Joint Computation Offloading and Prioritized Scheduling in Mobile Edge Computing

With the rapid development of smart phones and Internet of Things (IoT) devices, enormous amounts of data have been generated. These data usually require real-time, intensive computation. Yet, Meeting the required Quality of Service (QoS) remains a challenge due to the tension between resource-limited devices and computation-intensive applications. Mobile-edge computing (MEC) emerges as a promising technique to cope with the stringent requirements of mobile applications. This paper proposes a joint computation offloading and prioritized task scheduling scheme in a MEC system. We investigate an energy-minimizing task offloading strategy in mobile devices, and also develop an effective dynamic priority-based task scheduling algorithm at the edge server. The execution time, execution cost, and bonus score are adopted as performance metrics. We also defined a new performance metric, bonus score, which measures the benefit of finishing a task before its latency requirement deadline, and is a function of its completion time, priority level, and task size. Performance evaluation results show that the proposed algorithm significantly reduces both task completion time and edge server VM usage cost, and improves QoS in terms of bonus score. Moreover, dynamic prioritized task scheduling is also developed, with results showing that dynamic threshold setting further improves the performance. We believe that this work is significant to the emerging MEC paradigm, and can be applied to other IoT-edge applications in 5G systems.