An intelligent/cognitive model of task scheduling for IoT applications in cloud computing environment

In a cloud computing environment, it is not easy to schedule various Internet of Things (loT) application tasks due to the heterogeneity characterises of IoT. Efficient scheduling and load balancing of IoT applications is important to minimize the total execution time(makespan) while adhering to constraints like task dependencies. In this paper, a cognitive or intelligent model of bio-inspired approach is used to find the optimal solution of task scheduling for IoT applications in a heterogeneous multiprocessor cloud environment. Natural selection of genes and evolutionary foraging traits has proved that only the fittest species survive in nature. In this case, a fit schedule is considered as one which is efficient and follows the task ordering in the multiprocessor environment. A hybrid algorithm GAACO combining Genetic Algorithm (GA) and Ant Colony Optimization (ACO) has been used to select only the best combination of tasks at each stage. This unique combination of GA and ACO used ensures the appropriate convergence and optimality when GAACO is developed. Scheduling using GAACO is not pre-emptive and it is assumed that one task can be assigned to one processor. When tested on various sizes of task graphs and different number of processors, GAACO has proved to be competent with the conventional approaches of using GA and ACO in the heterogeneous multiprocessor environment. (C) 2018 Elsevier B.V. All rights reserved.