LU101400B1 - High-precision estimation method and apparatus for carrier frequency of amplitude modulation signal - Google Patents

Abstract

The present invention discloses a high-precision estimation method and apparatus for a carrier frequency of an amplitude modulation signal, comprising: sampling an input amplitude modulation signal at a known sampling rate, and collecting a total of 3N-1 sampling points; performing Nth-order all-phase FFT spectral analysis of first 2N-1 sampling points to obtain a spectral analysis result Y_1(k), k = 0,...,N-1; performing Nth-order all-phase FFT spectral analysis of subsequent 2N-1 sampling points to obtain a spectral analysis result Y_2(k), k = 0,...,N-1 ; finding peak spectral positions k* of the all-phase FFT spectral analysis results Y_1(k) and Y_2(k), and extracting phase spectrum values p_1(k*) and p_2(k*} of Y_1(k*) and Y_2(k*) ; and estimating a digital angular frequency w_c=(p_1(k*)-p_2(k*))/N of the carrier frequency, and converting the digital angular frequency into an analog frequency, namely, obtaining a final carrier frequency estimate f_c=(w_c/2pi)f_s . The present invention directly samples the amplitude modulation signal, and performs carrier frequency estimation by means of digital signal processing, without additional hardware facilities such as a down converter and a local oscillator.

Description

I 1 BL-5111 LU101400 HIGH-PRECISION ESTIMATION METHOD AND APPARATUS FOR CARRIER

FREQUENCY OF AMPLITUDE MODULATION SIGNAL Technical field The present invention belongs to the technical field of digital signal processing, and particularly relates to a problem of how to detect a carrier frequency of an amplitude modulation signal with a high precision, and more particularly to a high-precision estimation method and apparatus for the carrier frequency of the amplitude modulated signal. Technical background Amplitude Modulation (AM) is a commonly used signal modulation method. Compared with frequency modulation and phase modulation, amplitude modulation waves occupy a narrower bandwidth. Therefore, an AM modulation method is adopted in amplitude modulation radio stations, civil aviation VHF ground-to-air communication [1], maritime radio MF/HF wave band communication [2] and other long-distance communication, for example. Further, during an operation of a power system, there will be dynamic processes such as flicker and out-of-step, which often cause a certain low-frequency fault wave to fluctuate against the amplitude and frequency of the grid [3]. Therefore, estimation and spectral analysis of a carrier frequency of an amplitude modulation signal has a high engineering significance.

In a wireless channel, a baseband signal is modulated on a high frequency carrier for sending and transmitting, and at a receiving and demodulating end, the estimation precision of the carrier frequency directly determines the recovery performance of the baseband signal. Therefore, high-precision frequency estimation of carrier signals has high application value in both military and civilian fields. In the military field, the estimation of the carrier frequency is the premise of

| 2 BL-5111 LU101400 interference or interception of enemy communications.

Once the carrier frequency of the enemy communication signal is known, enemy information can be intercepted, so that reconnaissance and anti-reconnaissance strategies can be formulated in a targeted manner.

Applications in the civilian field include radio management for signal validation, interference identification, spectrum monitoring and the like.

However, due to the crystal stability of the system transmitter and receiver, the Doppler effect (e.g. in civil aviation ground-to-air communication, an aircraft may generate the Doppler effect during moving) and other effects, the carrier frequency often drifts, which will increase the difficulty of estimating the carrier frequency.

The estimation of the carrier frequency may be performed in the time domain or may be performed in the frequency domain.

A zero-crossing method based on RF signal is the most typical time domain estimation method, but the method is sensitive to noise.

The frequency domain estimation method includes a periodogram method and a frequency centering method.

The periodogram method [4] is based on maximum likelihood estimation.

The highest peak of the periodogram is estimated as the carrier frequency.

However, the periodogram method just obtains power spectra of signals and the phase information is lost, and thus the precision of frequency estimation is limited.

The frequency centering method is applied to symmetric signals, and the estimation effect for asymmetric signals is not good.

A document [5] proposes a method for estimating a carrier frequency by an instantaneous frequency obtained from an instantaneous phase containing a linear phase component.

It can be seen that calculating an instantaneous phase of a sampling point sampled by a demodulation end is the key to the carrier frequency estimation.

In order to solve the instantaneous phase, the document introduces Hilbert transformation, median filtering, multiple average mathematics expectation, etc., and the algorithm is more complicated.

However, a document [6] proposes a carrier frequency estimation method based on second-order cyclic cumulants.

However, the method is only applicable for digital communication and is not

BL-5111 LU101400 applicable for analog communication.

The present invention will utilize a spectral analysis method of all-phase FFT [7] [8].

The all-phase FFT has excellent spectral leakage suppression performance and "phase invariance", and by using these characteristics, an instantaneous phase of a signal can be measured with a high precision. The present invention applies the all-phase FFT to the analysis of the AM signal and theoretically analyzes the all-phase FFT phase spectrum features of the AM signal, and proposes an “all-phase time-shift phase difference carrier frequency estimation method", achieving a better effect.

[References]

[1] ZHU Dao-xian, JI Song-hai. Overview of Civil Aviation VHF Air/Ground Voice Communication System [J]. Air Traffic Management, 2004(3):46-48.

[2] YANG Yong-kang, MAO Qi-huang. Marine Radio Communication[M]. Version 2. Beijing, China Communications Press, 2009.

[3] ZHANG Dao-nong, YU Yue-hai. The Technology and Application of Synchronous Phasor Measurement in Power System[M]. Beijing, China Electric Power Press, 2017.

[4] So, H.C., Chan, Y.T., Ma, Q., Ching, P.C.. Comparison of Various Periodograms for Single Tone Detection and Frequency Estimation[J]. IEEE Trans. on Aerospace and Electronic Systems, 1999, 35(3):945-952.

[5] LIU Gui-jjang, FENG Xiao-ping. Carrier Frequency Estimation Method for Digitally Modulated Signal[J]. Systems Engineering and Electronics, 2004, 26(12):1787-17889.

[6] GONG Mu-dan, GUO Rong-hui. Carrier Frequency Estimation Based on

! 4 BL-5111 LU101400 Second-order Cyclic Cumulants[J]. Computer Engineering, 2011, 37(20):81-82+86.

[7] WANG Zhao-hua, HOU Zheng-xin, SU Fei. All-phase FFT spectral analysis[J]. Journal on Communications, 2003, 24(11A):6-19.

[8] HUANG Xiang-dong, WANG Zhao-hua. All-Phase Digital Spectral analysis Method[M]. Beijing: Science Press, 2017. Summary of the invention An object of the present invention is to overcome the deficiencies in the prior art, and to propose a high-precision estimation method and apparatus for a carrier frequency of an amplitude modulation signal, which directly samples the amplitude modulation signal, and performs carrier frequency estimation by means of digital signal processing without additional hardware facilities such as a down converter and a local oscillator. The carrier frequency estimation algorithm has high efficiency and low computational complexity, and can avoid heavy calculations. The object of the present invention is achieved by the following technical solution.

A high-precision estimation method for a carrier frequency of an amplitude modulation signal of the present invention, comprising the following steps: step 1, sampling an input amplitude modulation signal at a known sampling rate £, and collecting a total of 3N-1 sampling points x,,...,x;y_,; step 2: performing Nth-order all-phase FFT spectral analysis of first 2N-1 sampling points, i.e. x,,....x,,, to obtain a spectral analysis result Y(k), k=0,...,N-1; and also performing Nth-order all-phase FFT spectral analysis of subsequent 2N-1 sampling points, ie. x,,.….,x,y,, to obtain a spectral analysis result | |

! | BL-5111 LU101400 Y,(k), k=0,.,N-1; step 3, finding peak spectral positions k" of the all-phase FFT spectral analysis results Y(k) and Y,(k), and extracting phase spectrum values ¢(k’) and 5 @(k') of Y(k') and Y,(k'); and CL a x _ pk )-9,(k") ; step 4, estimating a digital angular frequency ©, = N of the carrier frequency, and converting the digital angular frequency into an analog frequency, namely, obtaining a final carrier frequency estimate f. ==, ;

TT The object of the present invention may also be achieved by the following technical solution. An apparatus for the high-precision estimation method for the carrier frequency of the amplitude modulation signal described above, comprising a digital signal processor, wherein an output terminal of the digital signal processor is connected with an output driver and a display module thereof; an I/O port of the digital signal processor is connected with an analog-to-digital converter; a clock input port of the digital signal processor is connected with a main clock module; and a clock output port of the digital signal processor is connected to the analog-to-digital converter. Compared with the prior art, the technical solutions of the present invention have the following beneficial effects: (1) The estimation method of the present invention does not depend on any additional analog circuits, and directly samples the amplitude modulation signal, and then measures the carrier frequency by constructing a digital signal processing

BL-5111 LU101400 algorithm. (2) The present invention utilizes a simple flow of the all-phase FFT to greatly reduce the computational complexity, so that the algorithm core is quickly realized, the calculation amount of subsequent steps is greatly reduced, and the timeliness is improved. (3) The present invention performs carrier frequency measurement at the highest spectral line corresponding to a carrier, the carrier frequency representing a portion with the highest energy of a wireless transmission signal, and it has high frequency measurement accuracy and strong anti-noise capability. (4) The carrier measurement and recovery of the present invention is a premise of communication, and carrier synchronization is established to facilitate demodulation of subsequent baseband signals.

Brief description of the drawings Fig.1 is an all-phase FFT spectral analysis diagram (N=4),

Fig. 2 is conventional FFT and all-phase FFT amplitude spectra and phase spectra of an AM signal, Fig. 3 is a flow chart of a phase difference frequency measurement method for eliminating phase blur; Fig. 4 is sampling waveform diagrams of a baseband voice signal and an AM signal, Fig. 5 is all-phase FFT amplitude spectra and phase spectra of an AM signal;

BL-5111 LU101400 Fig. 6 is a hardware implementation diagram of the present invention; and Fig. 7 is a flow chart of a program inside a DSP.

Detailed description of the embodiments The present invention will be further described below with reference to the accompanying drawings.

The present invention firstly sets forth an operation flow of a technical solution, next gives internal technical details and principles, then summarizes a processing flow chart of the technical solution, and finally gives an experiment to verify the technical solution.

A high-precision estimation method for a carrier frequency of an amplitude modulation signal of the present invention is specifically implemented by the following process. The carrier frequency of the amplitude modulation signal can be estimated by processing as follows.

Step 1, an input amplitude modulation signal is sampled at a known sampling rate £, and a total of 3N-1 sampling points x,,...,x,,_, are collected. Step 2: Nth-order all-phase FFT spectral analysis of first 2N-1 sampling points, i.e.

x,,...X,y_, IS performed to obtain a spectral analysis result Y,(k), k=0,...,N-1; and Nth-order all-phase FFT spectral analysis of subsequent 2N-1 sampling points, Le. Xp Xp, is also performed to obtain a spectral analysis result Y,(k), k=0,.,N-1.

| 8 BL-5111 LU101400 Step 3, peak spectral positions k" of the all-phase FFT spectral analysis results Y(k) and Y,(k) are found, and phase spectrum values ¢ (k’) and p,(k') of Y(k') and Y,(k') are extracted. A A p(k), (kK) . ; Step 4, a digital angular frequency ©, = N of the carrier frequency is estimated, and the digital angular frequency is converted into an analog frequency, namely, obtaining a final carrier frequency estimate f. = Ses. ;

A Brief description of all-phase FFT spectral analysis: A processing procedure of the all-phase FFT spectral analysis is shown in Fig. 1. As can be seen from Fig. 1, it is only necessary to use a convolution window w, of length (2N-1) to weight data of (2N-1) before and after a center sampling point x(0), next add the data with an interval between two of N to form N data, and then perform FFT with the number of points N, namely, obtaining the apFFT result. If the computational complexity is considered, it can be seen from Fig. 1 that the Nth-order apFFT only increases (2N-1) multiplication and accumulation operations compared with the direct FFT with the number of points N. The part of additional multiplication has small overhead, but it can bring about a great improvement in performance. The convolution window in Fig. 1 is convoluted by a front window f of length N and an overturned rear window b, namely: we(n)=f(n)*b(-n) -N+1<n< N-1 (1) For the convenience of research, fand b are usually taken as the same symmetric window, that is, f= b, and then the Fourier transform of window fis expressed as: i |

| ! 9 BL-5111 LU101400 - De F,(jo)=F,(w)e * (2) Let the FFT frequency resolution be Aw=2n/N, the signal digital angular frequency wo = BA (Note: ß may be a decimal). The document [7] proves that the conventional FFT spectrum X(k) plus the f window and the apFFT spectrum Y(k) of the complex exponential sequence {x(n)= e/@w"9), -N+1< n<N-1} are as follows: 641 sk) X(k)=e I) |rlg-Wrolre N-1] (3) Y(K)=e".F°{(B-K)Ao] From Equation (3), we will have: [Y(k)| =|X(k)*, ke[0,N-1] (4) From Equation (3), the phase spectrum px(k) of the FFT spectrum X(k) and the phase spectrum pxk) of the apFFT spectrum Y(k) can be expressed as: 1 EL ke[o,N-1] (5) py(k)=0 Two excellent properties of the apFFT can be derived from Equations (4) and (5).

Property 1: a normalized Nth-order all-phase FFT spectral amplitude value of the sequence {x(n)=e"«"*9} is equal to the square of the conventional FFT amplitude spectrum with the number of points N. Property 2: the apFFT has a "phase invariance property", and this property contains two layers of meaning: (1) the phase spectrum value on the apFFT main

' | 10 BL-5111 LU101400 line is equal to the theoretical phase value of the center sampling point of the input 2N-1 samples; and (2) there is no dependence between the phase spectrum value of the apFFT main line and the signal frequency offset.

The squared relationship described by Property 1 and Equation (4) is for all spectral lines, which inevitably makes the amplitude values of side lines of the all-phase FFT decay in a squared relationship relative to the amplitude value of the main line compared with the conventional FFT, so that the main spectrum is more prominent. Then, when a signal contains multiple frequency components, the inter-spectrum interference is greatly reduced compared with the conventional FFT, and thereby the apFFT has excellent spectral leakage suppression properties. Property 2 means that the apFFT block diagram shown in Fig. 1 can form a high-precision "phase measuring instrument" wherein: without any additional correction measures, the Y(k) phase value at the output peak spectral line is directly taken to "measure" the theoretical phase 0 of the center sampling point x(0). All-phase FFT spectral analysis of amplitude modulation signal: To briefly explain the principle of the full-phase FFT spectral analysis of the AM signal, it is assumed that the modulation signal is a single-frequency cosine signal with digital angular frequency of w; and an initial phase value of ¢;, and for a carrier, digital angular frequency is w., and an initial phase value is po; and a DC bias size is m. Then, the sampling sequence of the AM signal may be expressed as: x(n) = Acos(@.n + mo) [cos(ann +g) +m] (6) Further simplifying equation (6), we will have:

| 11 BL-5111 LU101400 x(m=5 {eos[(æ + an + go + 00] mi + cos[(æ, — a n+ (po — 91) |} + mA cos(w,n + po) Equation (7) shows that after modulating the carrier, a sum frequency term and a difference frequency term are generated, and another term of mAcos(wen+wo) still completely retains the carrier information.

From the viewpoint of spectral distribution, since the digital angular frequency «. of the AM carrier is much larger than the digital angular frequency w;, the spectral positions of three frequency components w., ww, and wo; in Equation (7) will be closely spaced.

If the spectrum analysis method is improper, the three frequency components are highly likely to cause the inter-spectrum interference and affect the accuracy of the spectrum analysis, thereby reducing the precision of extracting the carrier information.

If the all-phase FFT spectral analysis is performed on the sampling sequence of Equation (7), since the apFFT has excellent spectral leakage suppression performance, the apFFT can reduce the inter-spectrum interference of the three frequency components of Equation (7). In addition, according to Property 2, since the apFFT has the "phase invariance", the phase information po can be extracted very accurately as long as the phase value corresponding to the peak spectral line at the carrier œ is directly solved.

Let N=128, then the frequency resolution of the FFT is Aw=2n/128. Let the carrier frequency be w=40.2Awœ, the initial phase value be po=60°, the modulation frequency be w;=4.1Aw, the initial phase value be ¢,=30° 4=1, and m=1. The conventional FFT spectrum analysis plus a Hanning window and the all-phase FFT spectrum analysis of the AM sequence (n € [-N+1,N-1]) shown in Equation (6) are separately performed.

The comparison diagrams of corresponding amplitude spectra and phase spectra are shown in Fig. 2.

| 12 BL-5111 LU101400 It can be seen from Figs. 2(a) and 2(b) that, the amplitude spectra of the conventional FFT and the apFFT both correspond to three clusters of spectral lines of the sum frequency (centered at k=44), the carrier frequency (centered at k=40), and the difference frequency (centered at k=36), but the degree of interference between the three clusters of spectral lines of the apFFT is much lower than that of the conventional FFT (which is expressed as MG! at the inter-spectrum interference positions k=39, 42 and 43 being much smaller than |X(k)). In addition, it can be seen from the phase spectra that, the phase spectrum of the conventional FFT spectrum is very turbulent, and the carrier phase information cannot be extracted intuitively, whereas the phase spectrum of the all-phase FFT is not so, wherein the values of three phase spectral lines near the carrier frequency k=40 are almost equal to the theoretical value po=60 °, the values of multiple phase spectral lines near the sum frequency k=44 are equal to po+p1=90°, and the values of multiple phase spectral lines near the difference frequency k=36 are equal to @o-p1=30°. That is, the apFFT can very accurately extract the phase information of each frequency component term of Equation (7). Since the energy of the carrier frequency component is the largest, it corresponds to the highest spectral line position on the amplitude spectrum, and the phase spectrum precision measured from the phase spectrum is also the highest.

Principle for carrier frequency estimation of amplitude modulation signal: According to the first layer of meaning of the apFFT "phase invariance" that: the phase spectrum value on the apFFT main line is equal to the theoretical phase value of the center sampling point of the input 2N-1 data, for the sequence { x(n)=e/e"0, -N+1<n<N-1}, it is assumed that its apFFT main line position is Æ, and then the value of o,(k ) is equal to the theoretical phase value 6 of its center sampling point x(0). If the sequence is delayed by ny sampling intervals, a new sequence { x(n=e" "9, -N+1-no<n<N-1-no } can be obtained, and at this time, the

| 13 BL-5111 LU101400 value of p2 (k') is equal to the theoretical phase value 0-w.no of its center sampling point x(-no). Then, the difference between ¢;k* and pz(k*) is obtained, namely, obtaining a frequency estimation formula: pW)-0 =a _ ok) 0k) _Agp @) p,(k )=60-@" Ho My When ny=N, the estimation precision of Equation (8) is the most accurate, and the "phase blur phenomenon” of the phase difference method can also be eliminated. Then, the entire analysis process based on the "all-phase time-shift phase difference frequency measurement method" is shown in Fig. 3. Note that as shown in Fig. 1, the Nth-order apFFT requires 2N-1 sampling points, while Fig. 3 shows that the apFFT is performed on two sequences of length 2N -1 at a delay of N and N data are overlapping between the two sequences. Therefore, the frequency measurement of Fig. 3 requires a total of 3N-1 consecutive sampling points. Although the measurement flow chart of Fig. 3 is proposed for a single-frequency complex index signal, since the apFFT can suppress the inter-spectrum interference of each frequency component to a low level, and the AM signal carrier frequency component corresponds to the maximum amplitude spectrum value, the effects of the sum frequency, the difference frequency and other interference frequencies are small. Therefore, the frequency measurement flow chart of Fig. 3 is still applicable for the AM signal.

Experiment: In order to verify the AM carrier estimation performance based on the all-phase FFT, a simulation experiment is carried out. The modulation signal used in the | |

BL-5111 LU101400 experiment is a live voice recording of “Hello, everybody” (the audio sampling frequency is 22.05 kHz), wherein the carrier frequency is #=1x10°Hz, the carrier amplitude is 1, the DC offset m=0.5, the sampling frequency is /=4.400000x10°Hz, and the sampling waveforms of its baseband signal and AM signal are as shown in Fig. 4. Fig. 5(a) gives the apFFT amplitude spectrum of the AM signal (it can be seen that the carrier frequency is at k=233, and both sides are the voice baseband components, showing the harmonic distribution). Fig. 5(b) gives the amplitude spectrum near the carrier spectrum line. Figs. 5(c) and 5(d) give the apFFT phase spectra of two delay sequences near the carrier spectrum line, respectively. It can be seen that the phase spectra near the carrier spectrum line present a flat "phase invariance". Specifically, the apFFT phase value of the first sequence is 32.9537 degrees, and the apFFT phase value of the second sequence is -66.1044 degrees.

Note that the frequency calculated by Equation (8) is the digital angular frequency, which can be converted to the analog estimation frequency of the carrier according to Equation (9): 0 f=, (9) The simulation signal-to-noise ratio is set to 17.8dB, and Z=100 Monte Carlo simulation is performed for each N value. It is assumed that the kth carrier frequency is estimated to be f.(k), and according to Equations (10) and (11), the Root-of-Mean-Square Error (RMSE) is calculated: — 1 L f ==> FA) (10)

LS

| 15 BL-5111 LU101400 1 & — 2 RMSE = [F310 7] (11) Table 1 gives the estimation results of the carrier frequency when N takes different values.

Table 1. Root-of-Mean-Square Error of carrier frequency estimation based on the all-phase FFT N 512 1024 2048 Method of the 1.1046 0.9650 0.5566 present invention Conventional 1.4973 1.2259 0.6438 phase difference method lt can be seen from Table 1 that the precision of AM carrier estimation based on the all-phase FFT is higher, and the larger N is, the higher the precision is. Of course, the required storage amount and calculation amount also increase. The hardware for implementing the present invention will be briefly described below. An apparatus for the high-precision estimation method for the carrier frequency of the amplitude modulation signal of the present invention, of which a hardware implementation diagram is as shown in Fig. 6, includes a digital signal processor DSP, wherein an output terminal of the digital signal processor DSP is connected with an output driver and its display module; an VO port of the digital signal processor is connected with an analog-to-digital converter A/D; a clock input port of

! | 16 BL-5111 LU101400 the digital signal processor A/D is connected with a main clock module; and a clock output port of the digital signal processor DSP is connected to the analog-to-digital converter A/D. An acquired amplitude modulation signal x(s) is sampled by the analog-to-digital converter A/D to obtain a sample sequence x(n), and it enters the digital signal processor DSP in the form of a parallel digital input, and is processed by the internal algorithm of the DSP chip to obtain an estimate of a mixing matrix. Finally, the estimate value of the mixing matrix is displayed by means of the output driver and its display module. Herein, the digital signal processor DSP of Fig. 6 is a core device, and the following main functions are completed in the signal parameter estimation process: (1) calling a core algorithm to complete the all-phase FFT, peak spectrum search, phase difference extraction, digital angular frequency calculation, and carrier frequency conversion; (2) multiple sampling, using the carrier frequency calculated by the core algorithm multiple times, and taking their mean values to improve the estimation precision; and (3) outputting the result to the output driver and its display module.

The internal program flow of the digital signal processor DSP is shown in Fig. 7. The present invention embeds the proposed core estimation algorithm of the “high-precision estimation method and apparatus for a carrier frequency of an amplitude modulation signal’ into the digital signal processor DSP, and based on this, completes the high-precision, low-complexity and high-efficiency estimation of a source signal number and a mixing matrix.

|

| | 17 BL-5111 LU101400 The flow of Fig. 7 is divided into the following steps: 1) firstly, setting the number of sampling points of a signal to 3N-1 according to specific application requirements; 2) secondly, a CPU main controller in the digital signal processor DSP reading sampling data from the I/O port, and entering an internal RAM; and 3) finally, according to the processing procedure of the present invention in Fig. 2, performing a high-precision estimation of the carrier frequency of the amplitude modulation signal, and displaying the recovered signal by an external display apparatus. Although the functions and working processes of the present invention have been described above with reference to the accompanying drawings, the present invention is not limited to the specific functions and working processes described above, and the specific embodiments described above are merely illustrative and not limiting. Many forms may be made by a person ordinarily skilled in the art under the inspiration of the present invention, without departing from the gist of the present invention and the scope protected by the claims, and all of these are within the protection of the present invention.

| |

Claims (2)

| 18 BL-5111 LU101400 Claims

1. A high-precision estimation method for a carrier frequency of an amplitude modulation signal, characterized in that it comprises the following steps: step 1, sampling an input amplitude modulation signal at a known sampling rate £, and collecting a total of 3N-1 sampling points x,,...,x;, ,; step 2: performing Nth-order all-phase FFT spectral analysis of first 2N-1 sampling points, i.e. x,,...,x,,_, to obtain a spectral analysis result Y(t), k=0,..,N—1; and also performing Nth-order all-phase FFT spectral analysis of subsequent 2N-1 sampling points, ie. x,,..x;,,, to obtain a spectral analysis result Y,(k), k=0,.,N-1; step 3, finding peak spectral positions k’ of the all-phase FFT spectral analysis results Y(k) and Y,(k), and extracting phase spectrum values p(k') and p,(k") of Y(k') and Y,(k); and CL a x _ pK) =p, (K) ; step 4, estimating a digital angular frequency ©, = y of the carrier frequency, and converting the digital angular frequency into an analog frequency, namely, obtaining a final carrier frequency estimate f. = fi

TT

2. An apparatus for the high-precision estimation method for a carrier frequency of an amplitude modulation signal according to claim 1, characterized in that it comprises a digital signal processor, wherein an output terminal of the digital signal | |

BL-5111 LU101400 processor is connected with an output driver and a display module thereof; an 1/0 port of the digital signal processor is connected with an analog-to-digital converter; a clock input port of the digital signal processor is connected with a main clock module; and a clock output port of the digital signal processor is connected to the analog-to-digital converter. | |