Since ATM was identified by the CCITT in 1988 as the target transfer mode for broadband communications, there has been considerable research activity on the topic world-wide. Within Europe, the RACE programme of the European Community has brought together experts from a wide variety of organisations to work on several projects. This book results from the work of one of those projects. Aimed at those interested in ATM generally, or those needing to understand the issues in designing or implementing broadband networks, the book draws on the results of the research project to present a description of ATM from a network point of view. Starting with the principles of ATM, it goes on to cover topics such as network performance, network structure, evolution and interworking. It also discusses more general issues including numbering, charging and the need for intelligence in the network. It concludes by explaining the current position on traffic engineering for broadband ATM networks.
Inspec keywords: internetworking; asynchronous transfer mode; telecommunication network routing; telecommunication traffic
Other keywords: routeing technique; traffic control; resource management; interworking; traffic engineering; asynchronous transfer mode
Subjects: Communication switching; Communication network design, planning and routing; Computer communications; Switching centres and equipment; Computer networks and techniques
This chapter describes and explains the main features of a broadband asynchronous transfer mode (ATM) network, from the network point of view. It does not describe in detail the various techniques for switching ATM cells, particularly as many of the switching architectures are proprietary.
ATM has been chosen as the target transfer mode for B-ISDN, since it offers a flexible transfer capability common to all services because of its independence of the bit-rate and data structure of the services carried. For these reasons, additional functionalities are needed to accommodate the various services and these are added at the edge of the ATM network.
Although this book is mainly concerned with the network aspects of ATM, some regard must be paid to the applications that can make use of a broadband communications infrastructure. It cannot be emphasised too strongly that the customer will pay for services only if these services support the required application and are seen to be priced at a level that makes their use worthwhile.
The main purpose of the ATM network is to provide ATM connectivity, that is an end-to-end ATM link for the transport of the user's information in ATM cells. To achieve this requires switching and transmission, together with the necessary control and management. This chapter considers the functions for switching, transmission and signalling (control).
There are several performance measures that are specific for ATM, such as cell transfer delay and delay variation, cell-loss probability and cell-insertion rate. Cell-delay variation and cell loss arise from queuing and buffer overflow, respectively. They occur because of the characteristics and statistical variation of ATM traffic; they are therefore important measures of performance in ATM traffic engineering. ATM-layer phenomena depend on both ATM-layer and physical-layer performance. Physical-layer effects are not a subject of ATM traffic engineering per se. However, because random bit errors can cause cell loss (for example), it is important to be able to distinguish the reasons for cell losses when this phenomenon is measured on a real system. Thus, it is instructive to address the causes and effects of network performance in layers. The physical-layer performance is independent of the performance of the ATM layer; however, the latter is affected by the former. The performance of both layers will affect the performance of the AAL. Aspects of all three layers have an effect on the quality of service experienced by the user. This interdependence of cause and effect between layers is illustrated in Figure 5.1.
It is possible to introduce ATM first into the public network or into the private network and indeed there are already ATM products (such as ATM switches and interface cards for workstations) available for the customer's network. Private network use of ATM may well be the first use of the technology, but the public network must evolve in parallel to provide the wider-area connectivity that customers will require.
Interworking functions may be implemented in the broadband network, in the customer's equipment, in other types of networks, or in some combination of these. The three aspects of interworking that have to be considered are network interworking, service interworking and numbering and addressing.
Numbering and charging are two factors that are immediately obvious to the customer and charging in particular can have a significant impact on the take-up of a new service. Historically, in telephone networks, there has been a link between the two: different numbering sequences clearly identify local, national and international destinations, with the consequential differences in tariff. Recently, new, non-geographical codes have been introduced for services such as mobile, premium-rate and free calls thus breaking the obvious link between code and tariff. However, there has been some customer confusion with these new codes, illustrating the importance of numbering to the customer as well as to the network operator.
Network structures form the basis of a routeing plan. Different customer groups can be linked by different types of route, choosing between combinations of hierarchies, mesh or star structures, direct, alternative and transit routes. The average and maximum numbers of transmission links between any pair of customers are important parameters in the trade-off between optimum technical performance and economic efficiency.
The primary role of traffic control and resource management procedures is to protect the network so that it can achieve the required network performance objectives. The uncertainties of broadband traffic patterns and the complexity of resource management suggest a step-wise approach for defining these parameters and procedures. An initial set of traffic-control and resource-management capabilities is currently defined by the CCITT in Recommendation 1.371. Further sets of capabilities may subsequently be defined to achieve increased network efficiency. The following four functions are considered: network resource management, connection admission control, usage parameter control and network parameter control. Their location in an ATM network is also discussed.
This chapter discusses intelligence in the network. Network management is the act of controlling, supervising and maintaining communication networks with the ultimate aim of maximising the profit for the network operator, while providing the best possible services and quality of service for the customer.
This chapter treats the traffic engineering tasks of (1) traffic modelling, (2) performance measures, (3) performance evaluation and (4) network dimensioning in turn. The specific ATM traffic issues associated with traffic control and resource management have been dealt with earlier.
This Appendix briefly outlines the scope of current CCITT recommendations dealing with B-ISDN.