Designing fault-diagnosis and reintegration to prevent node redundancy attrition in highly reliable control systems based on FTT-Ethernet

Abstract

Distributed Embedded Control Systems (DECSs) used for Real-Time (RT) critical applications must satisfy stringent time requirements and attain high reliability. FTT-Ethernet provides nodes of DECSs with real-time communication capabilities, but does not include Fault Tolerance (FT) mechanisms. The FT4FTT project aims at proposing a complete FT architecture for RT critical DECSs. It uses a duplicated switched FTT-Ethernet star and active node replication with consistent distributed majority voting to respectively tolerate channel and node faults. However, FT4FTT, in its current state, still lacks mechanisms to prevent node redundancy attrition due to temporary faults affecting the nodes and channel, which are the most likely types of faults in DESs. This paper presents our ongoing work to complete the FT4FTT architecture with appropriate fault-diagnosis and reintegration mechanisms that overcome this limitation.