IPv6网络路径容量测量方法研究

Research on Path Capacity Estimation for IPv6 Networks

分组离散技术是常用的路径容量测量方法之一,它通常采用连续发送策略来发送测量分组。然而该策略对发送带宽的利用率较低,影响了测量的准确性。针对IPv6网络环境,提出了一种新的基于分片机制的路径容量测量方法。该方法通过构造长度大于路径最大传输单元的原始测量分组,迫使协议栈对该分组进行分片操作,生成长度相等的测量分片序列。实验结果表明,与经典的pathrate工具相比,该方法能够获得更小的源端分组离散以及更好的背景流量抑制能力。

Packet-dispersion technique is one of the useful path capacity estimation methodology and usually employs continuous sending policy to send probing packets.But the accuracy of measurement is affected by the lower utilization of transmission rate at the sender.Based on fragmentation mechanism,a novel scheme of path capacity estimation for IPv6 networks was proposed.An original probing packet with its size larger than Path Maximum Transfer Unit is generated at the sender,and then be split into a serial of equal-sized fragments by the protocol stack.It is shown by extensive experimental results that the proposed scheme can lead to lower dispersion of the probing packets at the sender and a greater capability against the cross traffic,compared with the classical tool called pathrate.