Method and device for frequency estimation of subsampled waveform under strong noise interference
A technology of frequency estimation and sampling frequency, which is applied in the field of frequency estimation of subsampled waveforms, can solve problems such as deviation, large error in reconstruction results, and unsuitable frequency estimation, and achieve high accuracy, strong adaptability, and improved anti-noise performance Effect
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0056] A frequency estimation method for undersampled waveforms under strong noise interference, see figure 1 , the frequency estimation method includes the following steps:
[0057] 101: respectively with f s1 ~ f sL For the sampling rate, L channels of low-speed sampling are performed on high-frequency signals containing noise interference to obtain L channels of sampling signals x i (n)(i=1,...,L);
[0058] 102: L channels of signal x obtained by sampling at a low rate i (n) Perform Fast Fourier Transform (FFT), and obtain f for each channel si (1≤i≤L) sample points and store them;
[0059] 103: Use the L-channel FFT spectrum distribution characteristics to extract the frequency corresponding to the peak position of the amplitude spectrum as the noise-containing remainder of each channel output, and set the residual error of each channel to be Δr i , that is Among them, r i is the signal frequency f 0 Modulo each sampling rate f si remainder obtained after.
...
Embodiment 2
[0066] Below in conjunction with specific calculation formula, example the scheme in embodiment 1 is described in detail:
[0067] 201: L-channel low-speed undersampling;
[0068] Let the received noisy signal be x(t)=a exp(j2πf 0 t)+ω(t), f 0 is the signal frequency, that is, the measured frequency. Perform L-way undersampling on it, and the sampling frequency is f s1 ~ f sL , sampling frequency f si (1≤i≤L) requires that the common divisor is M, and f si The factor Γ obtained after dividing by the common divisor M i It is pairwise prime. Then each sampling signal x i (n) is:
[0069]
[0070] 202: Perform FFT transformation on each channel of the sampled signal, analyze the spectrum characteristics, and extract the remainder;
[0071] For the L signal x obtained after sampling i (n), respectively do the sampling frequency f si Point FFT transformation, and obtain the magnitude spectrum of the i-th path, find out the frequency value corresponding to the largest...
Embodiment 3
[0094] Below in conjunction with concrete experimental data, accompanying drawing, the scheme in embodiment 1 and 2 is carried out feasibility verification, see the following description for details:
[0095] Analysis of results
[0096] Experiment 1 screening error remainder
[0097] It may be advisable to take L=6 channels of undersampling and K=4 channels with small error residuals as an example to illustrate the detailed process of screening error residuals. Assume that there is a large error in the remainder of the second path, that is, the error Δr 2 >M / 4, and set the sampling frequency as f s1 =18M,f s2 =19M, f s3 =23M, f s4 =25M, f s5 =29M,f s6 =31M, where M=801 is f s1 ~ f s6 greatest common divisor of . The signal frequency is set to f 0 =N=250000Hz.
[0098] According to the closed CRT algorithm of literature [10], the correct remainder set is r={4894 6496 10501 9700 177101690}, and the correct folded integer n i The set is n={17 16 13 12 10 10}, and th...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


