Dynamic service deployment for budget-constrained mobile edge computing

摘要

Currently, Mobile edge computing (MEC) is facing a great challenge that is how to make fulluse of edge resources to provide a seamless support for compute-intensive latency-sensitiveapplications. Prior studies often make a simple assumption that tasks can be executed uponevery edge server, but the assumption does not hold in practical scenarios. Because a specificapplication task often corresponds to a certain service that provides the corresponding runningenvironment, whereas an edge server only has limited resources and cannot offer too manyservices. How to decide service deployment of so many types of services among multiple edgeservers is also a big challenge. To address the challenge, we study dynamic service deploymentfor latency-sensitive applications. We first model the long-term budget-constrained latencyminimization problem as a multi-slot latency minimization problem based on the Lyapunovframework. By doing this, the hardness of a problem is significantly reduced, since we neverrequire future information to solve the long-term optimization. Furthermore, we extend ourstudy by joining the task scheduling optimization, where every edge server is fully utilized in aneven more efficient collaborative manner. Our extensive experiments show that the proposedalgorithms can bring short latency with low cost.