In the Shadows We Trust: A Secure Aggregation Tolerant Watermark for Data Streams

In many applications such as sensor networks, e-healthcare and environmental monitoring, data is continuously streamed and combined from multiple resources in order to make decisions based on the aggregated data streams. One major concern in these applications is assuring high trustworthiness of the aggregated data stream for correct decision-making. For example, an adversary may compromise a few data-sources and introduce false data into the aggregated data-stream and cause catastrophic consequences. In this work, we propose a novel method for verifying data integrity by embedding several signature codes within data streams known as digital watermarking. Therefore, the integrity of the data streams can be verified by decoding the embedded signatures even as the data go through multiple stages of aggregation process. Although the idea of secure data aggregation based on digital watermarking has been explored before, we aim to improve the efficiency of the scheme by examining several signature codes that could also decrease the watermark detection complexity. This is achieved by simultaneous embedding of several shifted watermark patterns into aggregated data stream, such that the contribution of each data-source is hidden in the relative shifts of the patterns. We, also, derive conditions to preserve the main statistical properties of data-streams prior to the embedding procedure. Therefore, we can guarantee that the embedding procedure does not compromise the usability of data streams for any operations that depends on these statistical characteristics. The simulation results show that the embedded watermarks can successfully be recovered with high confidence if proper hiding codes are chosen.