Abstract
Multiservice traffic is characterized by its stochastic nature, which significantly complicates the tasks of evaluating the capacity and performance of telecommunication networks. For effective planning and design of modern networks, it is necessary to use reliable methods for analyzing and evaluating the statistical characteristics of traffic, such as flow intensity, variation coefficients, autocorrelation, packet interarrival time distribution, and session duration. This paper presents an analysis of methods for collecting, processing, and interpreting multiservice traffic parameters. Approaches to network monitoring and the use of statistical data analysis under conditions of variable network load are considered. Special attention is given to formulating scientifically grounded hypotheses for modeling the behavior of different traffic types, such as voice, video, and data traffic. This allows for more accurate consideration of the specific characteristics of each traffic type when predicting network load. As a result of the study, recommendations were developed for selecting measurement and analysis methods to ensure optimal distribution of network resources. Approaches to improving the reliability and accuracy of performance evaluation under unpredictable load changes are also proposed. The results obtained will be applied to the development of a methodology for designing a prospective multiservice network in the Republic of Kazakhstan, which will enhance the efficiency of network infrastructure use, optimize routing processes, and improve the quality of service for end users. The findings will be useful for telecommunications specialists, as well as those involved in the design and operation of networks, in addressing tasks related to improving the performance and reliability of multiservice networks.