Optimal Resource Selection by using Fuzzy Numbers

Abstract

Grid environment is constructed in order to achieve better utilization of available resources. Selection of appropriate resources for executing the user jobs is one of the most important features of any grid environment because non-optimal resource selection may spoil the ultimate objective of the grid environment. User describes its request for resources using some resource specification language. These specifications lead to a lot of ambiguities in resource description which may lead to selection of undesired resources. The resources with attribute values too higher than the required values can be selected with existing matchmaking mechanisms. This is not optimal and may leads to wastage of resources. In existing matchmaking mechanisms, all the resource attributes are treated equally whether they have great impact on better execution of user job or not. This paper suggests an intelligent matchmaking mechanism which uses fuzzy logic to select most appropriate resources for user jobs according to user specified criteria. It also helps to select most reliable resources as well.