Wireless Video Streaming Transmission Based On Delay Aware Fountain Code

In recent years,video streaming on the Internet has received wide attention.More than 50% of the world's data traffic is consumed by video at present and will exceed 80% by the end of 2020.With the popularity of smart mobile devices,the demand for online video has increased rapidly.Therefore,it is very challenging to achieve high quality and low latency video transmission in wireless networks.Because of the stochastic nature of wireless channel,video streaming will have a high packet loss rate in wireless networks.The forward error correction code has been widely applied in the field of wireless video transmission because of the ability to recover packet loss.Fountain code is a state-of-the-art forward error correction code which is an ideal solution for wireless video streaming due to the characteristics of ratelessness.Fountain code can be applied to large file high-speed transmission,streaming media on demand or broadcast,but there are shortcomings such as high latency and low definition.Therefore,the domestic and foreign researchers propose various solutions for the problems of large delay and poor effect based on the coding and decoding principle of Fountain code.The thesis aims to implement a low latency and high definition delay aware Fountain code(DAF)scheme.The scheme combines the delay aware feature in video coding and the channel adaptive feature in rateless coding,and transmits the wireless video stream using Fountain code based on time sliding window method.The main work of the thesis is the following three aspects:1.A wireless video streaming transmission scheme based on DAF is implemented.The key technology of the scheme is the time based sliding window and the best window sampling strategy,which solves the problem of the inhomogeneous coding caused by the bit rate fluctuation of the video stream,and improves in-time decoding rate(IDR)of the video stream.2.An online algorithm for live wireless video streaming based on DAF and model predictive control(MPC)is implemented.Based on MPC,the algorithm predicts the bit rate of future video stream,optimizes the slope of each window from the predicted bit rate,and realizes a low complexity online algorithm.3.An unequal error protection(UEP)scheme based DAF is implemented.In the scheme,UEP is added to protect important bits information in DAF.According to the importance profile,the scheme samples data packets to achieve unequal protection.The experimental results show that the algorithm has a higher IDR and peak signal to noise ratio(PSNR)compared with equal error protection(EEP)under the same delay and bit rate.