Joint Optimization of Service Placement and Computation Offloading for Mobile Edge Computing

Abstract

Mobile Edge Computing (MEC) is emerging as a promising approach for enhancing the quality-of-service (QoS) of delay-sensitive applications in the B5G/6G era, via offloading certain computation tasks to the network edge that approximates to end-users. Existing researches on computation offloading in MEC mainly focused on the hardware resource constraint (e.g., CPU and storage) at the edge nodes, without considering the specific software service requirements of applications (e.g., runtime environment and operating system). In this work, we study the computation offloading for delay-sensitive applications under both constraints of hardware resources and software services, where each application can be offloaded to an edge node only if both the required hardware resources and software services have been deployed at that node. We formulate a Joint Service Placement and Computation Offloading (JSPCO) problem, aiming at minimizing the offloading delay cost and the service operation cost. The problem is challenging due to the inherent coupling between service placement and computation offloading. To solve the problem, we introduce several equivalent transformation methods that convert the original problem into a Mixed Integer Linear Programming (MILP) problem, which can be solved efficiently using classic algorithms. Simulation results show that our proposed joint optimization solution can reduce the total system cost, service operation cost, and UE delay cost by up to 63.06%, 62.90%, and 54.76%, respectively, compared to existing baseline solutions.