The paper introduces "Mobi-IoST", a collaborative framework for real-time edge/fog computing in IoT systems with a back-end cloud. It leverages the mobility patterns of IoT and edge devices in a 2-D space, using GPS logs and contextual data to predict their real-time locations through machine learning. The framework’s features include a hierarchical IoT-Edge-Fog-Cloud architecture for improved real-time QoS, mobility-based agent location prediction, and effective handling of delay and power consumption, resulting in around 93% prediction accuracy and 23–26% delay reduction and 37–41% power reduction compared to existing systems.
Shreya Ghosh, Anwesha Mukherjee, Soumya K. Ghosh, Rajkumar Buyya