Digital signal processing methods, frequency modulated continuous wave signal processing methods and devices

Abstract

A digital signal processing method, a frequency-modulated continuous wave signal processing method, and an apparatus are disclosed. The digital signal processing method includes: acquiring digital signals corresponding to multiple frequency points in the frequency domain; determining a window parameter value corresponding to each of the multiple frequency points, wherein the window parameter value for each frequency point minimizes the energy of the digital signal corresponding to that frequency point after windowing processing based on the window parameter value, and the time-domain window function corresponding to the window parameter value is non-negative; and performing windowing processing on the digital signals corresponding to the multiple frequency points according to the window parameter values. This disclosure embodiment can achieve windowing sidelobe suppression while simultaneously enabling the detection of distant weak targets and distinguishing between nearby targets.

Description

[0001] This application claims priority to Chinese Patent Application No. 202210219018.5, filed on March 3, 2022, entitled "A Method for Adding Windows", the contents of which shall be construed as incorporated herein by reference. Technical Field

[0002] This disclosure relates to, but is not limited to, the field of signal processing technology, and particularly to a digital signal processing method, a frequency modulated continuous wave signal processing method, and an apparatus. Background Technology

[0003] The Fast Fourier Transform (FFT) is a fast and efficient implementation of the Discrete Fourier Transform (DFT), capable of transforming a power of 2 number of sample points. FFT has the advantages of low computational cost and ease of hardware implementation, and has been widely used in the field of information processing technology. It is also one of the most important techniques for spectrum analysis.

[0004] However, since the sampled signal processed by FFT is a data segment of finite length, the finite data length may lead to the truncation of the sampling period, resulting in the appearance of frequency components in the transformed spectrum that are not present in the signal itself, i.e., spectral leakage.

[0005] In digital signal processing, sampled data can be windowed by multiplying it by a window function, such as a rectangular window, Hanning window, Hamming window, or Blackman window. Performing an FFT on the windowed data can reduce spectral leakage, but its performance still needs further improvement. Summary of the Invention

[0006] This disclosure provides a digital signal processing method, which may include:

[0007] Acquire digital signals corresponding to multiple frequency points in the frequency domain;

[0008] Determine the window parameter value corresponding to each of the plurality of frequency points. The window parameter value corresponding to each frequency point makes the energy of the digital signal corresponding to that frequency point after windowing processing based on the window parameter value minimized, and the time-domain window function corresponding to the window parameter value is non-negative.

[0009] Windowing processing is applied to the digital signals corresponding to the multiple frequency points based on the window parameter values ​​corresponding to those frequency points. For example, for any given frequency point, windowing processing is applied to the digital signal corresponding to that frequency point based on the window parameter values, thereby achieving dynamic windowing operation on the digital signals at each frequency point and effectively suppressing problems such as spectral leakage caused by the discrete spectrum analysis (such as FFT).

[0010] In some optional embodiments, the window parameters may include multiple parameters, and the time-domain window functions corresponding to the multiple window parameters are equal to 0 at both ends of the window. That is, the factor (i.e., window parameter) for windowing operation at each frequency point can be one, two or more, and can be selected according to actual needs.

[0011] In some optional embodiments, the window parameters may include a first window parameter and a second window parameter, wherein the second window parameter can be determined based on the obtained first window parameter; windowing processing of the digital signals corresponding to the plurality of frequency points according to the window parameter values ​​corresponding to the plurality of frequency points may include:

[0012] A convolution operation is performed on the digital signals of multiple consecutive frequency points centered at frequency point k to obtain the digital signal corresponding to frequency point k after windowing. The weights of the digital signals of the multiple consecutive frequency points in the convolution operation are determined according to the first window parameter value corresponding to frequency point k, where k = 0, 1, ..., N-1, and N is the number of frequency points.

[0013] In some optional embodiments, the plurality of consecutive frequency points include frequency point k-2, frequency point k-1, frequency point k, frequency point k+1, and frequency point k+2;

[0014] The first window parameter α corresponding to frequency point k 1k The value is obtained by the following formula:

[0015]

[0016] Where, β 1k To determine the parameter, β 1k It is determined based on the digital signals corresponding to frequency points k-2, k-1, k, k+1, and k+2.

[0017] In some alternative embodiments, β 1k We obtain it from the following formula:

[0018]

[0019] Where real represents finding the real part of a complex number, and X(k), X(k-1), X(k+1), X(k-2), and X(k+2) are the digital signals corresponding to frequency points k, k-1, k+1, k-2, and k+2, respectively.

[0020] In some optional embodiments, the digital signal X corresponding to the windowed frequency point k is... W (k) is:

[0021]

[0022]

[0023] Where real represents finding the real part of a complex number, and X(K), X(k-1), X(k+1), X(k-2), and X(k+2) are the digital signals corresponding to frequency points k, k-1, k+1, k-2, and k+2, respectively.

[0024] This application also provides a digital signal processing method, which may include:

[0025] Acquire digital signals corresponding to multiple frequency points in the frequency domain. These digital signals can be FFT processing results data in the radar signal processing process, such as range-dimensional FFT result data, velocity-dimensional FFT result data, and / or angle-dimensional FFT result data, etc.

[0026] For any given frequency point, dynamic windowing is applied to that frequency point based on the digital signals of multiple frequency points continuously distributed around that frequency point.

[0027] In some optional embodiments, the step of dynamically windowing a frequency point based on the digital signals corresponding to multiple frequency points continuously distributed around that frequency point includes:

[0028] For any given frequency point, the judgment parameters are obtained based on the real part of the digital signals of multiple frequency points continuously distributed around that frequency point.

[0029] Based on the comparison result between the judgment parameter and the preset threshold, the digital signals corresponding to multiple frequency points continuously distributed around the current frequency point and at least a portion of the judgment parameter are selected to obtain the windowed digital signal of the current frequency point.

[0030] In some optional embodiments, a judgment parameter β is obtained based on the real part of the digital signals of five consecutively distributed frequency points centered on that frequency point. 1k We obtain it from the following formula:

[0031]

[0032] Where real represents finding the real part of a complex number, and the five frequency points continuously distributed around the current frequency point are frequency point k-2, frequency point k-1, current frequency point k, frequency point k+1, and frequency point k+2 respectively. The digital signal corresponding to frequency point k-2 is X(k-2), the digital signal corresponding to frequency point k-1 is X(k-1), the digital signal corresponding to frequency point k is X(k), the digital signal corresponding to frequency point k+1 is X(k+1), and the digital signal corresponding to frequency point k+2 is X(k+2).

[0033] In some optional embodiments, the digital signal at the current frequency point is obtained by selecting at least a portion of the digital signals corresponding to five consecutively distributed frequency points centered on the current frequency point and the judgment parameters; the preset threshold is 0 and 4 / 3; the windowed digital signal X at the current frequency point k. W The expression for (k) is:

[0034]

[0035] Where, β 1k To determine the parameters, the five frequency points continuously distributed around the current frequency point are frequency point k-2, frequency point k-1, current frequency point k, frequency point k+1, and frequency point k+2, respectively. The digital signal corresponding to frequency point k-2 is X(k-2), the digital signal corresponding to frequency point k-1 is X(k-1), the digital signal corresponding to frequency point k is X(k), the digital signal corresponding to frequency point k+1 is X(k+1), and the digital signal corresponding to frequency point k+2 is X(k+2).

[0036] This application also provides a frequency-modulated continuous wave signal processing method, which may include:

[0037] The signal obtained by mixing the transmitted and echo signals of the frequency-modulated continuous wave is sampled to obtain a digital signal set including multiple chirps;

[0038] The digital signal sequence in the digital signal set is windowed according to any of the digital signal processing methods described above.

[0039] In some optional embodiments, the windowing process of the digital signal sequence in the digital signal set according to any of the digital signal processing methods described above includes at least one of the following:

[0040] Perform FFT on the digital signal sequence inside the Chirp to obtain the distance dimension digital signal corresponding to multiple frequency points in the frequency domain, and perform windowing processing on the distance dimension digital signal according to any of the digital signal processing methods described above.

[0041] Perform an FFT on the digital signal sequence between the Chirps to obtain velocity-dimensional digital signals corresponding to multiple frequency points in the frequency domain, and then perform windowing processing on the velocity-dimensional digital signals according to any of the digital signal processing methods described above.

[0042] Perform FFT on the angular dimension data to obtain angular dimension digital signals corresponding to multiple frequency points in the frequency domain, and then perform windowing processing on the angular dimension digital signals according to any of the digital signal processing methods described above.

[0043] In some optional embodiments, the windowing process of the digital signal sequence in the digital signal set according to any of the digital signal processing methods described above includes one of the following:

[0044] Perform an FFT on the digital signal sequence within the Chirp to obtain a distance-dimensional digital signal corresponding to multiple frequency points in the frequency domain, and then perform windowing processing on the distance-dimensional digital signal according to any of the digital signal processing methods described above; perform an FFT on the digital signal sequence between the Chirps to obtain a velocity-dimensional digital signal corresponding to multiple frequency points in the frequency domain, and then perform windowing processing on the velocity-dimensional digital signal according to any of the digital signal processing methods described above.

[0045] Perform an FFT on the digital signal sequence within the Chirp to obtain distance-dimensional digital signals corresponding to multiple frequency points in the frequency domain; perform an FFT on the digital signal sequence between the Chirps to obtain velocity-dimensional digital signals corresponding to multiple frequency points in the frequency domain; perform windowing processing on the distance-dimensional digital signals according to any of the above-described digital signal processing methods; perform windowing processing on the velocity-dimensional digital signals according to any of the above-described digital signal processing methods.

[0046] Perform an FFT on the digital signal sequence within the Chirp to obtain distance-dimensional digital signals corresponding to multiple frequency points in the frequency domain; perform an FFT on the digital signal sequence between the Chirps to obtain velocity-dimensional digital signals corresponding to multiple frequency points in the frequency domain; perform windowing processing on the velocity-dimensional digital signals according to any of the above-described digital signal processing methods; perform windowing processing on the distance-dimensional digital signals according to any of the above-described digital signal processing methods.

[0047] This application embodiment also provides a frequency-modulated continuous wave signal processing device, which may include: a sampling circuit and a baseband signal processing circuit, wherein:

[0048] The sampling circuit is configured to sample the signal after mixing the transmitted signal and the echo signal of the frequency-modulated continuous wave to obtain a digital signal set including multiple chirps;

[0049] The baseband signal processing circuit is configured to perform baseband processing on the digital signal set including multiple chirps, wherein the baseband processing includes windowing the digital signal sequence in the digital signal set according to any of the frequency modulated continuous wave signal processing methods described above.

[0050] In some optional embodiments, the baseband signal processing circuit includes: an FFT unit and a windowing processing unit, wherein:

[0051] The FFT unit is configured to perform FFT on the digital signal sequence within the Chirp to obtain distance-dimensional digital signals corresponding to multiple frequency points in the frequency domain; and to perform FFT on the digital signal sequence between the Chirps to obtain velocity-dimensional digital signals corresponding to multiple frequency points in the frequency domain.

[0052] The windowing processing unit is configured to perform windowing processing on the distance dimension digital signal and / or the velocity dimension digital signal according to any of the digital signal processing methods described above.

[0053] This application also provides a digital signal processing apparatus, which may include a Fourier transform module, a window parameter determination module, and a windowing processing module, wherein:

[0054] The Fourier transform module is configured to perform discrete Fourier transform on the digital signal to obtain digital signals corresponding to multiple frequency points in the frequency domain.

[0055] The window parameter determination module is configured to determine the window parameter value corresponding to each of the plurality of frequency points, wherein the window parameter value corresponding to each frequency point minimizes the energy of the digital signal corresponding to that frequency point after windowing processing based on the window parameter value, and the time-domain window function corresponding to the window parameter value is non-negative.

[0056] The windowing processing module is configured to perform windowing processing on the digital signals corresponding to the multiple frequency points according to the window parameter values ​​corresponding to the multiple frequency points.

[0057] In some optional embodiments, the window parameters include multiple parameters, and the time-domain window functions corresponding to the multiple window parameters are equal to 0 at both ends of the window.

[0058] In some optional embodiments, the window parameters include a first window parameter and a second window parameter, and the windowing processing module is configured as follows:

[0059] A convolution operation is performed on the digital signals of multiple consecutive frequency points centered at frequency point k to obtain the digital signal corresponding to frequency point k after windowing. The weights of the digital signals of the multiple consecutive frequency points in the convolution operation are determined according to the first window parameter value corresponding to frequency point k, where k = 0, 1, ..., N-1, and N is the number of frequency points.

[0060] In some optional embodiments, the plurality of consecutive frequency points include frequency point k-2, frequency point k-1, frequency point k, frequency point k+1, and frequency point k+2;

[0061] The first window parameter α corresponding to the frequency point k determined by the window parameter determination module 1k The value is obtained by the following formula:

[0062]

[0063] Where, β 1k It is determined based on the digital signals corresponding to frequency points k-2, k-1, k, k+1, and k+2.

[0064] In some optional embodiments, the window parameter determination module obtains β by the following formula. 1k :

[0065]

[0066] The windowing processing module obtains the digital signal X corresponding to the frequency point k after windowing processing using the following formula. W (k):

[0067]

[0068] Where real represents finding the real part of a complex number, and X(k), X(k-1), X(k+1), X(k-2), and X(k+2) are the digital signals corresponding to frequency points k, k-1, k+1, k-2, and k+2, respectively.

[0069] This application also provides a digital signal processing apparatus, which may include a memory and a processor;

[0070] The memory is used to store program instructions;

[0071] The processor is configured to invoke the program instructions stored in the memory to implement the digital signal processing method as described above.

[0072] The digital signal processing method, frequency modulated continuous wave signal processing method, and apparatus of this disclosure determine window parameter values ​​that correspond one-to-one with each of multiple frequency points, and use the window parameter values ​​corresponding to multiple frequency points to perform windowing processing on the digital signals corresponding to multiple frequency points respectively. This can achieve the effect of windowing sidelobe suppression, while also enabling the detection of weak targets at a greater distance, and distinguishing targets that are closer together.

[0073] Other features and advantages of this disclosure will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the disclosure. Other advantages of this disclosure may be realized and obtained by means of the methods described in the description and the accompanying drawings. Attached Figure Description

[0074] The accompanying drawings are used to provide an understanding of the technical solutions of this disclosure and form part of the specification. They are used together with the embodiments of this disclosure to explain the technical solutions of this disclosure and do not constitute a limitation on the technical solutions of this disclosure.

[0075] Figure 1 This is a schematic diagram of the transmitted and echo signals of an FMCW using sawtooth wave modulation.

[0076] Figure 2 This is a schematic diagram of a two-dimensional FFT processing flow for FMCW radar signals;

[0077] Figure 3a This is a schematic diagram illustrating the effect of adding windows to rectangular windows and Hanning windows in a simulated scenario.

[0078] Figure 3b for Figure 3a Enlarged view of region A in the middle;

[0079] Figure 4 This is a schematic diagram illustrating the effect of adding windows multiple times and then reducing them point by point.

[0080] Figure 5 This is a flowchart illustrating an exemplary embodiment of a digital signal processing method according to the present disclosure;

[0081] Figure 6a This is a flowchart illustrating an exemplary embodiment of a frequency-modulated continuous wave signal processing method according to the present disclosure.

[0082] Figures 6b to 6g This is a schematic flowchart of six frequency-modulated continuous wave signal processing methods according to embodiments of this disclosure;

[0083] Figure 7a A schematic diagram showing the comparison of the effects of using a rectangular window, a Chebyshev 80dB window, and the dynamic windowing method proposed in this disclosure to window the spectrum in the distance dimension;

[0084] Figure 7b for Figure 7a Enlarged view of region B in the middle;

[0085] Figure 8 This is a schematic diagram of the structure of a frequency-modulated continuous wave signal processing device according to an embodiment of the present disclosure;

[0086] Figure 9 This is a schematic diagram of the structure of a digital signal processing device according to an embodiment of the present disclosure;

[0087] Figure 10 This is a schematic diagram of another digital signal processing device according to an embodiment of the present disclosure. Detailed Implementation

[0088] To make the objectives, technical solutions, and advantages of this disclosure clearer, the embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. It should be noted that, unless otherwise specified, the embodiments and features described in this disclosure can be arbitrarily combined with each other.

[0089] Unless otherwise defined, the technical or scientific terms used in the embodiments of this disclosure shall have the ordinary meaning understood by one of ordinary skill in the art to which this disclosure pertains. The terms "first," "second," and similar terms used in the embodiments of this disclosure do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" indicate that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, but do not exclude other elements or objects.

[0090] The Discrete Fourier Transform (DFT) is an effective tool for signal spectrum analysis, and the Fast Fourier Transform (FFT) is a fast way to calculate the DFT. In Frequency Modulated Continuous Wave (FMCW) signal processing, it can be used to estimate the range and velocity information of a target. The DFT has a prior condition: the transformed time-domain sequence must be the principal value sequence of a discrete periodic sequence. When this prior condition is not met, spectral leakage occurs. Spectral leakage not only reduces the energy at the target's frequency peak but also results in higher leakage energy within the frequency band centered on the target's frequency, potentially masking smaller targets.

[0091] To address spectral leakage, the time-domain signal needs to be periodically truncated, which is difficult to achieve in practice. A common method to reduce the impact of spectral leakage is windowing. Windowing involves multiplying the truncated signal by a window function to ensure that the truncated signal meets the periodicity requirement of the DFT transform as much as possible. Windowed signals have more concentrated spectral energy, thus reducing spectral leakage.

[0092] Figure 1 This is a schematic diagram of the transmitted and echo signals of an FMCW using sawtooth wave modulation, as shown below. Figure 1 As shown, a single sawtooth pattern is typically referred to as a chirp. The time-domain expression of the signal after the transmit and echo signals are mixed and discretely sampled can be expressed as:

[0093]

[0094] See Table 1 for the meaning of each parameter.

[0095] R normal distance of the target f rd ]]> Target Doppler frequency f B ]]> Target range frequency [CAT D ]] Chirp duration [TECHNICAL FIELD] S ]] Sampling time l Chirp number, also known as pulse number n Sampling points within Chirp <![CDATA[f c ]]> Transmitted signal center frequency c speed of light

[0096] Table 1

[0097] like Figure 2 As shown, in the process of performing two-dimensional FFT processing on FMCW radar signals, we first window the samples within the chirp, called range windowing. Then, we perform FFT on the data within the chirp to obtain the range-pulse two-dimensional data spectrum, called range-FFT. The output is stored in a matrix in the form of consecutive rows. Next, we window between the chirp sequences, called velocity windowing. Then, we perform FFT on the chirp sequences, called Doppler FFT or velocity-dimensional FFT. This gives us the range-Doppler two-dimensional spectrum, i.e., the range-Doppler spectrum.

[0098] The windowing process for the distance dimension and the velocity dimension is similar. The following uses windowing for the distance dimension as an example to illustrate the implementation process of windowing.

[0099] The signal of the l-th Chirp is x(n) = cos(2π·(f)). B ·T S The windowing process for n) is as follows:

[0100] x w (n)=x(n)·w n ;

[0101] Where, x w (n) is the time-domain expression of the original signal sequence after windowing, w n It is a window function. Commonly used windows include rectangular windows, Hanning windows, Hamming windows, Chebyshev windows, etc.

[0102] After windowing in the time domain, a Fast Fourier Transform is performed to obtain the range domain spectrum. That is:

[0103]

[0104] Where X(k) represents the frequency domain expression of the original signal sequence after FFT, and W N For rotation factor, N is the number of sampling points. In this embodiment of the disclosure, N can also be the number of frequency points, that is, k = 0, 1, ..., N-1.

[0105] The spectrum of a window function is a continuous spectrum with one main lobe and several side lobes. The main lobe is the central part of the frequency components of the time-domain signal, and its width directly determines the radar system's target resolution capability. The height of the side lobes shows the influence of the window function on frequencies surrounding the main lobe; the side lobe response to a strong sinusoidal signal may exceed the main lobe response to a nearby weak sinusoidal signal. Generally, the lower the side lobes, the less spectral leakage, but low side lobes also increase the main lobe width. The wider the main lobe, the worse the radar system's target resolution capability. Therefore, the aforementioned windowing techniques require a trade-off between spectral width and side lobe suppression.

[0106] Figure 3a This is a schematic diagram illustrating the effect of adding windows to rectangular windows and Hanning windows in a simulated scenario. Figure 3b for Figure 3a A magnified view of region A in the middle. (See image below.) Figure 3a and Figure 3b As shown, in this simulation scenario, there are three single-frequency signals: f1, f2, and f3. Two frequency components, f1 and f2, are relatively close and have a high signal-to-noise ratio (SNR), while the other frequency component, f3, has a low SNR. To separate the two stronger frequency components, f1 and f2, a rectangular window can be used. However, in this case, the spectral leakage of f1 and f2 is quite severe, and the weaker frequency component f3 is almost completely obscured and difficult to detect. If the spectral leakage of f1 and f2 is to be suppressed, then a Hanning window is a better choice. In this case, the Hanning window can suppress the spectral leakage of f1 and f2 to a certain level, allowing the weaker target f3 to be detected. However, at this point, the main lobe width increases, making it difficult to distinguish between the two closely spaced targets, f1 and f2.

[0107] How to minimize sidelobe or spectral leakage while maintaining a narrow main lobe width is a technical problem that urgently needs to be solved by those skilled in the art.

[0108] One implementation method is to use 2 to 4 windowing steps, and then gradually reduce the size point by point. Figure 4 This is a schematic diagram illustrating an effect achieved by adding windows multiple times and then gradually reducing the size point by point. Figure 4 As shown, to obtain the final spectrum, a rectangular window, a Hanning window, a Hamming window, and a Chebyshev 80dB window are sequentially applied to the time-domain signal, and then the values ​​are reduced point by point. Figure 4 As shown, while this method can maintain a narrow main lobe width and achieve the desired sidelobes at positions far from the target frequency band, it suffers from the following problems:

[0109] 1) Irregular spectral shape near the main lobe leads to false alarms;

[0110] 2) The side lobes near the main lobe are relatively high, which may cause weak targets near the side lobes to be submerged;

[0111] 3) The same original data needs to be windowed multiple times and the results of each windowing need to be saved before the final windowing result is obtained by comparison. This results in a large amount of computation and consumes too much memory resources.

[0112] like Figure 5 As shown, this disclosure provides a digital signal processing method, including:

[0113] Step 501: Obtain the digital signals corresponding to multiple frequency points in the frequency domain;

[0114] Step 502: Determine the window parameter value corresponding to each of the multiple frequency points. The window parameter value corresponding to each frequency point makes the energy of the digital signal corresponding to that frequency point after windowing processing based on the window parameter value is minimized, and the time-domain window function corresponding to the window parameter value is non-negative.

[0115] Step 503: Window the digital signals corresponding to multiple frequency points according to the window parameter values ​​corresponding to multiple frequency points.

[0116] The digital signal processing method of this disclosure determines a window parameter value that corresponds one-to-one with each of multiple frequency points, and uses the window parameter values ​​corresponding to multiple frequency points to perform windowing processing on the digital signals corresponding to multiple frequency points respectively. This can achieve the effect of windowing sidelobe suppression, while also enabling the detection of weak targets at a greater distance, and distinguishing targets that are closer together.

[0117] In some exemplary embodiments, in step 501, the time-domain digital signal can be transformed by DFT, such as FFT, to obtain digital signals corresponding to multiple frequency points in the frequency domain. However, the embodiments disclosed herein do not limit this.

[0118] The Discrete Fourier Transform (DFT) presents a discrete form in both the time and frequency domains. It transforms samples of a time-domain signal into samples in the frequency domain of the Discrete-Time Fourier Transform (DTFT). Formally, the sequences at both ends of the transform (in the time and frequency domains) are of finite length, but in practice, both sets of sequences should be considered as principal value sequences of discrete periodic signals. Even when performing a DFT on a finite-length discrete signal, it should be viewed as a periodically extended signal before the transform. In practical applications, the Free-Form Fourier Transform (FFT) is typically used for efficient DFT computation.

[0119] In some exemplary embodiments, in step 502, the window parameters include multiple parameters, and the time-domain window functions corresponding to the multiple window parameters are equal to 0 at both ends of the window.

[0120] In some exemplary embodiments, the window parameters include a first window parameter α. 1kSecond window parameter α 2k .

[0121] In some exemplary embodiments, in step 503, windowing processing is performed on the digital signals corresponding to multiple frequency points according to the window parameter values ​​corresponding to multiple frequency points, including:

[0122] A convolution operation is performed on digital signals at multiple consecutive frequency points centered at frequency point k to obtain a windowed digital signal corresponding to frequency point k. The weights of the digital signals at the multiple consecutive frequency points in the convolution operation are based on the first window parameter α corresponding to frequency point k. 1k The value of k is determined, k = 0, 1, ..., N-1, where N is the number of frequency points.

[0123] In some exemplary embodiments, the multiple consecutive frequency points include frequency point k-2, frequency point k-1, frequency point k, frequency point k+1, and frequency point k+2. In step 502, the first window parameter α corresponding to frequency point k... 1k The value of is obtained by the following formula:

[0124]

[0125] Where, β 1k It is determined based on the digital signals corresponding to frequency points k-2, k-1, k, k+1, and k+2.

[0126] In some exemplary embodiments, β 1k The value is obtained by the following formula:

[0127]

[0128] Where real represents finding the real part of a complex number, and X(k), X(k-1), X(k+1), X(k-2), and X(k+2) are the digital signals corresponding to frequency points k, k-1, k+1, k-2, and k+2, respectively.

[0129] In some exemplary embodiments, in step 503, the digital signal X corresponding to the windowed frequency point k is... W (k) is:

[0130]

[0131] In some exemplary embodiments, the digital signal processing method of this disclosure applies windowing to digital signals at multiple frequency points in the frequency domain, and the window function w n In the time domain, it has the following form:

[0132]

[0133] Where α1 is the first window parameter and α2 is the second window parameter.

[0134] Since time-domain multiplication is equivalent to circular convolution in the frequency domain, this windowing process can be performed in the frequency domain using the spectrum of the digital signal and the window function w. n The frequency spectrum is convolved to achieve this. A five-point convolution is performed using frequency points k, k-1, k+1, k-2, and k+2, resulting in the windowed digital signal X. W (k) is obtained through the following formula:

[0135]

[0136] The digital signal processing method of this disclosure applies a dynamically changing window function for each frequency point k. That is, for each frequency point...

[0137]

[0138] Since N is determined, w k The parameters that vary include the first window parameter α. 1k Second window parameter α 2k Therefore, for each frequency point k, it is only necessary to find the first window parameter α. 1k Second window parameter α 2k This allows us to obtain the window function corresponding to that frequency point.

[0139] For each frequency point k, the first window parameter α 1k Second window parameter α 2k The following optimization problem is used to solve the problem.

[0140]

[0141] The optimization problem described above can be expressed as follows: for each frequency point k, the energy of the digital signal at that frequency point k is minimized after windowing, and the time-domain window function of that frequency point k is non-negative, and the time-domain window function of that frequency point k is equal to 0 at both ends of the window.

[0142] After a series of mathematical derivations, the above optimization problem can be transformed into one that only includes the first window parameter α. 1k The optimization problem, namely the second window parameter α 2k Based on the obtained first window parameter α 1k Determined. By taking the partial derivative of the above objective function, we can obtain the first window parameter α. 1k The explicit expression is:

[0143]

[0144] Here, real represents finding the real part of a complex number.

[0145] Thus, after calculating the above parameters for each frequency point, we only need to compare them with 0 and 4 / 3 to obtain the final first window parameter values. That is:

[0146] make but:

[0147]

[0148] Therefore, the final output after windowing for each frequency point can be given by the following expression:

[0149]

[0150] In summary, the windowing method proposed in this embodiment implements the dynamic windowing (DW) method.

[0151] This application also provides a digital signal processing method that can be applied to various discretized spectrum analysis result data to suppress problems such as spectral leakage caused during the discretized spectrum analysis, for example, for distance-dimensional FFT result data, velocity-dimensional FFT result data, and / or angle-dimensional FFT result data, specifically:

[0152] First, the digital signals corresponding to multiple frequency points in the frequency domain can be obtained through the discretized spectrum analysis process described above.

[0153] Then, for any current frequency point, dynamic windowing processing is performed on the digital signals of multiple frequency points (e.g., 3, 5, or 7, or an even number, such as 4, 6, or 8, can be selected, but in this case, the current frequency point is one of two frequency points located at the center, and the specific number can be set according to actual needs) that are continuously distributed around the current frequency point. This achieves the purpose of dynamic windowing processing on the digital signals corresponding to each frequency point, so that the windowing processing of the digital signals corresponding to each frequency point can be appropriate to the values ​​of the digital signals of the current frequency point and the digital signals corresponding to the neighboring frequency points, so as to suppress the spectral leakage and other problems caused by the above-mentioned spectral analysis and other operations to the greatest extent.

[0154] If the current frequency point is located at both ends of the frequency domain, the frequency point and the corresponding digital signal can be extended by extrapolation or fitting. The dummy frequency point and the corresponding dummy digital signal obtained by the extension can be used to dynamically window the frequency point located at the end in the same way as above.

[0155] In some optional embodiments, for any given current frequency point, a judgment parameter is obtained based on the real part of the digital signals of multiple frequency points continuously distributed around that frequency point. Based on the comparison result between the judgment parameter and a preset threshold, at least a portion of the digital signals corresponding to the multiple frequency points continuously distributed around that frequency point and the judgment parameter can be selected to obtain the windowed digital signal of the current frequency point. The preset threshold can be one or more numerical values, or a threshold range, and can be specifically set according to the expression of the digital signal and the requirements of the corresponding system parameters.

[0156] The following section will take three or five consecutively distributed frequency points centered on the current frequency point as examples to explain in detail the above judgment parameters and the corresponding windowing operations:

[0157] When three frequencies are consecutively distributed with the current frequency as the center:

[0158] First, for any current frequency point k, the judgment parameter β for that current frequency point k... k It can be obtained through the following formula:

[0159]

[0160] Where, β k The first judgment parameter is real, which means finding the real part of the complex number and conj means finding the conjugate of the complex number. The three frequency points continuously distributed in the frequency domain with the current frequency point as the center are frequency point k-1, the current frequency point k and frequency point k+1 respectively. The digital signal corresponding to frequency point k-1 is X(k-1), the digital signal corresponding to frequency point k is X(k), and the digital signal corresponding to frequency point k+1 is X(k+1). k can be an integer greater than or equal to 1.

[0161] Secondly, based on the judgment parameter β obtained above k Simultaneously, based on the analysis, preset thresholds can be set to 0 and 1, and the windowed digital signal X at the current frequency point k can be used for... W (k) can be obtained through the following formula:

[0162]

[0163] in,

[0164] Where, β k The frequency k, frequency k-1, and frequency k+1 are determined based on the corresponding digital signals, and β is... k Make the digital signal corresponding to frequency point k based on α k The energy is minimized after windowing.

[0165] When five frequencies are consecutively distributed with the current frequency as the center:

[0166] First, for any current frequency point k, the judgment parameter β for that current frequency point k... 1k It can be obtained through the following formula:

[0167]

[0168] Where, β 1k The second judgment parameter is real, which means finding the real part of the complex number. The five frequency points continuously distributed in the frequency domain with the current frequency point as the center are frequency point k-2, frequency point k-1, current frequency point k, frequency point k+1 and frequency point k+2 respectively. The digital signal corresponding to frequency point k-2 is X(k-2), the digital signal corresponding to frequency point k-1 is X(k-1), the digital signal corresponding to frequency point k is X(k), the digital signal corresponding to frequency point k+1 is X(k+1), and the digital signal corresponding to frequency point k+2 is X(k+2). k can be an integer greater than or equal to 2.

[0169] Secondly, based on the judgment parameter β obtained above 1k Simultaneously, based on the analysis, the preset threshold can be set to 0 and 4 / 3, and the windowed digital signal X at the current frequency point k can be used. W (k) can be obtained through the following formula:

[0170]

[0171] Where, β 1k As the second judgment parameter, the five frequency points continuously distributed around the current frequency point are frequency point k-2, frequency point k-1, current frequency point k, frequency point k+1, and frequency point k+2, respectively. The digital signal corresponding to frequency point k-2 is X(k-2), the digital signal corresponding to frequency point k-1 is X(k-1), the digital signal corresponding to frequency point k is X(k), the digital signal corresponding to frequency point k+1 is X(k+1), and the digital signal corresponding to frequency point k+2 is X(k+2).

[0172] This disclosure also provides a method for processing frequency-modulated continuous wave signals, such as... Figure 6a As shown, it includes:

[0173] Step 601: Sample the signal after mixing the transmitted signal and the echo signal of the frequency-modulated continuous wave to obtain a digital signal set including multiple chirps;

[0174] Step 602: Window the digital signal sequence in the digital signal set according to the digital signal processing method as described in any embodiment of this disclosure.

[0175] In some exemplary embodiments, in step 602, the digital signal sequence in the digital signal set is windowed according to the digital signal processing method as described in any embodiment of this disclosure, including at least one of the following:

[0176] (1) Perform FFT on the digital signal sequence inside Chirp to obtain the distance dimension digital signal corresponding to multiple frequency points in the frequency domain, and perform windowing processing on the distance dimension digital signal according to the digital signal processing method described in any embodiment of this disclosure.

[0177] (2) Perform FFT on the digital signal sequence between Chirps to obtain velocity dimension digital signals corresponding to multiple frequency points in the frequency domain, and perform windowing processing on the velocity dimension digital signals according to the digital signal processing method described in any embodiment of this disclosure.

[0178] (3) Perform FFT on the angle dimension data to obtain angle dimension digital signals corresponding to multiple frequency points in the frequency domain, and perform windowing processing on the angle dimension digital signals according to the digital signal processing method described in any embodiment of this disclosure.

[0179] The frequency-modulated continuous wave signal processing method of this disclosure first performs FFT on the digital signal sequence to obtain digital signals corresponding to multiple frequency points in the frequency domain, and then performs windowing processing on the digital signals corresponding to each frequency point in the frequency domain. That is, before FFT, there is no need to perform time-domain windowing operation on the digital signal sequence.

[0180] The frequency-modulated continuous wave signal processing method of this disclosure can perform dynamic windowing processing (as described in any embodiment of this disclosure) on any one of the distance-dimensional digital signals, velocity-dimensional digital signals, or angle-dimensional digital signals, using the digital signal processing method described in any embodiment of this disclosure. Figure 6b , Figure 6c and Figure 6d As shown), dynamic windowing processing can also be performed on any two or three of the distance-dimensional digital signal, velocity-dimensional digital signal, and angle-dimensional digital signal using the digital signal processing method described in any embodiment of this disclosure (e.g., Figure 6e , Figure 6f and Figure 6g As shown in the figure, the embodiments of this disclosure do not limit this. When dynamic windowing processing is performed using the digital signal processing method described in any embodiment of this disclosure only for any one of the distance dimension digital signal, velocity dimension digital signal, or angle dimension digital signal, or for any two of the distance dimension digital signal, velocity dimension digital signal, and angle dimension digital signal, the digital signals of other dimensions can be processed using any other windowing processing method, and the embodiments of this disclosure do not limit this.

[0181] In some exemplary embodiments, when dynamic windowing processing is performed on both the distance-dimensional digital signal and the velocity-dimensional digital signal using the digital signal processing method described in any embodiment of this disclosure, windowing processing of the digital signal sequence in the digital signal set according to the digital signal processing method described in any embodiment of this disclosure includes one of the following:

[0182] (I)If Figure 6e As shown, an FFT is performed on the digital signal sequence within a Chirp to obtain a distance-dimensional digital signal corresponding to multiple frequency points in the frequency domain. The distance-dimensional digital signal is then windowed according to the digital signal processing method described in any embodiment of this disclosure. An FFT is performed on the digital signal sequence between Chirps to obtain a velocity-dimensional digital signal corresponding to multiple frequency points in the frequency domain. The velocity-dimensional digital signal is then windowed according to the digital signal processing method described in any embodiment of this disclosure.

[0183] (II) such as Figure 6f As shown, an FFT is performed on the digital signal sequence within a chirp to obtain distance-dimensional digital signals corresponding to multiple frequency points in the frequency domain; an FFT is performed on the digital signal sequence between chirps to obtain velocity-dimensional digital signals corresponding to multiple frequency points in the frequency domain; windowing processing is performed on the distance-dimensional digital signals according to the digital signal processing method described in any embodiment of this disclosure; windowing processing is performed on the velocity-dimensional digital signals according to the digital signal processing method described in any embodiment of this disclosure.

[0184] (III) such as Figure 6g As shown, an FFT is performed on the digital signal sequence within a chirp to obtain a distance-dimensional digital signal corresponding to multiple frequency points in the frequency domain. An FFT is also performed on the digital signal sequence between chirps to obtain a velocity-dimensional digital signal corresponding to multiple frequency points in the frequency domain. The velocity-dimensional digital signal is then windowed according to the digital signal processing method described in any embodiment of this disclosure. The distance-dimensional digital signal is then windowed according to the digital signal processing method described in any embodiment of this disclosure.

[0185] Taking the range-dimensional window as an example, the dynamic windowing method proposed in this embodiment is performed in the range frequency domain, and its window function in the time domain has the following form:

[0186]

[0187] Where α1 is the first window parameter and α2 is the second window parameter.

[0188] Since time-domain multiplication is equivalent to circular convolution in the frequency domain, the above windowing process can be implemented in the distance-frequency domain through convolution. In the distance-frequency domain, the digital signals at frequencies k, k-1, k+1, k-2, and k+2 are selected for convolution operation, i.e.:

[0189]

[0190] The key point of the frequency-modulated continuous wave signal processing method in this disclosure is that the applied window function is dynamically changed for each frequency point k. That is, for each frequency point k, the window function w... k for:

[0191]

[0192] Since N is determined, w k The parameters that vary in the middle are α 1k and α 2k Therefore, for each frequency point k, it is only necessary to find the parameter α. 1k and α 2k This allows us to obtain the window function corresponding to that frequency point.

[0193] For each frequency point k, α 1k and α 2k The following optimization problem can be solved to obtain it:

[0194]

[0195] The optimization problem described above can be expressed as follows: For each frequency point k, we want the energy of the digital signal at that frequency point k to be minimized after windowing, while the time-domain window function of that frequency point k is non-negative and equal to 0 at both ends of the window.

[0196] After a series of mathematical derivations, the above optimization problem can be transformed into one that only involves the parameter α. 1k The optimization problem is as follows. By taking the partial derivative of the above objective function, we can obtain the parameter α. 1k The explicit expression is:

[0197]

[0198] Here, real represents finding the real part of a complex number.

[0199] Thus, after calculating the above parameters for each frequency point, we only need to compare them with 0 and 4 / 3 to obtain the final windowing parameter values. That is:

[0200] make but:

[0201]

[0202] Therefore, the final output after windowing for each frequency point can be given by the following expression:

[0203]

[0204] As an example, this disclosure uses a set of simulation results to illustrate the effect of dynamic windowing in this disclosure. The simulation parameter configuration is shown in Table 2, and the simulation target information is shown in Table 3, where Fc represents the center frequency of the transmitted signal, BandWidth represents the bandwidth of the transmitted signal, T_ramp_up represents the Chirp rise time, Tr represents the Chirp period, Fs represents the sampling rate, Rng_fft represents the number of FFT frequency points, and SNR represents the signal-to-noise ratio.

[0205] BandWidth (MHZ) 200 T_ramp_up(us) 25 Tr(us) 30 Fs(MHz) 20 Rng_fft 512

[0206] Table 2

[0207] Speed ​​(m / s) [0 0 0 0 0] SNR (db) [36 36 36 0]

[0208] Table 3

[0209] During the simulation, we first acquired two-dimensional ADC data converted by an analog-to-digital converter (ADC), and then followed... Figure 2 The process is as shown. Here we will compare the spectrum in the distance dimension using a rectangular window, a Chebyshev 80dB window, and the dynamic windowing method proposed in this disclosure. The results are as follows: Figure 7a and Figure 7b As shown, where, Figure 7b for Figure 7a A magnified view of region B in the middle.

[0210] like Figure 7a and Figure 7b As shown, the method proposed in this embodiment can achieve a sidelobe suppression effect similar to the Chebyshev 80dB window, enabling the effective detection of the fourth weak target f4' at a relatively far distance. Simultaneously, compared to the Chebyshev 80dB window, the method proposed in this embodiment can also effectively distinguish the second and third closely spaced targets f2' and f3', thereby achieving the resolution effect of a rectangular window.

[0211] In practical applications, the windowing method proposed in this disclosure can be used for windowing in either the distance or velocity dimension, or in both dimensions. For example, the following processing flows can be implemented:

[0212] 1) such as Figure 6b As shown, after ADC sampling, a distance-dimensional FFT is first performed, and then the dynamic windowing method of this disclosure embodiment is applied along the distance dimension. Next, velocity-dimensional windowing (Hanning window, Hamming window, or other windows) is performed between chirps, and finally, a velocity-dimensional FFT is performed.

[0213] 2) such as Figure 6cAs shown, after ADC sampling, a distance-dimensional window (Hanning window, Hamming window, or other window) is first applied, followed by a distance-dimensional FFT, then a velocity-dimensional FFT, and finally the dynamic windowing method of this disclosure is applied along the velocity dimension.

[0214] 3) such as Figure 6e As shown, after ADC sampling, a distance-dimensional FFT is first performed, and then the dynamic windowing method of this disclosure embodiment is applied along the distance dimension. Next, a velocity-dimensional FFT is performed between chirs, and finally, the dynamic windowing method of this disclosure embodiment is applied along the velocity dimension.

[0215] 4) such as Figure 6f As shown, after the ADC sampling is completed, the distance dimension FFT processing is performed first, then the velocity dimension FFT processing is performed between chirs, then the dynamic windowing method of the present disclosure embodiment is applied in the distance dimension, and finally the dynamic windowing method of the present disclosure embodiment is applied along the velocity dimension.

[0216] 5) such as Figure 6g As shown, after the ADC sampling is completed, the distance dimension FFT processing is performed first, then the velocity dimension FFT processing is performed between chirs, then the dynamic windowing method of the present disclosure embodiment is applied in the velocity dimension, and finally the dynamic windowing method of the present disclosure embodiment is applied along the distance dimension.

[0217] In practical applications, a range-doppler map can be obtained by following any one of the above processing steps 1), 2), 3), 4), or 5). If the array is a MIMO (Multiple-Input Multiple-Output) radar with uniformly divided subarrays, then angle measurement can be achieved using DFT. In this case, the spectrum obtained by DFT can also be obtained using the dynamic windowing method proposed in this disclosure.

[0218] For example, such as Figure 6dAs shown, firstly, the range Doppler image is combined, which mainly merges the two-dimensional FFT data of the virtual channel. Incoherent or coherent combining can be used. Then, constant false alarm rate (CFAR) is used to compare the received signal with the detection threshold to determine the presence of the target. Mean-based CFAR, sorted CFAR, and adaptive CFAR can be selected. Next, phase compensation is performed on the antenna array, mainly to compensate for the additional phase introduced by target motion between different transmit (TX) antennas in time-division MIMO mode. Then, the angle spectrum is obtained based on angle-dimensional FFT processing. After processing, the dynamic windowing method proposed in this embodiment is applied to the angle-dimensional digital signal.

[0219] CFAR (Cross-Crystal Array Ranging) technology refers to a radar system's ability to determine the presence of a target signal by distinguishing between the receiver's output signal and noise while maintaining a constant false alarm probability. CFAR first processes the input noise and then determines a threshold. This threshold is compared to the input signal; if the input signal exceeds the threshold, a target is detected; otherwise, no target is detected. CFAR methods include: mean-based CFAR (estimates background power by averaging sampled data within a reference window), sorting-based CFAR (sorts data within the reference window from smallest to largest and selects the k-th value as clutter background noise), and adaptive CFAR (uses different decision methods for different types of clutter).

[0220] Compared to single-input multiple-output (SIMO) radar, MIMO radar can achieve a larger-aperture virtual antenna array using a smaller antenna array, thereby improving the radar's angular resolution. Frequency-modulated continuous wave (FM-CHW) radar is characterized by low cost, simple structure, and small size. Simultaneously, it can accurately measure the distance and velocity of targets. Combined with antenna arrays, it can achieve target angle measurement. FM-CHW MIMO radar combines the advantages of the above two types of radar, achieving higher radar angular resolution using a simpler antenna array structure. However, the phase of the received signal from the virtual antenna array is determined not only by the target angle but also by the target velocity. Without compensation for this phase term, errors in target angle calculation will occur. Therefore, compensation is used to achieve phase correction for moving targets without reducing the radar's angular resolution.

[0221] This application also provides a radar signal processing method, which can be applied to the processing of discrete data of intermediate frequency signals obtained after mixing and analog-to-digital conversion of echo signals. Specifically, after performing range-dimensional FFT processing on the discrete data to obtain a range-chirp (i.e., pulse) data spectrum containing the target distance, the method then dynamically windows each digital data point in the range-chirp data spectrum along the range dimension to suppress spectral leakage caused by the preceding range-dimensional FFT processing. The dynamic windowing parameters for each digital data point are based on the digital data and a preset number of adjacent data points (e.g., 2, 3, 4). The method involves acquiring 5 digital data points (e.g., 1, 2, 3, etc.). For any data point to be windowed, the dynamic windowing parameters are determined along the aforementioned distance dimension based on the data point and its neighboring data. This allows the windowing operation of each digital data point to be more adapted to its data. After determining the dynamic windowing parameters, only one windowing operation is needed after the corresponding FFT operation, without the need for other windowing processing before or after the FFT. This method is not only more flexible but also ensures that the windowing effect of each target parameter can be optimized, and more effectively suppresses the spectral leakage caused by the corresponding FFT processing.

[0222] Specifically, the dynamic windowing operation based on the distance-dimensional FFT result described above is also applicable to the velocity-dimensional FFT result and the angle-dimensional FFT result. For details on the implementation, please refer to the relevant descriptions of the digital signal processing method in the embodiments of this application, which will not be repeated here. The dynamic windowing operation for the distance dimension can be performed after the distance-dimensional FFT and before CFAR (Constant False Alarm Rate) processing, while the dynamic windowing operation for the velocity dimension can be performed after the velocity-dimensional FFT processing and before CFAR. Furthermore, if the above-mentioned dynamic windowing operations for the distance and velocity dimensions are performed on the distance-velocity two-dimensional data spectrum after the velocity-dimensional FFT processing, these two windowing operations can be performed sequentially or simultaneously, and the order does not affect the final processing result. The dynamic windowing operation for the angle dimension can be performed after the angle-dimensional FFT and before the final result output.

[0223] like Figure 8 As shown in the figure, this disclosure also provides a frequency-modulated continuous wave signal processing device, including: a sampling circuit 801 and a baseband signal processing circuit 802, wherein:

[0224] The sampling circuit 801 is configured to sample the signal after mixing the transmitted signal and the echo signal of the frequency-modulated continuous wave to obtain a digital signal set including multiple chirps;

[0225] The baseband signal processing circuit 802 is configured to perform baseband processing on a digital signal set including multiple chirps. The baseband processing includes processing a digital signal sequence in the digital signal set according to the frequency modulated continuous wave signal processing method as described in any embodiment of the present disclosure.

[0226] In some exemplary embodiments, the baseband signal processing circuit 802 may include: an FFT unit 8021 and a windowing processing unit 8022, wherein:

[0227] FFT unit 8021 is configured to perform FFT on the digital signal sequence within a Chirp to obtain distance-dimensional digital signals corresponding to multiple frequency points in the frequency domain; and to perform FFT on the digital signal sequence between Chirps to obtain velocity-dimensional digital signals corresponding to multiple frequency points in the frequency domain.

[0228] The windowing processing unit 8022 is configured to process the distance dimension digital signal and / or the velocity dimension digital signal according to the digital signal processing method as described in any embodiment of the present disclosure.

[0229] In some exemplary embodiments, the baseband signal processing circuit 802 may further include: a combining unit (not shown in the figure), a CFAR unit (not shown in the figure), and a compensation unit (not shown in the figure), wherein:

[0230] The merging unit is configured to merge distance Doppler images.

[0231] The CFAR unit is configured to distinguish between the signal and noise output by the receiver while maintaining a constant false alarm probability to determine whether the target signal exists.

[0232] The compensation unit is configured to perform phase compensation on the antenna array.

[0233] FFT unit 8021 is also configured to perform FFT on angular dimension data to obtain angular dimension digital signals corresponding to multiple frequency points in the frequency domain.

[0234] The windowing processing unit 8022 is further configured to process the angular dimension digital signal according to the digital signal processing method as described in any embodiment of the present disclosure.

[0235] In some exemplary embodiments, the merging unit may perform incoherent or coherent merging processing on the distance Doppler image.

[0236] In some exemplary embodiments, the CFAR unit may use mean-based CFAR, sorting-based CFAR, and adaptive CFAR detection techniques, and this disclosure does not limit this.

[0237] like Figure 9As shown, this disclosure also provides a digital signal processing apparatus, including a Fourier transform module 901, a window parameter determination module 902, and a windowing processing module 903, wherein:

[0238] The Fourier transform module 901 is configured to perform discrete Fourier transform on digital signals to obtain digital signals corresponding to multiple frequency points in the frequency domain.

[0239] The window parameter determination module 902 is configured to determine the window parameter value corresponding to each of the multiple frequency points. The window parameter value corresponding to each frequency point minimizes the energy of the digital signal corresponding to that frequency point after windowing based on the window parameter value, and the time-domain window function corresponding to the window parameter value is non-negative.

[0240] The windowing processing module 903 is configured to perform windowing processing on the digital signals corresponding to multiple frequency points according to the window parameter values ​​corresponding to multiple frequency points.

[0241] In some exemplary embodiments, the window parameters include a first window parameter α. 1k Second window parameter α 2k The windowing processing module 903 is specifically configured as follows:

[0242] A convolution operation is performed on digital signals at multiple consecutive frequency points centered at frequency point k to obtain a windowed digital signal corresponding to frequency point k. The weights of the digital signals at the multiple consecutive frequency points in the convolution operation are based on the first window parameter α corresponding to frequency point k. 1k The value of k is determined, k = 0, 1, ..., N-1, where N is the number of frequency points.

[0243] In some exemplary embodiments, the plurality of consecutive frequency points include frequency point k-2, frequency point k-1, frequency point k, frequency point k+1, and frequency point k+2;

[0244] The window parameter α corresponding to the frequency point k determined by the window parameter determination module 902 1k We obtain it from the following formula:

[0245]

[0246] Where, β 1k It is determined based on the digital signals corresponding to frequency points k-2, k-1, k, k+1, and k+2.

[0247] In some exemplary embodiments, the window parameter determination module 902 obtains β by the following formula. 1k :

[0248]

[0249] The windowing processing module 903 obtains the digital signal X corresponding to the windowed frequency point k using the following formula.W (k):

[0250]

[0251] Where real represents finding the real part of a complex number, and X(k), X(k-1), X(k+1), X(k-2), and X(k+2) are the digital signals corresponding to frequency points k, k-1, k+1, k-2, and k+2, respectively.

[0252] like Figure 10 As shown, this disclosure also provides a digital signal processing apparatus, including a memory 1020 and a processor 1010;

[0253] Among them, memory 1020 is used to store program instructions;

[0254] The processor 1010 is configured to invoke the program instructions stored in the memory 1020 to implement the digital signal processing method as described in any embodiment of the present disclosure.

[0255] In one example, the digital signal processing device may include a processor 1010, a memory 1020, a bus system 1030, and a transceiver 1040. The processor 1010, memory 1020, and transceiver 1040 are connected via the bus system 1030. The memory 1020 stores instructions, and the processor 1010 executes the instructions stored in the memory 1020 to control the transceiver 1040 to transmit and receive signals. Specifically, the transceiver 1040, under the control of the processor 1010, can acquire digital signals corresponding to multiple frequency points in the frequency domain. The processor 1010 determines a window parameter value corresponding to each of the multiple frequency points. The window parameter value for each frequency point minimizes the energy of the digital signal at that frequency point after windowing processing based on the window parameter value, and the time-domain window function corresponding to the window parameter value is non-negative. Windowing processing is then applied to the digital signals corresponding to the multiple frequency points according to the window parameter values.

[0256] It should be understood that processor 1010 can be a Central Processing Unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0257] Memory 1020 may include read-only memory and random access memory, and provides instructions and data to processor 1010. A portion of memory 1020 may also include non-volatile random access memory. For example, memory 1020 may also store device type information.

[0258] In addition to a data bus, the bus system 1030 may also include a power bus, a control bus, and a status signal bus. However, for clarity, in... Figure 10 The general labeled all buses as Bus System 1030.

[0259] In implementation, the processing performed by the processing device can be accomplished through integrated logic circuits in the hardware of the processor 1010 or through software instructions. That is, the method steps of this embodiment can be executed by a hardware processor, or by a combination of hardware and software modules within the processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other storage media. This storage medium is located in memory 1020. The processor 1010 reads information from memory 1020 and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, further details are omitted here.

[0260] While the embodiments disclosed herein are as described above, the content is merely for the purpose of facilitating understanding of this disclosure and is not intended to limit this disclosure. Any person skilled in the art to which this disclosure pertains may make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed herein; however, the scope of protection of this disclosure shall still be determined by the scope defined in the appended claims.

Claims

1. A digital signal processing method, characterized in that, include: Acquire digital signals corresponding to multiple frequency points in the frequency domain; Determine the window parameter value corresponding to each of the plurality of frequency points. The window parameter value corresponding to each frequency point minimizes the energy of the digital signal corresponding to that frequency point after windowing processing by performing convolution operation based on the window parameter value, and the time-domain window function corresponding to the window parameter value is non-negative. Windowing processing is performed on the digital signals corresponding to the multiple frequency points according to the window parameter values ​​corresponding to the multiple frequency points.

2. The digital signal processing method according to claim 1, characterized in that, The window parameters include multiple parameters, and the time-domain window functions corresponding to the multiple window parameters are equal to 0 at both ends of the window.

3. The digital signal processing method according to claim 1, characterized in that, The window parameters include a first window parameter and a second window parameter. Windowing processing is performed on the digital signals corresponding to the plurality of frequency points according to the window parameter values, including: A convolution operation is performed on the digital signals of multiple consecutively distributed frequency points centered at frequency point k to obtain the digital signal corresponding to frequency point k after windowing. The weights of the digital signals of the multiple consecutive frequency points in the convolution operation are determined according to the first window parameter value corresponding to frequency point k, k=0,1,…,N-1, where N is the number of frequency points.

4. The digital signal processing method according to claim 3, characterized in that, The plurality of consecutive frequency points include frequency point k-2, frequency point k-1, frequency point k, frequency point k+1, and frequency point k+2; The first window parameter corresponding to frequency point k The value is obtained by the following formula: ； in, To determine the parameters, It is determined based on the digital signals corresponding to frequency points k-2, k-1, k, k+1, and k+2.

5. The digital signal processing method according to claim 4, characterized in that, We obtain it from the following formula: ； Where real denotes finding the real part of a complex number. These are the digital signals corresponding to frequency points k, k-1, k+1, k-2, and k+2, respectively.

6. The digital signal processing method according to claim 1, characterized in that, The digital signal corresponding to frequency point k after windowing processing for: ； ； Where real denotes finding the real part of a complex number. These are the digital signals corresponding to frequency points k, k-1, k+1, k-2, and k+2, respectively.

7. A digital signal processing method, characterized in that, include: Acquire digital signals corresponding to multiple frequency points in the frequency domain; Dynamic windowing is achieved by performing a convolution operation on digital signals of multiple frequency points continuously distributed around the current frequency point for any given frequency point.

8. The digital signal processing method according to claim 7, characterized in that, The step of dynamically windowing a frequency point based on the digital signals corresponding to multiple frequency points continuously distributed around that current frequency point includes: For any given current frequency point, a judgment parameter is obtained based on the real part of the digital signals of multiple frequency points continuously distributed around the current frequency point. Based on the comparison result between the judgment parameter and the preset threshold, the digital signals corresponding to the multiple frequency points continuously distributed around the current frequency point and at least a portion of the judgment parameter are selected to obtain the windowed digital signal of the current frequency point.

9. The digital signal processing method according to claim 8, characterized in that, For any given frequency point, a judgment parameter is obtained based on the real part of the digital signals of five consecutive frequency points centered on that current frequency point. We obtain it from the following formula: ； Where, real represents finding the real part of the complex number, and the five frequency points continuously distributed around this frequency point are frequency point k-2, frequency point k-1, the current frequency point k, frequency point k+1, and frequency point k+2, respectively. The digital signal corresponding to frequency point k-2 is... The digital signal corresponding to frequency point k-1 is The digital signal corresponding to frequency point k is The digital signal corresponding to frequency point k+1 is The digital signal corresponding to frequency point k+2 is .

10. The digital signal processing method according to claim 8, characterized in that, The digital signals corresponding to five consecutively distributed frequency points centered on the current frequency point and at least a portion of the judgment parameters are selected to obtain the windowed digital signal of the current frequency point; the preset thresholds are 0 and 4 / 3; the windowed digital signal of the current frequency point k The expression is: in, To determine the parameters, the five consecutive frequency points centered on the current frequency point are, in order, frequency point k-2, frequency point k-1, the current frequency point k, frequency point k+1, and frequency point k+2. The digital signal corresponding to frequency point k-2 is... The digital signal corresponding to frequency point k-1 is The digital signal corresponding to frequency point k is The digital signal corresponding to frequency point k+1 is The digital signal corresponding to frequency point k+2 is .

11. A method for processing frequency-modulated continuous wave signals, characterized in that, include: The signal obtained by mixing the transmitted and echo signals of the frequency-modulated continuous wave is sampled to obtain a digital signal set including multiple chirps; The digital signal sequence in the digital signal set is windowed according to any one of the digital signal processing methods described in claims 1 to 10.

12. The frequency-modulated continuous wave signal processing method according to claim 11, characterized in that, The windowing process of the digital signal sequence in the digital signal set according to any one of claims 1 to 10 includes at least one of the following: Perform FFT on the digital signal sequence inside the Chirp to obtain the distance dimension digital signal corresponding to multiple frequency points in the frequency domain, and perform windowing processing on the distance dimension digital signal according to any one of the digital signal processing methods described in claims 1 to 10. Perform FFT on the digital signal sequence between the Chirps to obtain velocity dimension digital signals corresponding to multiple frequency points in the frequency domain, and perform windowing processing on the velocity dimension digital signals according to any one of the digital signal processing methods described in claims 1 to 10. Perform FFT on the angular dimension data to obtain angular dimension digital signals corresponding to multiple frequency points in the frequency domain, and perform windowing processing on the angular dimension digital signals according to any one of the digital signal processing methods described in claims 1 to 10.

13. The frequency-modulated continuous wave signal processing method according to claim 11, characterized in that, The windowing process performed on the digital signal sequence in the digital signal set according to any one of claims 1 to 10 includes one of the following: Perform FFT on the digital signal sequence inside the Chirp to obtain the distance dimension digital signal corresponding to multiple frequency points in the frequency domain, and perform windowing processing on the distance dimension digital signal according to any one of the digital signal processing methods described in claims 1 to 10. Perform FFT on the digital signal sequence between the Chirps to obtain velocity dimension digital signals corresponding to multiple frequency points in the frequency domain, and perform windowing processing on the velocity dimension digital signals according to any one of the digital signal processing methods described in claims 1 to 10. Perform FFT on the digital signal sequence within the Chirp to obtain distance-dimensional digital signals corresponding to multiple frequency points in the frequency domain; perform FFT on the digital signal sequence between the Chirps to obtain velocity-dimensional digital signals corresponding to multiple frequency points in the frequency domain; and perform windowing processing on the distance-dimensional digital signals according to any one of the digital signal processing methods described in claims 1 to 10. The velocity-dimensional digital signal is windowed according to any one of claims 1 to 10; Perform FFT on the digital signal sequence within the Chirp to obtain distance-dimensional digital signals corresponding to multiple frequency points in the frequency domain; perform FFT on the digital signal sequence between the Chirps to obtain velocity-dimensional digital signals corresponding to multiple frequency points in the frequency domain; and perform windowing processing on the velocity-dimensional digital signals according to any one of the digital signal processing methods described in claims 1 to 10. The distance-dimensional digital signal is windowed according to any one of the digital signal processing methods described in claims 1 to 10.

14. A frequency-modulated continuous wave signal processing device, characterized in that, include: The sampling circuit and the baseband signal processing circuit include: The sampling circuit is configured to sample the signal after mixing the transmitted signal and the echo signal of the frequency-modulated continuous wave to obtain a digital signal set including multiple chirps; The baseband signal processing circuit is configured to perform baseband processing on the digital signal set including multiple chirps, wherein the baseband processing includes windowing the digital signal sequence in the digital signal set according to the frequency modulated continuous wave signal processing method as described in any one of claims 11 to 13.

15. The frequency-modulated continuous wave signal processing apparatus according to claim 14, characterized in that, The baseband signal processing circuit includes: an FFT unit and a windowing processing unit, wherein: The FFT unit is configured to perform FFT on the digital signal sequence within the Chirp to obtain distance-dimensional digital signals corresponding to multiple frequency points in the frequency domain; and to perform FFT on the digital signal sequence between the Chirps to obtain velocity-dimensional digital signals corresponding to multiple frequency points in the frequency domain. The windowing processing unit is configured to perform windowing processing on the distance dimension digital signal and / or the velocity dimension digital signal according to the digital signal processing method as described in any one of claims 1 to 10.

16. A digital signal processing apparatus, characterized in that, It includes a Fourier transform module, a window parameter determination module, and a windowing processing module, among which: The Fourier transform module is configured to perform discrete Fourier transform on the digital signal to obtain digital signals corresponding to multiple frequency points in the frequency domain. The window parameter determination module is configured to determine the window parameter value corresponding to each of the plurality of frequency points. The window parameter value corresponding to each frequency point minimizes the energy of the digital signal corresponding to that frequency point after windowing processing based on the window parameter value, and the time-domain window function corresponding to the window parameter value is non-negative, so that the shape of the window can be dynamically adjusted between a rectangular window and a Hanning window. The windowing processing module is configured to perform windowing processing on the digital signals corresponding to the multiple frequency points according to the window parameter values ​​corresponding to the multiple frequency points.

17. A digital signal processing apparatus, characterized in that, Including memory and processor; The memory is used to store program instructions; The processor is configured to invoke the program instructions stored in the memory to implement the digital signal processing method as described in any one of claims 1 to 10.

Images