{"title":"Study of the use of microprocessors in digital filtering","authors":"B. Farhang-Boroujeny, G. J. Hawkins","doi":"10.1049/IJ-CDT:19790032","DOIUrl":null,"url":null,"abstract":"In this paper, the application of microprocessors to digital filtering has been considered. To achieve the highest possible working frequencies when using microprocessors it has been found necessary to employ the method of distributed arithmetic. A new computational formulation is introduced which leads to methods of implementation which allow for better signal/roundoff-noise ratios at the expense of a few extra instructions in the iterative loops; this causes only a slight decrease from the optimum working frequencies. Also, the addition of bias values to precomputed constants is shown to lead to some improvement to the signal/roundoff-noise. These improvements have been verified by means of computer simulations. The methods suggested can be implemented by employing microprocessors with some dedicated hardware, and this helps to reduce the processing time. The employment of various facilities available on some microprocessors is considered, and tests have been carried out on a sample processor.","PeriodicalId":344610,"journal":{"name":"Iee Journal on Computers and Digital Techniques","volume":"34 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1979-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Iee Journal on Computers and Digital Techniques","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1049/IJ-CDT:19790032","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
In this paper, the application of microprocessors to digital filtering has been considered. To achieve the highest possible working frequencies when using microprocessors it has been found necessary to employ the method of distributed arithmetic. A new computational formulation is introduced which leads to methods of implementation which allow for better signal/roundoff-noise ratios at the expense of a few extra instructions in the iterative loops; this causes only a slight decrease from the optimum working frequencies. Also, the addition of bias values to precomputed constants is shown to lead to some improvement to the signal/roundoff-noise. These improvements have been verified by means of computer simulations. The methods suggested can be implemented by employing microprocessors with some dedicated hardware, and this helps to reduce the processing time. The employment of various facilities available on some microprocessors is considered, and tests have been carried out on a sample processor.