Priority-aware Static Task Mapping for Edge-Cloud Platforms
Jo Yoshimoto, Ittetsu Taniguchi, Hiroyuki Tomiyama, Takao Onoye · 2022
IoT (Internet-of- Things) technologies enable various applications, and these applications take an important role in our society. On the IoT systems, various tasks with priorities are mapped on the edge or the cloud, and these tasks are iteratively executed on them as communicating with each other. To realize such collaborative execution on edge-cloud platforms, effective task mapping is important to consider the task priority and deadline under various situations. Additionally, if the important tasks cannot be executed by keeping the deadline under the given situation, the limited resources should be assigned to the mandatory tasks to minimize the damage. In this paper, we propose the priority-aware static task mapping method for edge-cloud platforms. The proposed method finds the optimal task mapping to minimize the deadline miss and the task discarding under the given communication delay. Simulation is performed under the various communication delays, and the results show the effectiveness of the proposed method.