{"title":"Bounding error masking in linear output space compression schemes","authors":"S. Tarnick","doi":"10.1109/ATS.1994.367258","DOIUrl":null,"url":null,"abstract":"Based on the principle of linear output space compression we present a design method for concurrent checkers such that the masking probability of errors caused by faults of a given set of circuit faults is below a given bound, while keeping the space compression ratio, defined as the ratio of the number of circuit outputs to the number of outputs of the space compressor, as high as possible. Experiments performed on the ISCAS-85 benchmark circuits show that the compression ratios achieved with compression functions computed with this method can be very high, even for very low bounds for the error masking probability and large fault sets.<<ETX>>","PeriodicalId":182440,"journal":{"name":"Proceedings of IEEE 3rd Asian Test Symposium (ATS)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1994-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"13","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of IEEE 3rd Asian Test Symposium (ATS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ATS.1994.367258","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 13
Abstract
Based on the principle of linear output space compression we present a design method for concurrent checkers such that the masking probability of errors caused by faults of a given set of circuit faults is below a given bound, while keeping the space compression ratio, defined as the ratio of the number of circuit outputs to the number of outputs of the space compressor, as high as possible. Experiments performed on the ISCAS-85 benchmark circuits show that the compression ratios achieved with compression functions computed with this method can be very high, even for very low bounds for the error masking probability and large fault sets.<>