A time redundancy approach to TMR failures using fault-state likelihoods

Abstract

Failure to establish a majority among the processing modules in a triple modular redundant (TMR) system, called a TMR failure, is detected by using two voters and a disagreement detector. Assuming that no more than one module becomes permanently faulty during the execution of a task, Re-execution of the task on the Same Hardware (RSHW) upon detection of a TMR failure becomes a cost-effective recovery method, because 1) the TMR s p e m can mask the effects of one faulty module while RSHW can recover from nonpermanent faults, and 2) system reconfiguration-Replace the faulty Hardware, reload, and Restart (RHWR)-i s expensive both in time and hardware. We propose an adaptive Wovery method for TMR failures by "optimally" choosing eithp RSHW or RHWR based on the estimation of the costs involved. We apply the Bayes theorem to update the likelihoods of all possible states in the TMR system with each voting result. Upon detection of a TMR failure, the expected cost of RSHW is derived with these likelihoods and then compared with that of RHWR. RSHW will continue either until it recovers from the TMR failure or until the expected cost of RSHW becomes larger than that of RHWR. As the number of unsuccessful RSHW's increases, the probability of permanent fault@) having caused the TMR failure will increase, whicli will, in turn, increase the cost of RSHW. Our simulation results show that the proposed method outperforms the conventional reconfiguration method using only RHWR under various conditions.

Real-time computing: a new discipline of computer science and engineering

“…This analysis is useful for the one-shot delay model in [84], where a single event-a long-lasting failure-may cause a dynamic failure. The authors of [84] used this approach to derive deadlines for numerous example control systems and then applied the knowledge of hard deadline to the design of error recovery in a triple-modular-redundant (TMR) controller computer [83].…”

Energy efficient redundant configurations for real-time parallel reliable servers

“…For extremely high reliability, we may combine modular and temporal redundancy. That is, when no majority result is obtained on modular redundant processing units during the processing of an application, we can further explore temporal redundancy to re-execute it and increase the reliability instead of declaring a failure (Shin and Kim 1994).…”

An optimal single error locating TSC-TMRsystem

“…Current trends try to hybridize TMR systems, to improve on system reliability. Time redundancy can be incorporated on TMR system by re-execution of the current task on detection of a fault, thus treating non-permanent module failure (Shin and Kim 1994). Information redundancy in the form of coding techniques can be introduced to a TMR system to increase both reliability and safety (Vaidya and Pradhan 1993).…”