Ahmed A. Tarraf

Intelligent traffic control for ATM broadband networks

Performance results prove that a neural networks approach achieves better results, simpler and faster, than algorithmic approaches. The focus of this paper is to shed light on how neural networks (NNs) can be used to solve many of the serious problems encountered in the development of a coherent traffic control strategy in ATM networks. The main philosophy that favors neural networks over conventional programming approaches is their learning and adaptive capabilities, which can be utilized to construct adaptive (and computationally intelligent) algorithms for allocation of resources (e.g., bandwidth, buffers), thus providing highly effective tools for congestion control. >

A neural network controller for congestion control in ATM multiplexers

Abstract This paper presents and adaptive approach to the problem of congestion control arising at the User-to Network Interface (UNI) of an ATM multiplexer. We view the ATM multiplexer as a non-linear stochastic system whose dynamics are ill-defined. Real-time measurements of the arrival rate process and the queueing process, are used to identify, and minimize congestion episodes. The performance of the system is evaluated using a performance-index function which is a quantative measure of “how well” the system is performing. A three-layers backpropagation neural network controller generates a signal that attempts to minimize congestion without degrading the quality of the traffic. During periods of buffer over-load the control signal, adaptively, modulates the arrival process such that its peak-rate is throttled-down. As soon as congestion is terminated, the control signal is adjusted such that the coding rates are restored back to their original values. Adaptability is achieved by continuously adjusting the weights of the neural network controller such that the performance of the system, measured by its performance index function, is maximized over a certain optimization period. The performance index function is defined in terms of two main objectives: (1) to minimize the cell loss rate (CLR), i.e., minimize congestion episodes, and (2) to maintain the quality of the video/audio traffic by maintaining its original source coding rate. The neural network learning process can be viewed as a specialized form of reinforcement learning in the sense that the control signal is reinforced if it tends to maximize the performance index function. Performance evaluation results prove that this approach is effective in controlling congestion while maintaining the quality of the traffic.

Congestion control mechanism for ATM networks using neural networks

This paper presents a new approach to the problem of congestion control arising to the user-to-network interface (UNI) of the ATM-based broadband integrated services digital networks (B-ISDN). Our approach employs an adaptive rate based feedback control algorithm using reinforcement learning neural networks (NNs). The reinforcement learning NN controller provides an adaptive optimal control solution. This is achieved via the formulation of a performance measure function (cost function) that is used to, adaptively, tune the weights of the NN. The cost function is defined in terms of two main objectives: (1) to minimize the cell loss rate (CLR), i.e., control congestion and (2) to preserve the quality of the voice/video traffic via maintaining the original coding rate of the multimedia sources. The results show that the NN control system is adaptive in the sense that it is applicable to any type of multimedia traffic. Also, the control signal is optimal in the sense that it maximizes the performance of the system which is defined in terms of its performance measure function. Hence, our novel approach is very effective in controlling the congestion of multimedia traffic in ATM networks.

Characterization of packetized voice traffic in ATM networks using neural networks

Asynchronous transfer mode (ATM) broadband networks support a wide range of multimedia traffic (e.g. voice, video, image and data). Accurate characterization of the multimedia traffic is essential in ATM networks, in order to develop a robust set of traffic descriptors. Such a set is required by the ATM networks for the important functions of traffic enforcement (policing) and bandwidth allocation utilizing the statistical multiplexing gain. In this paper, we present a novel approach to characterize and model the multimedia traffic using neural networks (NNs). A backpropagation NN is used to characterize the statistical variations of the packet arrival process resulting from the superposition of a number of packetized voice sources. The NN is trained to characterize the arrival process over a fixed measurement period of time, based upon sampled values taken from the previous measurement period. The accuracy of the results were verified by matching the index of dispersion for counts and the variance of the arrival process to those of the NN output. The results reported show that the NNs can be successfully utilized to characterize the complex non-renewal process resulting from the aggregate voice-packet arrival process with extreme accuracy. Hence, NNs have excellent potential as traffic descriptors for the usage parameter control algorithm used in admission control and traffic enforcement in ATM networks.<<ETX>>

ATM multimedia traffic prediction using neural networks

Asynchronous transfer mode (ATM) broadband networks support a wide range of multimedia traffic (e.g. voice, video, image, and data). Accurate characterization of the multimedia traffic is essential, in ATM networks, in order to develop a robust set of traffic descriptors. Such set is required, by the usage parameter control (UPC) algorithm, for traffic enforcement (policing). In this paper, we present a novel approach to characterize and model the multimedia traffic using neural networks (NNs). A backpropagation neural network is used to characterize and predict the statistical variations of the packet arrival process resulting from the superposition of N packetized video sources and M packetized voice sources. The accuracy of the results were verified by matching the index of dispersion for counts (IDC), the variance, and the autocorrelation of the arrival process to those of the NN output. The reported results show that the NNs can be successfully utilized to characterize the complex non-renewal process with extreme accuracy.<<ETX>>

Reinforcement learning-based neural network congestion controller for ATM networks

This paper presents a new approach to the problem of congestion control arising at the user-to-network interface (UNI) of the ATM-based broadband integrated services digital networks (BISDN). Our approach employs an adaptive rate-based feedback control algorithm using reinforcement learning neural networks (NNs). The reinforcement learning NN controller provides an adaptive optimal control solution. This is achieved via the formulation of a performance measure function (cost function) that is used to, adaptively, tune the weights of the NN. The cost function is defined in terms of two main objectives: (1) to minimize the cell loss rate (CLR), i.e., control congestion and (2) to preserve the quality of the voice/video traffic via maintaining the original coding rate of the multimedia sources. The results show that the NN control system is adaptive in the sense that it is applicable to any type of multimedia traffic. Also, the control signal is optimal in the sense that it maximizes the performance of the system which is defined in terms of its performance measure function. Hence, our novel approach is very effective in controlling congestion of the multimedia traffic in ATM networks.

A novel neural network traffic descriptor for ATM networks

Accurate characterization of the multimedia traffic is essential, in asynchronous transfer mode (ATM) broadband networks, in order to develop a robust set of traffic descriptors. Such a set is required, by the usage parameter control (UPC) algorithm, for traffic enforcement (policing). The authors present a novel approach to characterize and model the multimedia traffic using neural networks (NNs). A backpropagation NN is used to characterize and predict the statistical variations of the packet arrival process resulting from the superposition of N packetized video sources and M packetized voice sources. The accuracy of the results is verified by matching the index of dispersion for counts (IDC), the variance, and the autocorrelation of the arrival process to those of the NN output. The results show that the NNs can be successfully utilized to characterize the complex nonrenewal process with extreme accuracy.<<ETX>>