Please use this identifier to cite or link to this item:
Title: Прогнозирование сетевого трафика при помощи авторегрессионных моделей
Authors: Овчинников, К. А.
Бушманов, В. С.
Issue Date: 2013
Publisher: ХНУРЭ
Citation: Овчинников К.А., Бушманов В.С. Прогнозирование сетевого трафика при помощи авторегрессионных моделей // Проблемы инфокоммуникаций. Наука и технологии (PIC S&T-2013): Сборник научных трудов первой международной научно-практической конференции, Харьков 9-11 октября 2013 г. / М-во образования и науки Украины, Харьковский национальный университет радиоэлектроники. –Харьков: ХНУРЭ, 2013. –С.177-179.
Abstract: The predictability of network traffic plays a significant role in many domains such as congestion control, admission control, and network management. An accurate traffic prediction model should have the ability to capture main traffic characteristics, such as long - and short - range dependences, self - similarity etc. In this paper two models of network traffic were analyzed and it is shown that time series simple autoregressive model (AR) provides good performance on short - term predictions. It is also shown that non - stationary time series representing number of packet on router’s port can be des cribed using first - order difference ARIMA model. It was also discovered that increasing number of lag variables of AR network traffic models leads to reduction of its performance. Suggested models can be used as a part of automatic management system for prediction - based routing problem solution.
Appears in Collections:Кафедра інфокомунікаційної інженерії (ІКІ)

Files in This Item:
File Description SizeFormat 
PICST_13-191.pdf1.38 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.