Privacy-homomorphism Based Privacy-preserving Data Aggregation Algorithm In WSNs

Data aggregation is a very important technique, which is designed to substantially reduce the communication overhead and energy expenditure of sensor node during the process of data collection in a Wireless Sensor Networks(WSNs). However, privacy-preservation is more challenging especially in data aggregation, where the aggregators need to perform some aggregation operations on sensing data it received. Since it allows direct computation on encrypted data, which means not required to decrypt the incoming ciphers before aggregating, the privacy homomorphism(PH) techniques have been wildly used to protect data privacy for data aggregation.Firstly, this thesis gives a state-of-the art survey of current development of PH-based privacy preserving aggregation methods from kinds of encryption, key distribution and algorithm performance. Also, comparison and inductive results are given based on that.Based on the analysis of current data aggregation privacy preserving scheme, this thesis present two privacy-preserving data aggregation schemes called Additive-Homomorphic based Private Data Aggregation(AHPDA) and Recoverable concealed Private Data Aggregation(HRCDA) respectively. Both AHPDA and HRCDA adopt privacy homomorphism technique, and they can preserve the privacy of a sensor data and check data integrity for data aggregation in WSNs. The AHPDA algorithm hides the sensing data through combining it with each node’s ID, the aggregator can simply add the receiving data and then transmit it to other node. In HRCDA algorithm, it reduces the communication overhead of the system by improving the ID of the transmission mechanism. Secondly, it adopts homomorphic Hash function to prevent replay attack. Finally, a base station can recover each sensing data generated by all sensors even if these data have been aggregated by aggregators.The performance analyses and simulation results denote that AHPDA and HRCDA can effectively preserve data confidentiality, check data integrity, satisfy data timeliness, and bring low communication overhead.