Heterogeneity-aware elastic provisioning in cloud-assisted edge computing systems

Edge computing is the provision of cloud services and IT environment services to application developers and service providers on the edge of the network. Edge computing faces some challenges, such as dealing with randomly varying workloads, which is an important issue. Thus, a cloud-assisted edge computing system (CAECS) is studied. A replica placement strategy is proposed to satisfy the diversity of user demands and reduce the response time. A data migration strategy is proposed to guarantee data reliability if there exist the released instances. A heterogeneity-aware elastic provisioning strategy is proposed to rent the cloud instances. Finally, the performance of the proposed algorithms is evaluated via extensive experiments. The results imply that the total tenanted cost of the heterogeneity-aware elastic provisioning algorithm can averagely achieve up to 19.23% and 9.50% reduction over that of ARP algorithm and MADRP algorithm, respectively. (C) 2020 Elsevier B.V. All rights reserved.