| Summary: | 碩士 === 國立成功大學 === 資訊工程學系碩博士班 === 95 === Continuous k-nearest neighbor (CKNN) search for moving objects is an important function in the applications of spatial databases and location based services (LBS). Many query processing approaches have been proposed in tradition environments. However, the CKNN query is still not addressed efficiently in wireless sensor environments. The challenges of addressing this issue in wireless sensor networks mainly come from fact that communication among sensors for continuous gathering and reporting object locations is highly expensive. In this paper, we propose a tracking-based distributed monitoring (TDM) approach for efficiently processing CKNN queries in wireless sensor networks. TDM establishes an answer space to determine the current KNN and monitor their movement continuously, so that the query answer at each time unit can be obtained with few communications. In addition, we also consider that different applications require different order precision of KNN, thus we extend TDM to address CKNN queries with various order precision. Finally, we conduct a comprehensive set of experiments to reveal that the proposed approach outperforms other existing methods.
|
|---|
