DOI: 10.4018/IJGHPC.2018010101
International Journal of Grid and High Performance Computing
Volume 10 • Issue 1 • January-March 2018
Copyright © 2018, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.
1
Resource Allocation in Grid
Computing Environment Using
Genetic–Auction Based Algorithm
Kuppani Satish, Department of Computer Science and Engineering, Sri Venkateswara University, Tirupati, India
A. Rama Mohan Reddy, Department of Computer Science and Engineering, Sri Venkateswara University, Tirupati, India
ABSTRACT
The main core functionality of Grid Computing is resource allocation and scheduling. With the
idea of genetic algorithms and microeconomics, it is proposed a Resource allocation method called
a genetic-auction based algorithm [GAAB]. This algorithm contains two modules, auction module
and genetic approach. Auction module find outs resource-trading price between resource provider
and resource buyer, and the resource allocation carried out by Genetic algorithm by considering
both time and cost constraints simultaneously. In this article, evaluations are made in the simulation
environment and the results show the effectiveness of the proposed model.
KEywoRdS
Auction, Cost, Genetic Algorithm, Resource Pricing, Time
1. INTRodUCTIoN
Grid computing is one of the emerging fields in parallel and distributed computing. Grids are evolving
as the compromising future generation of computational platforms for executing large-scale resource
concentrated applications arising in the field of science, engineering, and commerce (Foster and
Kesselman, 1999; Abraham et al., 2000; Xhafa et al., 2007). It supports the heterogeneous resources by
the creation of virtual organisations and enterprises. These virtual organisations enable the selection,
sharing, aggregation and exchange of information between heterogeneous resources. The customer
can access grid resources by maintaining a grid portal like Globus (Foster and Kesselman, 1997)
and Legion (Grimshaw and Wulf, 1997) and each resource owner share their resources by running a
grid portal too. Different customers’ demands different resources, different resources have different
capabilities and availabilities based on their policies, which are selected. At any point of time, the
resources may enter or exit from the grid. On the other side, customers with varying resources (Babu
and Krishna, 2014) can enter in to the grid. As a result, the environment of the grid is highly dynamic,
uncontrollable, and heterogeneous across different domains.
The traditional methods cannot be applied simply to the grid resource management because
they adopt total control over requests and resources. Meanwhile the grid resources are belonging to
different domains and are distributed in different geographical regions so decentralized method is
an appropriate solution for resource management in grid. A suitable resource management for grid
exploits the resources capability effectively and satisfies the customer requests. In the past few years,
the growth of market based resource management takes place.