Research on RTOS-Integrated TMR for Fault Tolerant Systems

Safety and availability are issues of major importance in many critical systems. A RTOS (realtime operating system)-integrated fault-tolerant system using TMR technology is presented in this paper. The system incorporates three homogeneous microcomputers and provides the fault-tolerant function through system-APIs to applications. As it is integrated with RTOS, the system is more general-purpose, and programmers need not pay too much attention to the fault tolerance technology. This system works in normal and degraded (duple or even single modular) modes, and can tolerate transient or permanent faults. The system also provides MultiTask-support fault-tolerant function, and reconfiguration after a fault occurs is transparent to applications. Meanwhile, a novel seamless software upgrade method through intelligent state-transition-control is brought forward.