Network phase transitions and critical behaviors under self-similar traffic model

Along with the expanding of network and the increase of network users, Internet has been developed into a huge complex non-linear system. Research of network non-linear dynamics has attracted great attention in the fields of computer science and statistical physics. In this paper, a model of computer network is put forward, and then various simulation tests are made under self-similar traffic. In the process of analyzing the simulation results, it is found that order parameter-(d) plays an important role on the whole network phase transition from sparse congestion to intermittence congestion and to durative congestion. To clearly illustrate the fluctuation of buffer length, a power spectrum analysis is made, in which the relations between order parameter-(d) and 1/f, 1/f(2) power law are revealed.