Bernd Heinrichs

Research: High performance transfer services to support multimedia group communications

This paper highlights the need for sophisticated multicast mechanisms to be provided by transfer protocols to support group communication. This need becomes even more evident in the light of the special QoS requirements imposed by multimedia applications. Todays protocols do not provide the broad range of functionality required by these upcoming applications. From several sample applications we derive functionality and performance requirements that have to be provided by the transfer systems. Taking XTP as an example, we show how some identified demands are met by this protocol. Furthermore, necessary enhancements are suggested. Usefulness of these enhancements is proved by simulation results. Our implementation strategy, based on a high modularization of the protocol software, is illustrated.

Adaptive QOS driven Communication Architecture

The RACE II project EuroBridge has introduced a concept for a flexible and adaptive Quality of Service (QOS) driven communication architecture. This paper will give an overview on the protocol structures within EuroBridge and will highlight the protocol selection mechanism which identifies suitable combinations of protocols in accordance to the requested QOS from the application. The main area will be the mapping from the demands onto the protocols that are available in the EuroBridge project. Furthermore we describe the integration of XTP into the platform and recommend a new distributed jitter control mechanism. The EuroBridge XTP implementation provides a modular design of the various protocol features. In accordance to the demands specified by the selector, only those parts are activated that are needed to fulfil the desired functionality.

Flexibles Design von Hochleistungsprotokollen mittels höherer Petri-Netze

Zur direkten Herleitung einer Protokollimplementierung aus einer formalen Spezifikation existieren derzeit kaum Verfahren, die eine fur Hochleistungsprotokolle adaquat leistungsstarke Implementierung erzeugen. In diesem Papier werden eine Implementierungssprache, die auf hoheren Petri-Netzen basiert und eine dedizierte parallele Controller-Architektur vorgestellt. Sie ermoglichen eine geradlinige Implementierung eines Hochleistungsprotokolles aus einer formalen Spezifikation. Da das Markenspiel von Petri-Netzen inharent parallele Ausfuhrungen aufweist, gewinnt man leicht parallele Implementierungen. Dieses Softwarekonzept wird beispielhaft am eXpress Transfer Protocol (XTP) vorgestellt. Insbesondere in einer Designphase, in der die Protokollarchitektur noch sehr variabel ist (wie bei XTP), aber auch in Hinblick auf die sich rasch andernden Anforderungen neuer Anwendungen sind Software-Ansatze reinen Hardware-Losungen vorzuziehen.

OSI communication services supporting CSCW applications

Mostcreatorsasui nmdersof these documents have access to highly reliable local ama networks (LAM), giving them access to a bandwidth of typically 10 Mbps. These LANsmay be interconnected locally via high-speed LANs/MANs (metropolitan area networks, transmission rates typically in the range of 100 Mbps), which in turn axe intemomected via WANs(widearea networks, transmission rates form 64 Kbps up to 2 Mbps) and will soonbeintmortnected viahigh-speed WAN (aseg.ATMnetworks), providingbandwidth in the range of, say, 1 Gbps.

Telefax-Kommunikation über Lokale Netze

In diesem Artikel wird die Integration des Gruppe 4 — Telefaxdienstes in Lokale Netze mit CSMAI CD-Medienzugang untersucht. Eine simulative Analyse liefert Leistungskenngrosen wie z.B. Verzogerungszeiten fur Nachrichten bzw. Pakete, Wartezeiten auf Zugang zum Netz und den mittleren Durchsatz je Telefax- bzw. Grundlaststation. Der Verkehr auf dem Netz wird durch saturierte Telefax Gruppe 4 — Lastgeneratoren und sogenannte „Grundlastgeneratoren“ generiert. Letztere werden zum einen durch den klassischen und zum anderen durch einen unterbrochenen Poisson-Prozes modelliert. Zusatzlich wird eine Sensitivitatsanalyse bzgl. der Intensitat der angebotenen Telefax- und Grundlast, sowie der eingesetzten Fluskontrollmechanismen durchgefuhrt.

A transfer systems for multi-media group communication

This paper highlights the need for sophisticated multicast mechanisms to be provided by transfer protocols in order to support group communication. This need becomes even more evident in the light of the special QOS requirements imposed by multi-media applications. Todays protocols do not provide the broad range of functionality required by these upcoming new applications. Taking XTP as an example, we show how some of the identified demands are met by this protocol. Furthermore, some severe shortcomings of XTP are identified and appropriate enhancements are suggested.<<ETX>>

Optimierung Von Transfersystemen

Die zunehmende Vielzahl neuer Applikationen motiviert aufgrund ihrer gehobenen Leistungsund Funktionalitatsanforderungen ein Redesign von Netzwerk-Protokollarchitekturen. Eine angeforderte Dienstgute (Quality of Service, QoS) soll nicht nur in lokalen homogenen Netzen, sondern auch uber Netzwerkgrenzen hinweg in einem globalen heterogenen Netz gewahrleistet werden. Eine Analyse der Syntax und Semantik existierender Transfersysteme (TCP/IP, OSI TP4/CLNP und XTP) hat ergeben, das diese nicht in der Lage sind, fur jede Anwendung die geeignete Protokollfunktionalitat zur Verfugung zu stellen. Es fehlen vor allem effiziente Protokollmechanismen zur Unterstutzung von Gruppenkommunikation und zur Vermeidung und zum Abbau von Uberlast. In diesem Artikel werden daher Ansatze zur Realisierung von Multi- casting basierend auf dem XTP-Bucket-Algorithmus sowie Erweiterungen und Untersuchungen eines auf Slow-Start/Tri-S basierenden Uberlastkontrollverfahrens vorgestellt.

The Eurobridge service platform (solution session): an approach to integrated multimedia

The Commission of the European Communities has established anumberof mearchprograrnmes inthe area of information technology. RACE (Reseamh into Advanced Communication in Europe) is one of t.hw2 programs Despite the “R”, RACE is strongly market oriented. That is, RACE developments need not only be technically sound and novel, but must in the fimt place prove a commercial potential. Within this program, EuroBridge is supposal to provide advanced communication services which may be exploited by other, more application-oriented projects.

Timer Handling in High-Performance Transport Systems

Abstract Timer handling has a strong impact on various transport layer mechanisms. For instance to achieve an efficient provision of multicast and congestion control timer management is crucial. In this paper we describe and evaluate performance of the Slow Start / Tri-S flow control mechanism. The mechanism is source controlled and the window size adjustment is based on round trip times (RTTs). One main objective of Slow Start / Tri-S is the reduction of window oscillation and the provision of low delay compared to TCP/IP Slow Start. The second part of the paper depicts the error control scheme of XTP in multicast mode. Inefficiencies like redundant retransmissions are illustrated and the influence of respective timers on the performance is explained. To overcome the problems we introduce some enhancements.

The DYCE Concept - Architecture & Implementation Strategies

Abstract This paper presents an approach to design a powerful and flexible controller architecture in order to support new applications across various heterogeneous networks. The approach allows a run-time configuration of the protocol modules. In contrast to existing systems, the developer has the choice to apply different implementation strategies to a module, changing in this way its encapsulation and execution behaviour. Thus, one and the same module can meet different configuration requirements depending on its implementation form as: a typed dynamic library procedure, a typed unmovable thread, a typed movable thread, or a composed protocol graph. The proposed environment is a light-weight framework, called DYCE-2, which represents an implementation of the Dynamic Configuration Entity (DYCE) [Hein93]. DYCE is the central part of our flexible protocol architecture. To build such a flexible and also high-performance system we have to find out and uncover protocol functionalities which can operate concurrently. To achieve results on protocol inherent functional parallelism we have done extensive simulation studies using a petri net based specification technique.