Dynamic System Activity Profile Forecasting for Improved Resource Selection in Quality of Service based Grid Computing model

Quality of Service based grid computing model enables the resource seekers to place demands of deadlines, budget, security, etc. In such a scenario there exists a very real possibility of numerous resource providers meeting the demanded resource criteria on static parameters like number of processors, RAM storage, etc. This thus turns out to be a problem of "selecting" the best resource provider. In this paper we debate as to why static parameters are inadequate when it comes to selection of the best resource. It has been demonstrated as to how the producer’s local activity profile can be monitored dynamically over time to generate information that can be used to forecast the suitability of a producer for the day, week, month or year during which the resource is being sought for. Deeming it as a platform for future work we propose ideas on a framework in which rating of resources could he done by combining all the static and dynamic parameters of the resource in question to enable improved and quicker resource selection for better quality of service to consumers.