基于计算效率优化的无人机边缘计算卸载策略

Offloading strategy for UAV edge computing based on computational efficiency optimization

突发的自然灾害导致计算与传输设备受损无法提供可靠的通信和计算能力,使数据传输有效性较差与计算效率较低。面向上述问题,综合考虑任务、计算资源以及无人机轨迹等因素,提出一种多人机辅助用户任务计算卸载模型,根据服务质量参数自适应选择最优无人机完成边缘设备计算卸载。所提计算效率最大化问题非凸且复杂,是非线性的分式规划问题。为此用Dinkelbach方法将其转换为参数寻优问题,采用两阶段迭代算法获得次优解。数值结果表明,该方案能更好地提高系统的整体计算效率。

Sudden natural disasters lead to computing and transmitting equipment damage,reliable communication and computing capabilities can not be provided,causing low data transferring effectiveness and low computational efficiency.Aiming at the above problems,a multi-human-machine assisted user task computation offloading model was proposed by comprehensively considering the factors such as tasks,computing resources and UAV trajectories.The optimal UAV was adaptively selected according to the quality of service(Qos)parameters to complete the computation offloading of edge devices.The proposed computational efficiency maximization problem was non-convex and complex,and it was a nonlinear fractional programming problem.The problem was transformed into a parameter optimization problem using Dinkelbach method.A two-stage iterative algorithm was used to obtain the suboptimal solution.Numerical results show that the proposed scheme can better improve the overall computational efficiency of the system.