Cloudlet deployment for workflow applications in a mobile edge computing-wireless metropolitan area network

With the rapid development of mobile technology, the demands of mobile applications for computational resources are increasing. Limited by some factors, the computational capacities of mobile devices (MDs) cannot meet mobile application requirements. Mobile edge computing(MEC) has emerged in this context and has brought innovation into the working mode of traditional cloud computing. By deploying edge servers at the network edge, the computation resources of cloud center are sinking, and the enrich computational resources of edge servers make up for the lack of MDs. As a specific form of edge server, cloudlet has been widely concerned by academia and industry in recent years. However, the existing works mainly focus on the computation offloading of simple tasks under the condition of fixed cloudlet positions and ignore the impact of cloudlet deployment scheme and data dependence among components of workflow applications (WAs) on the result of computation offloading. In this paper, the cloudlet deployment for WAs in a MEC-wireless metropolitan area network (WMAN) is formulated. We propose a encode library-enabled particle swarm optimization using genetic algorithm operators (EL-PSOGA) algorithm to optimize the execution end time of all WAs. Numerical results show that our algorithm can effectively reduce the execution end time of all WAs in the system compared with several benchmark algorithms.