Hardware redundant fault-tolerant systems and the different design approaches are discussed. The reliability analysis of fault-tolerant systems is usually done under permanent fault conditions. With statistical data suggesting that up to 90% of system failures are caused by intermittent faults, the reliability analysis of fault-tolerant systems must concentrate more on this class of faults. In this work, a reconfigurable Triple Modular Redundancy (TMR) with spare system that differentiates between permanent and intermittent faults has been built. The reconfiguration process of this system depends on both the current status of its modules and their history. Based on this, a different approach for reliability analysis under intermittent fault conditions using Markov models is presented. This approach shows a much higher system reliability compared to other redundant and non-redundant configurations.
Library of Congress Subject Headings
Fault-tolerant computing; System design; Electronic digital computers--Reliability
Department, Program, or Center
Computer Engineering (KGCOE)
Al-Kofahi, Khalid A., "Reliability analysis of triple modular redundancy system with spare" (1993). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus