车辆边缘计算中多车辆协同任务计算方案

A Multi-vehicle Collaborative Task Calculation Scheme in Vehicle Edge Computing

针对车辆的移动和任务流量的动态变化会对周围智能车辆选择造成严重干扰的问题,提出了一种基于库恩-曼克尔斯(Kuhn-Munkres,KM)算法的多车辆协同任务计算方案。首先,通过对车辆的移动分析,建立了动态车辆选择模型;然后,根据层次分析法和KM算法,在通信时间限制和计算资源约束下,求解多任务最优匹配,以实现任务的总计算时间最小化。仿真实验结果表明,相比已有卸载方案,所提方案能够有效减少任务的总计算时间,并降低任务卸载计算的失败率。

For the problem that the dynamic change of vehicle movement and task traffic will cause serious interference to the selection of surrounding intelligent vehicles,a multi-vehicle collaborative task calculation scheme based on Kuhn-Munkres(KM)algorithm is proposed.Firstly,a dynamic vehicle selection model is established by analyzing the movement of vehicles.Then,according to the analytic hierarchy process and KM algorithm,the multi-task optimal matching is solved under the constraints of communication time and computing resources to minimize the total computing time of the task.The simulation results show that compared with the existing offloading schemes,the proposed scheme can effectively reduce the total computation time of the task and reduce the failure rate of the task offloading computation.