Computation Offloading Based on Improved Sparrow Search Algorithm in Edge Computing Scenario

Abstract

With the network changes brought by 5G, Mobile Edge Computing (MEC) has been deeply concerned as a prospective computing pattern. In MEC network, offloading the tasks to edge servers can address the problems of 5G mobile users' delay sensitivity and insufficient energy. To further decrease the delay and energy consumption, the offloaded task data can be reduced by compressing part of the task data before computing offloading. This paper investigates the problem of collectively optimizing computation offloading, data compression, and resource distribution aiming at minimizing the total system cost with limited MEC computational capacity. To solve the problem, an improved sparrow search algorithm (ISSA) is developed, which integrates circle chaotic mapping strategy, dynamic step factor strategy and Levy flight strategy, and lots of experiments verified the excellent performance of ISSA-based offloading scheme. Experimental results show the developed offloading scheme outperforms the sparrow search algorithm (SSA) based offloading scheme and particle swarm optimization (PSO) based offloading scheme in reducing the total system cost.