基于延迟敏感应用的边云协同方案

智能设备存在着存储能力以及计算能力不足的问题，导致无法满足计算密集型和延迟敏感型应用的服务质量要求。边缘计算和云计算被认为是解决智能设备局限性的有效方法。为了有效利用边云资源，并在延迟和服务失败概率方面提供良好的服务质量，首先提出了一种三层计算系统框架，然后考虑到边缘服务器的异构性和任务的延迟敏感性，在边缘层提出了一种高效的资源调度策略。三层计算系统框架可以根据应用程序的延迟敏感性提供计算资源和传输时延，保证了边缘资源的有效利用以及任务的实时性。仿真结果验证了所提资源调度策略的有效性，并表明该调度算法优于现有传统方法。

Edge cloud collaboration scheme based on delay sensitive application

The storage and computing capabilities of smart devices are insufficient. As a result, they cannot meet the service quality requirements of computation-intensive and delay-sensitive applications. The effective methods to solve the limitations of smart devices are mainly edge computing and cloud computing. In order to effectively use edge cloud resources, and in terms of delay and service failure probability to provide good service quality, this paper proposed a three-layer computing system framework, and then considering the heterogeneity and edge server task delay sensitivity, proposed an efficient resource scheduling strategy at the edge layer. The three-layer computing system framework could provide computing resources and transmission delay according to the delay sensitivity of applications, which ensured the effective utilization of edge resources and real-time performance of tasks. Simulation results verify the effectiveness of the proposed resource scheduling strategy, and show that the scheduling algorithm is better than the existing traditional methods.