Font Size: a A A

Chaotic Prediction Theory And Its Application In VBR Video Traffic

Posted on:2005-11-10Degree:MasterType:Thesis
Country:ChinaCandidate:H C LiFull Text:PDF
GTID:2168360125453350Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the development of chaos theory and application technology study, analysis and prediction of chaotic time series have become the emphasis of chaotic signal precessing research domain, and can solve a lot of nonlinear signal process questions in engineering practice, which are difficult to be done by linear signal processing methods.Multimedia traffic has become the chief source of ATM network gradually with the development of broadband network and digital compression technology, and most of them are transmitted through variable bit rate (VBR) mode. Recent researches indicate that VBR video traffic not only has short-term dependence, but also has long-term dependence. Consequently traditional models of variable bit rate video traffic are deficient in depicting variable bit rate video traffic, but fractal models can do it well. Gernerally, self-similar behaviors of fractal mean the existence of chaos, thereby analyzing variable bit rate video traffic through chaos theory will be likely to reveal intrinsic laws, which are not founded by traditional analysis method.Research works focus on prediction methods of chaotic time series and its application in VBR video traffic in this dissertation, which mainly include analysis and local prediction method of chaotic time series, nonlinear adaptive prediction of chaotic time series, chaotic characteristics analysis and chaotic prediction of VBR video traffic. The main research fruits are as follows:1 . Prediction capability and anti-noise performance of three kinds of local zero-order prediction methods are analysed completely. Local first-order prediction of chaotic time series based on Wiener solution is studied. DCT domain local quadratic polynomial prediction is presented, and simulation results indicate that this method not only can predict some low-dimension chaotic time series efficiently, but also can be implemented simply.2. Based on the analysis of polynomial nonlinear adaptive prediction methods existed already, a DCT domain quadratic predictor for real-time prediction of low-dimension chaotic time series is proposed. The feasibility and prediction performance of chaotic time series with neural adaptive prediction method with adaptive amplitude are studied. Time-delay neural Chebyshevorthogonal polynomial adaptive nonlinear prediction model is constructed. Experimental results indicate that three kinds of prediction models can predict some low-dimension chaotic 'time series efficiently. Moreover nonlinear adaptive prediction structure and algorithm of chaotic time series are developed further.3. The positive maximum Lyapunov exponents of variable bit rate video traffic are calculated on any time scale using Wolfs algorithm, which indicates that variable bit rate video traffic has chaotic characteristics. Furthermore, chaotic local first-order prediction method is used to predict two typical variable bit rate video traffics. The prediction results show that chaotic local prediction methods can predict the variable bit rate video traffic efficiently, and are more accurate than adaptive linear prediction method without time delay.
Keywords/Search Tags:chaotic time series, phase-space reconstruction, chaotic characteristiec, nonlinear adaptive prediction, varable bit ratevideo traffic
PDF Full Text Request
Related items