(An increasing number of studies have shown that network traffic is not a Markov process, but has a bursty, self-similar property at any time scale. Mathematical models describing network traffic mainly include self-similar and long-correlation structures. Certain parameters of the network obey heavy-tailed distributions, which then lead to bursty properties on the time scale of network traffic. This paper analyzes the harm of traditional network traffic model and functional analysis, and describes the new network traffic)