Mobile edge computing (MEC) provides more computing and storage capability for the network edge devices to meet the latency-critical mobile applications running in mobile terminals. However, the computation/storage resource at the edge servers are limited, and only the important application data can be cached in the edge server. Thus, it is necessary to figure out the wise caching decision to minimize edge computing latency and energy consumption. This paper considers a system in which most mobile devices migrate duplicate computation tasks to the edge servers and share the data contents requested for computation tasks. To reduce the overall latency of all mobile devices, we study the collaborative task offloading and data caching models. In addition, we propose an efficient Lyapunov online algorithm that can perform joint task offloading and dynamic data caching strategies for computation tasks or data contents. The simulation results show that the proposed algorithm outperforms the traditional strategy in task offloading and data caching, which can effectively decrease the overall service latency of all mobile devices.

移动边缘计算（MEC）为网络边缘设备提供更多的计算和存储功能，以满足在移动终端中运行的对延迟至关重要的移动应用程序的需求。但是，边缘服务器上的计算/存储资源有限，并且只有重要的应用程序数据可以缓存在边缘服务器中。因此，有必要找出明智的缓存决策，以最大程度地减少边缘计算延迟和能耗。本文考虑了一个系统，其中大多数移动设备将重复的计算任务迁移到边缘服务器并共享计算任务所需的数据内容。为了减少所有移动设备的总体延迟，我们研究了协作任务卸载和数据缓存模型。此外，我们提出了一种有效的Lyapunov在线算法，该算法可以对计算任务或数据内容执行联合任务卸载和动态数据缓存策略。仿真结果表明，该算法在任务分载和数据缓存方面优于传统策略，可以有效降低所有移动设备的整体服务等待时间。