A new model for video traffic originating from multiplexed MPEG-4 videoconference streams

Call Admission Control for wireless videoconference traffic based on the users' willingness to pay

“…We have investigated the possibility of modeling the traces with a number of well-known distributions (gamma, lognormal, log-logistic, exponential, geometric, Weibull, Pearson V). Our results (derived with the use of Q-Q plots [10], Kolmogorov-Smirnov (KS) tests [10] and Kullback-Leibler (KL) tests [8]) have shown that, similarly to our work in [16] on modeling MPEG-4 videoconference traffic, the best fit among these distributions for modeling a single movie is achieved for all traces examined with the use of the Pearson type V distribution (also known as the inverted gamma distribution). The data for each trace consists of a sequence of the number of cells per video frame.…”

“…The first is that the estimation provided by our video traffic modeling approach yields an overestimation of the actual bandwidth requirements of the superposed sources. The reason for this is the slight overestimation of the I frames size with the use of the DAR model and are explained in our modeling work in [16,18]. Still, this overestimation is small and ranges in all simulated scenarios from a minimum of 4.06% to a maximum of 6.76%.…”

“…The reason for which our mechanism in [2] excelled in comparison to the "equivalent bandwidth" approach was the use of the accurate MPEG-4 videoconference traffic model from [16] in order to precompute various traffic scenarios and combine that knowledge with online simulation, in order to be able to make accurate decisions on the acceptance or rejection of a new call. The precomputation, along with the online simulation, was made in [2] based on the traffic parameters declared by the video sources at call setup.…”

“…Although widely used, "equivalent bandwidth"-based schemes are known to significantly overestimate the sources' actual bandwidth requirements and therefore to provide quite conservative CAC mechanisms, which fail to use efficiently all the available bandwidth [7,9]. Instead, we proposed the use of our modeling approach [16,17] for traffic originating from MPEG-4 and H.263 videoconference sources, in order to propose a new CAC mechanism for videoconference traffic transmission over wireless cellular networks. Our CAC mechanisms in [2,15] use the traffic parameters (peak, mean, standard deviation) which the video source either declares at call setup or has agreed on in its contract with the wireless provider, in order to precompute or compute online a large number of traffic scenarios for their decision-making.…”

Video scene identification and classification for user-tailored QoE in GEO satellites

“…The work in [3] used the Discrete Autoregressive Model of order one that was shown in [6] to be highly accurate. However, this accurate prediction is not possible for all types of video sequences, and even when it is, it often involves a higher degree of complexity which would incur additional computational requirements for an OBP system.…”