A novel improved hybrid optimization algorithm for efficient dynamic medical data scheduling in cloud-based systems for biomedical applications

The fluctuating workloads like cloud requests and the unpredictable resource usage of Virtual machines (VMs) with variable resource characterizations might lead the servers to a non-equilibrium condition. It is thereby causing low resource utilization and the performance degradation of servers. This paper integrates a Genetic Algorithm (GA) and JAYA algorithm to propose a hybrid metaheuristic technique named GAYA for scheduling dynamically independent biomedical data (tasks) to mitigate the above challenges. JAYA is a simple yet powerful population-based parameter-less optimization technique used to surmount the limitations of GA by expediting the convergence rate. In this work, it first uniformly disperses loads (medical data) among VMs through a load balancing strategy, and second, it schedules the tasks (data) among heterogeneous resources by mapping onto the best possible VMs using the GAYA. This algorithm notably meliorated the exploration capability by creating a balance between exploration and exploitation. The efficacy of the proposed approach is evaluated in MATLAB using standard benchmark functions. A real-world dataset consisting of disparate specifications of tasks, like the ones encountered often in biomedical data, has been utilized and simulated in CloudSim to evaluate the effectiveness of the proposed approach. The proposed work has been compared with other metaheuristics and task scheduling techniques such as bird swarm optimization (BSO), GA, JAYA, and Q-learning based modified particle swarm optimization (QMPSO). The Friedman test is conducted to determine the statistical importance of the performance of the algorithms. Simulation results show significant improvement by an increase in resource utilization with 36. 74% (GA), 19.75% (JAYA), 14.31% (QMPSO) and 12.17% (GA), 9.10% (JAYA), 6.02% (QMPSO) and a reduction in makespan by 10.45% (GA), 4.35% (JAYA), 2.31% (QMPSO) and 4.17% (GA), 1.44% (JAYA), 1.03% (QMPSO) in both homogenous and heterogeneous environments respectively.