TF-TEN Document Library

In Deliverable D11.1 a set of ATM experiments was specified as part of the TEN-34 Phase 1 Test Programme over the JAMES ATM network. This deliverable describes the interim results of these experiments. The main emphasis in this phase lies in examining the underlying technology for its suitability to support advanced applications. Some of the experiments concentrate on fine-tuning systems to maximise performance; others investigate the usability of new technologies and ATM traffic classes. The interim results of the experiments have shown that basic ATM services such as CBR are well understood and can be used in a production environment. In the area of the more advanced ATM services the interim results demonstrate that the services vary from being usable only under very idealistic circumstances (e.g., switching), to potentially providing a stable infrastructure, pending further investigation (e.g., VBR). The most important result of the experiments so far is that switching proves to be very difficult to use in the way it was created for, which is between end-user applications. There are no major problems with switching in a LAN environment, where bandwidth is close to unlimited. But over a public wide area network, with significant bottlenecks and policing, SVCs can only be used under severe limitations. Due to the number of diverse problems it is not expected to have SVCs available in a general production environment within the next two to three years.

In Deliverable D11.1 a set of ATM experiments was specified as part of the TEN-34 Phase 1 Test Programme over the JAMES ATM network. D11.2 reported on interim results, and this deliverable, D11.3, describes the final results of these experiments. The main emphasis in this phase lies in examining the underlying technology for its suitability to support advanced applications. Some of the experiments concentrate on fine-tuning systems to maximise performance; others investigate the usability of new technologies and ATM traffic classes. The results of the experiments have shown that basic ATM services such as CBR are well understood and can be used in a production environment. There is also a better understanding of VBR services and their suitability for IP traffic. The result of our research into more advanced services varied, and in most areas further research is clearly needed. There was significant progress with using signalling for setting up SVCs towards the end of the test phase, and despite the fact that the test environment was not stable enough to be used for production services, it was possible to find a solution to some of the principal problems outlined in D11.2. With the conclusion of the first phase of the experiments in the TEN-34 project we have proven the basic ATM services to work reliably for IP traffic. We also have gained greater understanding on how to use the more advanced features of ATM. Despite these positive results the overall conclusion of this phase is that advanced ATM features do not work reliably enough yet to be used on a production network.

This deliverable specifies the experiments that will be carried out in the TEN-34 Phase 2 Test Programme over the JAMES ATM network. In Phase 1 of the Test Programme (work package 11) the experiments focused mainly on the examination of various network technologies on their suitability to support advanced services. These experiments have shown that most advanced technologies such as SVCs are not yet stable enough to support operational services in a multi-vendor environment. In Phase 2 of the Test Programme the emphasis will be on providing additional services, albeit with initially limited functionality. Methods of providing such services will be investigated and trialled over the JAMES network. The tests will also include new traffic classes and technologies such as ABR and ATM routing. This deliverable specifies the experiments and activities planned for the Phase 2 Test Programme. A wide range of activities is specified for experimentation, but the intention is, based on initial experience, to focus on those technologies that might be applicable in the near future, and to provide details on how to use them in an operational environment.

Workpackage 11 of the TEN-34 carried out the first phase of the TEN-34 experimentation programme. Workpackage 14, the second phase of the test programme, deals with more complex experiments such as ATM routing, and advanced some of the experiments from phase 1 further. In deliverable D14.1 the experiments of this phase were defined and this deliverable, D14.2 describes the results. D14.3 will provide a summary of the complete TEN-34 testing programme. The experiments of this phase cover a wide range of wide area networking technologies such as ATM routing and signalling, and related issues like call admission control. Technologies to map IP services onto ATM were also discussed, such as label based switching and NHRP. As in phase 1 the experiments were carried out over the JAMES network, which will be described in more detail in this deliverable. The JAMES network provided a European ATM network facility, and the TF-TEN used an overlay network consisting of VPs on this network for most of their experiments. The results of this workpackage show that the basic technologies to use ATM services and high-level IP services are working in a test environment. However, the mechanisms to control and manage these advanced technologies are not fully developed yet, which makes the technologies unusuable in a production network. New ATM features such as point-to-multipoint SVCs are becoming available, but need more investigation. The integration of IP and ATM has made significant progress over the last year with the development of label based switching technology.

Michael Behringer, May 1997, compiled for JENC8, the 8th annual Joint European Networking Conference

Michael Behringer, June 1997, INET'97, the Annual Conference of the Internet Society, Kuala Lumpur, Malaysia

By T.Ferrari, A.Ghiselli, C.Vistoli; presented at INET98, July 1998