Leveraging User Mobility and Mobile App Services Behavior for Optimal Edge Resource Utilization

Edge computing has seen a great progress nowadays eliminating the network latency risks by placing data and functionality in low-end devices closer to the end user is central in application domains such as the support of the backend of mobile apps. When this problem is combined with big data requirements linked to the number of mobile app users, the optimization of edge resource utilization becomes extremely important cost- and quality-wise. This work explores approaches to model resource demand based on user mobility and application characteristics. The Application and User Context Resource Predictor (AUCORP) is introduced that performs performance analysis employing machine learning methods. The evaluation is based on real data from the deployment of a real mobile app for a large music festival.