Title
A Fast And Accurate Multi-Cycle Soft Error Rate Estimation Approach To Resilient Embedded Systems Design
Abstract
In this paper, we propose a very fast and accurate analytical approach to estimate the overall SER and to identify the most vulnerable gates, flip-flops, and paths of a circuit. Using such information, designers can selectively protect the vulnerable parts resulting in lower power and area overheads that are the most important factors in embedded systems. Unlike previous approaches, the proposed approach firstly does not rely on fault injection or fault simulation; secondly it measures the SER for multi cycles of circuit operation; thirdly, the proposed approach accurately computes all three masking factors, namely, logical, electrical, and timing masking; fourthly, the effects of error propagation in re-convergent fan-outs are considered in the proposed approach. SERs estimated by the proposed approach for some ISCAS89 circuit benchmarks are compared with that estimated by the Monte Carlo (MC) simulation based fault injection approach. The results show that the proposed approach is about four orders of magnitude faster than the MC fault injection approach while having an accuracy of about 97%. This level of fastness and accuracy makes the proposed approach a viable solution to measure the SER of very large size circuits used in industry.
Year
DOI
Venue
2010
10.1109/DSN.2010.5544952
2010 IEEE-IFIP INTERNATIONAL CONFERENCE ON DEPENDABLE SYSTEMS AND NETWORKS DSN
Keywords
Field
DocType
benchmark testing,monte carlo simulation,monte carlo,monte carlo methods,embedded systems,embedded system,information design,logic design,logic gates,error propagation
Logic synthesis,Logic gate,Monte Carlo method,Propagation of uncertainty,Soft error,Computer science,Real-time computing,Electronic circuit,Fault injection,Benchmark (computing),Embedded system
Conference
ISSN
Citations 
PageRank 
1530-0889
7
0.74
References 
Authors
14
4
Name
Order
Citations
PageRank
Mahdi Fazeli127131.45
Seyed Ghassem Miremadi253150.32
Hossein Asadi332331.94
Seyed Nematollah Ahmadian4321.68