CN115902391A - Dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition - Google Patents
Dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition Download PDFInfo
- Publication number
- CN115902391A CN115902391A CN202211296553.7A CN202211296553A CN115902391A CN 115902391 A CN115902391 A CN 115902391A CN 202211296553 A CN202211296553 A CN 202211296553A CN 115902391 A CN115902391 A CN 115902391A
- Authority
- CN
- China
- Prior art keywords
- power spectrum
- signal detection
- morphological filtering
- eigenvalue decomposition
- threshold
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention provides a dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition, which comprises the following steps: receiving and sampling signals, and calculating a sampling sequence of the received signals according to an average periodogram method to obtain an original power spectrum of the signals; carrying out minimum filtering and morphological filtering on the original power spectrum in sequence, and interpolating to the length of the original power spectrum to obtain a noise floor estimation sequence; subtracting the noise-floor estimation sequence from the original power spectrum, constructing a Hankel matrix, performing eigenvalue decomposition to obtain each order of component of the power spectrum without the noise floor, and taking the first order of component as a reconstructed power spectrum; calculating the reconstructed power spectrum according to an iterative approximation algorithm to obtain a signal detection threshold; and monitoring the frequency spectrum according to the reconstructed power spectrum and the signal detection threshold, judging whether the signal exists or not and analyzing the frequency and the bandwidth of the signal. The invention improves the accuracy rate under the conditions of too close carrier frequency distance, large noise floor fluctuation and weak signals.
Description
Technical Field
The invention relates to the field of signal detection, in particular to a dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition.
Background
The multi-signal detection technology under the broadband receiving condition is widely applied to the civil and military fields of electromagnetic spectrum monitoring, electronic warfare and the like. Has important theoretical value and practical significance for the research of the related technology. The wide frequency band range usually includes multiple paths of signals, the carrier frequency and the bandwidth of each path of signal are different, even a time-frequency aliasing phenomenon exists, and due to the influence of the impedance mismatching between the complicated electromagnetic environment and the analog front end part of the broadband receiver, the noise floor of the broadband signal obtained by the broadband receiver has larger fluctuation.
The traditional fixed threshold signal detection method is easy to have the conditions of missing detection and false alarm under the conditions that the carrier frequency distance of the signal is too close and the noise floor fluctuation is large. Usually, the power spectrum needs to be smoothed before signal detection, signals with too close carrier frequencies are easily smoothed into one signal, and weak signals are smoothed as noise, so that detection errors are caused. Based on the above background and similar problems encountered in practical engineering, it is desirable to provide a method capable of improving signal detection performance.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: aiming at the technical problems in the prior art, the invention provides a dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition, which solves the problem of difficult signal detection caused by uneven in-band, weak signals, too small signal carrier frequency interval and the like.
In order to solve the technical problems, the technical scheme provided by the invention is as follows:
a dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition comprises the following steps:
s1) receiving and sampling signals, and calculating a sampling sequence of the received signals according to an average periodogram method to obtain an original power spectrum of the signals;
s2) uniformly segmenting the original power spectrum sequence, extracting the minimum value of each segment, and taking the extracted sequence as the power spectrum after minimum value filtering extraction;
s3) performing morphological filtering opening operation on the power spectrum after minimum filtering extraction to obtain a morphologically filtered power spectrum;
s4) interpolating the power spectrum after morphological filtering to the length of the original power spectrum to obtain a noise floor estimation sequence;
s5) subtracting the noise-floor estimation sequence from the original power spectrum to obtain a power spectrum without a noise floor;
s6) constructing a Hankel matrix based on the power spectrum without the noise floor, performing eigenvalue decomposition on the Hankel matrix to obtain each order of component of the power spectrum without the noise floor, and taking the first order of component as a reconstructed power spectrum;
s7) calculating the reconstructed power spectrum according to an iterative approximation algorithm to obtain a signal detection threshold;
and S8) carrying out frequency spectrum monitoring according to the reconstructed power spectrum and the signal detection threshold, judging whether the signal exists or not and analyzing the frequency and the bandwidth of the signal.
Further, the structural element of the opening operation of the morphological filtering in step S3) is an element smaller than a preset threshold in the sequence of the power spectrum.
Further, the iterative approximation algorithm in step S7) specifically includes the following steps:
s71) acquiring a power spectrum and a threshold reference value;
s72) calculating the maximum value and the minimum value of the power spectrum, and initializing the data point count and the threshold value near the threshold;
s73) calculating the number of data points within a first preset value range of the maximum value of the power spectrum;
s74) if the number of the data points is larger than the data point count, taking the number of the data points as a new data point count, and taking the maximum value of the power spectrum as a new threshold value;
s75) the maximum value of the power spectrum is automatically reduced according to a second preset value, and the calculation result is used as a new maximum value of the power spectrum;
s76) returning to step S73) until the power spectrum maximum is less than the power spectrum minimum;
s77) adding the threshold value and the threshold reference value to output as a signal detection threshold.
Further, the first preset value is 0.5, and the second preset value is 0.1.
Further, in step S1), 262144 signal sample points are obtained by sampling, and the original power spectrum of the signal is obtained by performing power spectrum estimation on the 262144 signal sample points by using an average periodogram method in which the single frame 65536 points are overlapped to 50%.
Further, the sequence of the original power spectrum is uniformly divided into 1024 segments in step S2).
Further, the Hankel matrix in the step S6) is a 100 × 262045 order matrix.
The invention also provides a dynamic threshold signal detection device based on morphological filtering and eigenvalue decomposition, which comprises a computer device programmed or configured to execute the dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition.
The present invention further proposes a computer-readable storage medium storing a computer program programmed or configured to perform the method for dynamic threshold signal detection based on morphological filtering and eigenvalue decomposition.
Compared with the prior art, the invention has the advantages that:
the method of the invention extracts the minimum value of the power spectrum of the signal and carries out morphological filtering to remove the noise base, carries out characteristic value decomposition on the power spectrum on the basis of removing the noise base, reconstructs the power spectrum and estimates the signal detection threshold by using an iterative approximation method, thereby improving the signal detection accuracy rate under the conditions of too close carrier frequency distance, larger noise base fluctuation and weak signal.
Drawings
Fig. 1 is a block diagram illustrating a specific process according to an embodiment of the present invention.
Fig. 2 is a block diagram illustrating a noise floor estimation process according to an embodiment of the present invention.
Detailed Description
The invention is further described below with reference to the drawings and specific preferred embodiments of the description, without thereby limiting the scope of protection of the invention.
As shown in fig. 1, the present invention provides a dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition, which includes the following steps:
s1) receiving and sampling signals, and calculating a sampling sequence of the received signals according to an average periodogram method to obtain an original power spectrum of the signals;
s2) uniformly segmenting the original power spectrum sequence, extracting the minimum value of each segment, and taking the extracted sequence as the power spectrum after minimum value filtering extraction;
s3) performing morphological filtering opening operation on the power spectrum after minimum filtering extraction to obtain a morphologically filtered power spectrum;
s4) interpolating the power spectrum after morphological filtering to the length of the original power spectrum to obtain a noise floor estimation sequence;
s5) subtracting the noise-floor estimation sequence from the original power spectrum to obtain a power spectrum without a noise floor;
s6) constructing a Hankel matrix based on the power spectrum without the noise floor, performing eigenvalue decomposition on the Hankel matrix to obtain each order of component of the power spectrum without the noise floor, and taking the first order of component as a reconstructed power spectrum;
s7) calculating the reconstructed power spectrum according to an iterative approximation algorithm to obtain a signal detection threshold;
and S8) carrying out frequency spectrum monitoring according to the reconstructed power spectrum and the signal detection threshold, judging whether the signal exists or not and analyzing the frequency and the bandwidth of the signal.
In this embodiment, the noise floor is estimated and flattened by using morphological filtering in steps S3) to S5), and the morphological operation is a nonlinear transformation theory based on a shape, which has an advantage of being able to change local characteristics of a signal. Let signal F be a discrete function defined on F = {0,1, \8230;, N }, the structuring element G be a function defined on G = {0,1, \8230;, M } and N > M. Then four basic operations of morphology can be defined:
the expansion operation is
The corrosion operation is
The turn-on operation is
Closure operation is
The dilation operation reduces the valley bottom of the signal and expands the peak, and the erosion operation can reduce the peak of the signal and widen the valley bottom. The opening operation and the closing operation are a combination of the expansion operation and the corrosion operation, the opening operation can eliminate the signal peak, and the closing operation can inhibit the signal valley.
Although the estimation of the noise floor can also be completed by directly performing morphological filtering on the original power spectrum, when the signal bandwidth is wide, a larger structural element is needed, so that the required calculation amount is larger, and the larger the structural element is, the easier the peak and the valley at the bottom of the noise is to be eliminated. In order to reduce the operation time and improve the estimation accuracy, the present embodiment further performs minimum value filtering extraction on the original power spectrum before performing morphological filtering through step S2), as shown in fig. 1 and 2.
In this embodiment, the structural element of the opening operation of the morphological filtering in step S3) is an element smaller than the preset threshold in the sequence of the power spectrum, that is, in this embodiment, a smaller structural element is selected from the sequence of the power spectrum to perform the morphological opening operation, so as to improve the accuracy of the estimation.
In this embodiment, the power spectrum is reconstructed through step S6), and the specific implementation process is as follows:
for a power spectrum X = [ X (1), X (2), \8230;, X (N) ], a Hankel matrix D can be constructed as follows:
wherein 1<n<N, let m = N-N +1, then D ∈ R m*n ，R m*n A set of Hankel matrices constructed for the power spectral energy of a signal.
D is subjected to singular value decomposition:
in the above formula u i ∈R m*1 (u i Column vector R as matrix D m*1 Subset of (v), v i ∈R n*1 (v i Is the row vector R of the matrix D n*1 Subset of (d), i =1,2, \8230;, r r = min (m, n).
Is provided with
(σ i Is a singular value of order i), i.e.:
if at all i ＝[X i (1),X i (2),…,X i (n)]，E i ＝[X i (n+1),X i (n+2),…,X i (N)] T I.e. the ith order component of the component singular value can be constructed:
for the Hankel matrix D, let:
S＝[X(1),X(2),…,X(N)] (9)
Y＝[X(n+1),X(n+2),…,X(n)] T (10)
according to formula (6):
S＝O 1 +O 2 +…+O R (11)
known from the Hankel matrix construction process, the power spectrum X = [ S, Y = T ]The binding formulae (8), (11) and (12) give:
X＝P 1 +P 2 +…+P r (13)
the above formula is a decomposition formula of the power spectrum X by adopting a Hankel matrix and performing singular value decomposition. We observe a first order component P 1 Is the main component of the power spectrum X, can well embody the envelope of the power spectrum, and can well distinguish weak signals from adjacent signals, thereby leading the first-order component P to be 1 As a reconstructed power spectrum after decomposition.
After the power spectrum is preprocessed in the steps S3) to S6), when there is no signal in the frequency band, the noise floor is relatively flat, and the energy of the frequency band range with the signal is obviously higher than that of the noise region. The signal can be detected by a reasonable single threshold, and the center frequency and the bandwidth of the signal are roughly estimated for signal processing such as subsequent down-conversion. Usually, when the power spectrum of a signal changes, a fixed threshold is difficult to correctly distinguish the signal from noise, and we observe that the number of points of the power spectrum in an energy interval of a boundary between the signal and the noise is the largest, so that in step S7), an iterative approximation algorithm is provided to estimate a noise floor and a signal boundary of the power spectrum to realize threshold adaptation, and the iterative approximation algorithm specifically includes the following steps:
s71) acquiring a power spectrum and a threshold reference value;
s72) calculating the maximum value and the minimum value of the power spectrum, and initializing the data point count and the threshold value near the threshold;
s73) calculating the number of data points in a first preset value range of the maximum value of the power spectrum, wherein the first preset value is 0.5 in the embodiment;
s74) if the number of the data points is larger than the data point count, taking the number of the data points as a new data point count, and taking the maximum value of the power spectrum as a new threshold value;
s75) the maximum value of the power spectrum is automatically reduced according to a second preset value, and the calculation result is used as a new maximum value of the power spectrum, wherein the second preset value is 0.1;
s76) returning to step S73) until the power spectrum maximum is less than the power spectrum minimum;
s77) adding the threshold value and the threshold reference value to output as a signal detection threshold.
The specific algorithm flow is as follows:
the preferred embodiment is further described below in conjunction with FIG. 1:
step one, executing step S1), receiving signals and sampling to obtain 262144 signal sample points, performing power spectrum estimation of an average periodogram method with single-frame 65536 point overlapping of 50% on the 262144 signal sample points to obtain an original power spectrum of the signals, and performing calculation by using the parameters to obtain higher calculation accuracy with less calculation amount;
step two, executing step S2), uniformly dividing the original power spectrum sequence into 1024 sections, extracting the minimum value of each section, and filtering the extracted power spectrum by taking the extracted sequence as the minimum value;
step three, executing step 3), performing morphological filtering opening operation on the power spectrum after minimum filtering extraction to obtain a morphologically filtered power spectrum;
step four, executing step S4), interpolating the power spectrum after the morphological filtering to the length of the original power spectrum to obtain a noise base estimation sequence;
step five, executing step S5), subtracting the noise floor estimation sequence from the original power spectrum to obtain a power spectrum with the noise floor removed;
step six, executing step 6), constructing a 100 × 262045 order Hankel matrix based on the power spectrum without the noise base, performing eigenvalue decomposition on the Hankel matrix to obtain each order component of the power spectrum without the noise base, and taking the first order component as a reconstructed power spectrum;
step seven, executing step 7), and calculating the reconstructed power spectrum according to an iterative approximation algorithm to obtain a signal detection threshold;
step eight, executing step 8), performing spectrum monitoring according to the reconstructed power spectrum and the signal detection threshold, judging whether the signal exists, and analyzing the frequency and the bandwidth of the signal.
The invention also proposes a dynamic threshold signal detection device based on morphological filtering and eigenvalue decomposition, comprising a computer device programmed or configured to execute said method of dynamic threshold signal detection based on morphological filtering and eigenvalue decomposition.
The present invention also proposes a computer readable storage medium storing a computer program programmed or configured to perform the method for dynamic threshold signal detection based on morphological filtering and eigenvalue decomposition.
The foregoing is considered as illustrative of the preferred embodiments of the invention and is not to be construed as limiting the invention in any way. Although the present invention has been described with reference to the preferred embodiments, it is not intended to be limited thereto. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention shall fall within the protection scope of the technical solution of the present invention, unless the technical essence of the present invention departs from the content of the technical solution of the present invention.
Claims (9)
1. A dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition is characterized by comprising the following steps:
s1) receiving and sampling signals, and calculating a sampling sequence of the received signals according to an average periodogram method to obtain an original power spectrum of the signals;
s2) uniformly segmenting the original power spectrum sequence, extracting the minimum value of each segment, and taking the extracted sequence as the power spectrum after minimum value filtering extraction;
s3) performing morphological filtering opening operation on the power spectrum after minimum filtering extraction to obtain a morphologically filtered power spectrum;
s4) interpolating the power spectrum after morphological filtering to the length of the original power spectrum to obtain a noise floor estimation sequence;
s5) subtracting the noise-floor estimation sequence from the original power spectrum to obtain a power spectrum without a noise floor;
s6) constructing a Hankel matrix based on the power spectrum without the noise base, performing eigenvalue decomposition on the Hankel matrix to obtain each order of component of the power spectrum without the noise base, and taking the first order of component as a reconstructed power spectrum;
s7) calculating the reconstructed power spectrum according to an iterative approximation algorithm to obtain a signal detection threshold;
and S8) carrying out frequency spectrum monitoring according to the reconstructed power spectrum and the signal detection threshold, judging whether the signal exists or not and analyzing the frequency and the bandwidth of the signal.
2. The dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition according to claim 1, wherein the structural element of the opening operation of the morphological filtering in step S3) is an element in the sequence of the power spectrum that is smaller than a preset threshold.
3. The dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition according to claim 1, wherein the iterative approximation algorithm in step S7) specifically comprises the following steps:
s71) acquiring a power spectrum and a threshold reference value;
s72) calculating the maximum value and the minimum value of the power spectrum, and initializing the data point count and the threshold value near the threshold;
s73) calculating the number of data points within a first preset value range of the maximum value of the power spectrum;
s74) if the number of the data points is larger than the data point count, taking the number of the data points as a new data point count, and taking the maximum value of the power spectrum as a new threshold value;
s75) the maximum value of the power spectrum is automatically reduced according to a second preset value, and the calculation result is used as a new maximum value of the power spectrum;
s76) returning to step S73) until the power spectrum maximum is less than the power spectrum minimum;
s77) adding the threshold value and the threshold reference value to output as a signal detection threshold.
4. The dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition according to claim 3 wherein said first preset value is 0.5 and said second preset value is 0.1.
5. The dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition of claim 1 wherein, in step S1), 262144 signal sample points are obtained by sampling, and the original power spectrum of the signal is obtained by performing a single frame 65536 point overlapping of 50% on the average periodogram power spectrum estimation on the 262144 signal sample points.
6. The dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition according to claim 5 characterized in that the sequence of the original power spectrum is evenly divided into 1024 segments in step S2).
7. The dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition of claim 6 wherein in step S6) the Hankel matrix is a 100 x 262045 order matrix.
8. A dynamic threshold signal detection apparatus based on morphological filtering and eigenvalue decomposition comprising a computer device programmed or configured to perform the dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition of any of the claims 1 to 7.
9. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program programmed or configured to perform the dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition of any of the claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211296553.7A CN115902391A (en) | 2022-10-21 | 2022-10-21 | Dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211296553.7A CN115902391A (en) | 2022-10-21 | 2022-10-21 | Dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115902391A true CN115902391A (en) | 2023-04-04 |
Family
ID=86473472
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211296553.7A Pending CN115902391A (en) | 2022-10-21 | 2022-10-21 | Dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115902391A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116707675A (en) * | 2023-08-03 | 2023-09-05 | 兰州交通大学 | Method and device for detecting radio signal and method and device for detecting abnormality of radio signal |
-
2022
- 2022-10-21 CN CN202211296553.7A patent/CN115902391A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116707675A (en) * | 2023-08-03 | 2023-09-05 | 兰州交通大学 | Method and device for detecting radio signal and method and device for detecting abnormality of radio signal |
| CN116707675B (en) * | 2023-08-03 | 2023-11-03 | 兰州交通大学 | Method and device for detecting radio signal and method and device for detecting abnormality of radio signal |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Craigmile et al. | Trend assessment in a long memory dependence model using the discrete wavelet transform | |
| CN115902391A (en) | Dynamic threshold signal detection method based on morphological filtering and eigenvalue decomposition | |
| CN109490862B (en) | Carrier frequency estimation method based on phase difference statistical spectrum | |
| CN114492538B (en) | Urban medium-voltage distribution cable partial discharge signal denoising method | |
| CN112014692A (en) | Partial discharge ultrahigh frequency signal blind source separation denoising method based on principal component analysis | |
| CN113325277A (en) | Partial discharge processing method | |
| Ready et al. | Automatic noise floor spectrum estimation in the presence of signals | |
| CN111046791A (en) | Current signal filtering and denoising method based on generalized S transform containing variable factors | |
| CN103308829A (en) | GIS (gas insulated switchgear) single-time partial discharge signal extraction and trigger moment adjusting method | |
| CN114186596A (en) | Multi-window identification method and device for spectrogram peaks and electronic equipment | |
| CN115508618B (en) | Quasi-synchronous harmonic analysis device and method based on time domain Hermite interpolation | |
| CN111125625A (en) | Spectrum baseline calculation method based on embedded system | |
| CN112269182A (en) | Method and device for determining target radar signal, storage medium and electronic device | |
| CN107704831A (en) | A kind of gas density data noise reduction based on singular value decomposition median method | |
| Kazakov et al. | Algorithm of Robust Frequency Estimation in a Channel with White Gaussian Noise and Pulse Interferences | |
| CN106330361B (en) | Weak Signal Detection Method based on connectivity of graph analysis | |
| CN112817056B (en) | Magnetotelluric signal denoising method and system | |
| CN115378776A (en) | MFSK modulation identification method based on cyclic spectrum parameters | |
| CN112285793B (en) | Magnetotelluric denoising method and system | |
| CN114513226A (en) | Method and device for estimating parameters of frequency hopping network station, frequency hopping monitoring equipment and storage medium | |
| CN113705347A (en) | Space charge noise suppression method and device based on time-frequency analysis | |
| CN113824518A (en) | Non-cooperative signal detection method based on frequency spectrum | |
| CN113705335A (en) | Time-frequency super-resolution extraction method for low-frequency sparse line spectrum signal | |
| CN116805291B (en) | Unmanned aerial vehicle signal frequency band detection method and device | |
| CN113702772B (en) | High-resistance grounding fault detection method for power distribution network and related device thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |