Fundamental contributions to Traffic Theory and pioneering work in the definition of protocols for packet-switched telecommunications networks

Contributions to Communication Traffic Theory

Professor Paul Kühn is, among the living traffic theorists, the most prominent person in Europe. It is mainly his hard, relentless work in this area that contributed to his international career and made him well-known worldwide. Due to his many outstanding scientific contributions, his name stands in line with those of other pioneers such as Erlang, Engset, Pollaczek, and Khintchine. It should be mentioned that these persons all died long before the establishment of the Eduard Rhein Foundation. They are, hence, not eligible for a posthumous ERF award. The Board of Curators and the Executive Board proudly confer the Basic Resarch Award upon a native of Germany who has been spending most of his career in our country. Unfortunately, we are now used to the fact that most top-notch researchers are based outside of Germany and Europe.

Now, what is Traffic Theory, and why is it significant enough to provide us with this year’s award-winner? Information and Communications Technology both draw heavily upon the laws of physics to initiate new research activities and to develop new technical systems. These systems have attained unbelievably high degrees of complexity. Their successful implementations are largely based on mastery of miniaturization of components for signal processing and data storage. Engineers design and develop systems that are made of these components relying on theoretical models. Most of these systems would never be realized without accurate modelIing approaches and some mathematical predictability of their behavior. All basic transmission and signal processing parameters of mobile phones were, for instance, accurately modelIed and optimized using methods known from System Theory. Other communications units of similar complexity as a mobile phone terminal are, e.g., base stations of mobile systems, copper-based or optical wireline systems, and switching nodes in telecommunications networks.

Needless to mention that any form of technical communication in its present form, i.e., one with guaranteed and predictable performance, would be impossible without model-based analyses and subsequent optimization of technical systems, parameters, and procedures. Examples of these highly complex tasks include the proper cooperation between multiple signal transmission systems and switching nodes of a communications network, the management of rivalries concerning transmission and switching capacity resources, and the design of safety measures to be taken against overloading of components. This special area of work may be termed “System Theory at large”. It is also known amongst experts as “Communication Traffic Theory”. Paul Kühn¹s fundamental contributions to the above area are concerned with stochastic modelling, performance analyses of system structures and operational modes of systems with multi-layered traffic-theoretical models, and with the analysis of life-cycle processes of complete systems or elements of decomposed systems.

He didn’t just explore and formally establish basic principles in an abstract manner; he also tailored his methods, wherever possible, to solutions of real-life problems that happen to occur in communications networks. Resulting from these endeavors are important contributions to the design and performance evaluation of signalIing protocols of computer and communication systems, congestion and overload control, traffic flow control, and to the design of algorithms enabling a network to support quality-of-service differentiation of traffic message flows. Traffic flows in communication networks consist of a mix of traffic classes some of which require real-time support while others have only less restrictive requirements, e.g., on throughput and its continuity, packet delay and its variance, probability of packet loss, etc. The switching nodes of a network are specialized process control computers with sophisticated operating systems. The creative work of a traffic theorist consists of creating accurate models that represent, as closely as possible, the real-life message flows in existing communication networks. Moreover, these models shall accurately describe the various interactions between all processes involved and predict the network’s traffic performance.

Appropriate computational models describe complex queuing and service systems; they allow for relative prioritization of traffic flows which compete for capacity with one another in traffic nodes and transmission channels. One may also infer parameters related to the correlation between these traffic flows at any point of the network or along a given transmission path. In the context of stationary and transient performance analysis, Paul Kühn proposed fundamentally new methods for the decomposition of large network problems into smaller ones. By solving these smaller problems separately and subsequently aggregating the individual results, exact (and approximate) solutions are obtained to the original problem.

His name is closely linked to the consolidation of traffic theoretical models that existed in the mid ’70s. These works mainly focused on systematic comparisons and significantly more powerful traffic models with variable load categories, modes of operation, and various optimization goals. His “Tables on Stochastic Queuing Systems” then, for more than a decade, became invaluable to many traffic engineers ­ mostly because of the very limited computer resources available at that time. Digital packet-switched broadband networks and, in particular, signalling protocols have always been at the center of Paul Kühn¹s work. Many of today¹s traffic theoretical parameters, models, and algorithms are rightfully attributable to him and his research group at the University of Stuttgart. Most notably, various contributions in the context of service quality control and traffic performance optimization in ATM (synchronous Transfer Mode, an ITU-T Standard for packet switching based on fixed sized packets, called cells) networks came from this group of researchers. It is worth mentioning that he always provided validations of his analytical results based on event-driven stochastic simulations which, in turn, led to broad acceptance of his work. Although standard conformant ATM networks have been mainly introduced in Europe only, the methods developed by Paul Kühn have proven to be fully applicable also to the analysis, optimization, and the further development of Internet protocols and functions towards a quality-of-service supporting network. This underlines the value of his fundamental contributions to packet-switched telecommunication networks that support realtime applications ­ one of the reasons why he was nominated for this award. Paul Kühn apparently still enjoys an unlimited physical and intellectual energy. Accordingly, his colleagues worldwide have continuously admired his over the decades never ending energy and his tenacity when following the ambitious goals set by himself. He is awarded today with the Eduard Rhein Prize, and we sincerely hope that his personal energy resources will continue to be devoted to the domain of Traffic Theory for many years to come.

Prof. Dr. Bernhard Walke,
RWTH Aachen