Research Area:  Cloud Computing
Since the cloud users are expanding, their voluminous amount of data, along with their demands, are also growing alarmingly. To tackle these rising demands, Cloud computing has been evolved as cutting-edge technology. Cloud computing felicitates an efficient task scheduling mechanism that can schedule the cloud requests on the most compatible heterogeneous virtual machines. Since the involvement of various scheduling parameters, task scheduling algorithms fall under an NP-hard problem. Therefore, a metaheuristic approach is implemented in this work to mitigate the aforementioned challenges. In this regard, this work proposes a nature-inspired improved variant of the JAYA (IJAYA) optimization algorithm for scheduling tasks on heterogeneous resources. A binary IJAYA is adopted in order to convert the continuous values into discrete values. The CloudSim is used as a simulation tool to conduct the experiments. To assess the effectiveness of the algorithm, a real-world dataset is taken into consideration. The proposed algorithm is compared over Bird Swarm Optimization (BSO), Particle Swarm Optimization (PSO), Salp Swarm Algorithm (SSA), and the standard JAYA for gauging the performance. The simulation results depict notable improvements for makespan, degree of imbalance (DOI), response time and resource utilization over other compared algorithms.
Author(s) Name:  Uttam Kumar Jena; Pradipta Kumar Das; Kaushik Mishra; Manas Ranjan Kabat
Conferrence name:  12th International Conference on Computing Communication and Networking Technologies
Publisher name:  IEEE
Paper Link:   https://ieeexplore.ieee.org/document/9580109