Research On 3D Video Traffic Model And Its Dynamic Bandwidth Allocation Optimization

With the rapid development of network video services in the past few years,video traffic has become the largest proportion in Internet traffic.3D video technology is widely used in various fields due to the stereoscopic visual experience it provides for viewers,but the huge data volume of 3D video brings challenges to video storage,encoding and transmission.The traffic model established for 3D video can generate 3D video simulation traffic for network evaluation,and can also be used to implement dynamic resource allocation based on traffic prediction to reduce 3D video transmission delay.The research on traditional 2D video traffic model is relatively mature,but there are few discussions about 3D video traffic model.The existing two 3D video traffic models mainly simulate video sources,but cannot implement dynamic resource allocation based on traffic prediction.The adaptive LSTAR model can predict the video traffic in real time,but the traffic model has low accuracy when the adaptive LSTAR is used to establish the traffic model for the left-view of the 3D video.In this paper,the two characteristics of 3D video traffic are summarized and a new left-view traffic model of 3D video is proposed based on the short-term linear regression and wavelet transform.The experimental results show that the proposed traffic model is better than the adaptive LSTAR model.Static bandwidth allocation or simple dynamic bandwidth allocation may cause an increase in data transmission delay because 3D video traffic is large and bursty.In order to reduce the 3D transmission delay,a dynamic bandwidth allocation algorithm based on traffic prediction is proposed in this paper to ensure efficient transmission of 3D video data.The OPNET network simulation model is established for the three types of resource allocation methods,the simulation results show that the dynamic resource allocation algorithm proposed in this paper effectively reduces the 3D video transmission delay and reduces the delay jitter caused by sudden traffic changes.