Karl M. Göschka: Past Projects

[ Back to Göschka's Home Page ] [ Last update: 19.April 2011 ]

This list provides an overview of the most important research projects I initiated, was in charge of or took part in during the last years. The effort figures include employed staff only, numerous students contributed additional effort during their diploma or master's theses or through their practical work. Numerous smaller student projects where I was involved as supervisor, advisor or coach, are not enlisted here.

TRADE - Trustworthy adaptive quality balancing through temporal decoupling
Total Effort: 109 person months
Local Effort: 84 person months
Partner:IRIAN GmbH
Local Staff: Lorenz Froihofer (02/2008-02/2011),
Günther Starnberger (04/2008-04/2011)
My responsibility: Project co-ordinator (project leader), scientific&technology manager, and senior researcher.

This research project funded by the Austrian Federal Ministry for Transport, Innovation, and Technology aims at a concept of secure and dependable temporal decoupling: By relaxing the requirements for timeliness, the system properties integrity and availability can be adaptively balanced or traded against each other, thereby increasing the reliability of the system. It comprises:

COMPASS - Component Based Automotive System Software
Duration: 09/2005-12/2007
Local Effort: 30 person months (out of 133)
Partners: Decomsys and UAS Technikum Wien
Local Staff: Dietmar Schreiner (09/2005-12/2007),
Wolfgang Forster (09/2005-12/2007)
My responsibility: Project leader and scientific supervision of the Institute of Information Systems' part in the project.

This research project funded by the Austrian Federal Ministry for Transport, Innovation, and Technology aims at a component based hardware software infrastructure, metrics, and benchmarks to assess cost, performance and dependability aspects for an automotive application development. It comprises:

DeDiSys - Dependable Distributed Systems
Duration: 10/2004-11/2007
Total Effort: 399 person months
Local Effort: 82 person months
Partners:Faculty of Computer Science and Management, Wroclaw University of Technology, Poland,
Real-Time Systems Laboratory, Department of Computer and Information Science, Linköpings Universitet, Sweden,
Instituto Tecnologico de Informatica, Universidad Politecnica de Valencia, Spain,
Frequentis, Austria,
Etra Investigacion Y Desarrollo, S.A., Spain,
Cosylab d.o.o.,Laboratorij za kontrolne sisteme, Slovenia,
XLAB Razvoj programske opreme in svetovanje d.o.o., Slovenia.
Local Staff: Lorenz Froihofer (10/2004-01/2008),
Johannes Osrael (11/2004-12/2007)
My responsibility: Project co-ordinator (project leader), scientific&technology manager, and senior researcher.

DeDiSys is a specific targeted research project (STREP) funded by the European Commission under the FP6. This international project with eight partner organisations in five countries invests an effort of some 33 person years into a concept for balancing dependability in distributed software systems by partially or temporarily relaxing integrity (correctness) in order to increase availability. This concept comprises an architecture, rules for technology integration including interaction and composition standards for component integration, the definition of required and provided open platform services, and well-defined metrics and evaluation methods for such a system proven by prototype implementations. By promoting the idea of open systems, the project does not aim at a new type of middleware, but rather at the integration with existing component infrastructures. Technically, the project focuses on the following two key issues:

VBness - Virtual Business
Duration: 03/2001-12/2003
Effort: 72 person months
Partner: Siemens Austria
Staff: Jürgen Falb (03/2001-10/2003),
Rudolf M. Liebhart (03/2001-06/2003),
Roman Popp (01/2003-12/2003)
My responsibility: Supervision and Scientific Lead

VBness, a project in cooperation with Siemens Austria, investigated the process of user interface design for distributed, heterogeneous applications. Such applications tend to be both multi-user and multi-device applications (possible devices are desktop PC, PDA, or mobile phone). This poses problems for the developers regarding the consistency among different user interfaces throughout the software lifecycle, especially when the application changes and all the corresponding user interfaces need to be updated. The project's approach was to capture the semantics of the user interface and describe it in a platform independent way. In contrast to common user interface design, the project used parts of the human communication theory (speech act theory) to capture the semantics (e.g. intention) of the user interaction, rather than the elements or widgets of the user interface. Compared to traditional approaches, the project's solution provides more flexibility and is applicable even for non-graphical user interfaces. However, this comes at the cost of increased complexity for the description and the generation process.

DTMS - Distributed Telecommunication Management System
Duration: 10/2000-09/2004
Effort: 48 person months
Partner: Frequentis
Staff: Robert Smeikal (10/2000-09/2004)
My responsibility: Project Manager, Scientific Lead, and Software Architect/Designer

High availability is not only demanded for safety-critical networked voice communication systems but also for the management systems controlling them. Major requirements for such a Distributed Telecommunication Management System (DTMS) comprise fault-tolerance against site or network failures, transactional safety, and reliable persistence. In order to provide distribution and persistence both transparently and fault-tolerant the project introduced a two-layer architecture facilitating an asynchronous replication algorithm. Among the lessons learned are: component based software engineering poses a significant initial overhead but is worth it in the long term; a fault-tolerant naming service is a key requirement for fail-safe distribution; the reasonable granularity for persistence and concurrency control is one whole object; asynchronous replication on the database layer is superior to synchronous replication on the instance level in terms of robustness and consistency; semi-structured persistence with XML has drawbacks regarding consistency, performance and convenience; in contrast to an arbitrarily meshed object model, a accentuated hierarchical structure is more robust and feasible; a query engine has to provide a means for navigation through the object model; finally the propagation of deletion operation becomes more complex in an object-oriented model. Incorporating these lessons learned will help to provide a platform for highly available distributed systems.

VOILA - Voice Over IP Laboratory for Air traffic control
Duration: 08/2000-06/2004
Effort: 98 person months
Partner: Frequentis
Staff: Paul Smutny (08/2000-09/2002),
Christian Keusch (09/2000-08/2001),
Klaus Darilion (07/2001-06/2004),
Christoph Kurth (07/2002-06/2004)
My responsibility: Supervision and Scientific Lead

The main task of this project was to contribute significant research results to a new VoIP (SIP) based productline of voice communication systems for air traffic control (ATC). It comprised the following key issues:

EAI - Enterprise Application Integration
Duration: 09/1999-12/2003
Effort: 107 person months
Partner: Ericsson
Staff: Wolfgang Radinger (09/1999-12/2003),
Martin Jandl (07/2001-12/2003),
Alexander Szep (05/2001-09/2002, part time, and 07/2003-12/2003),
Vivek Kuttapan (10/2002-05/2003)
My responsibility: Project Manager (later Supervision), Software Architect and Scientific Lead

Distributed telecommunication applications and management systems often provide complex and extensive interfaces. Even if these interfaces consist of standardized middleware API definitions (e.g. CORBA - Common Object Request Broker Architecture - IDL Interface Description Language), they still require too much in-depth knowledge for easy third party integration, which typically only requires just a small subset of the interface's functionality. Therefore, this project aimed at distributed, dependable, and robust software architectures for accessing complex back-end legacy software systems in a generic manner. The involved technologies include distributed software engineering technologies (CORBA), Directory services (LDAP), web service technologies (SOAP, WSDL, UDDI), and XML for interface definition.

OEBB - New Architecture for the Distributed Ticket Sales System
Duration: 06/1999-09/2000
Effort: 5.5 person months
Partner: Austrian Federal Railways (ÖBB)
Staff: Jürgen Falb, Wolfgang Radinger, Bernd Petrovitsch
My responsibility: Project Manager, Consultant, and Software Architect

This technical assessment investigates new architectures for the distributed ticket sales system of the Austrian Federated Railways (ÖBB). Based on the most important requirements availability, transactional safety, extensibility and maitainability, a particular focus was on the flexible integration of different client software: Counter, travel agent, mobile sales machines (online and offline), Web clients, and ticketing via mobile phone. he study resulted in implementations performed by Siemens Austria and Ascom Switzerland.

VIPAS - A virtual personal assistant
Duration: 12/1998-02/2001
Effort: 27 person months
Partner: Siemens Austria
Staff: Jürgen Falb (12/1998-02/2001)
My responsibility: Supervision and Scientific Lead

When the project started, location independent computing was already a common reality, but most users encountered difficulties handling and synchronizing multiple mobile devices like mobile phones, portable digital assistants (PDAs) or laptop computers. The approach of this project was based on the idea of decoupling the functionality of a mobile application from the physical device. Therefore, a "virtual device" was developed, which can be migrated between different physical devices. The aim was to provide a service, which is not only location independent, but also independent from the underlying infrastructure. This service belongs to the user rather than to the device, providing him with a single, personalizable communication endpoint. For the technical implementation a mobile software agent architecture was used, where different VoIP (Voice over IP) signaling protocols and multimedia frameworks were integrated. These agents can "travel" with the user, thus enabling communication hand-over between different devices, even during ongoing calls.

QoS - Quality of Service in VoIP networks
Duration: 04/1997-01/2002
Effort: 68 person months
Partner: Ericsson
Staff: Joachim Fabini (05/1997-04/1998),
Gerhard Fetty (12/1997-07/1999),
Wolfgang Kampichler (02/1999-01/2002)
My responsibility: Project Manager and Scientific Lead

Internet and mobility were pushing the rapid development in the telecommunications sector. Heterogeneous communication networks based on IP started to provide the basis for integrated data and multi-media communications. This so-called IP-Telephony or "Voice over IP" was standardized in a competition between ITU-T (H.323) and the IETF (SIP). Integral to VoIP are the mechanisms to guarantee or control the quailty of service (QoS). This project developed methods and tools to allow the non-intrusive measuring of QoS, even before the VoIP equipment has been installed.

Demenet - Database-backed Web Application for the Demeter Project
Duration: 12/1996-06/1998
Local Effort: 3 person months
Partner: EU Project Demeter (FP4)
Staff: no employed staff
My responsibility: Contribution with software and database design, as well as implementation

The project DEMETER (Distance Education, Multimedia Teleservices and Telework for Farmers) was funded by the European Commission under the 4th Framework Programme TURA (Telematics for Urban and Rural Areas). It was specifically tailored to farmers' needs, aimed at enhancing their quality of life and finding new sources for additional income, among others by providing a database for product and tourism marketing and sales. The Demenet project contributed to Demeter with a Web-based database for product marketing and tourism information. Most importantly, the Web interface was restricted to pure HTML in order to avoid consistency issues. On the other hand, it had to be multilingual and provide means for sophisticated user interaction. This was achieved by a new method to integrate databases and Web clients, which was based on state machines for robust client interaction.

This space is left blank intentionally.