Energy-aware and Deadline-constrained Task Scheduling in Fog Computing Systems

We investigate a deadline-constrained task scheduling problem in the fog computing environments where tasks can be offloaded to heterogeneous resources. Three kinds of resources are involved: mobile device, fog device and cloud server. The objective is to schedule all the tasks with minimum energy consumption. We develop an energy-aware strategy and propose a critical path based iterative algorithm which can obtain the optimal solution in polynomial time complexity. We also discuss the cases when no feasible solution exists. Experimental results show that the proposal is robust and effective for the problems under study.