Аналітична модель інтелектуальної надбудови NGN з урахуванням самоподібності трафіку

Autor: Князєва, Н. О., Шестопалов, С. В., Кунуп, Т. В.
Zdroj: Refrigeration Engineering & Technology; 2018, Vol. 54 Issue 4, p73-80, 8p
Abstrakt: With the advent of multiservice networks there were intelligent network services (INS) and, accordingly, a new type of traffic. For a long time, it was thought that network traffic corresponds to Poisson processes, but further research has shown that the traffic of some networks has the effect of self-similarity. Because of the properties of self-similar traffic, traditional methods for calculating the characteristics of the functioning of networks give too optimistic results and lead to an underestimation of the real load. There is an urgent question of determining the presence of the effect of selfsimilarity of traffic, containing requests for INS, and also taking into account this effect when forming an analytical model of the NGN intelligent superstructure. It is this question devoted to this work. Based on the analysis of existing methods for calculating the Hurst index, which allows you to determine the nature of traffic, the R/S method is chosen because its use allows to analyze a large amount of data and also does not contain too much computations. This method is implemented using the AutoSignal program. Based on the analysis of the results obtained, it can be argued that traffic containing requests for INS is a self-similar process. The effect of self-similarity manifests itself in a wide range of time - from several hours to a year. The conducted studies of the nature of traffic identified the possibility of solving the actual problem - the development of an analytical model of the NGN intelligent superstructure, which is responsible for managing the provision of INS, taking into account the self-similarity of traffic. To construct an analytical model of intelligent superstructure, the apparatus of queuing theory was used. The proposed analytical model of intelligent superstructure, which takes into account the self-similarity of the flow of requests for INS, provides the opportunity to determine the necessary network resources to provide the necessary value of the effectiveness of managing the provision of INS. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index