Resource Scheduling for Delay Minimization in Multi-Server Cellular Edge Computing Systems

This paper studies resource scheduling for delay minimization in multi-server cellular edge computing systems. The traditional method defines queue length-based Lyapunov functions and designs scheduling algorithms which solve the corresponding queue length control problem. Different from the traditional method, this paper defines the delay-based Lyapunov function. Specifically, the formulas of the communication delay and computing delay in cellular edge computing systems are derived without needing the assumptions about traffic's statistics. Then, a resource scheduling algorithm which directly minimizes the weighted sum of the communication delay and computing delay is proposed. The simulation results show that the total delay of the proposed scheduling algorithm is decreased as compared to that of the traditional method. The impact of parameters on the delay performance is also evaluated.