Autoregressive video conference models

Abstract

Video conferencing is an important application that has been extensively used in IP, ATM networks, and TV broadcasting as a means of interactive communications. Teleconferencing video traffic consists of video scenes in which one or more people are talking with low to medium motion and almost unchanged background. To reflect the properties of video signals in the design of communication and transmission networks, modeling this teleconference video has received considerable attention.For over a decade, there has been an enormous amount of interest and research in traffic modeling of compressed variable bit rate (VBR) video. An essential class of video models that has received much attention lately is regression models. This paper presents a survey of the state of the art regression traffic studies that have been proposed in the literature to model teleconference video. After an introduction to the basic definition of the regression process, the third section begins with a discussion of the earlier regression models. Then, a number of recent regression models are discussed. The paper concludes with a comparison of all the models, recommendations, and open problems.

An extended discrete autoregressive model for variable bit rate video traffic encoded by scalable video codec

“…In addition, the transition matrix for each value of p is in form of equation (1)(2)(3)(4)(5)(6)(7)(8)(9). Furthermore, the overall transition matrix can be determined as the mean of different values of p that correspond to every value of n, where ρ is the autocorrelation coefficient, I is the identity matrix, and Q is a matrix with Q ij = π(j), for i, j ∈ S.…”

“…For example, Izquierdo and Reeves [5] reviewed video source models, including a reasonable collection of existing models. Tanwir and Perros [6] presented a more complete evaluation of the following VBR video models: Autoregressive (AR) models [7,8], discrete autoregressive (DAR) models [9,10], Markov renewal processes (MRP) [11], and transform expand sample (TES) [12]. The main focus of the current study was DAR models.…”

Estimation of Modelling Parameters for H.263-Quantized Video Traces

“…Under the above expectation, it is evident that a statistical model for this type of traffic would be very useful to predict network usage and estimate resources. For this reason, a lot of traffic models exist mainly as autoregressive (see [17] for a review of such models). Newer studies of video traffic modelling, for example, [18][19][20] reinforce the general conclusions obtained by the above earlier studies by evaluating and extending the existing models and also proposing new methods for successful and accurate modelling.…”

Analysis and Modeling of H.264 Unconstrained VBR Video Traffic

“…This detailed traffic study and mathematical quantitative approach of H.264 VBR video is necessary for understanding the properties of its characteristics, which will be used for generating synthetic H.264 traffic by an appropriate video model of H.264 Unconstraint VBR traffic. Thus, in order the proposed model to be efficient, it must fulfill the following criteria [1]:  It must adopt specific statistical characteristics of the real video traffic, like PDF and ACF.  The characteristics of the synthetic video model must be similar to the ones of the real video, so that it can be used instead of real video traffic for predicting desired performance metrics.…”