DMCA
Comparative Survey on Load Balancing Techniques in Computational Grids
BibTeX
@MISC{Kasthuri_comparativesurvey,
author = {R. Rajeswari Dr. N. Kasthuri},
title = {Comparative Survey on Load Balancing Techniques in Computational Grids},
year = {}
}
OpenURL
Abstract
Abstract — Grid is the system which provides a new, powerful and innovative platform that caters the need of massively computational or data intensive applications from its pool of resources like processors, memory, data, services etc. It differs from traditional computing systems because of its heterogeneous nature and back ground workloads. Performance and utilization of the grid rests on the optimal balancing of load among the available nodes which is very complex and highly dynamic in nature. Finding optimal solution in load balancing for such an environment using the traditional method is an NP-hard problem whereas heuristic approaches will provide near optimal solutions. Algorithms that could capture the dynamic need and complexity have to be developed for solving wide range of load balancing scenarios. Heuristic and artificial life techniques have the power of providing near by solutions from large search spaces since it deals real world scenarios with the capability of handling very large dataset and combinations. In this study, suitability and performance comparison are discussed with various heuristic and agent based techniques. Genetic Algorithm, Tabu Search, Ant Colony Optimization, Particle swarm Optimization are analyzed with their merits, demerits, solutions, issues and improvements towards load balancing in computational grid. Similarity in their nature towards load balancing motivates the attempts in the experimentation to get near optimal solutions from unpredictable information. Performance comparison is analyzed with algorithms like min-max, max-min and Sufferage embedded with Genetic Algorithm and Tabu search. Another heuristic method, Ant Colony Optimization algorithm is suitable for scheduling in grid environment which in tern balances the load. For the same purpose particle swarm optimization algorithm is also adopted. Particle Swarm Optimization is one of the latest evolutionary optimization techniques by nature which has the better ability of global searching leading to minimal makespan time due
Keyphrases
computational grid optimal solution comparative survey load balancing technique particle swarm optimization performance comparison genetic algorithm tabu search purpose particle swarm optimization algorithm real world scenario wide range artificial life technique np-hard problem whereas heuristic approach nature towards load grid environment data intensive application improvement towards load unpredictable information tern balance grid rest makespan time abstract grid ant colony optimization algorithm ant colony optimization large dataset innovative platform large search space ground workload heterogeneous nature optimal balancing dynamic need heuristic method evolutionary optimization technique traditional method available node