2007Transducer and Microsystem TechnologiesRequires access

Design and reliability analysis of self-repair TMR systems based on evolvable hardware

Gao Gui-jun

Open publisher page 0 citations

Abstract

A new design method of self-repair triple-modular redundancy systems is proposed by introducing evolvable hardware into the traditional triple-modular redundancy system.There are multi-mechanisms of fault-tolerance and repair in the system:the system has a structure of triple-modular redundancy in the mass,which can check the fault modular autonomously;each modular is made up of subassemblies with spare parts,and can recover from fault fast by switching to the spare parts;while each subassembly in the modular can be repaired through evolution.The design method is proof-tested by a triple-modular redundancy 2 bit multiplier on-chip.Moreover,the reliability model of the system is given,and the reliability formula is educed.The reliability of the system is analyzed in theory.It is shown that the self-repair triple-modular redundancy system can recover from stuck-at faults effectively,and thus has longer life span and higher reliability than the conventional triple-modular redundancy system.

About this research paper

What this paper is about

A new design method of self-repair triple-modular redundancy systems is proposed by introducing evolvable hardware into the traditional triple-modular redundancy system.There are multi-mechanisms of fault-tolerance and repair in the system:the system has a structure of triple-modular redundancy in the mass,which can check the fault modular autonomously;each modular is made up of subassemblies with spare parts,and can recover from fault fast by switching to the spare parts;while each subassembly in the modular can be repaired through evolution.The design method is proof-tested by a triple-modular redundancy 2 bit multiplier on-chip.Moreover,the reliability model of the system is given,and the reliability formula is educed.The reliability of the system is analyzed in theory.It is shown that the self-repair triple-modular redundancy system can recover from stuck-at faults effectively,and thus has longer life span and higher reliability than the conventional triple-modular redundancy system.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

A new design method of self-repair triple-modular redundancy systems is proposed by introducing evolvable hardware into the traditional triple-modular redundancy system.There are multi-mechanisms of fault-tolerance and repair in the system:the system has a structure of triple-modular redundancy in the mass,which can check the fault modular autonomously;each modular is made up of subassemblies with spare parts,and can recover from fault fast by switching to the spare parts;while each subassembly in the modular can be repaired through evolution.The design method is proof-tested by a triple-modular redundancy 2 bit multiplier on-chip.Moreover,the reliability model of the system is given,and the reliability formula is educed.The reliability of the system is analyzed in theory.It is shown that the self-repair triple-modular redundancy system can recover from stuck-at faults effectively,and thus has longer life span and higher reliability than the conventional triple-modular redundancy system.

Key concepts: Triple modular redundancy, Redundancy (engineering), Modular design, Spare part, Fault tolerance, Reliability engineering, Embedded system, Computer science

Related papers

Back to paper searchBrowse research topicsOriginal source
Design and reliability analysis of self-repair TMR systems based on evolvable hardware — Research Paper | ScholarLens