On Efficient Processing of Multiple KNN Queries in Constrained Spatial Networks

In the real world applications, application severs often receive a lot of KNN requests. To achieve better processing performance, the efficient processing of multiple KNN queries becomes a challenging research issue. This paper studies the multiple KNN queries processing techniques in constrained spatial networks. We propose an efficient cluster-bound-refine algorithm that clusters both the multiple queries and the objects in constrained spatial networks. After the queries and objects in spatial networks being compressed, we could perform the searching of the KNN of the multiple queries in batch. Moreover, we provide a theoretic analysis of when to stop the search for KNN. The main ideas in the paper are to compress the queries and objects in constrained spatial networks, to perform the search operations in block and to refine the primary results based on proved theorem. Experiments on synthetic data sets demonstrate the scalability, effectiveness and efficiency of our methods.