中文摘要：

在分析网络可用带宽测量算法性能及存在问题的基础上，提出一种轻负载、高精度、自适应的端到端可用带宽测量算法iChirp。基于自拥塞理论采用近似指数分布的动态探测队列结构，以可用带宽预测值为中心设置关键域，在其中加快采样频率，增加报丈密度，并能根据反馈自适应的对探测范围和关键域进行调整。仿真实验结果表明，该算法对网络干扰性较小，能够快速准确地测量出端对端可用带宽。

英文摘要：

After analyzing the performance and problems of network available bandwidth estimation algorithms, iChirp, a light overhead,high accuracy, self adaptive end to end available bandwidth estimation algorithm is presented. Based on the concept of ＂self-induced congestion＂, iChirp features a dynamic approximate exponential flight pattern ofprobes. The probing stream consists ofa critical region around the predictive value of available bandwidth, where the probing rates are sampled more frequently and the density of probing packets is higher. According to the feedback, iChirp also can adjust probing range and critical region adptively. Experiments show that iChirp provides accurate estimation of the available bandwidth rapidly and low-intrusively.