Data Aggregation Based On Privacy Preserving For Wireless Sensor Networks

Wireless Sensor Network can accomplish data collection, aggregation and transmissionintelligently. In the process of data aggregation, privacy preserving is most important. Dataaggregation algorithm based on privacy preserving can realize the data aggregation withoutdisclosing one node’s privacy data to the other nodes using the technology of privacy preserving.After having researched of Wireless Sensor Network and its characteristics, this dissertationdivides the whole process of data aggregation into two parts: data clustering and the meaningfuldata aggregation based on clustering, raises improved algorithms, makes the theoretical andexperimental analysis, and then designs a prototype system of privacy preserving data aggregation.The work is mainly shown in the following three aspects:(1) This paper raises a privacy preserving clustering algorithm (SMCPCA) based on principalcomponent analysis and secure multi-party computation. The algorithm applies the principalcomponent analysis to reduce the dimensions of data attributes collected from sites, uses thesecure multi-party computation protocol to protect data privacy in the clustering process.Simulation results show that SMCPCA has a smaller computational complexity and error rate.(2) Then this paper proposes an integrity-protecting Homomoriphic Encryption Private DataAggregation (iHEPDA) to solve the problems of privacy and integrity protection. The algorithmintroduces random seeds to disturb the original data, uses the data’s head through homomoriphicencryption algorithm to accomplish integrity detection. By simulation experiments, iHEPDA hascharacteristics of more simple calculation, less communication and better accuracy.(3) On the basis of (1) and (2), the thesis designs a prototype system of privacy preservingdata aggregation. It makes a visual application’s interface which simulates the whole process ofprivacy preserving data aggregation under the application scenarios of environment monitoring inWireless Sensor Network, and displays the processes of privacy preserving clustering,aggregation and integrity detection.In summary, this thesis makes a deep research on the privacy preserving data aggregation inWireless Sensor Network, put forward two appropriate improvement algorithms, and then designsthe prototype system of privacy preserving data aggregation.