RELIABILITY OF CHECKPOINTED REAL-TIME SYSTEMS USING TIME REDUNDANCY

Real-time computers are often used in embedded, life-critical, applica tions where high reliability is important. A common approach to making such systems dependable is to vote on redundant processors executing multiple copies of the same task. The processors which make up such vo ted systems are subject not only to independently occurring permanent & transient failure, but also to correlated transients brought about b y electromagnetic interference from the operating environment. To coun teract these transients, we require checkpointing and time redundancy, in addition to processor redundancy. White many models in the literat ure deal with processor redundancy, no other authors, to our knowledge , have studied the effectiveness of time redundancy against correlated failures in real-time systems. This paper analyzes the use of time & device redundancy in systems subject to correlated failure, We study t radeoffs in the design of embedded systems for optimal reliability.