Yue Liu and Laurie Cuthbert
adaptive intra update, H.264, Cognitive radio, Rate-Distortion model, Finite state Markov chain
The channel variations and restrictions to protect primary users in Cognitive Radio (CR) networks bring challenges to video transmission for secondary users. Researchers try to find a trade-off between the video compression and video quality to satisfy the requirement for video streaming. For all DCT-based video coding such as H.264 (MPEG-4 part 10), the video is encoded using Macroblock (MB) motion compensation and transform coding of the prediction residual in order to compress the video size as well as retain acceptable audiovisual quality during error-prone transmission. However, this kind of compression allows error spreading along the motion compensation path to the following frames, which can cause catastrophic quality degradation. Adaptive Intra Update (AIU) offers a good solution since the intra-coded MB is coded only using the spatial information within the same frame and can stop the error propagation and hence can improve the error resillience. Instead of transmitting a whole intra-frame intensively, which is rather rate consuming, AIU enables the refreshing process to take place in a distributed manner. In this paper, several intra update schemes for video transmission over CR networks have been investigated. The optimality is measured by computational complexity, video quality improvement and error resilience. The CR system is viewed as a Finite State Markov Chain.
Important Links:
Go Back
