A agent-based decentralized algorithm for resource semantic discovery in economic grid

Abstract

Grid computing offers geographically dispatched heterogeneous resources in order to perform complex computations. Besides, issues such as sharing, allocation and integration of resources as well as resource discovery are of high importance in grid computing systems. In regards to resource discovery in such systems with such characteristics, various methods have been presented. Among them are methods which use semantic description to display resources. In economic grid computing systems, in order to use shared resources for task implementation, some costs must be paid. In these systems, every user determines a budget and asks for his/her desired task. Aiming at responding to the task, a budgeting algorithm should perform the practice of allocating the grid resources to heterogeneous tasks in a way that the tasks are finalized with the same or less than the pre-determine budget, given their costs and capabilities. This paper proposes a new method for resource discovery based on semantic description in an economic grid.