Secure Provenance Transmission for Streaming Data

On the Properties of Non-Media Digital Watermarking: A Review of State of the Art Techniques

“…Summarizing, the invisibility definition requires a good understanding of the data model features as well as the target application. Taking example of provenance [86] where the information is embedded within time differences between consecutive packets, the watermark said to be invisible if it does not introduce large delays in the network. For marking graph structures such as social network data whereby a few vertices or edges are added to the original graph according to watermark values, the watermark is invisible if the number of modified vertices and edges are small and therefore the main properties of the graph such as average path length or its diameter are preserved [58].…”

“…The authors argued that blind watermarking schemes can be used in sensor networks for answering the question "Where did the data come from" whereas non-blind techniques can answer the question of "Is this data mine?". Later on, Sultana et al [86] considered the problem of secure provenance transmission in a data streaming fashion. The presented method constructs an audit trail of data by embedding provenance information within inter-packet delays of consecutive packets from a sensor node over the routing tree.…”

“…6) Sultana et al [86] proposed a SS watermarking technique for securely transmitting provenance information in sensor networks, whereby watermarks (provenance information) are spread over the inter-packet delays rather than sensory data itself. In other words, the inter-packet delays (IPDs) are dithered according to watermark values.…”

“…As a result, other places for watermarking should be investigated. For example, in [86], the authors embedded provenance information in timing differences between consecutive packets in a sensor network. While this might betray the watermarking definition as the embedding procedure involves change of timing information that is part of the header not within the data itself, we argue that this could also be considered as a watermarking scheme if one is collecting headers and may not care about the data.…”

Current and Emerging Trends in Cyber Operations

“…Also, it is acknowledged that data trustworthiness is an issue greater than security, and the proposed data-exchange infrastructure will allow solutions combining different approaches and techniques including data integrity, data quality, record linkage (Inan et al 2012), and data provenance (Sultana, Shehab and Bertino 2013). For example, the proposed cyber-security data-exchange infrastructure should allow integration of tools that enable the refinement and vetting of data.…”