Toward Adaptive Energy Management for Mobile Edge Networks

Abstract

Mobile edge computing (MEC) is one of the promising solutions for 5G networks, which provides computing resources on network edges for end users. In these networks, MEC servers are densely deployed to meet the requirements of computation-intensive tasks. Besides, the traffic distribution in the MEC network is heterogeneous due to the spatial and temporal dynamics. It is already known that idle power consumption of MEC servers takes up a large portion of the total network energy consumption. Thus, we address a sleep control problem to optimize the energy consumption in a dense MEC network. First, we formulate the energy optimization problem under delay constraint. Then, the problem is addressed by using the lateral induction and inhibition mechanism which is one of the bio-inspired methods. We propose a sleep control method through the delta-notch signaling among neighboring MEC servers. The experimental results show that the proposed algorithm can reduce energy consumption effectively under delay constraints.