Optimal K-Nearest-Neighbor Query in Data Grid

The paper proposes an optimal distributed k Nearest Neighbor query processing algorithm based on Data Grid, called the opGk NN. Three steps are incorporated in the opGk NN. First when a user submits a query with a vector Vq and a number k, an iDistance[3]-based vector set reduction is first conducted at data node level in parallel. Then the candidate vectors are transferred to the executing nodes for the refinement process in which the answer set is obtained. Finally, the answer set is transferred to the query node. The experimental results show that the performance of the algorithm is efficient and effective in minimizing the response time by decreasing network transfer cost and increasing the parallelism of I/O and CPU.