall efficiency of task processing in the Internet of Vehicles at the expense of a small amount of clustering effects.