Aggregation-tolerant steganography for data-stream integrity verification

Techset Com Data outsourcing can make data-integrity protection a challenging task, especially when the trustworthiness of a third-party is unproven. A novel auditing process for integrity verification of data stream, whose storage and handling is outsourced to a third-party, is explored. For this purpose, the hidden information within this data that provides support for in-network data aggregation environments, such as sensor networks is masked. This mask is achieved by simultaneous embedding of several shifted watermark patterns into multiple data streams. The nature of this watermark allows it to be aggregated into a single data stream with minimal loss of this information. A great benefit of the proposed scheme is that the embedded watermarks are invariant to linear time-sequential or cross-stream aggregation operations, such as summation or averaging. Therefore, multiple data streams can be merged and at the same time, and the existence of each individual watermark within allowable bounds can still be verified. The simulation results show that the embedded watermarks can successfully be recovered with high confidence if proper hiding codes are chosen.