基于拥塞控制的视频网络传输质量保证算法

Quality Assurance Algorithm of Video Network Transmission Based on Congestion Control

针对视频在网络传输过程中的时延、拥塞和丢包率高等问题,提出了一种基于非平衡多描述编码的视频网络传输质量保证算法,在检测网络链路可用带宽的基础上,通过马尔科夫模型预判网络链路的拥塞状态并据此调整视频传输路径。实验表明,与RED(random early detection)算法相比,该算法更能有效地估计网络拥塞状况、降低视频丢包率和减少网络延迟,从而更有效地保证视频网络传输质量。

The drop- rate and net delay during video network transmission are two important factors in ensuring video net-work transmission quality. The purpose of network congestion control is to reduce the bad quality impact of video network transmission caused by net delay and the drop - rate. A video network transmission quality assurance algorithm was presented based on unbalanced multiple description coding scheme. Markov model was employed to predict the state of network congestion based on available bandwidth detection; and video transmission path was changed on the basis of the prediction. Experiments show that, compared with RED algorithm, the algorithm is much more effective to estimate network congestion, reduce video packet loss rate and net delay, thus the video network transmission quality can be ensured more effectively.