A new offloading method in the green mobile cloud computing based on a hybrid meta-heuristic algorithm

Abstract

Mobile Cloud Computing (MCC) results from the evolution of several Internet-based technologies that allow mobile consumers to gain cloud computing advantages and attain green computing by utilizing their cellphones. MCC's use may be restricted because of increasing energy utilization and restricted access to energy supplies. This challenge may be tackled using the green approach, which entails the development of several algorithms or approaches to attain the highest possible degree of service quality. Several techniques in this idea allow consumers to offload data processing and storage to cloud-based servers to reduce the energy utilization of mobile nodes. Offloading is an essential element that helps MCC reduce task load and optimize data storage utilizing a cloud resource pool. Regarding the advantages of the offloading procedure, issues such as reliability (node failure), energy efficiency (energy consumption, task processing), and time management (execution time and task deadline) must be handled. This work presents a unique offloading approach in green MCC depending on a hybrid meta-heuristic algorithm because of the NP-Hard nature of the issue. The African Wild Dog Algorithm (AWDA) and the cellular learning automata technique are merged to generate better responses. AWDA is based on the cooperative hunting behavior of African wild dogs. The quantity of cooperative hunters determines the efficacy of hunting. The optimization procedure is analogous to this collective hunting behavior. Furthermore, a learning automaton in the cellular learning automata technique describes each of these components. An element's learning automaton seeks to discover which action best complements its surrounding components. The simulation findings illustrated that the suggested hybrid technique outperforms the other algorithms regarding energy consumption, delay time, and cost. The proposed algorithm has improved the delay time, cost, and energy metric by 0.43%, 3.17%, and 8.37%, compared to other algorithms.