21-08-2013, 02:49 PM
Optimal Job Scheduling using Ant Colony Algorithm in Grid Computing
Optimal Job Scheduling.pptx (Size: 351.72 KB / Downloads: 54)
ABSTRACT
Grid computing is a combination of shared reliable and non reliable resources
In grid the job scheduling is performed by various optimization to control the programs execution
It assigns the currently running process efficiently completes the job in a short period.
The scheduling of jobs is efficiently proposed by an ant colony algorithm for allocating optimal resources to each job in a minimal makespan at run time.
OBJECTIVE
To find an exact resource allocation by choosing shortest and optimal path for a required specific job.
With minimized schedule length of jobs with minimum makespan at run time.
ISSUES IN EXISTING SYSTEM
Machines can require a certain gap between jobs
Jobs and machines have mutual constraints as jobs can be scheduled on some machines only
Set of jobs can relate to different set of machines
MOTIVATION AND PROBLEM STATEMENT
Motivation: Due to increase in number of jobs requested the jobs may have constraints, a job needs to finish before another job starts which may lead to maximum lateness. Hence an exact shortest path is evaluated to schedule the job
Problem Statement: To manage the jobs efficiently which is required for functioning of job scheduling
PROPOSED SYSTEM
The system find exact resource allocation by choosing shortest and optimal path for a required specific job
A scheduler selects the best path
Schedule length of jobs are minimized with
better quality of service
minimum make span at run time
Input: Job requests are arrived to allocate resources for particular job
Output: Scheduled jobs are completed with minimum makespan
CONCLUSION
The proposed system can be optimized using ACO to allocate resources for specific assigned jobs.
The advantages of this proposed system is
Can be used in dynamic applications (adapts to changes such as new distances, etc.)
Has been applied to a wide variety of applications
As with GAs, good choice for constrained discrete problems