W.J. Liu,∗ ,∗∗ J. Gong,∗∗ and W. Ding∗∗
Packet sampling, IP flows, network measurement, statistic inference
Routers have the ability to output statistics about packets and flows of packets that traverse them. Passive traffic measurement increasingly employs sampling at the packet level. However, knowing the number and length of the original flows is necessary for some applications. This paper provides an algorithm that uses flow statistics formed from sampled packet stream to infer the absolute frequencies of lengths of flows in the unsampled stream. We achieve this through statistical inference and by exploiting heavy-tailed feather. We also investigate the impact on our results of different packet sampling rate. The theoretical analysis demonstrates that the computational complexity is well under control, and the experiment results show the inferred distributions are accurate in most cases.
Important Links:
Go Back
