Keynote Speech

Research and Development Test-Beds for Future Networks

Erol Gelenbe
Dennis Gabor Chair
Head of Intelligent Systems and Networks
Dept. of Electrical and Electronic Engineering
Imperial College
London SW7 2BT
e.gelenbe@imperial.ac.uk

Networking research is entering a new phase in which many of the taboos of the last twenty years are being questioned. Industry is introducing and proposing a wide range of new communication technologies, many of which are based on the wireless and mobile revolution. Governments and funding agencies (e.g. EU, DARPA) seek new paradigms that better exploit these opportunities and respond to the needs for ubiquitous networking in both the civilian and defence sphere; they expect an improved exploitation of the dramatically increased potential for intelligence within networks. Users are now taking better advantage of the convergence between the Internet and the telecommunications industry. Finally, the research community is freeing itself from the intellectual constraints often imposed by standards bodies, by government and industry, so as to fully exploit the convergence of computing and communications. In this new and liberated environment, radical ideas are emerging, such as cognitive radio, and cognitive and autonomic networks. As new ideas emerge, and a broad debate is engaged, it becomes increasingly important that research ideas be illustrated by experimental implementations and test-beds. The test-beds that are used will need to converge to common frameworks that are easily accessible to larger research and industrial communities. In the recent past, experimental networking research has seldom been placed in a framework where experiments are reproducible. Contrary to common practice in the experimental sciences, publications describing network measurements are seldom being reproduced or validated by independent experimental evidence. Our presentation will suggest some principles that can be followed to provide the networking research community with a common set of test-beds and tools, including compatibility issues, open access by the user community, full understanding of point to point communications and channel characteristics, workload characterisation, and the possibility of interfacing test-beds directly with systems running the IP protocol. The discussion will be illustrated by our experience, both positive and negative, with wired and wireless test-beds over the last five years at the University of Central Florida and at Imperial College [http://san.ee.ic.ac.uk].

References

1] E. Gelenbe, "Cognitive Packet Network", U.S. Patent 6,804,201, October 11, 2004.
[2] E. Gelenbe, M. Gellman, R. Lent, P. Liu, Pu Su, "Autonomous smart routing for network QoS", Proc. First International Conference on Autonomic Computing (IEEE Computer Society), ISBN 0-7695-2114-2, 232-239, May 17-18, 2004, New York.
[3] E. Gelenbe, R. Lent, "Power aware ad hoc Cognitive Packet Networks", Ad Hoc Networks Journal, 2 (3), 205-216, 2004 (ISN: 1570-8705).
[4] E. Gelenbe, "Sensible decisions based on QoS", Computational Management Science, 1 (1), 1-14, 2004.
[5] E. Gelenbe, R. Lent, A. Nunez, "Self-aware networks and quality of service", Proc. IEEE, 92 (9), 1479-1490, 2004.
[6] E. Gelenbe, "Review of experiments in self-aware networks", Invited Paper, 18th International Symposium on Computer and Information Sciences (ISCIS2003), LNCS 2869, 1-8, Springer, 2003.
[7] E. Gelenbe, A. Nunez, "Adaptive Web Service for QoS Improvement", IADIS International WWW/Internet 2003 Conference, Algarve, Portugal, Nov. 5-8, 2003.
[8] E. Gelenbe, R. Lent, Z. Xu, "Design and performance of cognitive packet networks", Performance Evaluation, 46, 155-176, 2001.
[9] E. Gelenbe, R. Lent, Z. Xu, "Cognitive Packet Networks: QoS and Performance", Opening Keynote, IEEE Computer Society MASCOTS02 Symposium, ISBN 0-7695-1840-0, 3-12, Fort Worth, TX, Oct. 14, 2002.
[10] E. Gelenbe, K. Hussain, "Learning in the multiple class random neural network", IEEE Trans. on Neural Networks, 13 (6), 1257-1267, 2002.
[11] E. Gelenbe, Z.-H. Mao, Yan-Da Li, "Function approximation by random neural networks with a bounded number of layers", J. Differential Equations and Dynamical Systems, 12 (1&2),143-170, 2004.

Bio:

Erol Gelenbe (FIEEE, FACM, FIEE) is a graduate of the Middle East Technical University (Ankara, Turkey). He received his MS and PhD degrees from Brooklyn Poly (New York) and the DSc degree from Université Pierre et Marie Curie. Over the last four years he was awarded two US patents for network protocol design, and one for neural network algorithms. In the 1970's he pioneered the mathematical analysis of communication protocols, and he is known for contributions to the performance evaluation of computer systems and networks. His four authored books have appeared in French, English, Japanese and Korean, and he has published over 110 articles in the leading journals of computer science and communications, including JACM, CACM, IEEE Trans. on Comms., IEEE Comms. Letters, IEEE Trans. on Neural Networks, Neural Computation, IEEE Trans. on Computers, J. Applied Probability, IEEE J. Sel. Areas in Comms., SIAM J. Computing, Acta Informatica, Performance Evaluation, etc. His many successful former PhD students, including Guy Pujolle, Francois Baccelli, Jacques Labetoulle, Catherine Rosenberg, Andreas Stafylopatis, Zhen Liu, Salvatore Tucci, Brigitte Plateau, Jean-Michel Fourneau, Alain Jean-Marie, Xiaowen Mang, Vijay Srinivasan, Alain Kurinckx, Gérard Hébuterne, Anoop Ghanwani, Philippe Nain, Ali Labed, Georges Hébrail and others in industry and academia. Erol serves on the Editorial Board of several international and European journals including Acta Informatica, The Computer Journal (BCS), Performance Evaluation, Annales des Télécommunications, Telecommunication Systems, and Recherche Opérationnelle. He is a past Associate Editor of the IEEE Trans. on Software Engineering and the Proceedings of the IEEE. He has served as the President of the French Chapter of the IEEE Computer Society, as programme chair and committee member of many IEEE, ACM and IFIP conferences, and in academic administrative and leadership positions in France and the US. His technical awards include the Grand Prix France Telecom of the French Academy and two honorary doctorates. The French government has decorated him with the Chevalier and Officier of the Ordre National du Mérite and Chevalier des Palmes Académiques. Over the years he has been a consultant to, and collaborated with, many companies including IBM, Bull, Lucent Technologies, France Telecom, and General Dynamics. In the US from 1993 to 2003 he was funded by DoD, NSF, IBM, Lucent Technologies and other companies. His current work is funded by EPSRC, MoD, EU FP6 and the US Army and he collaborates with BT, QinetiQ and General Dynamics UK.