Achieving Private, Scalable, and Precise Data Collection in Wireless Sensor Networks

Wireless Sensor Networks (WSN) become increasingly popular to collect data over a large area. Given the collected data set, the network manager can extract various kinds of aggregate statistics from the set to characterize the physical space. On the collection of the data, three requirements should be imposed: (1) Privacy: as sensor nodes are source limited and often deployed in an open environment, the sensed data suffer from privacy vulnerabilities. Secure mechanism should be provided to protect data privacy, (2) Communication efficiency: collecting data from large-scale sensor networks often involves large-volume data generation and transmission, which may quickly consume the energy of the WSN. To prolong the lifetimes of the sensor nodes, the sensed data should be transmitted in lightweight manner, (3) Accuracy: the sensed data should be recovered accurately at the base station (BS) so that the manager can manipulate them freely to achieve any precise aggregate statistic he prefers. To satisfy these requirements, we propose two novel privacy-preserving data collection schemes based on compressive sensing techniques. Our schemes address the privacy, communication efficiency and accuracy issues simultaneously. Detailed theoretical analysis and simulation results confirm the high performance of the proposed schemes.