Ontology based resource usage policy matching in computational grid for pervasive computing applications

The dynamic integration of autonomous, clustered and heterogeneous networked components are dynamically integrated for effective resource utilization of computing resources. This paper proposes a mechanism for expressing resource usage policies in computational grid for heterogeneous device with varying resource request. It extends the WS-Policy specification to represent grid resource policy requirements and implements APIs to compare and negotiate the usage policies. It also enhances the usage policy matching by using the concept of ontology. Grid resource usage policy ontology is proposed to support semantic based matching of resource usage policy against the user's requirements. This semantic based grid resource usage policy mechanism is then integrated with Gridway meta-scheduler to implement the resource selection based on the capability matching as well as usage policy matching in the grid environment.