Low power modular redundancy: a power efficient fault tolerant approach for digital circuits

Author:

Ansari M. Saeed,Mahani Ali,Mohammadi Karim

Abstract

Purpose To increase protection level against transient faults, circuit designers usually take advantage of redundant structures like Triple Modular Redundancy (TMR). Since redundancy compel a significant power overhead, proposing a low power fault tolerant technique in digital circuits is the main objective of this research work. Design/methodology/approach In order to moderate power consumption, we use a dual to triple modular redundancy. In fact, we put one of the modules in a TMR system in sleep mode while the other two operating modules are producing the same outputs. Once a mismatch is detected, the third one exits the sleep mode and the dual modular redundancy (DMR) approach turns into a conventional TMR. Also a novel stoppable clock generator is proposed to handle the sleep mode of the third module. Finally, a new three-input majority voter, compatible with our proposed technique, is presented. Findings Power analysis of combinational circuit benchmarks, ISCAS85, and ISCAS89 as sequential circuit benchmarks are depicted. Simulation results show the power reduction of up to 30% in comparison with the conventional modular redundancy approach. Originality/value Since modular redundancy is the most effective and the most well-known fault tolerant approach which is widely used in reliable circuits designs, it is important to reduce its power consumption. In this paper configuring the sleep mode operation of a circuit and stoppable clock generator lead to a new TMR technique in which the power consumption is strongly reduced.

Publisher

Emerald

Subject

Applied Mathematics,Electrical and Electronic Engineering,Computational Theory and Mathematics,Computer Science Applications

Cited by 6 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3