Patent · US Expired

Filtering a signal with the use of approximate arithmetical division

US5325319A · kind A · utility

1Cited by
8References
3Claims
0Family size

Assignee

Inventors

Key dates

Filing dateNov 2, 1993
Grant dateJun 28, 1994
Priority date
Expiry dateNov 2, 2013

Classification

  • Technology area (CPC H)Electricity
  • CPC primaryH03H21/0012
  • WIPO fieldBasic communication processes
  • WIPO sectorElectrical engineering

Abstract

Filtering a signal with an adaptive digital filter includes updating the filter by performing an approximate division of a constant number by a variable number in binary form. The variable number represents a magnitude of energy of the signal to be filtered, and consists of a character bit and a plurality of bits which define the absolute value of the number. The variable number is first transformed by substituting with logic zeros any logic ones that have a lower significance than the most significant logic one of the bits. The first transformed number is then further transformed by reading the character bit of the variable number as a character bit and by reading the bits in the digital word in a reversed order. The resulting value is applied to adjust the adaptive digital filter which filters the signal.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.