Talks and Poster Presentations (with Proceedings-Entry):
"Identifying Bus Failures in a Time-Triggered Communication System Containing Redundant Communication Channels";
Talk: International Conference on Communications in Computing,
Las Vegas, USA;
- 2000-06-29; in: "Proceedings of the 2000 International Conference on Communications in Computing (CIC 2000)",
Preventing fault accumulation is a key issue in ultra-dependable real-time systems that rely on redundancy to meet the dependability requirements. Fault accumulation among redundant components can lead to spare exhaustion causing the system to crash even if the mean time between faults is long. High error detection coverage with low error detection latency is thus of major importance. In this paper a strategy for identifying bus failures within the time-triggered architecture TTA is presented. The system consists of spatially distributed nodes that are connected to redundant communication channels based on broadcast buses. While the nodes in such a system execute error detection mechanisms and built-in self-tests, the communication channels as shared passive resources cannot perform self-diagnosis. Thus the nodes must periodically assess the fault-free state of each channel.
Electronic version of the publication:
Created from the Publication Database of the Vienna University of Technology.