Fault-toleration by Duplication and Debugging for Distribution Real-time Systems

Abstract

Translate Article

Distributed real-time systems have continually being developed for many real applications. It is very important to assure the reliability of such systems. However, developing a fault-free distributed real-time system is very difficult. Duplication is one of the fault-tolerant technologies for improving the reliability of distributed real-time computer systems. In this paper, we present a dual-cluster distributed real-time system model with four states to improve the reliability of distributed real-time systems. The system employs cluster switch, output arbitration and fault detection to increase its reliability. The system's availability is evaluated according to the model. We also introduce a dynamic approach for visualizing and debugging the timing behavior of distributed real-time systems.