Research On Adaptive Rate Control Strategy For Ultra-low Delay Multipath Parallel Transmission

With the rapid construction and development of 5G networks,remote driving,telemedicine and other video applications requiring low latency and high bandwidth continue to land.The most important feature of these applications is the need to provide stable low latency and high image quality in video transmission.Even 5G networks are still inseparable from the time-varying characteristics of wireless,and the bandwidth and delay of wireless channels will dynamically change due to user movement.On the other hand,when a large number of users access the network at the same time,the available resources of a single user will also become limited,and if the node is at the edge of the cellular network,the speed of the node will drop rapidly.In the next generation of wireless networks,users not only rely on 5G networks,but 4G,WIFI and other networks will coexist for a long time.How to make full use of the characteristics of heterogeneous wireless networks and improve the efficiency of video transmission under these heterogeneous wireless networks is a hot topic of current research.In this paper,the video transmission technology under heterogeneous wireless networks is studied,and an adaptive rate control strategy for ultra-low-latency multi-channel parallel transmission is proposed to solve the problem of how to ensure ultra-low-latency video transmission in the next-generation heterogeneous wireless network.First of all,in view of the link asymmetry and time-varying characteristics of heterogeneous wireless networks,this paper proposes a joint rate control and target delay multi-objective optimization algorithm.In the traditional video transmission strategy,in order to ensure low latency,the video bit rate is usually adjusted based on the network bandwidth to ensure ultra-low latency video transmission under the fixed end-to-end target delay constraint.However,when the wireless network performance is poor,the video bit rate has to be lowered to meet the system delay constraint,so that the continuous video transmission quality cannot be guaranteed.For this reason,this paper proposes a multi-objective optimization model of joint rate control and target delay,in which the target delay and video bit rate are dynamically adjusted jointly.In order to reduce the complexity problem caused by the multi-objective and multi-parameter characteristics,the multi-objective optimization problem is further decomposed into independent sub-problems to solve.The experimental results show that the proposed scheme can dynamically adjust the video bit rate and target delay constraints in real time based on the multi-link time-varying characteristics,thereby ensuring that more video frames reach the server in time under the ultra-low delay constraints,and improving video transmission quality.Secondly,considering that network packet loss will significantly reduce the transmission quality of real-time video,this article uses scalable video coding(SVC)to improve the error recovery performance of video transmission.Fully combining the characteristics of SVC encoding and multipath parallel transmission,a transmission strategy is proposed which ensure ultra-low delay constraints through the base layer video stream,and to improve the video transmission quality through the enhancement layer video stream.To this end,this paper optimizes the joint stream distribution strategy synchronously,allocates different levels of SVC code streams under different link quality,and ensures the basic layer transmission through resource reservation for the SVC basic layer,so as to achieve the basic transmission quality under the ultra-low delay constraint.Experimental results show that the proposed algorithm can prioritize the allocation of basic layer data to a more reliable transmission path when the link is packet loss,thereby ensuring that more basic layer data is completely received under ultra-low latency constraints,and improving video transmission quality.Finally,the algorithms proposed in this paper have been verified on the actual test platform.The related research in this article has certain reference value for the research of real-time high-definition video multi-channel parallel transmission.