Research On Wireless Channel Modeling And QoE Improvement Techniques For Video Applications

With the rapid development of wireless communication technology and multimedia technologyin recent years, multimedia communication mainly composed of video application is becoming themain service in wireless communication networks and the Internet. Though the H.264video codecstandard can achieve high compression ratio, meanwhile it also makes the video stream moresensitive to the transmission error. At this stage the radio channel has limited network bandwidthand can be interfered easily. It is difficult to provide a reliable service for the video applicationswhich has large amount of data.In this paper the packet loss characteristic of the wireless network are studied first, and thensome methods to improve the quality of experience in video applications are discussed. The maincontents are below:(1) By investigating the packet loss characteristic in3G network, an improved networksimulation model based on the Gap is proposed. The model can be used for network simulationsoftware. The model can not only describe the packet loss status of the network for a long time, butalso reflects the relationship between the adjacent packet losses. It describes the packet losscharacteristic of3G network more accurately with little calculating amount.(2) After studying the QoS technology for video applications in UMTS network, a mediaaware active queue management mechanism is proposed to improve the transmission quality ofvideo businesses. The mechanism is achieved under the AM mode in the RLC layer of the UMTSnetwork. The method manages the packet queue according to the importance of the packets, thepackets with data of I frames are more import than the ones with P and B frames. The QoE of thevideo business in the UMTS network can be improved by reducing the loss of I frames packets.(3) A control scheme based on QoE for wireless video transmission is also realized in thispaper. A model is used to predict video quality according to a set of objective parameters. In orderto optimize the end user’s visual experience, the codec rate at the sender side will be adjustedaccording to the model. The proposed scheme was tested in real network. The result shows that thismethod can achieve video transmission with better quality.