Conference Home

Final Program

Tutorials

Workshops

FastAbstracts

Student Forum

Keynote Speaker

Registration

Venue

Local Information

Hotel Reservations

News

Registration Desk at DSN 2004 will be open on June 28 during the entire afternoon and the Welcome Reception. We strongly suggest that attendees register on June 28, and only latecomers use June 29 from 7.30 to 8.00.
The Opening Session will start on June 29 at 8.00 sharp

Angelo Marino of EU will present during the Business Meeting the latest information on IST Programme and on the cooperation with NSF

Tutorial A "Creating Strategy and Tactics Cyberspace " and Tutorial C "Distributed Denial of Service Attacks Background, Diagnosis and Mitigation" have been cancelled all, other three Tutorials are confirmed

Dr. Leslie Lamport Keynote Speaker at DSN 2004

Student Grants

Committees

Original Call for Contributions

Preliminary information about DSN 2005

Tutorials

Tutorial B: Survivability of Telecommunications Systems - Concepts, Architectures and Analysis

Presenter(s): Veena Mendiratta, Bell Labs, Lucent Technologies, USA, Yun Liu, Kishor Trivedi, Duke U., USA

E-mail: veena@lucent.com

Attendee type: The target audience includes students, academic/industrial researchers, and practitioners in the area of telecommunications, fault tolerance, and computer networks.  The assumed background is basic knowledge of computer networks, probability, and stochastic modeling.

Short description: This tutorial will provide an introduction to the subject of telecommunications system survivability in terms of survivability concepts, architectures and analysis. Our focus is on the voice network for wireline, cellular and voice over packet   implementations   though   the   concepts  and   analysis  methods  are   general  enough   to  be  applicable  to  other architectures. Survivability definitions, both qualitative and quantitative, are presented and alternative architectures for survivability for wireline, cellular and voice over packet networks are defined. Survivability models with performance, availability and composite performance-availability measures and their traditional solution/interpretation are presented. The survivability analysis evaluates the survivability of the alternative architectures for certain failure scenarios. We summarize with what is required for architecting a survivable voice telecommunications system and what survivability analysis is needed for evaluating the effectiveness and economics of different alternative architectures.

Tutorial D: Architecture Analysis & Design Language (AADL)

Presenter(s): David Gluch, Embry-Riddle Aeronautical U., USA, Peter Feiler, SEI, CMU, USA, Bruce Lewis, Army AMCOM SED, USA

E-mail: gluchd@erau.edu

Attendee type: A typical attendee would have basic knowledge of real-time and dependability design issues and techniques for software intensive systems (e.g., knowledge of scheduling, communications, redundancy, partitioning) and an interest in understanding the use of architecture-driven and model-based design and analysis in the development of these systems.

Short description: The tutorial is aimed at creating awareness and understanding of the emerging SAE Architecture Analysis & Design Language (AADL) standard and its application to fault-tolerant and high dependability systems design.  The AADL's precisely defined semantics
can describe and facilitate the analysis of important performance-critical and dependability aspects such as timing, schedulability, fault and error behaviors, time and space partitioning,
and safety properties. Attendees will gain insight into the benefits of using the AADL in the engineering of highly dependable and performance-critical systems, through examples of its use in the analysis and development of actual embedded real-time systems.  An avionics system design is used as the principal example application throughout the tutorial.

Tutorial E: Detecting Crash Failures in Asynchronous Distributed Systems: What? Why? How?

Presenter: Michel Raynal, IFSIC-IRISA, U. of Rennes, France

E-mail: raynal@irisa.fr

Attendee type: This tutorial is mainly designed for people who want to understand the fundamental issues raised by crash failures and how these issues can be mastered during the design of fault-tolerant, asynchronous, distributed software.  This includes PhD students working on fault-tolerant systems, researchers who are not familiar with the failure-detector approach and want to understand its basics, and engineers who have to cope with dependability problems related to failure detection.

Short description: The tutorial focuses on the problems caused by failures in asynchronous, distributed systems. Indeed, the net effect of asynchrony and process crashes makes the design of provably-correct, reliable middleware services far from trivial.

 The structure of the tutorial is the following:

 (i) We first review fundamental definitions and concepts related to crash-failure detection in asynchronous systems. The presentation is based on the notion of "failure detectors" (introduced by Sam Toueg and Tushar Chandra in 1991).

 (ii) We then consider basic distributed computing problems (such as atomic broadcast, non-blocking atomic commit) and show what sorts of failure-detector modules an asynchronous distributed system has to be equipped with in order to be able to solve these problems in the presence of process-crash failures.

 (iii) In the last part of the tutorial, we focus on the implementation of the required failure-detector modules.

 The existing literature on this topic is very technical and appears mainly in theoretically-inclined journals and conferences. This tutorial offers a unique opportunity to survey the fundamentals of the failure detector approach and to quickly understand its aim, its power, its benefits and its limitations.