Worst case timing requirement of real-time tasks with time redundancy

The improvement of system reliability can be achieved through the use of time redundancy techniques including retry, checkpointing and recovery block. The objective of this paper is to analyze the worst case timing requirement of real-time tasks with time redundancy in the presence of multiple transient faults. The timing behavior of system level checkpointing is modeled with additional tasks that save the state of all tasks. The timing information derived in this paper makes time redundancy techniques applicable to real-time systems while keeping the validity of schedulability check developed in the past