Energy-Efficient Resource Allocation for Slicing-enabled Multi-access Edge Computing

Abstract

Multi-access edge computing (MEC) provides a distributed computing environment and cloud-like services at the network edge, aiming to relieve high network latency and enormous backbone network pressure. However, with the diversification of mobile devices and increased demand for mobile services, the shortage of edge computing resources has become a new challenge. In this context, researching more efficient resource management methods has become the key to addressing this challenge. This paper investigates the energy-efficient computing resource allocation problem in a slicing-enabled MEC system. First, we establish mathematical models to characterize the MEC system and characterize each edge server (ES) slice as an M/M/m queueing model, such that the average power consumption (APC) of the ES and the average response time (ART) of different tasks can be calculated analytically. Then, we design numerical methods to obtain the optimal resource allocation scheme for each ES slice to reduce the ES's APC while ensuring that the ART meets the latency constraints of different tasks. Finally, we provide several numerical experiments to show the effectiveness of the proposed methods.