Fast computing answer to query streams with wavelet deconstruction

Continuous and real-time data streams exist in large-scale database systems and Internet applications. Fast query and analysis are often desirable in order to efficiently process the streams. In this paper, we first investigate the aggregate computation in Stream queries by analysing the features of and the relation between aggregate query language and aggregate query rewriting. Based on the analysis, we present an efficient fast query computational approach, in which equivalent subset synopsis is designed based on wavelet approach. Finally, by deploying an implementation of the fast query model in a network monitoring and analysis system, we test and verify the efficiency of aggregate computation on stream queries. The result shows that the wavelet-based approach has distinct advantage to traditional method.