The authors survey a number of important research topics in ATM (asynchronous transfer mode) networks. The topics covered include mathematical modeling of various types of traffic sources, congestion-control and error-control schemes for ATM networks, and priority schemes to support multiple classes of traffic. Standard activity for ATM networks and future research problems in ATM are also presented. It is shown that the cell-arrival process for data sources can be modeled by a simple Poisson process. However, voice sources or video sources require more complex processes because of the correlation among cell arrivals. Due to the effects of high-speed channels, preventive control is more effective in ATM networks than reactive control. Due to the use of optical fibers in ATM networks, the channel error rate is very small. The effects of propagation delay and processing time become significant in such high-speed networks. These fundamental changes trigger the necessity to reexamine the error-control schemes used in existing networks. Due to the diversity of service and performance requirements, the notion of multiple traffic classes is required, and separate control mechanisms should be used according to the traffic classes. The priority scheme is shown to be an effective method to support multiple classes of traffic. >

Survey of traffic control schemes and protocols in ATM networks

This paper proposes a neural fuzzy approach for connection admission control (CAC) with QoS guarantee in multimedia high-speed networks. Fuzzy logic systems have been successfully applied to deal with traffic-control-related problems and have provided a robust mathematical framework for dealing with real-world imprecision. However, there is no clear and general technique to map domain knowledge on traffic control onto the parameters of a fuzzy logic system. Neural networks have learning and adaptive capabilities that can be used to construct intelligent computational algorithms for traffic control. However, the knowledge embodied in conventional methods is difficult to incorporate into the design of neural networks. The proposed neural fuzzy connection admission control (NFCAC) scheme is an integrated method that combines the linguistic control capabilities of a fuzzy logic controller and the learning abilities of a neural network. It is an intelligent implementation so that it can provide a robust framework to mimic experts' knowledge embodied in existing traffic control techniques and can construct efficient computational algorithms for traffic control. We properly choose input variables and design the rule structure for the NFCAC controller so that it can have robust operation even under dynamic environments. Simulation results show that compared with a conventional effective-bandwidth-based CAC, a fuzzy-logic-based CAC, and a neural-net-based CAC, the proposed NFCAC can achieve superior system utilization, high learning speed, and simple design procedure, while keeping the QoS contract.

https://ir.nctu.edu.tw:443/bitstream/11536/31526/1/000079621900010.pdf

A QoS-Provisioning neural fuzzy connection admission controller for multimedia high-speed networks

We solve the adaptive call admission control (CAC) problem in multimedia networks via reinforcement learning (RL). The problem requires that network revenue be maximized while simultaneously meeting quality of service (QoS) constraints that forbid entry into certain states and use of certain actions. We show that RL provides a solution to this constrained semi-Markov decision problem and is able to earn significantly higher revenues than alternative heuristics. Unlike other model-based algorithms, RL does not require the explicit state transition models to solve the decision problems. This feature is very important if one considers large integrated service networks supporting a number of different service types, where the number of states is so large that model-based optimization algorithms are infeasible. Both packet-level and call-level QoS constraints are addressed, and both conservative and aggressive approaches to the QoS constraints are considered. Results are demonstrated on a single link and extended to routing on a multilink network.

Adaptive call admission control under quality of service constraints: a reinforcement learning solution

A general overview of resource management aspects of ATM networks is presented, and mechanisms of load control have been considered. A general framework of resource management is first addressed. Two main aspects of traffic characterization are then examined from the traffic engineering point of view: teleservice characterization and user characterization. These form the basis of the analysis that follows. In particular, a bandwidth assignment proposal is presented and evaluated by simulation. Policing issues are discussed. The advantages of the proposed assigned method over the peak bit rate allocation method is proved for two typical dimensioning cases. >

Resource management and dimensioning in ATM networks

It is demonstrated that dynamic routing control based on a genetic algorithm can provide flexible real-time management of the dynamic traffic changes in broadband networks. A string structure is proposed, each of whose elements represents paths between each pair of origin and destination terminal nodes, together with a new technique using the past solutions as the initial data for new searches. These techniques dramatically improve the efficiency and convergence speed of the genetic algorithm. Computer simulations show that the genetic algorithm using the proposed techniques can generate the exact solution of path arrangement and can find a routing arrangement that keeps the traffic loss-rate below a target value, even after changes in traffic. >

A dynamic routing control based on a genetic algorithm

A learning control method using neural networks for service quality control in the asynchronous transfer mode (ATM) network is described. An ATM network is a high-speed packet-switching network for the data transmission layer of B-ISDN (broadband integrated services digital network) which provides multimedia services, including voice, data and video. Service quality control is one of the most crucial issues in realizing a flexible ATM network. It is a challenging research task to build an efficient network controller that can control the network traffic even when the precise characteristics of the source traffic are not known and the service quality requirements change over time. The proposed ATM network controller is flexible in function and simple in implementation because neural networks using backpropagation learn the relations between the offered traffic and service qualities. A training data selection method called leaky pattern tables is proposed for learning the accurate relations. The performance of the proposed controller is evaluated by simulation of a basic call regulation model. >

Hiramatsu

Papers

ATM communications network control by neural network