| With the rapid development of the network technology and multi-medium technology, demands for video communication services such as video conference and monitoring are increseing as well. Despite the growing tendency of bandwidth, only best-effort transmission service decided by IP network system could be provided. Due to the shortage of service quality guarantee, stable bandwidth required by video transmission cannot be ensued, thus, network congestion is caused, video data are lost in transmission and the quality of video received is low. How to improve transmission quality of video image under IP network environment are a hot point of network video studies and a key problem that video transmission services should slove as well.Network video communication is a challenging study subject, involved in video coding, speed rate control, congestion control, etc. Currently, congestion control studies are numerous based on protocol. However, congestion control study is at the theoretical stage, effective congestion control mechanism is short in most video transmission systems.In this paper, the transmission control is deeply analysed based on the background of the application on network video transmission, further, the video flow self adaptation communication technology is extensively studied. This technology consists of network self- adaptable congestion control and video self-adaptable speed rate control technology. Network self-adaptable congestion control technology uses -friendly congestion control, offers smooth congestion window for establishing data transmission bandwidth. Video self-adaptable speed rate control technology controls transmission rate according to transmission bandwidth. Video transmission prototype system using video flow self-adaptation technology is automatically adaptable to network condition, ease congestion and improve video transmission quality.Following points are solved in this paper:(1) On researching the congestion control based on network, this paper we adopted a Diffserv transmission strategy based on frame type, solving the problem that the RED can not judge the frame type and introduce the unfair transmission for some importance video frames.(2) On researching the congestion control based on the source, we study the current congestion control based on the source, and study the congestion control based on window deeply. Aiming at the shortage in the congestion control, we adopted a self-adaptable congestion control and study its implementation. It introuduce the novel of"Random"to control the sending time, also use the"gentle"congestion window mode to control the congestion window . Experiment results show that the congestion control we adopted can reduce the packets loss while smooths the video transmission rate, and also -friendly, very suitable for video transmission.(3) The self-adaptable congestion control has been applied in the network video transmission system, thereby puts forward a network video transmission system based on NS2. In order to reduce the packets and the network congestion further, the self-adaptable buffer management has also been applied in this system. It introduces the loss policy to ensure the transmission of the important video data and control the output transmission rate. Finally the self-adaptable video transmission technology has been realized in the source and receiver. Experiment results validate the validity of this system; also show that the subjective and objective video quality has been improved.The study findings and design thinking of this paper are highly meaningful to network self-adaptable video transmission design and are useful to apply to video conference and monitoring. |
PDF Full Text Request