Mobile Edge Computing Network Control: Tradeoff between Delay and Cost

As mobile edge computing (MEC) finds widespread use for relieving the computational burden of compute- and interaction-intensive applications on end user devices, understanding the resulting delay and cost performance is drawing significant attention. While most existing works focus on single-task offloading in single-hop MEC networks, next generation applications (e.g., industrial automation, augmented/virtual reality) require advance models and algorithms for dynamic configuration of multi-task services over multi-hop MEC networks. In this work, we leverage recent advances in dynamic cloud network control to provide a comprehensive study of the performance of multi-hop MEC networks, addressing the key problems of multi-task offloading, timely packet scheduling, and joint computation and communication resource allocation. We present a fully distributed algorithm based on Lyapunov control theory that achieves throughput-optimal performance with delay and cost guarantees. Simulation results validate our theoretical analysis and provide insightful guidelines on the interplay between communication and computation resources in MEC networks.

I. Introduction

Resource- and interaction-intensive applications such as real-time computer vision and augmented reality will increasingly dominate our daily lives [1]. Due to the limited computation capabilities and restricted energy supply of end user equipments (UEs), many resource-demanding tasks that cannot be executed locally end up being offloaded to centralized cloud data centers. However, the additional delays incurred in routing data streams from UEs to distant clouds significantly degrade the performance of real-time interactive applications. To address this challenge, mobile edge computing (MEC) emerges as an attractive alternative by bringing computation resources to edge servers deployed close to the end users (e.g., at base stations), striking a good balance between cost efficiency and low latency access.