{"title":"Stabilizing Liao's absorbing boundary conditions using single-precision arithmetic","authors":"M. Moghaddam, W. Chew","doi":"10.1109/APS.1991.174867","DOIUrl":null,"url":null,"abstract":"The boundary conditions given by Z.P. Liao et al. (Scientia Sinica, vol.XXVII, no.10, p.1065, 1984) for finite difference time-domain analysis have been shown to yield excellent results in scalar as well as vector electromagnetics problems using double-precision arithmetic. However, it is more desirable to use single-precision arithmetic if possible, since the time required to perform a real, double-precision multiplication is two to three times that of a single-precision operation. But with a single-precision implementation, it is observed that the solution becomes unstable as the number of time steps grows. The authors present a procedure for remedying this problem.<<ETX>>","PeriodicalId":264366,"journal":{"name":"Antennas and Propagation Society Symposium 1991 Digest","volume":"15 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1991-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"31","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Antennas and Propagation Society Symposium 1991 Digest","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/APS.1991.174867","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 31
Abstract
The boundary conditions given by Z.P. Liao et al. (Scientia Sinica, vol.XXVII, no.10, p.1065, 1984) for finite difference time-domain analysis have been shown to yield excellent results in scalar as well as vector electromagnetics problems using double-precision arithmetic. However, it is more desirable to use single-precision arithmetic if possible, since the time required to perform a real, double-precision multiplication is two to three times that of a single-precision operation. But with a single-precision implementation, it is observed that the solution becomes unstable as the number of time steps grows. The authors present a procedure for remedying this problem.<>