HYBRID ACO-IWD OPTIMIZATION ALGORITHM FOR MINIMIZING WEIGHTED FLOWTIME IN CLOUD-BASED PARAMETER SWEEP EXPERIMENTS

R. Angel Preethima, Margret Johnson

Abstract: Scientists and engineers conduct several experiments by executing the same coding against the various input data, which is achieved by the Parameter Sweep Experiments (PSEs). This may finally results in too many jobs with high computational requirements. Therefore the distributed environments, particularly clouds, are used in-order to fulfill these demands. Since it is an NP-complete problem the job scheduling is much changeling. Now the proposed work is determined by the Cloud scheduler based on the bioinspired techniques, since it works well in approximating problems with little input. But in existing proposals the job priority is ignored; which in turn it is the important aspect in PSEs because it accelerates the result of the PSE and visualization of scientific clouds. The weighted flow time is minimized with the help of the cloud scheduler based on Ant Colony Optimization (ACO). All matching recourses of the job requirements and the routing information are defined by the Intelligent Water Drops (IWDs) in order to reach the recourses. Among all matching resources of the job the Ant colony optimization is determined as the best resources. The main aim of this approach is to converge to the optimal scheduler faster, minimize the make span of the job, improve load balancing.

Keywords: Ant Colony Optimization, Intelligent Water Drops, Parameter Sweep Experiments, Weighted Flowtime

DOI: https://doi.org/10.15623/ijret.2014.0303058