Urbana-Champaign

Data aggregation is an efficient mechanism to collect statistics of data in wireless sensor networks. However, the shared-medium na-ture of communication facilitates eavesdropping and packet tam-pering/injection by adversaries. Hence, wireless sensor networks pose both privacy and security concerns on data aggregation. In this paper, we present SPDA—a secure and privacy-preserving data ag-gregation technique. We utilize disjoint aggregation paths/trees to collect data of interest, thus the base station can verify the integrity of the aggregation result by comparing the result from different paths/trees. To address data privacy, we cloak the privacy-sensitive data by sliding and reassembling technique. We evaluate the SPDA scheme in terms of capacity of privacy-preservation, communica-tion overhead, and data aggregation accuracy. The evaluation is conducted through both theoretical analysis and simulation. 1