Adjustment of adaptive digital filter coefficients in modular codes

Abstract

The main advantage of adaptive digital filters is the minimum error of their output response. It is reached by recalculation of values of coefficients of the filter. The conducted researches have shown that the known algorithms of adjustment of coefficients of the adaptive digital filter are inexpedient to be used at their implementation in the RNS codes. The new consecutive algorithm of recalculation of coefficients which application allows to carry out adjustment of coefficients of modular digital filters much quicker, than the existing algorithms is presented in article. At the same time for carrying out recalculation of each coefficient of the modular digital filter it is necessary to execute one operation of subtraction, one operation of multiplication and one transaction of addition of the module, i.e. time of recalculation is proportional to a filter order. Thus, it is obvious that use of the developed algorithm allows to reduce time expenditure in comparison with earlier known algorithms of adaptation.