基于流量负载的改进型几何抽样算法

为了解决在流量高峰时期网络节点由于超负荷工作导致报文丢失的问题,在几何抽样的基础上,设计了一种能自动适应流量负载变化的抽样算法.该算法在流量高峰时会根据负载情况动态地计算最佳的抽样概率来匹配节点的处理能力从而降低节点丢失的报文数,在流量负载轻时会自动地提高抽样概率以充分利用节点的处理性能.通过真实网络流量数据的实验分析表明,改进后的几何抽样算法不仅能有效降低节点丢弃的报文数,同时还提高了网络测量的精度,证明了该改进算法对流量负载具有良好的适应性.

A Improved Geometric Sampling Algorithm Based On Traffic Load

In order to solve the problem that the network node may lead to the loss of packets due to work overload during the period of peak traffic, a sampling algorithm based on geometric sampling is designed, which can adapt to changes in the traffic load automatically. The algorithm working in the period of peak traffic will reduce the sampling rate to adapt to the network node's processing capability while increasing the sampling rate in order to fully utilize the processing performance of the node when the traffic load is light. The experiment analysis through the real network data shows that the improved geometric sampling algorithm has a lower lost packets quantity and a higher measurement accuracy than the geometric sampling algorithm, which proves that the improved algorithm can adapt to the changes of the traffic load better.