CoDiCom: Distributed Real-time Computer Control System

From Navigators

Revision as of 18:07, 13 November 2012

• Research Line(s): Fault and Intrusion Tolerance in Open Distributed Systems (FIT), Timeliness and Adaptation in Dependable Systems (TADS)
• Sponsor: STRIDE, Jnict (since 1993)
• Coordinator: INESC
• Start Date: Oct. 1991
• Duration: 36 months
• Keywords: real-time, reliable multicast, fault-tolerance, I/O, control
• Team at FCUL: 4 researchers, including Paulo Verissimo

Distributed real-time system architectures are still in fast evolution. Still, there are challenging fields yet unexplored: the most efficient combination of time-triggered and event-triggered behaviours; highly concurrent and decentralised real-time processing, based on object-oriented paradigms other than client-server via RPC (eg. reliable Casts, peer object conversations); the application of distributed fault-tolerance techniques (eg. those pursued in Delta-4) to fault-tolerant I/O processing, exploring different flavours of synchronism and the advantages of taking an embedded approach.

Aims

Exploring concepts which may open unexpected doors in fields such as manufacturing, process control, real-time cooperative work, transactional systems, etc.

Developing fault-tolerant input-output solutions relying on techniques otherwise used for distributed computation, based on the paradigms of groups and reliable multicast.

Approach and Methods

Look at architecture from the viewpoint of an embedded distributed real-time system, to match the kind of applicationms typically found in field-buses.

Progress and Results

Definition of the system architecture. Development of a PC-farm style testbed with remote download. Work published on the underlying concepts expressed above.

Potential

The issues under research in this project may lead to the development of new techniques in process control and automation.

Consultants

• Prof. H. Kopetz, T.U. Vienna
• Prof. R. Schlichting, U. Arizona

Publications

• Paulo Verissimo, “Causal Delivery Protocols in Real-time Systems: a Generic Model”, 1996. Journal of Real-Time Systems, vol 10, n 1, pp: 45-73. KluwerAcademic Publishers, Jan. 1996.

• Paulo Verissimo, “Ordering and timeliness requirements of dependable real-time programs”, Journal of Real-Time Systems, Kluwer Eds., (7): 105--128, 1994, Jan. 1994.

• Luís Rodrigues, Paulo Verissimo, José Rufino, “A low-level processor group membership protocol for LANs”, in Proceedings of the 13th International Conference on Distributed Computing Systems. Pittsburgh, Pennsylvania, USA, May 1993., May 1993.

• José Rufino, “Rede de Difusão Abstracta: um modelo para construção de protocolos portáteis sobre rede local”, Master’s thesis, Technical University of Lisbon, Instituto Superior Técnico, Lisbon, Portugal, Apr. 1993. In Portuguese

• Paulo Verissimo, Luís Rodrigues, “A posteriori Agreement for Fault-tolerant Clock Synchronization on Broadcast Networks”, in July 1992. Digest of Papers, The 22th International Symposium on Fault-Tolerant Computing. Boston, USA, Jul. 1992.

• José Rufino, Paulo Verissimo, “Minimizing Token-Bus inaccessibility through network planning and parameterizing”, in Proceedings of the EFOC/LAN92 Conference, IGI. Paris, France, June 1992., Jun. 1992.

• José Rufino, Paulo Verissimo, “A study on the inaccessibility characteristics of ISO 8802/4 Token-Bus LAN”, in Proceedings of the IEEE INFOCOM'92 Conference on Computer Communications. Florence, Italy, May 1992., May 1992.

• José Rufino, Paulo Verissimo, “A study on the inaccessibility characteristics of the FDDI LAN”, INESC, Tech. Rep. RT-25/92, Mar. 1992.

• José Rufino, Paulo Verissimo, “A study on the inaccessibility characteristics of ISO 8802/5 Token-Ring LAN”, INESC, Tech. Rep. RT/24-92, Feb. 1992.

BibTeX