Resilient Rights Protection for Sensor Streams

Protecting rights over outsourced digital content becomes essential when considering areas where the data is sensitive and valuable. This chapter addresses the issue of rights protection for sensor streams. It proposes a watermarking solution, based on novel ideas, such as on-the-fly labeling and watermark encoding, resilient to important domain-specific transforms. Streaming data sources represents an important class of emerging applications. These applications produce a virtually endless stream of data that is too large to be stored directly. The method proves to be extremely resilient to all considered transforms, including sampling, summarization, random alterations, and combined transforms. The chapter also analyzes streams of categorical data to investigate other aggregates in the summarization process.