Offloading Strategy for End-to-Edge Collaboration under Limited MEC Computing Resources

Abstract

The emergence of mobile edge computing (MEC) is to meet the requirements on computing capacity raised by massive Internet of Things devices. Task offloading strategy is one of the methods in MEC to schedule computing resources. In this paper, an End-to-Edge collaborative resource allocation model is established. Firstly, the system model with multiple user equipments and MEC servers is established and an End-to-Edge connection model is built, then considering the limited computing resources of servers, time delay and energy consumption are formulated according to queuing theory. The task offloading problem is transformed into an optimization model to minimize the total cost of the system. To manage the non-convex problem, an improved particle swarm optimization algorithm is proposed to formulate task offloading strategies, which can jointly reduce time delay and energy consumption and solve the problem of uneven computing resource allocation. Simulations demonstrate that the proposed algorithm has less possibility of falling into lo-cal optimum and has better performance in reducing system cost, by comparing with the existing algorithm and local processing strategy.