Reliability-aware and Deadline-constrained workflow scheduling in Mobile Edge Computing

The rapid development of mobile communication technologies and mobile devices have improved our daily lives a lot, and it also provides us a new perspective that the computational task can be solved locally where is closer to data source. Mobile edge computing is such a computing paradigm that considers nearby computing resources are more suitable to handle mobile tasks than remote cloud. However, various challenges, especially the reliability and quality-of-service of such a workflow scheduling over mobile edge computing, are yet to be properly tackled. In this paper, we propose a novel method to evaluate the resource reliability in mobile edge computing environment, then formulate the workflow scheduling into an optimization problem and develop a Krill-based algorithm to solve it. Experiments based on real-world scientific workflow templates and mobile user contract traces are conducted, and the results show that our approach outperform traditional ones significantly in terms of success rate and makespan.