Verifiable Privacy-Preserving Sensor Network Storage for Range Query

We consider a hybrid two-tiered sensor network consisting of regular sensors and special sensors with large storage capacity, called storage nodes. In this structure, regular sensors "push” their raw data to nearby storage nodes and the sink diffuses queries only to storage nodes and "pull” the reply from them. We investigate security and privacy threats when the sensor network is deployed in an untrusted or hostile environment. The major concern is that storage nodes might easily become the target for the adversary to compromise due to their important role. A compromised storage node may leak the data stored there to the adversary breaching the data privacy. Also, it may send wrong information as the reply to a query breaking the data integrity. This paper focuses on range query, a fundamental operation in a sensor network. The solution framework includes a privacy-preserving storage scheme which utilizes a bucketing technique to mix the data in a certain range, and a verifiable query protocol which employs encoding numbers to enable the sink to validate the reply. We further study the performance of event detection, an application implemented by range query. Our simulation results illustrate that our schemes are efficient for communication and effective for privacy and security protection.