Collaborating Video Transmission Based On Neighbor Node

Video application has become a killer application of users in current wireless network. According to a research report from Cisco, half of Internet traffic has been contributed by video application and the ratio will be increased continuously in the future.At the same time, the wireless network is undergoing enormous development. On one hand, even though the throughput of wireless access network has been improved greatly by widely deployed of LTE networks, it is still difficult to transmit video traffic with satisfied quality because of limited bandwidth. On the other hand, due to the increasing development of wireless access technology and the fierce market competition, more and more users choose to use wireless terminals and the wireless terminals generally have multiple interfaces. Aggregating multiple upload bandwidths of neighbor nodes can realize transmission bandwidth required by the video application.Based on the huge demand of video transmission and the development of present wireless network, we propose two kinds of video cooperative transmission schemes in this paper. These two schemes can improve the transmission quality of video without changing the network infrastructure. Specifically, we propose the following two schemes in this paper:Collaborative video transmission scheme based on redundancy. Its basic idea is as follow:the video source node duplicates the video packets into multiple copies, and each copy is relayed by a neighbor node. Finally, the video packet will be uploaded by neighbor node to the Internet. We build a model to explore the relationship between video QoE and network setting. According to the model, we obtain the optimal video encoding rate in order to maximize the video quality under a specific network condition, such as the probability of packet loss of physical channel, channel bandwidth and the number of transmission links.Redundant transmission scheme based on the characteristics of real-time video. Its basic idea is as following:Based on the observation that different frames in a GOP has different transmission delay requirement, we propose that different frames of the video should be transmitted with different number of copies according to their delay requirement. Then, we build a continuous optimization model and a discrete optimization model to obtain the optimal encoding rate and optimal redundancy configuration in order to maximize the real-time video quality. Numerical result shows that our scheme can effectively improve the real-time video quality.