A scalable semantic routing architecture for grid resource discovery

Grids technology enables the sharing and collaborating of wide variety of resources. To fully utilize these resources, effective discovery techniques are necessities. However, the complicated and heterogeneous characteristics of the grid resource make sharing and discovering a challenging issue. In this paper, we propose a comprehensive semantic-based resource discovery framework, which performs an effective searching according to the semantic properties of what is searched. In the framework, nodes are grouped into clusters according to some criteria. Resources are indexed and aggregated with a highly compressed format. The summarized index can act as network knowledge to guide routing in the network. Intra-cluster and inter-cluster routing strategies are proposed to support scalable and efficient searching. Results from simulation demonstrate that this architecture is very effective in grid resource discovery.