A Method for Modeling and Generating Normal Network Traffic Based on the Features of Length and Arrival Time of Packets Using the Zipf's Law

Today, modeling and generating normal network traffic is a very important. In existing works, the features of network traffic are modeled using probabilistic distributions. In this paper, a new method is proposed for modeling the features of network traffic. The proposed method is based on the Zipf’s law. The Zipf's law is an empirical law that provides the relationship between the frequency and rank of each category in data set. In this paper, we will show that the Zipf’s law can model different features of network traffic in a good manner. For this propose, two important features of network traffic, i.e., length and inter-arrival time of TCP and UDP packets, are examined. The proposed method for modeling the features of network traffic can use in various applications areas, such as, simulation or generation of the normal network traffic. The advantage of this law is that it can provide high similarity using less information. Furthermore, the Zipf’s law can model different features of network traffic that may not follow from probalistic distributions. The simple approach of this law can provide accuracy and lower limits from existing methods. Furthermore, the proposed method can provide good times for modeling and simulation.
In this paper, we will show that by classifying the feature values and obtaining their ranks, we can create an accurate modeling of features. In other words, the rank of each category will be the model resulting from the feature values that can be used in simulation.