Research And Implementation Of Distributed Data Management In Wireless Sensor Networks

With the developments of wireless sensor networks and Internet of Things, skyline queries,as a part of distributed data management, has drawn tremendous attention form academic com-munity. Skyline queries were designed to help users making decisions and normally applied intomultidimensional data process. However, since the traditional skyline queries were used in acentralized network environment, they are difficult to be adopted into a wireless sensor networkcase.The purpose of this thesis is offering a possible strategy to apply skyline queries into a wire-less sensor network system. Consequently, skyline query processing in wireless sensor networksbased clustering (CLSKY) was developed in this thesis to resolve the low efficiency issue fromdata flooding. CLSKY started with a clustering architecture and followed by a pruning strategy.The clustering architecture was used to separate partitions sensor readings to several clusters andto identify the sensor node in skyline points. The skyline points might dominate many datapoints in the early stage of the skyline query processing and also can be used to discard unquali-fied data points for further examination. The pruning strategy is designed to help the amount ofquery forwarding controlling, limiting the number of sensor nodes involved in query processing.The simulation results suggested that CLSKY significantly improved the query efficiency ratherthan other traditional methods.