Prediction-Awareness Edge User Allocating in Edge Based Intelligent Video Systems Driven by Priority

Abstract

Recently, to mitigate tremendous damage caused by various accidents, edge-cutting technologies are utilised to protect lives and properties continuously. Specifically, edge based intelligent video systems have been proved to be an effective tool to monitor and regulate these public security accidents. In these systems, Edge User Allocation (EUA) problem focuses on allocating edge resources to various calculating tasks efficiently, which attracts much attention with multiple approaches proposed. However, in these existing approaches, the priorities of tasks and the varieties of these priorities are not fully considered. Furtherly, these tasks’ priorities are not immutable, which depends on these previous moving persons in the evacuation process. In this regard, we take these concerns into consideration and formulate a Priority-Awareness Edge User Allocation (PA-EUA) problem. Then, we propose our novel prediction-based approaches called UGP and CCGP. Lastly, three series of extensive experiments are conducted on a widely-used real-world data to evaluate our approaches against four representative approaches, and the results show that our novel approaches dominate the performances.