Designing a fault-tolerant architecture for real-time distributed control system

Distributed control systems play major roles in real-time applications. In some circumstances human lives may depend on these systems. Hence, they should be highly dependable. Since there is no specific way to forecast a failure, the systems should have fault-tolerant features to allow them to continue to operate in the presence of fault. In this research, a fault-tolerant architecture for real-time distributed control system was developed. A fault-tolerant node was designed and its performance was evaluated using the Markov chains model. All these nodes communicate via a network. To choose the most suitable network for this system, several networks were analyzed qualitatively. Several medium access control protocols were also compared quantitatively using OMNET++. From the results obtained, a new fault-tolerant architecture is proposed. This system possesses high reliability not only at the node stage but also at the network stage, hence increasing the reliability of the overall system.