AN IMPROVED ALGORITHM FOR TIME DOMAIN AVERAGING BASED ON FRACTIONAL DELAY FILTERING

Abstract

For reducing the influence of truncation error on the accuracy of time domain averaging（ TDA） induced by noninteger-period sampling,an improved algorithm of TDA based on fractional delay filtering（ FDF） is presented.On the basis of analyzing the disadvantages of current methods of TDA and the characteristics of FDF,a presently improved method of signal truncation was used in the method to avoid the accumulation of truncation error and obtain a delay factor on each of the truncated signals,the reconstruction and resample for the truncated signals were implemented with FDF,and finally the arithmetic mean were computed after all of the truncated signals were added.The theoretical analysis and experimental results demonstrate that the method which fully makes using of the advantages of FDF on the change of delay factor,apparently reduces the truncation error of signals with non-integer-period sampling,and could obviously improve the accuracy of TDA without key-phase signal.

Keywords

• Fractional delay filtering;Time domain averaging;Truncation error;No key-phase signal