Network Traffic Generation: A Combination of Stochastic and Self-similar

Network traffic generation is a vital part of traffic research as the exponential growth of the number of servers, as well as the number of users. Various researchers have reported traffic analysis that demonstrates different results of traffic modeling, such as Poisson distribution or considerable burstiness on a range of time scales with properties of self-similarity. Due to the distinct standpoint about the network traffic distribution, traffic generators have been developed dissimilar. In order to simulate the network traffic all-around, we present a technology of traffic generation which compose of stochastic and self-similar, and provide details on algorithm and implementation. In this paper, a new model for multi-patterns network traffic generation is presented. This model is based on three elements including the latency of network frames, Hurst exponent and network traffic types. This paper analyses the three parameters and finds a way to describe the relations among these. We choose multifractal wavelet model as the basis method, and perfect it applicable to multi-patterns network traffic generation. In this research, a network traffic generation system based on programming multi-core processor is build and the test result is given.