Multi-Level Adaptive Video Streaming Over Wireless Channels

Abstract

Delivery of digital video over wireless networks offers added functionalities and advantages over traditional wired video systems. However, reliable video streaming over wireless channels is fraught with several challenges. Video streaming has strict requirements on bandwidth, delay, and loss rate. In addition, wireless channels are dynamic and error-prone by nature. In this thesis, multi-level adaptive approaches are proposed to mitigate these challenges. The objective of these approaches is to ensure continuous playback with acceptable video quality. First, bitstream switch- ing is combined with adaptive playback to accommodate variations in the channel condition and to avoid interrupted video playback. Second, scalable coding, adap- tive modulation, and adaptive channel coding are integrated to achieve e cient video streaming. A probabilistic approach is used to adequately scale video frames to ensure successful delivery within a bu er-state-dependent budget time. Adaptive modula- tion and channel coding help reduce the amount of required scaling, hence, improving the quality of the received video. Moreover, in this thesis, new temporal quality metrics are introduced for the evaluation of video quality. These quality metrics are the skip length (SL) and inter- starvation distance (ISD). A quality assessment system which takes into account these two metrics is implemented. Experimental results show that the proposed system is capable of estimating the peak signal to noise ratio (PSNR) of the received video frames without the need of the reference video. Hence, it provides a better alternative to the conventional PSNR calculation approach.