Traffic model in heterogeneous networks based on experimental data

Abstract

Almost all modern networks provide data transfer between different subscriber devices types using subnets operating in different standards. Therefore, they are heterogeneous. To obtain correct estimates of the load and throughput of heterogeneous networks, it is necessary to be able to simulate such networks traffic close to real. Therefore, the traffic model development in heterogeneous networks based on experimental data is relevant. To do this, the article conducted an experimental study of real traffic in the operator “New Technologies of the XXI Century” heterogeneous network, analyzed the traffic generation main models in wireless networks, mathematical models IPP, 4IPP, IDP and 2IRP were selected for generating traffic in heterogeneous networks and determined the necessary for them application settings. Subscribers are classified depending on the subscriber devices types, the traffic types they use, and the data transferred volume. Based on the experimental data obtained, subscribers are classified depending on the subscriber devices types, the traffic types they use, as well as the data volume they transmit. A list of the subscribers’ main classes has been generated, indicating the models used to generate their traffic, and their parameters have been determined. A traffic mathematical model in heterogeneous networks has been developed, the transmission simulation of real and generated according to the developed model traffic over a heterogeneous network has been carried out, on the basis of which the correctness of the developed model has been shown. The resulting model allows us to obtain correct estimates of heterogeneous networks main characteristics using the mathematical modeling method.