Information entropy theoretic approach to traffic sampling measurement in high-speed IPv6 networks
Changxing Pei · Journal of Jilin University · 2009
A novel traffic sampling measurement method in high-speed IPv6 networks is proposed.The method is to sample IPv6 traffic based on the information entropy theory.Firstly,we present a distributed IPv6 network sampling measurement framework including communication and sampling modules.Secondly,through computing the bit entropy in real IPv6 traffic traces,we discuss the randomicity of each byte in IPv6 packet headers,and select those bytes having a good randomicity as the input of our traffic sampling method.Finally,an experiment is performed to prove that the output of our method has sufficiently uniform and the sample's curve of cumulative distribution function of packet size is fully conformed to the parent's.