A greedy randomized adaptive search procedure for scheduling IoT tasks in virtualized fog-cloud computing

Abstract Virtualized fog-cloud computing (VFCC) has become an ideal platform for processing the growing number of emerging Internet of Things (IoT) applications. VFCC resources are provided to IoT applications as a set of virtual machines (VMs). How VMs may be used efficiently for performing IoT tasks with various requirements is a primary challenge considering that they are heterogeneous in terms of processing power, communication delay, and energy consumption. In order to deal with this problem, we initially present a system model for scheduling IoT tasks in VFCCs, that considers not only each task's deadline but also the system's energy consumption. Then, a greedy randomized adaptive search process (GRASP) is utilized to determine the optimal assignment of IoT tasks among VMs. GRASP is a metaheuristic-based technique that provides several appealing characteristics, such as simplicity and ease of implementation, a limited number of tuning parameters, and the capability of parallel implementation. We conducted comprehensive experiments to evaluate the efficacy of the proposed method and compared its performance to that of the most advanced algorithms. Extensive experiments show that the proposed technique is superior to the baseline method in terms of deadline satisfaction ratio, average response time, energy consumption, and makespan.