U-MEC: Energy-Efficient Mobile Edge Computing for IoT Applications in Ultra Dense Networks

With the development of the Internet of Things (IoT) technologies, many kinds of IoT devices as well as many colorful IoT applications are emerging and absorb great attention. In this paper, we propose U-MEC, a Mobile Edge Computing framework deployed in ultra dense networks, in order to narrow the resource gap between the constantly increasing demand of IoT applications and the restricted supply of IoT devices. To improve the energy efficiency of this framework, we present a comprehensive model and formulate a mixed integer programming problem to capture task offloading, user association and base station switching. We propose an online scheduling algorithm, which exploits current system information only by invoking Lyapunov optimization, Lagrange multiplier, and sub-gradient techniques. The simulation results show that our framework achieves a better performance in energy consumption compared with several benchmark schemes.