| In this thesis, we look closely at two fundamental problems that arise within the context of multimedia blind watermark decoding and timing channels steganalysis. The central problem considered, loosely speaking, is that of implementing optimal (or near-optimal) strategies at the receiver, which is typically tasked to perform reliable decoding or detection, depending on the application at hand, in the presence of numerous unavoidable statistical uncertainties that are rather unique to the problem setup. A typical question we will be asking is, "Can we perform reliable decoding of hidden data in spite of the presence of unknown channel parameters?" or "How best can we detect presence of hidden data with unknown, and rather arbitrary, host and observation statistics?" While such questions are naturally relevant from a practical viewpoint, we draw additional inspiration for our study from profound theoretical insights arising from our recent research.;As our solution to the first problem, we propose a new paradigm for blind watermark decoding in the presence of various signal distortion operations. Employing Forney-style factor graphs to model the watermarking system, we cast the blind watermark decoding problem as a probabilistic inference problem on a graph, and solve it via messagepassing. We study a wide range of moderate to strong distortions including scaling, amplitude modulation, fractional shift, arbitrary linear and shift invariant (LSI) filtering, and blockwise filtering, and show that the graph-based iterative decoders perform almost as well as if they had exact knowledge of the distortion channel parameters. Other desirable features of the graph-based decoders include the flexibility to adapt to other types of distortions and the ability to cope with the "curse of dimensionality" problem that seemingly results when the distortion channel parameters' space has high dimensionality. These properties are unlike most blind watermark decoders proposed to date, and close an important computational gap in favor of deploying joint estimation-decoding strategies (shown to be theoretically optimal in our earlier work) to cope with common signal distortions.;For the second problem, we propose new tools for steganalysis of queuebased stegocodes over covert timing channels. We propose a universal estimator for the Kullback-Leibler (KL) divergence-rate between the covertext process and the stegotext process. We empirically illustrate the performance of our estimator on some simple queue-based stegocodes and study its convergence properties. |
