CN112134818A - Underwater sound signal modulation mode self-adaptive in-class identification method - Google Patents
Underwater sound signal modulation mode self-adaptive in-class identification method Download PDFInfo
- Publication number
- CN112134818A CN112134818A CN202011007851.0A CN202011007851A CN112134818A CN 112134818 A CN112134818 A CN 112134818A CN 202011007851 A CN202011007851 A CN 202011007851A CN 112134818 A CN112134818 A CN 112134818A
- Authority
- CN
- China
- Prior art keywords
- signal
- modulation
- underwater sound
- sound signal
- modulation mode
- 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.)
- Withdrawn
Links
- 238000000034 method Methods 0.000 title claims abstract description 29
- 230000005236 sound signal Effects 0.000 title claims abstract description 26
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 20
- 102100026758 Serine/threonine-protein kinase 16 Human genes 0.000 claims abstract description 14
- 101710184778 Serine/threonine-protein kinase 16 Proteins 0.000 claims abstract description 14
- 238000012545 processing Methods 0.000 claims abstract description 14
- 230000003044 adaptive effect Effects 0.000 claims abstract description 12
- 238000011156 evaluation Methods 0.000 claims abstract description 12
- 238000010586 diagram Methods 0.000 claims abstract description 7
- 238000013528 artificial neural network Methods 0.000 claims description 17
- 230000009466 transformation Effects 0.000 claims description 12
- 238000001914 filtration Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 7
- 230000008859 change Effects 0.000 claims description 6
- 230000009467 reduction Effects 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000011176 pooling Methods 0.000 claims description 4
- 230000003595 spectral effect Effects 0.000 claims description 4
- 238000012847 principal component analysis method Methods 0.000 claims description 3
- 238000000605 extraction Methods 0.000 abstract description 5
- 238000012850 discrimination method Methods 0.000 abstract description 2
- 238000001228 spectrum Methods 0.000 description 19
- 238000004891 communication Methods 0.000 description 11
- 230000005540 biological transmission Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 108010076504 Protein Sorting Signals Proteins 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L27/00—Modulated-carrier systems
- H04L27/0012—Modulated-carrier systems arrangements for identifying the type of modulation
-
- 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
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Complex Calculations (AREA)
- Digital Transmission Methods That Use Modulated Carrier Waves (AREA)
Abstract
The invention discloses a method for identifying the self-adaptive class of an underwater sound signal modulation mode, which comprises the following steps: firstly, identifying underwater sound signal modulation modes: if the underwater sound signal of the MFSK modulation mode is identified among the classes, performing Hilbert transform on the underwater sound signal; then, calculating the order M of the MFSK modulation by using a fuzzy evaluation algorithm; if underwater acoustic signals of MPSK or MQAM modulation mode are identified among the classes, the signals are processed to obtain a two-dimensional signal constellation diagram; and processing the signal constellation diagram by using a quadratic clustering algorithm, and judging the order M of MPSK or MQAM modulation. The method can be applied to the identification in various modulation modes, and solves the problem that the identification performance of other modulation modes in the prior art is limited; and different discrimination methods are adopted for adaptive feature extraction and aiming at the characteristics of extracted features, so that the accuracy of in-class identification is improved, and targeted and high-accuracy underwater sound signal modulation mode adaptive in-class identification is finally realized.
Description
Technical Field
The invention belongs to the technical field of underwater acoustic communication, and particularly relates to an underwater acoustic signal modulation mode self-adaptive in-class identification method.
Background
The underwater wireless data transmission technology is a key technology for acquiring ocean state information and realizing ocean monitoring, and underwater acoustic communication is small in propagation loss and long in transmission distance, so that a main way for acquiring underwater data needs to depend on the underwater acoustic communication technology. Adaptive Modulation Coding (AMC) applied to an underwater acoustic communication system greatly improves the efficiency of data transmission, and the technology can adaptively select the optimal modulation coding mode according to the change of the channel condition of the communication system, thereby fully improving the information transmission efficiency of the underwater acoustic communication. However, in the communication system based on multiple modulation modes, two communication parties need to determine the currently adopted modulation mode through the handshake signals, and the complicated underwater channel environment may cause the handshake signals to be in error, possibly causing the receiving end to adopt a demodulation mode with an incorrect order. And the modulation mode identification can help a receiving end to automatically identify the modulation mode of the signal.
The modulation mode identification comprises two contents of inter-class identification and intra-class identification, wherein the inter-class identification is used for identifying the modulation type of a signal, and the intra-class identification is used for further identifying the modulation order. In-class recognition is more difficult than inter-class recognition because the recognition features are not obvious. At present, the method for identifying the underwater acoustic signal in the modulation mode class comprises the following steps: the designed automatic modulation classification system is used for identifying the modulation mode, and the in-class identification rate of MPSK and MFSK signals is kept at a higher level under the condition that the signal-to-noise ratio is-10 dB-40 dB; the convolutional neural networks trained on different data sets identify signals, and the modulation order of QAM can be effectively identified under low signal-to-noise ratio; the basic principle that the M power spectrum of the MPSK signal has line spectrum characteristics at M times of carrier frequency is utilized, and the analytic signal is adopted to construct the higher order spectrum aiming at the defect of directly estimating the suppression effect of the higher order spectrum on the small signal, so that the processing gain is improved, and the identification performance in PSK signals under low signal-to-noise ratio is improved.
The method can be only used for identifying a few modulation modes or even one type of modulation modes, and has limited identification performance for other modulation modes, thereby greatly limiting the application of the identification method.
Disclosure of Invention
The invention aims to provide an underwater sound signal modulation mode self-adaptive in-class identification method to make up for the defects of the prior art.
In order to realize the purpose of the invention, the invention is realized by adopting the following technical scheme:
an underwater sound signal modulation mode self-adaptive in-class identification method comprises the following steps:
s1: firstly, identifying underwater sound signal modulation modes:
s2: if the underwater sound signal of the MFSK modulation mode is identified among the classes, performing Hilbert transform on the underwater sound signal; then, calculating the order M of the MFSK modulation by using a fuzzy evaluation algorithm;
s3: if underwater acoustic signals of MPSK or MQAM modulation mode are identified among the classes, the signals are processed to obtain a two-dimensional signal constellation diagram; and processing the signal constellation diagram by using a quadratic clustering algorithm, and judging the order M of MPSK or MQAM modulation.
Further, the S1 specifically includes: calculating the spectral characteristic and entropy characteristic of the underwater sound signal; performing dimensionality reduction and denoising treatment on the extracted features by adopting a principal component analysis method; selecting a dense neural network as a training model, removing a pooling layer of the dense neural network, pre-training by using the existing real data, and adjusting parameters of a full connection layer in the neural network according to an actual underwater sound signal to enable the neural network to adapt to a target sea area, thereby obtaining the trained model; and then, carrying out normalization and dimension change processing on the features subjected to the dimension reduction and denoising processing, and inputting the features into the trained model to finish inter-class identification.
Of course, if other existing manners are adopted to complete the classification between classes, the technical solution of the present invention is also considered.
Further, the hilbert transform in S2 is specifically: computingOutputting a signal instantaneous frequency f (t); taking absolute value of instantaneous frequency f (t) and normalizing to obtain signal normalized instantaneous frequency fu(t); at the same time, the signal is wavelet transformed, denoted CHWTMFSK(ii) a And (4) taking an absolute value of the signal after the wavelet transform and performing median filtering to obtain a signal wavelet transform frequency histogram.
Furthermore, the signal normalization instantaneous frequency is calculated by,
wherein f (t) is the instantaneous frequency of the signal, abs is an absolute value taking function, max is a maximum value taking function, and N is the maximum value of the independent variable f (t);
the wavelet transform of the signal is calculated by,
where a is the scaling, i.e., transform scale, τ is the time shift,is the power of the amplitude signal, ωiFor the ith signal without adjusting the carrier angular frequency, omegacFor modulating the carrier angular frequency, thetacIs the carrier initial phase.
Further, the fuzzy evaluation algorithm in S2 specifically includes: normalizing the instantaneous frequency f of the underwater acoustic signal in MFSK modulation modeu(t) carrying out median filtering, and obtaining signal normalization instantaneous frequency f by statisticsu(t) number of stepsCounting the number F of peak values of the wavelet transform frequency histogram of the signalMUsing fuzzy evaluation algorithm to convert the peak value F of wavelet transform frequency histogram of signalMAnd normalized instantaneous frequency statistics of the number of step levelsAnd comparing and judging the set A with a characteristic value set B of the underwater sound signal corresponding to the MFSK modulation under the theoretical condition, and calculating the order M of the MFSK modulation.
Further, the euclidean closeness on which the fuzzy evaluation algorithm is based is defined as,
wherein, A (u)i) U representing A setiElement, B (u)i) Represents the corresponding u in the B setiElements, wherein K is the number of the elements in the set A and the set B; calculating the Euclidean closeness of the signal to be identified and each order of ideal FSK modulation signal, if the Euclidean closeness satisfies the following conditions:
(A,Bi)=Max[N(A,B1),N(A,B2),...,N(A,Bn)]
max is a function for finding the maximum value;
then consider A to be closest to BiAnd identifying A as BiThe modulation mode.
Further, in S3, the underwater acoustic signal is down-converted, matched filtered, and the like to obtain a symbol complex baseband signal, and the signal sequence is further processed to obtain a two-dimensional signal constellation.
Further, in S3, the signal constellation is processed by using a quadratic clustering algorithm, first, an initial clustering number S is obtained by using signal-to-noise ratio-based adaptive subtraction clustering, then, the clustering performance is further improved by using a C-means clustering method, and finally, the clustering number is counted, and the order M of MPSK or MQAM modulation is determined.
The invention has the advantages and beneficial effects that:
the invention adopts a differentiation in-class identification method aiming at different sensitivities of modulation modes to phase and frequency characteristic changes, extracts characteristics according to the characteristics of different modulation modes, adopts a fuzzy evaluation method to identify an MFSK modulation mode, and adopts a quadratic clustering algorithm to identify an MPSK modulation mode and an MQAM modulation mode so as to realize accurate in-class identification of various common modulation modes of underwater acoustic communication.
The method can be applied to the identification in various modulation modes, and solves the problems that the prior art method is only effective for one modulation mode and has limited identification performance for other modulation modes; and different discrimination methods are adopted for adaptive feature extraction and aiming at the characteristics of extracted features, so that the accuracy of in-class identification is improved, and targeted and high-accuracy underwater sound signal modulation mode adaptive in-class identification is finally realized.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
FIG. 1 is a flow chart of an embodiment of the present invention.
Fig. 2 is a signal frequency histogram of an embodiment of the present invention.
Fig. 3 is a signal constellation diagram in an embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings and examples.
The identification step among modulation modes comprises the following steps:
s11, the characteristic extraction and processing step for the inter-class identification comprises the following steps:
s111, calculating power spectrum, singular spectrum, envelope spectrum, square spectrum and fourth power spectrum of the modulation signal, and calculatingCalculating the spectral characteristic and the entropy characteristic of the signal; the spectral characteristics are: maximum value gamma of Q parameter, peak number of power spectrum, R parameter and zero-center normalized instantaneous amplitude spectrum densitymax(ii) a The entropy characteristics are as follows: the entropy of the power spectrum Shannon entropy, the entropy of the power spectrum index, the entropy of the singular spectrum Shannon entropy, the entropy of the singular spectrum index, the entropy of the frequency spectrum amplitude Shannon entropy, the entropy of the frequency spectrum amplitude index, the entropy of the phase spectrum Shannon entropy and the entropy of the phase spectrum index.
And S112, performing dimensionality reduction and denoising treatment on the extracted features by adopting a principal component analysis method.
S12, training a neural network, comprising:
s121, pre-training the improved dense neural network by using a greedy algorithm by using a small amount of existing sea test data and simulation data, and removing a pooling layer for simplifying features on the basis of the original dense neural network by using the network model to avoid losing important features caused by pooling;
and S122, putting the pre-trained improved dense neural network into the target sea area, and finely adjusting parameters of a full connection layer in the neural network according to the actual underwater sound signal to enable the neural network to adapt to the target sea area.
S13, inter-class identification, including:
s131, normalizing and changing dimensions of the features subjected to dimension reduction and denoising processing to enable the features to meet the input type requirement of a neural network; normalizing and changing the dimension of the extracted features;
the characteristic normalization formula is as follows:
x is original characteristic data, X' is normalized data, and Max and Min are respectively the maximum value and the minimum value of the characteristic data;
the dimension change is specifically as follows:
the original feature data of n x m dimensions is converted into feature data of n x m x 1 dimensions suitable for the dense neural network, the feature quantity of the feature data does not change, but logically changes from two dimensions to three dimensions.
And S132, inputting the characteristics into the trained improved dense neural network, and outputting the MFSK, MPSK or MQAM modulation types by the network according to the characteristics to finish the identification among modulation modes.
Based on the inter-class identification result, the modulation mode intra-class identification step comprises the following steps:
s21, a step of adaptive feature extraction, as shown in fig. 1, including,
s211, performing hilbert transform on the underwater acoustic signal of the MFSK modulation scheme identified between classes, and calculating instantaneous frequency f (t), f (t) of the signal as follows:
let the signal to be identified be s (t), and the expression of the analytic signal thereof be
z(t)=s(t)+j·y(t)
Wherein the content of the first and second substances,
wherein H represents Hilbert transform, P represents Cauchy's principal value interval, and τ is time shift. The definition of the method is that,
the instantaneous frequency can be expressed as:
taking absolute value of instantaneous frequency f (t) and normalizing to obtain signal normalized instantaneous frequency fu(t),fuThe calculation method of (t) is as follows:
wherein f (t) is the instantaneous frequency of the signal, abs is the absolute value taking function, max is the maximum value taking function, and N is the maximum value of the independent variable f (t).
At the same time, the signal is wavelet transformed, denoted CHWTMFSKThe calculation method is as follows:
the wavelet transform of signal s (t) is defined as follows:
where a is the scaling, i.e. transformation scale,. tau.is the time shift,. tau.is the complex conjugate,. psi (t) is the wavelet mother function,. psiaAnd (t) the child wavelet function obtained by scaling and translating the mother function.
Because the scaling scale a restricts the signal-to-noise ratio gain after wavelet transformation, the higher the signal-to-noise ratio of the signal is, the larger the amplitude variance of the signal is, the clearer the peak value in the histogram of the wavelet transformation modulus sequence is, and the higher the accuracy of modulation order identification is. Therefore, before wavelet transform is performed on the signal, the scaling a of the wavelet transform needs to be determined first. Suppose R and RWTRespectively, the snr before and after transformation, the relationship between the snr gain and the transformation scale a can be expressed as:
wherein f iscAnd fsRespectively the center frequency and the sampling rate of the system. After wavelet transformation, if GSNR>1, the signal-to-noise ratio after transformation is increased, and the modulation identification accuracy is improved. This example is based on the formula GSNRThe illustrated relationship calculation finds the optimal transformation scale a.
After the wavelet transformation scale is determined, the wavelet transformation is carried out. Within one symbol period, the passband MFSK modulated hydroacoustic signal can be expressed as:
wherein, iTMFSK≤t≤(i+1)TMFSK;Is the power of the amplitude signal; omegacIs a modulated carrier angular frequency; omeganM Δ ω, M1, 2, Λ, M, Δ ω is the amount of change of the symbol with respect to the frequency; thetacIs the initial phase of the carrier; gTIs a rectangular function of unit amplitude and T is the symbol period. The expression of the signal after wavelet transformation is as follows:
wherein, a is a scaling scale,is the power of the amplitude signal, ωcFor modulating the carrier angular frequency, thetacIs the initial phase, ω, of the carrieriThe carrier angular frequency is not adjusted for the ith signal.
Taking absolute values of the wavelet-transformed signals and performing median filtering to obtain a wavelet transform frequency histogram of the signals, wherein the frequency histogram after 2FSK and 4FSK wavelet transform is shown in FIG. 2;
for underwater acoustic signals of MPSK and MQAM modulation modes identified between classes, the symbol complex baseband signal obtained by processing the signals by down-conversion, matched filtering, etc. can be expressed as:
wherein R iskIs the signal amplitude, nkN is the number of sequences for the corresponding noise samples.
After further processing the sequence, a two-dimensional signal constellation is obtained, and the constellation of MQAM is shown in fig. 3.
S22, an identification step, comprising:
s221, normalizing the signal into a transient time-frequency signal aiming at the underwater sound signal of the MFSK modulation modeRate fu(t) carrying out median filtering, and obtaining signal normalization instantaneous frequency f by statisticsu(t) number of stepsCounting the number F of peak values of the wavelet transform frequency histogram of the signalM. Utilizing fuzzy evaluation algorithm to convert the peak value number F of wavelet transform frequency histogram of signalMAnd normalized instantaneous frequency statistics of the number of step levelsAnd comparing and judging the set A with a characteristic value set B of the underwater sound signal corresponding to the MFSK modulation under the theoretical condition, and calculating the order M of the MFSK modulation.
Further, the euclidean closeness on which the fuzzy evaluation algorithm is based is defined as,
wherein, A (u)i) U representing A setiElement, B (u)i) Represents the corresponding u in the B setiAnd K is the number of elements in the set A and the set B.
Further, calculating the Euclidean closeness of the signal to be identified and each order of ideal FSK modulation signal, if the Euclidean closeness satisfies:
(A,Bi)=Max[N(A,B1),N(A,B2),...,N(A,Bn)]
where Max is a function for finding the maximum.
Then consider A to be closest to BiAnd identifying A as BiThe modulation mode.
S222, processing the signal constellation using a quadratic clustering algorithm, assuming that R ═ Rn,kAnd k is 1,2, Λ, N, which is a symbol complex signal sequence to be identified. The quadratic clustering algorithm firstly adopts the self-adaptive subtraction clustering based on the signal-to-noise ratio to obtain the initial clustering number S, and then utilizes the C mean value clustering method to further improve the clustering performance. First, each signal-to-noise ratio-based adaptive subtractive clustering is calculatedThe density index value of the sample point has the following density index function:
wherein, SNR is signal-to-noise ratio; gamma raya(SNR) is based on the density index field of the signal-to-noise ratio, and represents the aggregation degree of signal points and the size of a clustering radius; kaFinding the relation between the clustering radius and the noise by adjusting the value as a performance adjusting coefficient; pnRepresenting the average noise power of the signal; and N is the length of the code element complex signal sequence.
After the density index of each sample point is calculated, the sample point corresponding to the maximum index value is selected as a clustering center, and then the density index of each sample point is corrected. Let this point be rnc,1The density index is Dnc,1The density index is corrected as follows:
wherein, KbIs a constant, and represents the adjustment coefficient of the neighborhood radius with the remarkably reduced density index, which is generally larger than KaTo avoid cluster centers that are very close together. PnRepresenting the average noise power of the signal. After the density index of each point is corrected, the next clustering center r is selectednc,2The density index of all sample points is corrected again. This process is repeated until all sample points have been classified.
It should be noted that, starting from the second time of selecting the cluster center, it is necessary to determine whether the selected cluster center is the center of the new class. Suppose that 1 cluster center has been found to be rnc={rnc,1,rnc,2,Λ,rnc,lThen the l +1 th sample point with the maximum density index is calculated as rnc,l+1Then, whether the point is the center of the new class is judged, and the judgment formula is as follows:
min{||rnc,k-rnc,l+1||2}>KcPn,k=1,2,Λ,l
wherein min is a function of minimum value, KcTo adjust the coefficient, PnRepresenting the average noise power of the signal. After the judgment, if the point is not the new class center, the repeated process is ended.
The initial clustering number S can be obtained by the subtractive clustering, and then the performance is improved by performing the second C-means clustering. Let mi(i ═ 1,2, Λ, S) is the center of each cluster, NiIs composed ofiNumber of samples in a class. The second C-means clustering step is:
average noise power P based on distance squared minimum principlenAnd cluster effective radius KdAll sample points are classified into the class represented by the initial cluster center:
i={rn,j,j=1,2,Λ,Ni},i=1,2,Λ,S
wherein r isn,jIt should satisfy:
taking the mean value of all sample points in the class to obtain a new clustering center:
r isiSample points of class, NiThe number of corresponding sample points.
Finding the minimum value among all cluster center points:
dmin=||ma-mb||2
where a, b ≠ 1,2, Λ, i, and a ≠ b.
Judging whether the clustering centers are too close according to the clustering ending radius according to the following judgment basis:
dmin<KuPn
where Ku is a determination coefficient and Pn is an average noise power of the signal.
If the above formula is satisfied, the average value of the two kinds of cluster centers is used as a new cluster center after combination, and the total cluster number is changed into S-1.
And repeating the steps until the formula is not satisfied. Solving the square sum of each sample and the cluster center in all classes:
wherein r isiSample points of classes, m is the cluster center, and S is the total cluster number.
Cost function JeWhen the slope of the curve is approximately 0, the loop is ended, a final clustering center point is given, the clustering number is finally counted, and the order M of MPSK or MQAM modulation is judged.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions.
Claims (8)
1. An underwater sound signal modulation mode self-adaptive in-class identification method is characterized by comprising the following steps:
s1: firstly, identifying underwater sound signal modulation modes:
s2: if the underwater sound signal of the MFSK modulation mode is identified among the classes, performing Hilbert transform on the underwater sound signal; then, calculating the order M of the MFSK modulation by using a fuzzy evaluation algorithm;
s3: if underwater acoustic signals of MPSK or MQAM modulation mode are identified among the classes, the signals are processed to obtain a two-dimensional signal constellation diagram; and processing the signal constellation diagram by using a quadratic clustering algorithm, and judging the order M of MPSK or MQAM modulation.
2. The in-class identification method according to claim 1, wherein the S1 is specifically: calculating the spectral characteristic and entropy characteristic of the underwater sound signal; performing dimensionality reduction and denoising treatment on the extracted features by adopting a principal component analysis method; selecting a dense neural network as a training model, removing a pooling layer of the dense neural network, pre-training by using the existing real data, and adjusting parameters of a full connection layer in the neural network according to an actual underwater sound signal to enable the neural network to adapt to a target sea area, thereby obtaining the trained model; and then, carrying out normalization and dimension change processing on the features subjected to the dimension reduction and denoising processing, and inputting the features into the trained model to finish inter-class identification.
3. The in-class identification method according to claim 1, wherein the hilbert transform in S2 is specifically: calculating the instantaneous frequency f (t) of the signal; taking absolute value of instantaneous frequency f (t) and normalizing to obtain signal normalized instantaneous frequency fu(t); at the same time, the signal is wavelet transformed, denoted CHWTMFSK(ii) a And (4) taking an absolute value of the signal after the wavelet transform and performing median filtering to obtain a signal wavelet transform frequency histogram.
4. The in-class identification method of claim 3 wherein said signal normalized instantaneous frequency is calculated by,
wherein f (t) is the instantaneous frequency of the signal, abs is an absolute value taking function, max is a maximum value taking function, and N is the maximum value of the independent variable f (t);
the wavelet transform of the signal is calculated by,
5. The method according to claim 1, wherein the fuzzy evaluation algorithm in S2 is specifically: normalizing the instantaneous frequency f of the underwater acoustic signal in MFSK modulation modeu(t) carrying out median filtering, and obtaining signal normalization instantaneous frequency f by statisticsu(t) number of stepsCounting the number F of peak values of the wavelet transform frequency histogram of the signalMUsing fuzzy evaluation algorithm to convert the peak value F of wavelet transform frequency histogram of signalMAnd normalized instantaneous frequency statistics of the number of step levelsAnd comparing and judging the set A with a characteristic value set B of the underwater sound signal corresponding to the MFSK modulation under the theoretical condition, and calculating the order M of the MFSK modulation.
6. The method of in-class identification according to claim 5 wherein the Euclidean closeness on which the fuzzy evaluation algorithm is based is defined as,
wherein, A (u)i) U representing A setiElement, B (u)i) Represents the corresponding u in the B setiElements, wherein K is the number of the elements in the set A and the set B; calculating the Euclidean closeness between the signal to be identified and each order of ideal FSK modulation signal ifSatisfies the following conditions: (A, B)i)=Max[N(A,B1),N(A,B2),...,N(A,Bn)]
Then consider A to be closest to BiAnd identifying A as BiThe modulation mode.
7. The method according to claim 1, wherein in S3, the two-dimensional signal constellation is obtained by further processing the symbol complex baseband signal obtained by down-converting and matched filtering the underwater acoustic signal.
8. The method according to claim 1, wherein in S3, the signal constellation is processed by using quadratic clustering algorithm, first, an initial clustering number S is obtained by adaptive subtraction clustering based on signal-to-noise ratio, then, the clustering performance is further improved by using C-means clustering method, and finally, the clustering number is counted to determine the order M of MPSK or MQAM modulation.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011007851.0A CN112134818A (en) | 2020-09-23 | 2020-09-23 | Underwater sound signal modulation mode self-adaptive in-class identification method |
CN202110214166.3A CN112737992B (en) | 2020-09-23 | 2021-02-26 | Underwater sound signal modulation mode self-adaptive in-class identification method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011007851.0A CN112134818A (en) | 2020-09-23 | 2020-09-23 | Underwater sound signal modulation mode self-adaptive in-class identification method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112134818A true CN112134818A (en) | 2020-12-25 |
Family
ID=73842787
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011007851.0A Withdrawn CN112134818A (en) | 2020-09-23 | 2020-09-23 | Underwater sound signal modulation mode self-adaptive in-class identification method |
CN202110214166.3A Active CN112737992B (en) | 2020-09-23 | 2021-02-26 | Underwater sound signal modulation mode self-adaptive in-class identification method |
Family Applications After (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110214166.3A Active CN112737992B (en) | 2020-09-23 | 2021-02-26 | Underwater sound signal modulation mode self-adaptive in-class identification method |
Country Status (1)
Country | Link |
---|---|
CN (2) | CN112134818A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112910813A (en) * | 2021-04-10 | 2021-06-04 | 青岛科技大学 | LDA-KNN-based underwater sound signal automatic modulation identification method |
CN113259288A (en) * | 2021-05-05 | 2021-08-13 | 青岛科技大学 | Underwater acoustic communication modulation mode identification method based on feature fusion and lightweight hybrid neural network |
CN114024804A (en) * | 2021-12-10 | 2022-02-08 | 西安交通大学 | Maximum likelihood detection method and system assisted by overlapping clustering |
Family Cites Families (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN100521670C (en) * | 2006-10-27 | 2009-07-29 | 清华大学 | Detecting and analyzing method for multi system frequency shift key control signal |
CN101764786B (en) * | 2009-12-11 | 2012-06-20 | 西安电子科技大学 | MQAM signal recognition method based on clustering algorithm |
CN102497343A (en) * | 2011-11-25 | 2012-06-13 | 南京邮电大学 | Combined modulation recognition method based on clustering and support vector machine |
CN102916917B (en) * | 2012-09-25 | 2014-12-17 | 哈尔滨工程大学 | Individual identification method of FSK (frequency-shift keying) signal based on slice bi-spectrum and wavelet transformation |
CN105119862B (en) * | 2015-07-22 | 2019-02-19 | 中国电子科技集团公司第三十六研究所 | A kind of identification of signal modulation method and system |
CN107707497B (en) * | 2017-05-09 | 2020-06-02 | 电子科技大学 | Communication signal identification method based on subtraction clustering and fuzzy clustering algorithm |
CN107612867B (en) * | 2017-07-29 | 2020-06-09 | 西安电子科技大学 | MQAM signal modulation order identification method |
CN107948107B (en) * | 2017-11-16 | 2021-04-20 | 成都玖锦科技有限公司 | Digital modulation signal classification method based on joint features |
CN108540202B (en) * | 2018-03-15 | 2021-01-26 | 西安电子科技大学 | Satellite communication signal modulation mode identification method and satellite communication system |
CN110086554B (en) * | 2018-11-16 | 2021-09-28 | 中国西安卫星测控中心 | Signal identification method based on spectrum sensing |
CN109597043B (en) * | 2018-11-16 | 2023-05-26 | 江苏科技大学 | Radar signal identification method based on quantum particle swarm convolutional neural network |
CN109802905B (en) * | 2018-12-27 | 2022-01-14 | 西安电子科技大学 | CNN convolutional neural network-based digital signal automatic modulation identification method |
CN111010356A (en) * | 2019-11-08 | 2020-04-14 | 西北工业大学 | Underwater acoustic communication signal modulation mode identification method based on support vector machine |
-
2020
- 2020-09-23 CN CN202011007851.0A patent/CN112134818A/en not_active Withdrawn
-
2021
- 2021-02-26 CN CN202110214166.3A patent/CN112737992B/en active Active
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112910813A (en) * | 2021-04-10 | 2021-06-04 | 青岛科技大学 | LDA-KNN-based underwater sound signal automatic modulation identification method |
CN112910813B (en) * | 2021-04-10 | 2022-09-06 | 青岛科技大学 | LDA-KNN-based underwater sound signal automatic modulation identification method |
CN113259288A (en) * | 2021-05-05 | 2021-08-13 | 青岛科技大学 | Underwater acoustic communication modulation mode identification method based on feature fusion and lightweight hybrid neural network |
CN113259288B (en) * | 2021-05-05 | 2023-08-08 | 青岛科技大学 | Underwater sound modulation mode identification method based on feature fusion and lightweight hybrid model |
CN114024804A (en) * | 2021-12-10 | 2022-02-08 | 西安交通大学 | Maximum likelihood detection method and system assisted by overlapping clustering |
Also Published As
Publication number | Publication date |
---|---|
CN112737992A (en) | 2021-04-30 |
CN112737992B (en) | 2023-02-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112737992B (en) | Underwater sound signal modulation mode self-adaptive in-class identification method | |
CN109120563B (en) | Modulation recognition method based on neural network integration | |
CN111814777B (en) | Modulation pattern recognition method based on characteristic quantity grading | |
CN108718288A (en) | Recognition of digital modulation schemes method based on convolutional neural networks | |
CN112733811B (en) | Method for identifying underwater sound signal modulation modes based on improved dense neural network | |
CN114567528B (en) | Communication signal modulation mode open set recognition method and system based on deep learning | |
CN108052956A (en) | Wireless light communication subcarrier modulation constellation recognition methods under a kind of atmospheric turbulance | |
CN100521670C (en) | Detecting and analyzing method for multi system frequency shift key control signal | |
CN113037663B (en) | Improved code element rate estimation method suitable for non-constant envelope signal | |
CN113225282A (en) | Communication signal modulation identification method based on BP neural network | |
CN113452637B (en) | Underwater acoustic communication signal modulation identification method based on feature selection and support vector machine | |
Wong et al. | Emitter identification using CNN IQ imbalance estimators | |
CN116708108A (en) | Communication signal modulation recognition system and method | |
CN112202696B (en) | Underwater sound signal automatic modulation identification method based on fuzzy self-encoder | |
CN115378776A (en) | MFSK modulation identification method based on cyclic spectrum parameters | |
Mihandoost et al. | Introducing an efficient statistical model for automatic modulation classification | |
CN113076925B (en) | M-QAM signal modulation mode identification method based on CNN and ELM | |
CN115913849A (en) | Electromagnetic signal identification method based on one-dimensional complex value residual error network | |
CN116359851A (en) | Radar active interference detection and identification method and device based on converged network | |
Batra et al. | A Study of Automatic Modulation Classification for Rayleigh Fading Channel | |
CN104009951B (en) | Signal modulation mode recognition methods and system based on multi-scale wavelet transformation | |
CN114785649B (en) | Satellite communication signal identification method based on multiport neural network | |
CN112910813B (en) | LDA-KNN-based underwater sound signal automatic modulation identification method | |
CN111935044B (en) | PSK (phase Shift keying) and QAM (Quadrature amplitude modulation) signal modulation identification method based on EMD (empirical mode decomposition) | |
CN117614467B (en) | Underwater sound signal intelligent receiving method based on noise reduction neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
WW01 | Invention patent application withdrawn after publication | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20201225 |