Computation offloading and service allocation in mobile edge computing

Abstract

The intensive mobile data traffic poses a great challenge for energy-constrained mobile devices. In the mobile edge environment, effective computing offloading and resource allocation can improve the service performance of edge computing systems. Therefore, a dynamic computation offloading model based on genetic algorithm is proposed in this paper. In this strategy, a task weight cost model based on processing delay and energy consumption is built, which can optimize processing delay and energy consumption simultaneously. Moreover, in view of the limited computing resources of edge servers, a resource allocation model based on utility maximization is proposed. In this strategy, the bidding strategies of users and edge nodes are studied and the resource is allocated to the high-unit bidding users based on the greedy strategy during the double auction process. A large number of experimental results show that the proposed computation offloading algorithm can significantly reduce task processing delay and energy consumption. For instance, the proposed offloading algorithm can save energy up to 14.81% and reduce processing delay up to 7.71% compared with the COPSO algorithm. Besides, the proposed resource allocation algorithm can promote the number of successful auction users and maximize the utility of the users and the edge nodes.