A Mobile Edge Computing Task Offloading Framework Based on Improved Beetle Antennae Search
Zhi Fang, Xin Li, Rundong Fan · 2021
As an indispensable key technology in 5G Internet of Things (IoT), Mobile edge computing (MEC) can provide low-latency and low-energy computing requirements for mobile devices. Moreover, computing offloading is an important part of MEC. It determines how the device offloads tasks to the server and affects the system revenue of edge computing system. However, existing algorithms usually have high complexity and computational time during obtaining a calculation offloading strategy. These would increase the overall time overhead of MEC system. Thus, in the computing offloading scenario for one server and multiple users, this paper proposes a low-complexity computing offloading strategy based on improved Beetle Antennae search (IBAS) to maximize system revenue. First, considering the system delay and energy consumption as optimization objectives, an edge computing system is remodeled into a system revenue optimization with constraints of the maximum number of server cores and longest task delay. Finally, by introducing multi-channel exploration and simulated annealing, an improved Beetle Antennae search algorithm is suggested to solve the optimization problem. Through simulation experiments of different equipment scales, they demonstrate that the proposed framework has lower delay than traditional algorithm without reducing global optimization capabilities and algorithm stability. So, it can effectively improve the user experience.