面向移动边缘计算的动态规划算法研究

移动边缘计算可以将用户任务卸载至边缘服务器，以减少移动设备的能耗与时延。通过研究边缘计算场景，提出了一种自适应动态规划算法，以优化用户的卸载决策。所提算法采用创新的比特流填表方式以节省计算时间，同时在满足时间约束的条件下减小能耗与时延。结果表明，该算法可在满足应用程序执行时间约束的前提下找到近似最优解，同时在不损失计算效率的前提下处理较大的卸载问题。

Research on Dynamic Programming Offload Optimization Algorithm for Mobile Edge Computing

Mobile edge computing offloads user''s tasks to edge servers to reduce the power consumption and delay of mobile devices.By studying edge computing scenarios, an adaptive dynamic programming algorithm is proposed to optimize users'' offloading decisions.The proposed algorithm uses an innovative bitstream filling method to save calculation time while reducing power consumption and delay under the condition of satisfying time constraints.The results show that the algorithm can find the approximate optimal solution under the premise of satisfying the application execution time constraint, and can be used for the processing of large unloading problems without losing computational efficiency.