Broadband signal detection method and system
By calculating the extreme value of the average power spectrum density of the signal spectrum, the rapid extraction and real-time improvement of the broadband signal detection method are achieved, and the problems of prior information, calculation complexity and noise influence in the prior art are solved.
Patent Information
- Application Number
- CN202510190922.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-20
- Publication Date
- 2025-06-03
AI Technical Summary
The existing broadband signal detection methods are affected by prior information, calculation complexity and noise in non-cooperative communication reconnaissance, making it difficult to quickly extract signal parameters in broadband scan signals.
By calculating the extreme value of the average power spectrum density of the signal spectrum, the upper and lower edge detection of the signal on the spectrum is extracted, without requiring prior information or relying on noise estimation, which reduces the computational complexity and improves the real-timeness of signal detection.
It realizes the rapid extraction of signal parameters in broadband scanning signals, improves the real-time and efficiency of detection, and reduces system overhead.
Smart Images
Figure CN120090736A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of electronic information technology, and particularly relates to a broadband signal detection method and system. Background Art
[0002] In communication reconnaissance, generally, after collecting signals by an ultra-wideband receiver based on the idea of software radio, signal extraction and subsequent reconnaissance processing are carried out. Therefore, broadband signal detection has become an essential technology. In non-cooperative communication reconnaissance, it is necessary to scan and detect the positions (center frequencies and bandwidths) of narrowband signals in a wide frequency band. For example, in a frequency band of several hundred megahertz, there are multiple narrowband signals with bandwidths of several megahertz or more than ten megahertz scattered. Therefore, how to effectively search and detect narrowband signal parameters has become an essential part of reconnaissance. The three most widely used signal detection algorithms currently are: the matched filter detection method, the cyclostationary detection method, and the energy detection method.
[0003] The matched filter detection method is often used for the detection of digital communication signals and radar signals. The matched filter detection method is theoretically an optimal signal detection method, but it requires a large amount of prior information such as carrier frequency, bandwidth, modulation method, etc., and has high requirements for phase synchronization, with poor flexibility. It is difficult to implement in practice and is not suitable for non-cooperative communication.
[0004] The cyclostationary detection method uses the cyclostationary characteristics of signals for detection, and can estimate basic parameters such as signal code rate, carrier frequency, and sampling frequency at the same time, with strong anti-interference ability. However, the calculation complexity of cyclostationary characteristics is relatively high. The cyclostationary detection method has high detection performance, but it is necessary to calculate the entire cyclic spectrum of the signal, with high complexity and long detection time, and is not suitable for occasions with high real-time requirements.
[0005] The energy detection method is currently the most widely used signal detection algorithm. It determines whether there is a signal according to the energy or power of the signal in the channel. Noise has a certain impact on the detection effect, and the traditional energy detection method will increase the system overhead while improving performance.
[0006] In summary, the existing broadband signal detection methods are affected by prior information, calculation complexity, or noise detection performance to varying degrees. Summary of the Invention
[0007] The technical problem to be solved by the present invention: Aiming at the above problems of the prior art, a broadband signal detection method and system are provided, which can quickly extract signal parameters from wide-band scanned signals.
[0008] To solve the above technical problems, the technical solution adopted by the present invention is as follows:
[0009] A broadband signal detection method includes the following steps:
[0010] Obtain the signal of the target frequency band and perform mean filtering to obtain the power spectrum;
[0011] According to the specified multiple of the target frequency band, calculate the first lower frequency limit and the first upper frequency limit. Determine the first upper frequency limit position and the first lower frequency limit position on the power spectrum based on the first lower frequency limit and the first upper frequency limit. Calculate the average power spectral density within the first upper frequency limit position and the first lower frequency limit position, and find all the maxima of the average power spectral density curve within the target frequency band, which are used as the upper frequency limit positions of each signal in the target frequency band;
[0012] Flip the power spectrum. According to the specified multiple of the target frequency band, calculate the second lower frequency limit and the second upper frequency limit. Determine the second upper frequency limit position and the second lower frequency limit position on the flipped power spectrum based on the second lower frequency limit and the second upper frequency limit. Calculate the average power spectral density within the second upper frequency limit position and the second lower frequency limit position, then flip the average power spectral density curve, and find all the maxima of the flipped average power spectral density curve within the target frequency band, which are used as the lower frequency limit positions of each signal in the target frequency band;
[0013] Calculate the corresponding signal parameters according to the upper frequency limit positions and the lower frequency limit positions of each signal in the target frequency band.
[0014] Furthermore, the expressions of the first lower frequency limit and the first upper frequency limit are as follows:
[0015] f o = Fc - 0.5 * k * Bw
[0016] f t = Fc + 0.5 * k * Bw
[0017] where f o is the first lower frequency limit, f t is the first upper frequency limit, Fc is the center frequency of the target frequency band, Bw is the bandwidth of the target frequency band, and k is the specified multiple.
[0018] Furthermore, the expressions of the second lower frequency limit and the second upper frequency limit are as follows:
[0019] f o1 = Fc + 0.5 * k * Bw
[0020] f t1 = Fc - 0.5 * k * Bw
[0021] where f o1 is the second lower frequency limit, f t1 is the second upper frequency limit, Fc is the center frequency of the target frequency band, Bw is the bandwidth of the target frequency band, and k is the specified multiple.
[0022] Further, when finding all the maximum values of the average power spectral density curve within the target frequency band, it includes:
[0023] Obtain all the points of the average power spectral density curve within the target frequency band, and calculate the corresponding elements in the matrix of the local maximum scale map;
[0024] Calculate the sum of each row in the matrix of the local maximum scale map, and calculate the minimum value in the sum of each row as the optimal local maximum scale, and calculate the local maximum scale according to the optimal local maximum scale and the local maximum scale map;
[0025] Select, from all the local maximum scales, the local maximum scales whose differences from the values of adjacent local maximum scales are all greater than zero as the maximum values.
[0026] Further, the expression of the elements in the matrix of the local maximum scale map is as follows:
[0027]
[0028] where m k,i represents the element in the k-th row and the i-th column of the matrix of the local maximum scale map, i ∈ {1, 2,..., N}, k ∈ {1, 2,..., L}, N represents the number of all points within the target frequency band, x i-1 , x i-k-1 , x i+k-1 represent the (i - 1)-th, (i - k - 1)-th, and (i + k - 1)-th points among all the points respectively.
[0029] Further, the expression of the sum of each row is as follows:
[0030]
[0031] where m k,i represents the element in the k-th row and the i-th column of the matrix of the local maximum scale map, i ∈ {1, 2,..., N}, k ∈ {1, 2,..., L}, N represents the number of all points within the target frequency band.
[0032] Further, the expression of the local maximum scale is as follows:
[0033]
[0034] where m k,i represents the element in the k-th row and the i-th column of the matrix of the local maximum scale map, i ∈ {1, 2,..., N}, k ∈ {1, 2,..., λ}, N represents the number of all points within the target frequency band, and λ represents the row number corresponding to the minimum value in the sum of each row.
[0035] The present invention also provides a broadband signal detection system, including a memory, a processor, and a computer program stored on the memory, and the processor executes the computer program to implement the steps of any one of the broadband signal detection methods.
[0036] The present invention also provides a computer-readable storage medium, on which a computer program / instructions are stored, and when the computer program / instructions are executed by a processor, the steps of any one of the broadband signal detection methods are implemented.
[0037] The present invention also provides a computer program product, including a computer program / instructions, and when the computer program / instructions are executed by a processor, the steps of any one of the broadband signal detection methods are implemented.
[0038] Compared with the prior art, the advantages of the present invention are as follows:
[0039] Based on the extraction of the extreme values of the average power spectral density of the signal spectrum, the present invention realizes the upper and lower edge detection of the signal in the spectrum. By calculating the average power spectral density of the broadband spectrum and extracting the extreme points of the average power spectral density curve as the signal boundaries, neither prior information is required nor is there a dependence on the noise floor estimation, reducing the computational complexity and improving the real-time performance of signal detection. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] Figure 1 It is a flowchart of the method according to an embodiment of the present invention.
[0041] Figure 2 It is a schematic diagram of the power spectrum of the target detection frequency band.
[0042] Figure 3 It is a schematic diagram of the power spectrum after mean filtering.
[0043] Figure 4 It is a schematic diagram of the average power spectral density.
[0044] Figure 5 It is a schematic diagram of the extraction of the peak value of the upper limit of the spectrum.
[0045] Figure 6 It is a schematic diagram of the extraction of the peak value of the lower limit of the spectrum. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0046] The present invention will be further described below in conjunction with the accompanying drawings of the specification and specific preferred embodiments, but the protection scope of the present invention is not limited thereby.
[0047] Embodiment 1
[0048] This embodiment proposes a broadband signal detection method, which realizes the upper and lower edge detection of the signal in the frequency spectrum based on the extraction of the extreme values of the average power spectral density of the signal spectrum. By calculating the average power spectral density of the broadband spectrum and extracting the extreme points of the average power spectral density curve as the signal boundary, the signal detection algorithm neither requires prior information nor depends on the noise floor estimation, and can quickly extract the signal parameters (carrier frequency and bandwidth) in the wideband scanning signal. The specific principle is as follows:
[0049] Suppose that additive white Gaussian noise (AWGN) is introduced during signal transmission, and there are two signals with non-overlapping spectra in the received signal. The real signal expression is:
[0050] x(t) = s 1 (t) + s 2 (t) + n(t) #(1)
[0051] Where n(t) is Gaussian white noise with a mean of 0 and a variance of N 0 , and is uncorrelated with the signals s 1 (t) and s 2 (t).
[0052] Then the power spectral density of x(t) can be expressed as:
[0053] X(f) = S 1 (f) + S 2 (f) + N(f) #(2)
[0054] If the sampling rate of the signal is f s , the frequency range of the signal s 1 (t) is [f L1 , f H1 , the frequency range of the signal s 2 (t) is [f L2 , f H2 . Take the frequency point f o (-∞ < f o < f L1 ) as the lower limit of the frequency spectrum, f t (f H2 < f t < ∞) as the upper limit of the frequency spectrum, and any frequency point f r (f r ≥ f o ) as the upper limit of the frequency. Calculate the average power spectral density in this frequency range as:
[0055]
[0056] When f r ≤ f L2 , discretize the above formula, where Δf is the minimum frequency interval, and we get:
[0057]
[0058] Similarly, in f r > f L2 When discretized, we get:
[0059]
[0060] For two adjacent frequency points f L1 greater than f L2 and less than f k , f k+1 , the difference in their average power spectral density is:
[0061]
[0062] Analyzing the above formula, due to the non-negativity of the power spectral density, when S 1 (f k+1 ) ≠ 0 and (k - o) is large enough (this condition is easily achieved), then Diff k must be positive. Similarly, Diff k+1 can be expressed as:
[0063]
[0064] Combining the previous two formulas, we get:
[0065]
[0066] It can be seen that when Diff k > 0, if [S 1 (f k+2 ) - S 1 (f k+1 )] ≥ 0, then Diff k+1 must also be greater than 0.
[0067] Similarly, it is not difficult to obtain that for two adjacent frequency points f L2 greater than f k , f k+1 , the difference in their average power spectral density is:
[0068]
[0069] Analyzing the above formula, due to the non-negativity of the power spectral density, when S 2 (f k+1 ) ≠ 0 and (k - o) is large enough, then Diff k must be positive. Similarly, Diff k+1 can be expressed as:
[0070]
[0071] It can be seen that when Diff k > 0, if [S 1 (f k+2 ) - S 1 (f k+1 )] ≥ 0, then Diff k+1 must also be greater than 0.
[0072] Based on the above analysis, it is not difficult to draw the following conclusions:
[0073] 1) When f o < f k < f L1 , Diff k = 0;
[0074] 2) When f k = f L1 , Diff k = Diff L1 > 0;
[0075] 3) When f L1 < f k ≤ f c1 , the conditions Diff L1 > 0 and [S(f k+2 ) - S(f k+1 )] ≥ 0 are satisfied, then Diff k > 0;
[0076] 4) When f c1 < f k ≤ f H1 , since S(f k ) is monotonically decreasing, Diffi i may not be greater than 0, then there must exist an integration lower limit f o (-∞ < f o < f L1 ), which can make Diff k > 0 (k < H1), Diff H1 = 0, Diff k < 0 (k > H1);
[0077] 5) When f H1 < f k < f L2 , Diff H1 < 0 and S(f k+1 ) = S(f k ), Diff k < 0;
[0078] 6) When f k = fL2 , Diff k = Diff L2 > 0;
[0079] 7) When f L2 < f k ≤ f c2 When, the condition Diff L2 > 0 and [S(f k+2 ) - S(f k+1 )] ≥ 0 are satisfied, then Diff k > 0;
[0080] 8) When f c2 < f k ≤ f H2 When, since S(f k ) is monotonically decreasing, Diff i may not be greater than 0. Then there must exist an integral lower limit f o (-∞ < f o < f L2 ) such that Diff k > 0 (k < H2), Diff H2 = 0, Diff k < 0 (k > H2);
[0081] 9) When f H2 < f k < f t When, Diff H < 0 and S(f k+1 ) = S(f k ), Diff k < 0.
[0082] The above conclusion shows that there indeed exists f o such that the average power spectral density is monotonically increasing within [f L1 , f H1 , monotonically decreasing within [f H1 , f L2 , monotonically increasing again within [f L2 , f H2 , and monotonically decreasing again within [f H1 , f s / 2]. That is, f H1 and f J2 correspond to two maxima within [f L1 , f s / 2]. This shows that f H1 and f H2 can be solved by the above method of finding the maximum value., Based on the above method, detect and calculate from back to front on the broadband spectrum (which can also be understood as flipping the spectrum). The original lower limit f of the spectrum o and the upper limit f of the spectrum t become the upper limit f of the spectrum t1 (i.e., f o ) and the lower limit f of the spectrum o1 (i.e., f t ). Then, f L1 and f L2 can be obtained. Based on the above derivation, it is not difficult to get that when there are more groups of non-overlapping signals on the spectrum, the above conclusion still holds.
[0083] Based on the above principle, as Figure 1 shown, the method of this embodiment includes the following steps:
[0084] S1) Obtain the signal of the target frequency band and perform mean filtering to obtain the power spectrum;
[0085] S2) Calculate the first lower frequency limit and the first upper frequency limit according to the specified multiple of the target frequency band. Determine the first upper frequency limit position and the first lower frequency limit position on the power spectrum according to the first lower frequency limit and the first upper frequency limit. Calculate the average power spectral density within the first upper frequency limit position and the first lower frequency limit position, and find all the maximum values of the average power spectral density curve within the target frequency band as the upper frequency limit positions of each signal in the target frequency band;
[0086] S3) Flip the power spectrum. Calculate the second lower frequency limit and the second upper frequency limit according to the specified multiple of the target frequency band. Determine the second upper frequency limit position and the second lower frequency limit position on the flipped power spectrum according to the second lower frequency limit and the second upper frequency limit. Calculate the average power spectral density within the second upper frequency limit position and the second lower frequency limit position, then flip the average power spectral density curve, and find all the maximum values of the flipped average power spectral density curve within the target frequency band as the lower frequency limit positions of each signal in the target frequency band. According to the lower frequency limit positions and the upper frequency limit positions of each signal in the target frequency band, the signal parameters such as the frequency range, center frequency, and bandwidth of each signal can be determined.
[0087] Next, take the broadband power spectrum with the center frequency being zero frequency as a demonstration to illustrate each step.
[0088] As Figure 2 shown, assume that the number of points of the power spectrum X(k) is W = 8192 points, the unit is dB, the sampling rate is 128 kHz. Since preprocessing such as down-conversion has been performed, the target detection frequency band has been shifted to near zero frequency, and there are 4 signals in the target frequency band.
[0089] In step S1 of this embodiment, when performing mean filtering, specifically, mean filtering is performed on X(k). The mean filter is a rectangular window low-pass filter with all coefficients being 1. The order of the filter is determined according to the number of power spectrum points W, and the empirical calculation formula is filter order = W / 256 + 1 (the power spectrum of broadband signal detection is generally a 2 N point FFT that is much larger than 256). After mean filtering, the power spectrum is as Figure 3 shown.
[0090] In step S2 of this embodiment, when calculating the first frequency lower limit and the first frequency upper limit according to the specified multiple of the target frequency band, specifically, the frequency lower limit f o and the frequency upper limit f t are calculated with a width of 2 to 4 times that of the target frequency band to be detected. Let the center frequency of the target frequency band be Fc and the bandwidth be Bw, then the expressions are as follows:
[0091] f o = Fc - 0.5 * k * Bw, k ∈ [2, 4] #(11)
[0092] f t = Fc + 0.5 * k * Bw, k ∈ [2, 4] #(12)
[0093] After calculating the frequency lower limit f o and the frequency upper limit f t , the frequency upper limit position and the frequency lower limit position on the power spectrum can be determined according to the frequency lower limit and the frequency upper limit, and the frequency lower limit f o and the frequency upper limit f t correspond to the frequency upper limit position O and the frequency lower limit position T on the power spectrum, that is, the boundaries can be used as the frequency upper and lower limits.
[0094] In step S2 of this embodiment, when calculating the average power spectral density within the first frequency upper limit position and the first frequency lower limit position, specifically, the average power spectral density within k ∈ [O, T] is calculated according to the content of the aforementioned formulas (1) to (5), and the result is as Figure 4 shown.
[0095] In step S2 of this embodiment, when finding all the maximum values of the average power spectral density curve within the target frequency band, specifically, all the maximum values (relatively obvious peaks) within the above target frequency band are found and denoted as H1, H2,..., Hn, as Figure 5 shown. The peak extraction algorithm uses an automatic multi-peak detection algorithm, and the specific implementation process is as follows:
[0096] S21) Obtain all the points of the average power spectral density curve within the target frequency band, and calculate the corresponding elements in the matrix of the local maximum scale map;
[0097] Assume that all points of the average power spectral density curve within the target frequency band are \(x = [x 1 ,x 2 ,…,x i ,…,x N \). The matrix for calculating the local maximum scale map is to calculate the -point local maximum scale map. The expression of the elements in the matrix of the local maximum scale map is as follows:
[0098]
[0099] where \(m k,i \) represents the element in the \(k\)-th row and \(i\)-th column of the matrix of the local maximum scale map, \(i\in\{1,2,\cdots,N\}\), \(k\in\{1,2,\cdots,L\}\), \(N\) represents the number of all points within the target frequency band, \(x i-1 ,x i-k-1 ,x i+k-1 \) represent the \((i - 1)\)-th, \((i - k - 1)\)-th, and \((i + k - 1)\)-th points among all points respectively;
[0100] S22) Calculate the sum of each row in the matrix of the local maximum scale map, and calculate the minimum value in the sum of each row as the optimal local maximum scale, and calculate the local maximum scale according to the optimal local maximum scale and the local maximum scale map;
[0101] The expression of the sum of each row is as follows:
[0102]
[0103] where \(m k,i \) represents the element in the \(k\)-th row and \(i\)-th column of the matrix of the local maximum scale map, \(i\in\{1,2,\cdots,N\}\), \(k\in\{1,2,\cdots,L\}\), \(N\) represents the number of all points within the target frequency band;
[0104] The sum of each row forms a set \(\gamma = [\gamma 1 ,\gamma 2 ,\cdots,\gamma i ,\cdots,\gamma L \). Assume that the row number corresponding to the minimum value is \(\lambda\). Then the expression of the local maximum scale is as follows:
[0105]
[0106] where \(m k,i \) represents the element in the \(k\)-th row and \(i\)-th column of the matrix of the local maximum scale map, \(i\in\{1,2,\cdots,N\}\), \(k\in\{1,2,\cdots,\lambda\}\), \(N\) represents the number of all points within the target frequency band, and \(\lambda\) represents the row number corresponding to the minimum value in the sum of each row;
[0107] S23) From all local maximum scales p = p 1 , p 2 , …, p q , …, p N Detect peaks, that is, select the local maximum scale whose difference from the values of adjacent local maximum scales is greater than zero as the maximum value.
[0108] In step S3 of this embodiment, after flipping the power spectrum, when calculating the second lower frequency limit and the second upper frequency limit according to a specified multiple of the target frequency band, specifically, in the opposite direction of the power spectrum in step S2, modify the rules for calculating the lower frequency limit and the upper frequency limit, so that the original upper frequency limit f t becomes the lower frequency limit f o1 , and the original lower frequency limit f o becomes the upper frequency limit f t1 , and the expressions are as follows:
[0109] f o1 = Fc + 0.5 * k * Bw, k ∈ [2, 4] #(16)
[0110] f t1 = Fc - 0.5 * k * Bw, k ∈ [2, 4] #(17)
[0111] The subsequent content of step S3 of this embodiment is basically the same as that of step S2, except that on the basis of step S2, the average power spectrum density curve is flipped. The subsequent content of step S3 will not be elaborated here. In step S3, after flipping the average power spectrum density curve within the second upper frequency limit position and the second lower frequency limit position as Figure 6 shown, after finding all the maximum values of the flipped average power spectrum density curve within the target frequency band, denoted as L1, L2, … Ln, then the frequency range corresponding to signal k is [L k , H k × f s / W, and the calculation formulas for the center frequency center_freq and the bandwidth band_width are
[0112]
[0113] Embodiment 2
[0114] This embodiment proposes a broadband signal detection system, including a memory, a processor, and a computer program stored on the memory. The processor executes the computer program to implement the steps of the broadband signal detection method described in Embodiment 1.
[0115] This embodiment also provides a computer-readable storage medium, on which a computer program / instructions is stored. When the computer program / instructions is executed by a processor, the steps of the broadband signal detection method described in Embodiment 1 are implemented.
[0116] This embodiment also provides a computer program product, including a computer program / instructions. When the computer program / instructions is executed by a processor, the steps of the broadband signal detection method described in Embodiment 1 are implemented.
[0117] The above are only the preferred embodiments of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements should also be regarded as the protection scope of the present invention.
Claims
1. A broadband signal detection method, characterized in that: The following steps are involved: Get the signal of the target frequency band and perform mean filtering to obtain the power spectrum; According to the specified multiple of the target frequency band, the first frequency lower limit and the first frequency upper limit are calculated, and the first frequency upper limit position and the first frequency lower limit position on the power spectrum are determined according to the first frequency lower limit and the first frequency upper limit, and the average power spectrum density within the first frequency upper limit position and the first frequency lower limit position is calculated, and all the maximum values of the average power spectrum density curve within the target frequency band are found as the frequency upper limit position of each signal in the target frequency band; Flip the power spectrum, calculate the second frequency lower limit and the second frequency upper limit according to the specified multiple of the target frequency band, determine the second frequency upper limit position and the second frequency lower limit position on the flipped power spectrum according to the second frequency lower limit and the second frequency upper limit, calculate the average power spectrum density within the second frequency upper limit position and the second frequency lower limit position, then flip the average power spectrum density curve, and find all maximum values of the flipped average power spectrum density curve within the target frequency band as the frequency lower limit position of each signal in the target frequency band; The corresponding signal parameters are calculated according to the upper frequency limit position and the lower frequency limit position of each signal in the target frequency band.
2. The broadband signal detection method according to claim 1, characterized in that: The expressions for the first lower frequency limit and the first upper frequency limit are as follows: f o =c-0.5*k*Bw f t =c+0.5*k*Bw Among them, f o is the first frequency lower limit, f t is the first frequency upper limit, Fc is the center frequency of the target frequency band, Bw is the bandwidth of the target frequency band, and k is the specified multiple.
3. The broadband signal detection method according to claim 1, characterized in that: The expressions for the second lower frequency limit and the second upper frequency limit are as follows: f o1 =c+0.5*k*Bw f t1 =c-0.5*k*Bw Among them, f o1 is the second frequency lower limit, f t1 is the second frequency upper limit, Fc is the center frequency of the target frequency band, Bw is the bandwidth of the target frequency band, and k is the specified multiple.
4. The broadband signal detection method according to claim 1, characterized in that: When finding all the maxima of the average power spectral density curve within the target frequency band, including: Get all points of the average power spectral density curve within the target frequency band and calculate the corresponding elements in the matrix of the local maximum scale map; Calculate the sum of each row in the matrix of the local maximum scale map, calculate the minimum value of the sum of each row as the best local maximum scale, and calculate the local maximum scale according to the best local maximum scale and the local maximum scale map; From all local maximum scales, the local maximum scale whose difference with the value of the adjacent local maximum scale is greater than zero is selected as the maximum value.
5. The broadband signal detection method according to claim 4, characterized in that: The expression of the elements in the matrix of the local maximum scale map is as follows: Among them, m k,i The element in the kth row and ith column of the matrix representing the local maximum scale map, i∈{1,2,…,N},k∈{1,2,…,}, Represents the number of all points in the target frequency band, x i-1 、x i-k-1 、x i+k-1 They respectively represent the i-1, ik-1, and i+k-1 points among all the points.
6. The broadband signal detection method according to claim 4, characterized in that: The expression for the sum of each row is as follows: Among them, m k,i The element in the kth row and ith column of the matrix representing the local maximum scale map, i∈{1,2,…,N},k∈{1,2,…,}, Represents the number of all points within the target frequency band.
7. The broadband signal detection method according to claim 4, characterized in that: The expression of the local maximum scale is as follows: Among them, m k,i The element in the kth row and ith column of the matrix representing the local maximum scale map, i∈{1,2,…,N},k∈{1,2,…,λ},N represents the number of all points in the target frequency band, and λ represents the row number corresponding to the minimum value in the sum of each row.
8. A broadband signal detection system, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the broadband signal detection method according to any one of claims 1 to 7.
9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instruction is executed by a processor, the steps of the broadband signal detection method according to any one of claims 1 to 7 are implemented.
10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the steps of the broadband signal detection method according to any one of claims 1 to 7 are implemented.