The mean field model and analysis on the dynamical behavious for network traffic

In a recent study, the chaotic phenomena of traffic has became an attractived topic in the filed of computer network. Based on theory of mean field, the mathematics model for describing the dynamic behavious of network traffic is presented, dynamical properties of this model are also analysed. It is shown that dynamical properties of the model is closely linked to the combination control parameter coupled by the some key factors including the average of the shortest-path, the average degree of nodes, the load of traffic, and the proportion of hosts; It is indicated, with the change of the combination control parameter, traffic behavious can transfer periodicity property to chaos property by the way of multi-periodicity bifurcation. Results further testify the chaos behavious is an intrinsic nature in the system of network traffic.