Edge Selection Strategies for Human-enabled Sensing Architectures

This thesis focuses on synergies between Mobile CrowdSensing (MCS) and Multi-access Edge Computing (MEC). Specifically, the thesis proposes a social mobile edge architecture composed by interoperable fixed (FMEC) and mobile (M2EC) units in the MEC middleware layer. After introducing a method of task assignment based on MCS mobile resources, the thesis proposes some optimization methods to share contents between users leveraging their mobility and sociability. The central part of the thesis presents an algorithm for the selection of M2ECs to be used in synergy and/or in places of the fixed ones. The M2EC social-aware selection algorithm is performed by leveraging the links between users and the fact that they tend to form cohesive communities. The thesis proposes two selection criteria: the first one is based on the users’ sociability, while the second criteria exploits the attitude of the users in performing the assigned tasks. The last part of the thesis introduces a probabilistic model for the estimation of the optimal number of M2ECs to be selected in order to achieve a specific coverage. The thesis also includes a comprehensive evaluation of the performance that can be obtained by combining FMECs and M2ECs. The solutions proposed have been tested with a real-world MCS dataset which provides meaningful mobility traces of students in urban areas for over one year of data collection.