The rise of the future grid (FG) largely depends on the efficient integration of Internet of Things (IoT) and Cloud computing technologies. By utilizing information and control flows, FG can deliver power more effectively and be capable to handle events occurring anywhere in the grid network. However, maintaining such functions consumes a great deal of computational resource which brings an enormous operational cost to the grid owner. In this paper, we propose an integrated task scheduling and resource provisioning model for dynamically operating an IoT-Cloud system to reduce the overall operational cost. Our proposed approach uses a bipartite graph to model the communication pattern between sensor groups and decentralized cloud data centers and a Pareto distribution-based method to estimate the required resources considering capacity limitation and failure of the system in each data center. We formulate the integrated model as a constraint optimization problem over all sensor groups and data centers. We solve the problem with genetic algorithms due to problem complexity, and our extensive computer simulations and comparisons demonstrate the correctness and effectiveness of the proposed model in minimizing operational cost while satisfying system performance requirements.
|Journal||Journal of Energy Engineering|
|Early online date||15 Jul 2019|
|Publication status||Published - 1 Oct 2019|