STATISTICAL METHODS FOR STUDYING NETWORK TRAFFIC

Abstract

Introduction: The single multiservice network concept which involves the integration of voice, data and multimedia communication has prompted interest in studying the nature of network traffic. Studies of traffic traces recorded on a large scale show the presence of a self-similar structure in it, which requires a revision of the results of modeling infocommunication networks under the assumption of a Poisson data flow. Purpose: Studying a sequence of application methods for studying the network traffic nature, identifying the self-similar traffic nature in the form of statistical estimates and the Hurst index. This should offer tools for generating artificial traffic which adequately reflects a real network traffic, taking into account the revealed self-similarity properties. Results: The self-similarity properties of the considered 3G traffic were checked on different time scales obtained by aggregation of 5, 10, 15 and 20 minutes on the available daily traffic of 3G data. An estimate of the tail severity for self-similar traffic distribution was obtained by constructing a regression line for the additional distribution function on a logarithmic scale. The self-similarity parameter value, determined by the severity of the distribution “tail”, made it possible to confirm the assumption of 3G traffic self-similarity. A review of models simulating real network traffic with a self-similar structure was performed. Tools were implemented for generating artificial traffic in accordance with the considered models. Various artificial network traffic generators were compared, according to the least squares method criterion, for approximating the artificial traffic point values by the approximation function of 3G traffic. Qualitative assessments of the traffic generators were taken into account, in the form of the their software implementation complexity, which, however, can be a subjective assessment. Comparative characteristics allow you to choose a generator which most faithfully simulates real network traffic. Practical relevance: The proposed sequence of methods to study the network traffic properties is necessary for understanding its nature and for developing appropriate models which simulate real network traffic.