The freshness of information in real-time monitoring systems has received increasing attention, with Age of Information (AoI) emerging as a novel metric for measuring information freshness. In many applications, update packets need to be computed before being delivered to a destination. Mobile edge computing (MEC) is a promising approach for efficiently accomplishing the computing process, where the transmission process and computation process are coupled, jointly affecting freshness. In this paper, we aim to minimize the average peak AoI (PAoI) in an MEC system. We consider the generate-at-will source model and study when to generate a new update in two edge server setups: 1) computing preemption, where the packet in the computing process will be preempted by the newly arrived one, and 2) non-preemption, where the newly arrived packet will wait in the queue until the current one completes computing. We prove that the fixed threshold policy is optimal in a non-preemptive system for arbitrary transmission time and computation time distributions. In a preemptive system, we show that the transmission-aware threshold policy is optimal when the computing time follows an exponential distribution. Our numerical simulation results not only validate the theoretical findings but also demonstrate that: 1) in our problem, preemptive systems are not always superior to non-preemptive systems, even with exponential distribution, and 2) as the ratio of the mean transmission time to the mean computation time increases, the optimal threshold increases in preemptive systems but decreases in non-preemptive systems.

翻译：实时监测系统中的信息新鲜度日益受到关注，信息年龄（AoI）成为衡量信息新鲜度的新型指标。在许多应用中，更新数据包需先经过计算处理才能传输至目的地。移动边缘计算（MEC）因其可高效完成计算过程而成为一种有前景的方案，其中传输过程与计算过程相互耦合，共同影响信息新鲜度。本文旨在最小化MEC系统中的平均峰值AoI（PAoI）。我们采用“按需生成”信源模型，研究在两种边缘服务器配置下何时生成新更新：1）计算抢占模式，即计算过程中的数据包将被新到达的数据包抢占；2）非抢占模式，即新到达的数据包需在队列中等待，直至当前数据包计算完成。我们证明：在任意传输时间与计算时间分布的非抢占系统中，固定阈值策略为最优策略；在计算时间服从指数分布的抢占系统中，传输感知阈值策略为最优策略。数值仿真结果不仅验证了理论结论，还表明：1）即使服从指数分布，抢占系统在本问题中并非始终优于非抢占系统；2）随着平均传输时间与平均计算时间比值增大，抢占系统的最优阈值递增，而非抢占系统的最优阈值递减。