DocumentCode
2609920
Title
Automated Generation of Built-In Self-Repair Architectures for Random Logic SoC Cores
Author
Dobai, Roland ; Balaz, Marcel ; Fischerova, Maria
Author_Institution
Inst. of Inf., Bratislava, Slovakia
fYear
2012
fDate
5-8 Sept. 2012
Firstpage
73
Lastpage
78
Abstract
Built-in self-repair (BISR) architectures and methods are widely used for memory cores of system-on-chips (SoCs), where the area-efficient fault detection and repair are crucial in order to meet the high quality requirements. Research of BISR architectures for logic cores has begun as well. However, the irregular structure of logic cores represents a serious limitation and therefore, currently only ad hoc methods exist. Automated generation of BISR architectures for random logic SoC cores is proposed in this paper. The generation is guided by the characteristics of the architecture: mean time to failure (MTTF) and area overhead. The main contribution is the fully automated handling of arbitrary random logic cores and the possibility to generate architectures based on various BISR principles. The proposed method was implemented and evaluated over benchmark circuits, and the experiments confirmed that BISR architectures can be successfully generated for random logic cores. The MTTFs of the generated architectures have been improved at the cost of relatively low area overhead.
Keywords
built-in self test; logic circuits; logic design; system-on-chip; BISR architectures; MTTF; ad hoc methods; area-efficient fault detection; automated generation; benchmark circuits; built-in self-repair architectures; mean time to failure; memory cores; random logic SoC cores; system-on-chips; Boolean functions; Computer architecture; Logic gates; Reliability; Sociology; Statistics; System-on-a-chip; built-in self-repair; genetic algorithms; logic core; reconfigurable logic; reliability; system-on-chip;
fLanguage
English
Publisher
ieee
Conference_Titel
Digital System Design (DSD), 2012 15th Euromicro Conference on
Conference_Location
Izmir
Print_ISBN
978-1-4673-2498-4
Type
conf
DOI
10.1109/DSD.2012.29
Filename
6386872
Link To Document