多用户多边缘的公平卸载及优化策略研究

Research on Multi-Edge Fair Offloading and Optimization Strategy for Multi-user

针对物联网设备计算密集和时延敏感的特点,边缘云可为物联网设备提供计算资源,对多个物联网设备在多个边缘云中的计算卸载进行研究。通过优化细粒度的卸载决策,解决时延敏感设备的计算卸载问题,在降低时延的同时考虑用户公平性,即优化物联网设备的时延并最小化其最大用户的任务完成时间。该优化问题是一个混合整数非线性规划问题,为了解决这个NP-hard问题,提出了基于潜在博弈论的分布式算法来降低运算复杂度。仿真结果验证了所提算法的收敛性,而且对公平性有显著提升的同时,对物联网时延影响较小。

To complete compute-intensive and delay-sensitive tasks,edge cloud could provide computing resources for IoT devices.It was studied the computing offloading of multiple IoT devices in multiple edge clouds.By optimizing fine-grained offloading decisions,the problem of computing offloading for delay-sensitive devices was solved,and user fairness was considered while reducing delay,that was,optimizing the delay of IoT devices and minimizing the task completion time of the largest user.The optimization problem was a mixed integer nonlinear programming problem.In order to solve the NP-hard problem,a distributed algorithm based on potential game theory was adoped to reduce computational complexity.The simulation results verified the convergence performance of the proposed algorithm,and the results showed that it had a significant improvement in fairness and little impact on the IoT devices delay.