Intelligence Packet Scheduling for optimized video transmission over wireless networks

This paper presents an intelligent packet scheduling scheme based on an analytical distortion model for Multi-Referenced H.264 video over 802.11 wireless networks. This model considers that video distortion in H.264 video codecs is based on multi-reference video frames, as opposed to previous video standards. The model has been verified with actual video distortion measurements and it has been compared with a simple additive model that omits the correlation between the frames. The packet dropping that is required in the case of the available transmission rate is limited, takes into consideration the impact of the loss of the frame on the resulting video distortion that depends not only on the distortion of the particular frame but on the distortion effect that propagates to the correlated frames. Furthermore, simulation results show that the proposed model accurately captures the distortion effect for reference based H.264 coding and the improves the reconstructed video quality when is implemented in rate adaptation via packet scheduling.