How to Increase the Maximum Frequency of DFT While Keeping the Speed of Sampling

概要

論文の詳細を見る
In harmonic analysis by sampling, the maximum frequency that can be defined depends on the representative number of samples. Frequencies higher than the Nyquist frequency are aliased with those in the lower frequency range. Consequently, this aliasing effect gives us inferior information. The work in this paper is put on the mathematical basis that a sequence of frequency components on the DFT is folded back and forth regularly with the space by a factor of the Nyquist frequency. The filters are used prior to sampling so that only information of interest is contained in the filtered data. Then, the filtered data are separately analyzed with the FFT algorithm. Correct information of the inherent frequencies is obtained by solving a matrix equation in terms of the transfer characteristics of the filters and the DFT coefficients of the filtered data. The correct information about the inherent frequencies above the Nyquist frequency is thus obtained while keeping the speed of sampling.
一般社団法人情報処理学会の論文
1985-02-05