Mobile Edge Computing (MEC) is a promising distributed computing paradigm, where the service providers deploy their computing power at the communication base stations close to mobile users. By providing task offloading at the network edge devices, the edge service providers can significantly reduce end-to-end latency and improve user satisfaction. However, there usually exists multiple providers in one edge, mobile users will face the choice of which edge service provider to offload their computing tasks after user allocation and offloading decision to a certain base station. In this study, from the perspective of MEC system, we investigate the online task offloading with edge service provider selection problem. We model it as a stochastic optimization problem, aiming to maximize the long-term time average user utility under limited resources, while ensuring the stability of the MEC system. We propose an online algorithm based on Lyapunov optimization to achieve a good trade-off between user satisfaction, energy consumption and system stability, then give theoretical proof of its performance bound. Experiments have been conducted based on a real-world dataset, and the results show that OEPS shows full-range performance advantages over three baseline methods.

Mobile edge computing (MEC) reduces data service latency by pushing data to the network edge. However, due to the dynamic and diverse requests of mobile users, the problem of mobile edge caching is more complex than cloud caching. Therefore, the existing model-based caching strategies cannot be directly used in the mobile edge caching environment. Besides, when taking the cooperative storage relationship between neighbor edge servers into consideration, the caching problem becomes more difficult. To this end, we formulate an mobile edge caching problem to minimize the total latency in mobile edge computing. Firstly, a heuristic caching strategy is proposed to solve the mobile edge caching problem in the single-time-slot scenario. Then, with the consideration of users’ mobility and the correlation of files, we propose a caching strategy for the multiple-time-slot scenario based on multi-agent deep reinforcement learning. To address the cold start problem in deep reinforcement learning, we adopt the proposed heuristic caching strategy used in the single-time-slot scenario to further optimize the training results. Extensive experiments on generated data and real-world datasets are conducted to verify that the proposed edge caching strategies can achieve the minimum latency compared with the state-of-the-art strategies.