TKK | Tietoverkkolaboratorio | Tutkimus

Dr.Tech. Kalevi Kilkki





This short presentation tries to outline the contents and my goals as a acting professor (1995-1996) in a new professorship in Laboratory of Telecommunications Technology. The area of education of the professorship is "Broadband integrated services networks and their traffic theory". This specification combines three different aspects of telecommunications: technology, services and traffic theory. The main topic connecting all these aspects is the development of traffic management systems for ATM (Asynchronous Transfer Mode) networks.

Three viewpoints into the area

The traffic management of telecommunication networks has several dimensions which are necessary for understanding the whole complex of problems. One distinctive feature between the dimensions is the society which is used as a viewpoint. The most essential societies are universities and research institutes, the manufactures of switching and management systems, the operators of telecommunication networks, and finally the users of services. The research and education in a technical university should cover all these viewpoints, although the interest and competence of the each research group may lead to emphasize some aspects on the expense of the others.

The way of outlining the problem area and the method of analysis depend strongly on the society. For most users the traffic management is only a tool which makes possible to create telecommunication services which meet the user needs. The following grouping is practical when assessing the services on the user point of view:

  • data services, such as LAN interconnection,

  • video services, such as video conference and video on demand,

  • voice over ATM,

  • mobile and wireless ATM, and

  • multimedia and internet applications

    These services and applications, and possibly some others which are still out of our range of vision, form a diverse set of requirements for any broadband network. To satisfy these requirements several network services have been introduced. A reasonable selection and an adequate definition of these services are of great importance for the operators of broadband networks. Although the standardization of the services, such as constant bit rate (CBR), variable bit rate (VBR), available bit rate (ABR) and unspecified bit rate (UBR), is in progress, there still are many open issues which need both theoretical research work and practical assessment of different aspects.

    Because of the diversity of services the control and management systems of broadband networks tend to become very complicated. Therefore, in order to make the standardization and implementation of the services feasible control system should be build of relatively independent elements or functions. A typical list of issues connected with the traffic management is as follows:

  • Connection admission control

  • Usage parameter control

  • Routing functions

  • Reactive load control functions

  • Service categories and priorities

  • Network dimensioning

  • Traffic measurement and monitoring

  • Charging and accounting

  • Switching structures and their performance

    This type of viewpoint is essential for every manufacturer developing switching nodes and other components for broadband networks.

    Unfortunately, it is not sufficient to comprehend and define different services and functions separately, in addition we should profoundly understand the interdependencies between different aspects. For example, the ABR service is a complicated method by which the operator can offer an advanced service which meets user requirements in case of data transport. All the aspects, related to users, operators and manufacturers, should be covered in the developing processes of the services, the switching nodes and the management systems. The challenge is how we can integrate all the solutions for different problems in a way that a feasible and robust total management system can be constructed.

    Objectives of research projects

    In addition to the above short discussion of prevalent technical problems we should bear in mind the goal of research work. Especially in a technologically oriented society, there seems to be strong tendency to present the objectives and assess the results numerically. We can collect data concerning the number of master and doctoral thesis, publications, etc. Although this effort could be useful, there remain profound problems with the selection of measures, in the weighting and combination of different aspects, and measuring things like quality and importance. Therefore, I believe that we should express the ultimate goal of the research and education by using very general concepts like quality, significance and usefulness. All these terms should be applied in different context, such as scientific society, the benefit for users or the economy of state and environment. These general goals should be expressed adequately taking into account the special problems of each situation.

    On the one hand, the research work of the professorship should support practical development efforts both on the manufacturer and the network operator side. To attain this objective we need a comprehensive view into the problem area, as presented in the technical discussion above, and close cooperation with other organizations in the telecommunication area.

    On the other hand, the research work should also be a creative process. We get the best opportunity to invent new ideas by connecting knowledge from separate domains. For example, neural networks and genetic algorithms have proved to be useful approaches in many technical problems; they may even form the basis of the whole traffic management system of ATM networks.

    Another method to stimulate the innovation process is to put the goal far enough in the future. This process calls for extensive surveying of the research problems so that the amount of preliminary ideas is sufficient for a thorough selection. One example of an interesting problem area is full-optical telecommunication networks based on wavelength multiplexing and the dimensioning and traffic control of these networks. The combination of broadband and radio technologies may also generate a lot of new questions for traffic theory.

    Research methods

    In the previous chapters we have sketched the problems and goals of traffic management in broadband networks. The next question is which kind of research approach we should apply in order to obtain as valuable results as possible taking into account the resources of the professorship. For answering to this question we need an additional viewpoint which determines the objectives and ways of research work. In this respect the following grouping of approaches could be applied:

    1. Theoretical research, such as queuing theory

    2. The development and evaluation of traffic models for practical purposes

    3. The determination of real source parameters by traffic measurements

    4. The development of simulation and analyzing software

    5. The invention and development of new control methods

    6. Hardware development

    The area of professorship gives a chance of taking any of the above approaches. Actually, we must choose our objectives in a way that the human and financial resources will be utilized as efficiently as possible. At the end of the last year Telecommunication laboratory got funding for two new research projects connected with the professorship: MITTA (funded by Suomen Akatemia) and FASTER (Tekes). The focuses in these projects are in items 2 and 3. In addition, it could be possible to use the general resources to develop simulation and analyzing tools (item 4). In consequence, these three items will likely be the primary scopes of research activities in the beginning of the professorship. However, this assumption does not restrict the application of the items 1, 5 and 6 later.



    Abstracts of following papers are stored with other papers from this project.

    Full papers are either in PostScript or PDF form.


    • Doctoral Thesis: Traffic Charactersation and Connection Admission Control in ATM Networks, Espoo 1994.


    • Marko Luoma, Kalevi Kilkki, Hanna Karjalainen, Kauko Rahko: Experiences with ATM Equipment and Network Operation
      This paper was presented in Nordic Teletraffic Seminar in Trondheim, August 20-22, 1996.

    • Kalevi Kilkki: Traffic Management Tools for ATM Networks With Real-Time and Non-Real-Time Services
      This paper was presented in ITC Spesialists Seminar on CONTROL IN COMMUNICATIONS in Lund, September 17-19, 1996.

      This paper was presented in IEEE Africon '96, September 25-27, 1996.

    • J.Kankare, K.Kilkki: Recursive Estimators for Traffic Analysis and Prediction
      This paper was presented in International Conference on Telecommunications '97 held in Melbourne, Australia, in April 1996).


    • Mika Ilvesmäki, Kalevi Kilkki & Marko Luoma: Packets or ports: The decisions of IP switching.
      To be presented in Voice, Video & Data Communications '97 in Dallas, Texas, USA.



    Tietoverkkolaboratorio on nyt osa Tietoliikenne- ja tietoverkkotekniikan laitosta. Tällä sivulla oleva tieto voi olla vanhentunutta.

    Tämän sivun sisällöstä vastaavat ja Webmaster.
    Sivua on viimeksi päivitetty 03.09.1997 17:00.
    [ TKK > Sähkö- ja tietoliikennetekniikan osasto > Tietoverkkolaboratorio > Tutkimus ]