COM² - Computational Methods for the Performance Analysis of Broadband Communication Networks

Note! This project has already ended (31.12.2001).

1. General Information

The COM² is a 3 year project (1998-2001) funded by the Academy of Finland. It is a part of the Research Programme for Telecommunication Electronics (TELECTRONICS) , which was initiated by the Academy of Finland in 1997 to further advance scientific research in the field of tele- and data communications.

For results of the project see the following:

2. Background

Telecommunications industry has become one of the key growth factors in the Finnish economy. Advanced communications also provide the basic infrastructure which enhances the growth of other industries as well as the development of the society as a whole. Three technological trends will give new opportunities for continuous growth in the telecommunications industry:

The emerging broadband networks
The development of the Internet technology and applications
The growth of mobile communications and the introduction of new mobile systems and services

There are many technological challenges in these areas. One of them is an appropriate dimensioning of the systems and networks along with adequate traffic management methods. The importance of these problems has become evident e.g. in the history of the ATM, where traffical issues, such as provisioning adequate quality of service to different users in a heterogeneous environment, have turned out to be a much harder problem than originally anticipated. Also the congestion problems of the Internet, experienced by most users, exemplify the reality of the problem. The problem will become even more pronounced with the introduction of various real-time services, such as voice over Internet. Finally, in the area of mobile communications, the scarcity of the available bandwidth makes it important to utilise the resources to the full extent, which is only possible with adequate dimensioning methods and design procedures.

The solution to the problems of dimensioning and traffic management relies on the performance analysis of telecommunications networks. Performance analysis comprises modelling the system and the traffic offered, and solving the model by means of various mathematical methods. Despite of the great advances in the area of performance analysis of multiservice networks over the past ten years, many problems remain to be solved.

3. Goals and research areas

Goals

The main goal of the project is to develop mathematical and computational methods for the performance analysis of broadband multiservice networks. It can be divided into four subgoals:

to further the development of tools needed by the industry
to educate specialists at different levels (M.Sc., Tech.Lic., Tech.Dr.) in the field
to produce results of high scientific standard
to strengthen the collaboration on the international level

Research areas

The research will be directed to the following three main subfields:

Computational Methods for the Solution of Problems in Scheduling, Dynamic Resource Allocation, Routing and Network Optimization

The development of computational methods for the solution of queueing models pertinent to telecommunication networks, e.g. those of advanced scheduling mechanisms in switches and routers. Similarly, computational methods for the solution of problems in optimal routing, optimal resource allocation and network optimisation will be developed. Special attention will be paid on the robustness of the methods in order to make them as independent of specific modelling assumptions as possible. The result of this work will be in the form of mathematical methods as well as efficient computational algorithms. The potential of extensive use of symbolic computation in producing the results will also be explored.

Simulation Methods

The multiservice networks pose hard problems for the simulation approach because of the huge dimensionality of the system and/or the rarity of the interesting events (e.g. cell loss in an ATM network). The work on new simulation techniques is currently pursued in a companion project COST 257 and preliminary work on a particularly promising approach, so called Markov Chain Monte Carlo method with Gibbs sampling has been started. In the future, this work will be done within this project.

Traffic modelling

In particular, the modelling of the data traffic in the Internet is a central issue. Long-range dependent traffic models have become a hot topic in the area. The pioneering work by Prof. Ilkka Norros at VTT on the use of fractional Brownian motion as a model provides basic tools. The project will address important questions still requiring an adequate solution, such as the estimation of the parameters of the model on the basis of measured traffic. Also the use of the model to practical problems such as the cache design will be studied. As a spin-off application the model may be utilised in the area of image compression.

5. Persons

Attila Vidács, 1.9.98 - 31.5.99
Jani Lakkakorpi, 1.9.98 - 30.4.99
Pasi Lassila, 1.1.99 - 31.12.01
Vinod Sharma (visiting professor with Tekes funding), 1.9.98 - 30.6.99
Hannu Rummukainen, 1.6.99 - 30.4.00
Aleksi Penttinen, 1.6.99 - 15.8.99, 5.6.00-31.7.01
Jan Hlinovsky, 1.6.99 - 31.8.99, 10.6.00 - 31.12.00
Pirkko Kuusela, 18.10.99 - 31.07.01
Laura Lignell, 1.2.00 - 31.8.00
Elena Siren, 5.6.00 - 31.8.00
Jouni Karvo, 1.8.00-31.7.01

6. Publications and Reports

Note! Those publications for which the copyright has been transferred to the publisher may not be available on-line. If you need a copy of such a paper, simply contact via e-mail the first author of the paper (firstname.lastname@hut.fi).

All publications of the teletraffic theory group can be found through our on-line search engine

Journal Papers

A. Penttinen and J. Virtamo, Simulation of Two-Dimensional Fractional Gaussian Noise, Methodology and Computing in Applied Probability, vol. 6, no. 1, pp. 99-107, 2004 (link)(bib)(abstract)
V. Sharma and J. Virtamo, A finite buffer queue with priorities, Performance Evaluation, vol. 47, no. 1, pp. 1-22, 2002 (bib)
V. Sharma, J. Virtamo and P. Lassila, Performance Analysis of the Random Early Detection Algorithm, Probability in the Engineering and Informational Sciences, vol. 16, no. 3, pp. 367-388, 2002 (bib)
J. Karvo, O. Martikainen, J. Virtamo and S. Aalto, Blocking of dynamic multicast connections, Telecommunication Systems, vol. 16, 3,4, pp. 467-481, 2001 (link)(bib)(abstract)
H. Rummukainen and J. Virtamo, Polynomial Cost Approximations in Markov Decision Theory Based Least Cost Routing, IEEE Transactions on Networking, vol. 9, no. 6, pp. 769-779, 2001 (link)(bib)
P. Lassila and J. Virtamo, Nearly Optimal Importance Sampling for Monte Carlo Simulation of Loss Systems, ACM Transactions on Modeling and Computer Simulation, vol. 10, no. 4, pp. 326-347, 2000 (link)(bib)
S. Aalto, Optimal control of batch service queues with finite capacity and linear holding costs, Mathematical Methods of Operations Research, vol. 51, no. 2, pp. 263-285, 2000 (link)(bib)(abstract)
S. Aalto and W. Scheinhardt, Tandem fluid queues fed by homogeneous on-off sources, Operations Research Letters, vol. 27, no. 2, pp. 73-82, 2000 (link)(bib)(abstract)
I. Norros, E. Valkeila and J. Virtamo, An Elementary Approach to a Girsanov Formula and Other Analytical Results on Fractional Brownian Motion, Bernoulli, vol. 5, no. 4, pp. 571-587, 1999 (bib)

Conference Papers

A. Penttinen and J. Virtamo, A packet marking algorithm for congestion pricing, in Proceedings of the 2002 International Symposium on Performance Evaluation of Computer and Telecommunication Systems (SPECTS 2002), pp. 379-385, 2002, San Diego, California, USA (bib)(abstract)
J. Karvo, Efficient simulation of blocking probabilities for multi-layer multicast streams, in Proceedings of Networking 2002, pp. 1020-1031, Springer-Verlag, 2002, Pisa, Italy (bib)(abstract)
J. Karvo, S. Aalto and J. Virtamo, Blocking probabilities of multi-layer multicast streams, in Proc. HPSR 2002, pp. 268-277, 2002, Kobe, Japan (bib)(abstract)
S. Aalto, J. Karvo and J. Virtamo, Calculating blocking probabilities in multicast loss systems, in Proc. SPECTS 2002, pp. 833-842, 2002, San Diego, CA (bib)(abstract)
P. Kuusela, P. Lassila, J. Virtamo and P. Key, Modeling RED with Idealized TCP Sources, in Proceedings of IFIP ATM & IP 2001, pp. 155-166, 2001, Budapest, Hungary (bib)
P. Lassila, J. Karvo and J. Virtamo, Efficient Importance Sampling for Monte Carlo Simulation of Multicast Networks, in Proceedings of IEEE INFOCOM, pp. 432-439, 2001, Anchorage, Alaska (bib)
P. Kuusela, P. Lassila and J. Virtamo, Stability of TCP-RED Congestion Control, in Proceedings of ITC-17, pp. 655-666, Elsevier, 2001, Salvador da Bahia, Brazil (link)(bib)
A. Penttinen and J. Virtamo, Robust Adaptive Capacity Allocation Algorithm, in Teletraffic Engineering in the Internet Era: Proceedings of the International Teletraffic Congress - ITC-17, pp. 223-231, North Holland, 2001, Salvador da Bahia, Brazil (bib)(abstract)
J. Karvo, S. Aalto and J. Virtamo, Blocking probabilities of two-layer statistically indistinguishable multicast streams, in Proceedings of ITC-17, pp. 769-779, Elsevier, 2001, Salvador da Bahia, Brazil (link)(bib)(abstract)
A. Vidács and J. Virtamo, Parameter estimation of geometrically sampled fractional Brownian traffic, in Proceedings of IEEE Infocom, pp. 1791-1796, , 2000, Tel Aviv, Israel (bib)
P. Lassila and J. Virtamo, Modeling the dynamics of the RED algorithm, in Proceedings of Quality of Future Internet Services (QofIS), pp. 28-42, Sringer-Verlag, 2000, Berlin, Germany (link)(bib)
P. Kuusela and J. Virtamo, Modeling RED with Two Traffic Classes, in Proceedings of the 15th Nordic Teletraffic Seminar (NTS-15), pp. 271-282, 2000, Lund, Sweden (bib)
S. Aalto and J. Virtamo, Combinatorial algorithm for calculating blocking probabilities in multicast networks, in Proceedings of the 15th Nordic Teletraffic Seminar, pp. 23-34, 2000, Lund, Sweden (bib)(abstract)
P. Lassila and J. Virtamo, Inverse convolution approach to importance sampling in Monte Carlo simulation of loss systems, in Proceedings of the 15th Nordic Teletraffic Seminar, pp. 161-172, 2000, Lund, Sweden (bib)
V. Sharma and J. Virtamo, A finite buffer queue, in Proceedings of Globecom, pp. 1053-1065, 1999, Rio de Janeiro, Brazil (link)(bib)
A. Vidács and J. Virtamo, ML Estimation of the Parameters of FBM Traffic with Geometrical Sampling, in Proceedings of Broadband Communications'99 - Convergence of Network Technologies, pp. 51-62, Kluwer, 1999 (bib)
P. Lassila and J. Virtamo, Efficient Importance Sampling for Monte Carlo Simulation of Loss Systems, in Proceedings of ITC-16, pp. 787-796, Elsevier, 1999, Edinburgh, England (link)(bib)
J. Karvo and S. Aalto, Average Signalling Load for Multicast Group Management, in Proceedings of ITC-16, pp. 509-518, Elsevier, 1999, Edinburgh, UK (link)(bib)(abstract)
J. Karvo, J. Virtamo, S. Aalto and O. Martikainen, Blocking of dynamic multicast connections, in the 4th INFORMS Telecommunications Conference, 1998, Boca Raton, Florida, published later in Selected Proceedings in Telecommunication Systems (bib)(abstract)

Theses

A. Penttinen, Mathematical models for marking in congestion pricing, Helsinki University of Technology, 2001, Master's Thesis (bib)
P. Lassila, Methods for network performance evaluation: fast simulation of loss systems and analysis of Internet congestion control, Ph.D. thesis, Laboratory Report Series, Networking Laboratory, 3/2001, 2001 (link)(bib)
H. Rummukainen, On approximative Markov control of multiservice telecommunication links, M.Sc. thesis, Laboratory Report Series, Networking Laboratory, 2/2000, 2000 (bib)

COM2 - Computational Methods for the Performance Analysis of Broadband Communication Networks