Mobility-aware Predictive Computation Offloading and Task Scheduling for Mobile Edge Computing Networks

Abstract

With the development of information technology, mobile edge computing is regarded as a promising technology by pushing more computing resources to the edge of the network to meet the requirements of mobile applications. However, reliable and efficient computation offloading in high mobility scenarios such as expressways and high-speed trains still remains a challenge. In this paper, we take the user’s high mobility characteristics into consideration and propose a mobility-aware predictive computation offloading and task scheduling scheme based on genetic algorithm to solve the users’ task offloading and scheduling problem in high mobility scenarios. To reduce the offloading failure rate of the high mobility users, we put forward a prediction module to predict the users’ speed and estimate the user’s residence time to assist the offloading decision. Based on the prediction information, we formulate the computation offloading and task scheduling problem as a combinational optimization problem, and utilize a two-layer coding genetic algorithm to optimize the offloading decision and scheduling sequence of the high mobility users. Simulation results show that the proposed algorithm can effectively reduce the offloading failure rate and is superior to other traditional algorithms in performance improvement.