Research And Implementation On QoS Of Video Real-time Streaming Transmission

In recent years, with the rapid development and effective integration of networktechnology and multimedia technology, the variety and amount of network multimediaapplication have been increasing. The video data transport quality via network causes moreand more attention. The weakness of current IP network like serious fluctuation, limitedresources and instability, which can’t provide quality of service (QoS), will result in theproblems of delay, jitter and packet loss in video real-time streaming transmission. Therefore,it is very critical to study how to guarantee QoS to the video real-time network transmission.This thesis studies the streaming media theory of transmission and existing QoStechnology to comes up with an improved network status evaluation algorithm, and achievesto design an adaptive QoS guarantee strategy based on the terminal. In network statusevaluation algorithm, packet loss rate of video data is considered to be the main evaluationfactor and network is divided into under loading, normal loading and over loading three mainstatuses. When the network is normal loading, delay jitter factor is added to test if the networkis going to be over loading to achieve pre-control. The adaptive QoS guarantee strategyconsists of adaptive bitrate adjustment algorithm and video frame rate shaping algorithm. Theadaptive bitrate adjustment algorithm restrains the minimum output of bitrates from encoderto guarantee the consistency of video image and its quality. When network congestion stillappears under the minimum output of bitrates, the video frame rate shaping algorithm isadopted by QoS guarantee system. It shapes the flux from encoder via Abandon frame filterand further reduces the amount of data which will be transited to the network. Therefore, itcan control effectively the packet loss rate, delay and jitter to realize QoS of videotransmission.QoS guarantee system studied by this thesis for video real-time streaming transmissionis tested abundantly and data collected under IP public network environment. It proves thatthis system is practical, flexible, stable and friendly. It also verifies that this system canrapidly reduce the packet loss rate and jitter delayed in video data transmission, improve theutilization rate of network link and provide effective QoS guarantee for video real-timetransmission.