摘要
针对远程驾驶领域所需的高实时性网络传输应用场景,提出了自适应BBRv2拥塞控制算法,旨在优化BBRv2拥塞控制算法在远程驾驶网络中不同往返时延(round trip time,RTT)流共享瓶颈链路的公平性,使其工作点趋近最佳工作点,并降低因为偏离最佳工作点造成的高传输时延。所提算法通过添加一个以RTT为减函数的因子动态提高较小RTT流的竞争性,设置排队时延阈值,改善较大RTT流和较小RTT流的反应灵敏度,实现相对公平的带宽分配和低时延传输。通过网络仿真器3(network simulator,NS3)平台验证自适应BBRv2拥塞控制算法的有效性,结果表明:相比于BBRv2,在远程驾驶网络深度缓冲区,自适应BBRv2算法在公平性上提升39.4%,时延大幅降低。
Aiming at the high real-time network transmission application scenarios required in the remote driving field,an adaptive BBRv2 congestion control algorithm is proposed,which aims to optimize the fairness of the BBRv2 congestion control algorithm in sharing bottleneck links between different RTT flows in the remote driving network,make its working point approach the optimal working point,and reduce the high transmission delay caused by deviation from the optimal work point.The proposed algorithm improves the competitiveness of shorter RTT flows by adding a factor dynamic with RTT as the minus function,and improves the response sensitivity of longer RTT flows and shorter RTT flows by setting the queuing delay threshold to achieve relatively fair bandwidth allocation and low delay transmission.The effectiveness of the adaptive BBRV2 congestion control algorithm is verified through the Network Simulator(NS3)platform.The results show that,compared with BBRv2,the adaptive BBRv2 algorithm under the depth buffer of the remote driving network improves the fairness by 39.4%and significantly reduces the delay.
作者
唐雄
赵津
韩金彪
刘鹏
TANG Xiong;ZHAO Jin;HAN Jinbiao;LIU Peng(School of Mechanical Engineering,Guizhou University,Guiyang 550025,China)
出处
《重庆理工大学学报（自然科学）》
北大核心
2023年第9期198-207,共10页
Journal of Chongqing University of Technology：Natural Science
基金
国家自然科学基金项目(51965008)
黔科合支撑项目(〔2022〕045)。
