Key properties of multimedia traffic are characterized by the great diversity of correlation and burstiness. The authors construct the multimedia traffic by heterogeneous 2-state Markov chains (MC) and allow each 2-state MC to be defined on different transition intervals, representing the diversity of peak access rate from individual sources. Upon time renormalization, the queue is modeled by a multi-dimensional discrete Markov process. Generalizing the technique developed previously the authors are able to construct the entire solution of queue length distributions. It uses a generating function approach. By matrix spectral decomposition, one can decompose the evaluation of each individual vanishing/non-vanishing root. Each vanishing root is used to obtain the queue boundary values; each non-vanishing root constructs a geometric term in the expression of the queue length distribution. It is found that all the vanishing roots must be real and all the complex non-vanishing roots can be ignored. In numerical studies, the authors examine the effect of diverse source access rates and time-varying scales on ATM queues.>

A wideband source in high speed networks is typically represented by a binary random process. In this paper we characterize the second-order properties of each binary source by a multi-state Markov modulated Poisson process (MMPP). A comprehensive numerical study is carried out to identify the individual effect of the source second-order dynamics on the queue length and loss rate. The results can be used to verify the validity of the two-state Markov chain binary source assumption which is commonly made within the framework of input rate control and bandwidth allocation in high speed networks. The concept of input power spectrum is then developed as a unified source characterization for multimedia traffic queueing analyses.>

Uses the well known technique of power spectral representation to characterize the second order statistics of packet loss during congestion at a statistical multiplexer. Unlike the steady state statistics, the second order statistics measures the time correlation behavior of loss. Typically, the low frequency loss power is associated with long periods of highly consecutive loss. The effect of the loss rate power spectrum on transmission qualities is service dependent. The study indicates that, by selective packet discarding, one can properly change the distribution of the overall loss rate spectrum among the multiplexed traffic streams, which can significantly improve the transmission quality of individual services. Moreover, with the design of multi-level buffer overload control, one can tune the loss rate steady state distribution of different priority streams to a piece-wise step function, which minimizes the packet loss impact on service qualities.>

A wideband source in high speed networks is typically represented by a binary random process. The second-order properties of each binary source are characterized here by a multistate Markov-modulated Poisson process. A comprehensive numerical study is carried out to identify the individual effect of source second-order dynamics on queue length and loss rate. The concept of input power spectrum is then developed as a unified source measurement for multimedia traffic queuing analyses. The authors thoroughly explore the relationship between the source second-order dynamics in the time domain and the input power spectrum in the frequency domain, as well as its overall impact on system performance.>

A solution for multi-media traffic queueing analysis is presented. The key properties of multi-media traffic are characterized by a great diversity in correlation and burstiness. The multi-media traffic is constructed based on various two-state Markov chains. Upon a time index normalization, the queue is modeled by a multi-dimensional discrete Markov process. The entire solution of queue length distributions is constructed. It uses a generating function approach. By matrix spectral decomposition and diagonalization, one can decompose the evaluation of each individual vanishing/non-vanishing root. It is found that all the vanishing roots must be real and all the complex non-vanishing roots can be ignored. In numerical studies, the time index normalization in source modeling is first verified, and the effect of multi-media traffic properties on ATM queues is examined.>