基于BBR的NDN拥塞控制算法

摘要

随着网络传输带宽以及用户对实时应用需求的增加,如何在充分利用瓶颈带宽的同时降低缓存占用率以及传输时延,成为传输控制的一个新问题.提出了一种基于瓶颈带宽以及往返时延(round-trip time,RTT)的命名数据网络(named data networking,NDN)拥塞控制算法.该算法不使用传统的基于丢包的拥塞感知与调节方法,而是主动控制注入网络的流量,使其匹配链路的传输能力.通过在接收端对一定时间范围内反馈的即时带宽和往返时延进行统计,估计传输链路的瓶颈带宽以及物理链路延迟的值,配合由状态决定的增益来控制Interest包的发送速率以及窗口的大小.在ndnSIM模拟器中实现了该算法,并与ICP(interest control protocol)拥塞控制算法进行对比,证明了在充分利用瓶颈带宽的同时,该算法能够实现更低的传输时延以及更快的收敛速度.%With the increase of transmission bandwidth and the need of real-time application,it becomes a new problem as for how to minimize the cache usage and the transmission delay while making full use of the bottleneck bandwidth.To solve this problem,we propose a congestion control algorithm based on bottleneck bandwidth and round-trip time (RTT) in named data networking.Instead of using traditional loss-based congestion detection and control methods,our algorithm proactively controls the traffic to match the capacity of the links.The bottleneck bandwidth and physical link delay are estimated at receiver by statistically analyzing round trip time and instantaneous delivery rate in a certain time frame.These two parameters,along with the gain decided by the state,will be used to control the delivery rate and the size of window of Interest packets.Finally we implement this algorithm in ndnSIM and compare it to interest control protocol (ICP).It's demonstrated that the proposed algorithm can achieve lower transmission delay and faster convergence speed while making full use of bottleneck bandwidth.