Cloudlet Placement Algorithm Based on Ratio Cut and Its Kernel Solution

Abstract

In wireless metropolitan area networks, mobile edge computing provides a way to reduce cloud service latency by establishing small cloud infrastructures (such as cloudlets at the edge of the network). The distribution of mobile users is very complex, and mobile devices have many computationally intensive tasks that need to be processed by cloudlet. How to choose suitable placement locations for cloudlets and minimize the average access latency of mobile users to cloudlets has become research hotspots in mobile edge computing. The existing cloudlet placement algorithms have poor generalization ability and are prone to produce uneven partitions. To solve this problem, we proposes a cloudlet placement algorithm based on ratio cut and its kernel solution, which transforms the cloudlet placement problem into the ratio cut problem on the graph, and uses the kernel clustering method to optimize the objective function of cloudlet placement to reduce mobile users’ access delay and improve the service performance of mobile devices. Finally, simulation experiments prove the effectiveness of the proposed algorithm and it can meet the strict response time requirements of mobile users.