Vol. 10 Issue 2
Year:2015
Issue:Oct-Dec
Title: An Optimal ANT Algorithm for Balanced Scheduling in Computing Grids
Author Name:K. Jairam Naik,A. Jagan and N. Satya Narayana
Synopsis:
Grid computing relies on distributed heterogeneous resources to support convoluted computing problems. Grids are mainly classified into computing grid and data grid. Scheduling the jobs in computing grid is a very big obstacle. For efficient and incentive based use of grid resources, optimal mechanism in grids is needed that can distribute jobs to the prime resources and balance work load among them. In the real world, the ants have an unique ability to team up for finding an optimal path for food resources. The behavior of real ants was simulated by an Ant algorithm. In this paper, the authors proposed an optimized Ant algorithm for balanced job scheduling in the computing grid environment. The primary goal of the approach is to schedule and balance the entire system load, so that no resource is overloaded or under loaded under any circumstances. To achieve the primary goal, a Novel Balanced Ant Colony Optimization (NBACO) always calculates and updates local and global pheromone values, which makes the resource always normally loaded. The secondary aim is to considerably improve grid performance in terms of increased throughput, reduced makespan and resubmission time. To achieve this goal, NBACO always finds the most successive available resources in the grid and assigns the job to it. The successive resource is the one, which has the lower tendency to fail. According to the experimental results, NBACO can outperform other job scheduling and load balancing algorithms.
Year:2015
Issue:Oct-Dec
Title: An Optimal ANT Algorithm for Balanced Scheduling in Computing Grids
Author Name:K. Jairam Naik,A. Jagan and N. Satya Narayana
Synopsis:
Grid computing relies on distributed heterogeneous resources to support convoluted computing problems. Grids are mainly classified into computing grid and data grid. Scheduling the jobs in computing grid is a very big obstacle. For efficient and incentive based use of grid resources, optimal mechanism in grids is needed that can distribute jobs to the prime resources and balance work load among them. In the real world, the ants have an unique ability to team up for finding an optimal path for food resources. The behavior of real ants was simulated by an Ant algorithm. In this paper, the authors proposed an optimized Ant algorithm for balanced job scheduling in the computing grid environment. The primary goal of the approach is to schedule and balance the entire system load, so that no resource is overloaded or under loaded under any circumstances. To achieve the primary goal, a Novel Balanced Ant Colony Optimization (NBACO) always calculates and updates local and global pheromone values, which makes the resource always normally loaded. The secondary aim is to considerably improve grid performance in terms of increased throughput, reduced makespan and resubmission time. To achieve this goal, NBACO always finds the most successive available resources in the grid and assigns the job to it. The successive resource is the one, which has the lower tendency to fail. According to the experimental results, NBACO can outperform other job scheduling and load balancing algorithms.
No comments:
Post a Comment