An architecture for federated video processing and online streaming

Today's multimedia communication has been greatly shaped by the coexistence of a number of complementary as well as competing access, delivery, and consumption technologies. Rich media are now accessible via numerous multimedia enabled devices through a wide variety of network types. "What is missing" is so far a mechanism to ensure that multimedia users can receive different qualities of video proportional to their device capabilities and network conditions. In this thesis, we propose an online adaptive video streaming concept which takes into account different issues related to the Peer-to-Peer (P2P) content distribution paradigm such as peers' unreliability, as well as pragmatic aspects like receivers' heterogeneity. Our proposal also envelops compressed-domain video adaptation, watermarking, and perceptual encryption schemes where we utilize the MPEG-21 generic Bitstream Syntax Description (gBSD) as a content metadata. The proposed architecture aims at online video adaptation with streaming in the Application Layer Multicast (ALM) overlays to serve heterogeneous devices including small handhelds. In this proposed design, participating peers act as the stream source, adaptation engine, and perform the streaming tasks. We implemented a 3-in-1 adaptation-watermarking-encryption system to evaluate the compressed-domain adaptation performance. For the adaptive streaming, simulation is used to manifest that our design is robust, reliable, and suitable for multi-participant real-time collaboration and real-life deployment. The adaptive streaming system performance is also validated with the results found from an analytical model