Online Scheduling Strategy to Minimize Penalty of Tardiness for Real-Time Tasks in Mobile Edge Computing Systems

With the coming of big data era and 5G era, many tasks have higher and higher requirement for the latency and data, traditional cloud computing paradigm is gradually unable to handle such real-time scenarios with large number of tasks, tremendous data volume and high latency requirements. In order to solve these problems, mobile edge computing has become the focus of attention. However, although low latency is a major feature of mobile edge computing, some real-time tasks still cannot be completed on time due to the limitation of network and computing resources, which will affect the quality of experience (QoE) of users and lead the loss to the economy and reputation of enterprises.To reduce the penalty of tardiness of tasks, in this paper, we consider the mobile edge computing system as a soft real-time system and discuss how to assign the task to a server and how to schedule the task on the server. Considering the feature of real-time tasks, we have a formal description of this problem and discuss the online version of this problem. We propose a heuristic algorithm based on the urgency of deadline of tasks to reduce the loss caused by task timeout. Experiments show that our algorithm has better performance than the classic real-time task scheduling algorithm.