具有隐私保护的车联网边缘计算任务卸载资源分配策略

Privacy preserving resource allocation strategy for edge computing task offloading in internet of vehicles

针对车联网任务卸载的资源最优化问题,以无线供能移动边缘计算(WP-MEC)系统为基础,提出一种关于计算时间分配、能耗、本地计算能力和任务卸载的联合优化方案。在该系统中,将“收集然后传输”协议应用于车辆的能量采集和消耗阶段,确保车辆可以持续工作。为求解该最优化问题,提出一种基于模拟退火算法的系统能量效率最大化算法。实验结果表明,所提策略的平均电池电量比全卸载模式、仅本地计算模式提高了40%以上,有效降低了系统时延,验证了所提策略的有效性和高效性。

Aiming at the resource optimization problem of task unloading in the internet of vehicles, based on the wireless power supply mobile edge computing(WP-MEC) system, a joint optimization scheme on computing time allocation, energy consumption, local computing power and task unloading was proposed. In this system, the collect and then transmit protocol was applied to the energy acquisition and consumption stage of the vehicle, to ensure that the vehicle works continuously. To solve the optimization problem, a system energy efficiency maximization algorithm based on simulated annealing algorithm was proposed. Experimental results show that the average battery power of the proposed strategy is more than 40% higher than that of the full unloading mode and the local computing mode only, and the system delay is effectively reduced, which verifies the effectiveness and efficiency of the proposed strategy.