Task Allocation in Three Tier Fog IoT Architecture for Patient Monitoring System using Stackelberg Game and Matching Algorithm

Internet of Things (IoT) based personalized health monitoring system typically generates large amount of critical data that need to be analyzed and acted upon in real-time. Fog computing architecture has been proposed to handle this class of low-latency applications. Task allocation and scheduling on the fog nodes while satisfying stringent time constraint is a difficult problem that has been studied only in simple cases [1] [2]. In this paper, we propose a distributed optimization model for task allocation to fog nodes which can handle periodic as well as sporadic task allocation that maximize the utility of all the nodes participating in task execution process while satisfying delay constraint from the tasks. We have extended the heuristic based on Stackelberg game and matching algorithm [3] to handle sporadic and heterogeneous tasks as well. Proposed algorithm has been simulated with realistic health monitoring application parameters. Results show significant improvement over the existing algorithm.