CN110288025A - Frequency spectrum sensing method, device and equipment based on information geometry and spectral clustering - Google Patents
Frequency spectrum sensing method, device and equipment based on information geometry and spectral clustering Download PDFInfo
- Publication number
- CN110288025A CN110288025A CN201910564127.9A CN201910564127A CN110288025A CN 110288025 A CN110288025 A CN 110288025A CN 201910564127 A CN201910564127 A CN 201910564127A CN 110288025 A CN110288025 A CN 110288025A
- Authority
- CN
- China
- Prior art keywords
- matrix
- frequency spectrum
- spectral clustering
- covariance matrix
- obtains
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/382—Monitoring; Testing of propagation channels for resource allocation, admission control or handover
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Electromagnetism (AREA)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a kind of frequency spectrum sensing method based on information geometry and spectral clustering, comprising: obtain and treat each covariance matrix that perception primary user's signal progress preset times sampling and is calculated respectively by each secondary user；Each covariance matrix is mapped to matrix manifold using information geometry, obtains each covariance matrix corresponding coordinate points in matrix manifold；Obtain Riemann's mean value by obtaining in advance to noise covariance matrix training；Calculate separately the geodesic curve distance between each coordinate points and Riemann's mean value；Classified using spectral clustering to each geodesic curve distance, obtains the frequency spectrum perception result for treating perception primary user's signal.Using technical solution provided by the embodiment of the present invention, the detection efficiency of frequency spectrum perception is improved, significantly improves the accuracy of frequency spectrum perception.The invention also discloses a kind of frequency spectrum sensing device based on information geometry and spectral clustering, equipment and storage mediums, have relevant art effect.
Description
Technical field
The present invention relates to frequency spectrum perception technical fields, more particularly to a kind of frequency spectrum sense based on information geometry and spectral clustering
Perception method, device, equipment and computer readable storage medium.
Background technique
Radio communication frequency spectrum is a kind of resource of preciousness, and with the high speed development of wireless communication technique, frequency spectrum resource is poor
Weary problem is got worse, however the frequency spectrum resource utilization rate of overwhelming majority of countries but allows of no optimist.Frequency spectrum perception is to realize to recognize
Know the key technology of radio, frequency spectrum perception process nature is that time user (SU) basis believes primary user (PU) in frequency range to be detected
Number analysis, judge whether there is frequency spectrum cavity-pocket, just carry out the access and utilization of frequency spectrum if there is frequency spectrum cavity-pocket, it is on the contrary then after
It is continuous to find other idle frequency ranges.Traditional frequency spectrum sensing method mainly has energy measuring, matched filtering and cycle specificity detection
With random matrix detection method.But it includes noise that perceptual signal received by user is perceived in actual environment, this will affect biography
The detection performance of system frequency spectrum sensing method.
Mainly the method for information geometry is applied in frequency spectrum perception in the prior art, combining information geometry and is set in advance
Fixed decision threshold obtains treating the sensing results of perceptual signal, but the selection of its decision threshold is calculated by formula
, calculating process is relatively complicated.And using the method for decision threshold, often always there is deviation, and to detection performance
It impacts, the accuracy of frequency spectrum perception is low.
In conclusion the calculating process for how efficiently solving decision threshold is complicated, and there are deviation, the standards of frequency spectrum perception
The problems such as exactness is low is current those skilled in the art's urgent problem.
Summary of the invention
The object of the present invention is to provide a kind of frequency spectrum sensing method based on information geometry and spectral clustering, the method increase
The detection efficiency of frequency spectrum perception significantly improves the accuracy of frequency spectrum perception；It is a further object of the present invention to provide a kind of bases
In the frequency spectrum sensing method of information geometry and spectral clustering, device, equipment and computer readable storage medium.
In order to solve the above technical problems, the invention provides the following technical scheme:
A kind of frequency spectrum sensing method based on information geometry and spectral clustering, comprising:
It obtains and treats the progress preset times sampling of perception primary user's signal respectively by each secondary user and be calculated each
Covariance matrix；
Each covariance matrix is mapped to matrix manifold using information geometry, obtains each covariance matrix in institute
State corresponding coordinate points in matrix manifold；
Obtain Riemann's mean value by obtaining in advance to noise covariance matrix training；Wherein, the noise covariance square
Battle array noise signal of environment where primary user's signal to be perceived is constituted；
Calculate separately the geodesic curve distance between each coordinate points and Riemann's mean value；
Classified using spectral clustering to each geodesic curve distance, is obtained to primary user's signal to be perceived
Frequency spectrum perception result.
In a kind of specific embodiment of the invention, each geodesic curve distance is divided using spectral clustering
Class, comprising:
The similarity matrix of each geodesic curve distance is calculated, and corresponding degree square is calculated according to the similarity matrix
Battle array；
Corresponding Laplacian Matrix is calculated in conjunction with the similarity matrix and the degree matrix；
The characteristic value of the Laplacian Matrix is calculated, and each characteristic value is subjected to size sequence, obtains sequence knot
Fruit；
The first two characteristic value is chosen in the small one end of characteristic value from the ranking results, and calculates separately two features
It is worth corresponding feature vector, obtains the objective matrix being made of two described eigenvectors；Wherein, two described eigenvectors are
Column vector；
Each row vector of the objective matrix is chosen as target to be sorted, it will be described to be sorted using default clustering algorithm
Target clusters cluster.
In a kind of specific embodiment of the invention, each covariance matrix is mapped to matrix using information geometry
Manifold, comprising:
The corresponding probability-distribution function of each covariance matrix is calculated separately, probability-distribution function race is obtained；
The probability-distribution function race is mapped as matrix manifold using information geometry.
In a kind of specific embodiment of the invention, the multitude by obtaining in advance to noise covariance matrix training is obtained
Graceful mean value, comprising:
Obtain Riemann's mean value by obtaining in advance to the training of multiple noise signals using gradient descent algorithm.
A kind of frequency spectrum sensing device based on information geometry and spectral clustering, comprising:
Matrix obtains module, treats perception primary user's signal progress preset times respectively by each secondary user for acquisition and adopts
Sample and each covariance matrix being calculated；
Coordinate points obtain module and obtain for each covariance matrix to be mapped to matrix manifold using information geometry
Each covariance matrix corresponding coordinate points in the matrix manifold；
Riemann's mean value obtains module, for obtaining the Riemann's mean value for passing through and obtaining in advance to noise covariance matrix training；
Wherein, noise covariance matrix noise signal of environment where primary user's signal to be perceived is constituted；
Geodesic curve distance calculation module, for calculating separately the geodesic curve between each coordinate points and Riemann's mean value
Distance；
As a result module is obtained, for being classified using spectral clustering to each geodesic curve distance, is obtained to described
The frequency spectrum perception result of primary user's signal to be perceived.
In a kind of specific embodiment of the invention, it includes distance classification submodule that the result, which obtains module, described
Distance classification submodule includes:
Matrix calculation unit, for calculating the similarity matrix of each geodesic curve distance, and according to the similarity moment
Battle array calculates corresponding degree matrix；Corresponding Laplacian Matrix is calculated in conjunction with the similarity matrix and the degree matrix；
Ranking results obtaining unit, for calculating the characteristic value of the Laplacian Matrix, and by each characteristic value into
The sequence of row size, obtains ranking results；
The first two characteristic value is chosen in objective matrix obtaining unit, one end small for the characteristic value from the ranking results,
And the corresponding feature vector of two characteristic values is calculated separately, obtain the objective matrix being made of two described eigenvectors；
Wherein, two described eigenvectors are column vector；
It is clustered into cluster unit, each row vector for choosing the objective matrix is gathered as target to be sorted using default
The target to be sorted is clustered cluster by class algorithm.
In a kind of specific embodiment of the invention, it includes matrix manifold mapping submodule that the coordinate points, which obtain module,
Block, the matrix manifold mapping submodule include:
Family of functions's obtaining unit obtains general for calculating separately the corresponding probability-distribution function of each covariance matrix
Rate the class of distribution functions；
Matrix manifold mapping unit, for the probability-distribution function race to be mapped as matrix manifold using information geometry.
In a kind of specific embodiment of the invention, it is specially to obtain by using under gradient that Riemann's mean value, which obtains module,
The module for Riemann's mean value that drop algorithm in advance obtains the training of multiple noise covariance matrixes.
A kind of frequency spectrum perception equipment based on information geometry and spectral clustering, comprising:
Memory, for storing computer program；
Processor realizes the frequency spectrum based on information geometry and spectral clustering as previously described when for executing the computer program
The step of cognitive method.
A kind of computer readable storage medium is stored with computer program on the computer readable storage medium, described
The step of frequency spectrum sensing method as previously described based on information geometry and spectral clustering is realized when computer program is executed by processor.
Using method provided by the embodiment of the present invention, obtain by each secondary user treat respectively perception primary user's signal into
Each covariance matrix that row preset times are sampled and are calculated；Each covariance matrix is mapped to matrix stream using information geometry
Shape obtains each covariance matrix corresponding coordinate points in matrix manifold；It obtains by advance to noise covariance matrix
Riemann's mean value that training obtains；Wherein, the noise signal structure of noise covariance matrix environment where primary user's signal to be perceived
At；Calculate separately the geodesic curve distance between each coordinate points and Riemann's mean value；Using spectral clustering to each geodesic curve distance into
Row classification obtains the frequency spectrum perception result for treating perception primary user's signal.By using spectral clustering to passing through information geometry
The geodesic curve distance being calculated is classified, and avoids the complicated calculations to decision threshold, and can calculate by spectral clustering
Method carries out first dimensionality reduction to each geodesic curve distance and clusters again, reduces the calculation amount of cluster, improves the detection efficiency of frequency spectrum perception,
Significantly improve the accuracy of frequency spectrum perception.
Correspondingly, the embodiment of the invention also provides with the above-mentioned frequency spectrum sensing method phase based on information geometry and spectral clustering
The corresponding frequency spectrum sensing device based on information geometry and spectral clustering, equipment and computer readable storage medium have above-mentioned skill
Art effect, details are not described herein.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is a kind of implementing procedure of the frequency spectrum sensing method based on information geometry and spectral clustering in the embodiment of the present invention
Figure；
Fig. 2 is another implementing procedure of the frequency spectrum sensing method based on information geometry and spectral clustering in the embodiment of the present invention
Figure；
Fig. 3 be utilized respectively frequency spectrum sensing method provided by the embodiment of the present invention based on information geometry and spectral clustering with
Existing frequency spectrum sensing method carries out the frequency spectrum detection performance map of frequency spectrum perception；
Fig. 4 is a kind of structural block diagram of the frequency spectrum sensing device based on information geometry and spectral clustering in the embodiment of the present invention；
Fig. 5 is a kind of structural block diagram of the frequency spectrum perception equipment based on information geometry and spectral clustering in the embodiment of the present invention.
Specific embodiment
In order to enable those skilled in the art to better understand the solution of the present invention, with reference to the accompanying drawings and detailed description
The present invention is described in further detail.Obviously, described embodiments are only a part of the embodiments of the present invention, rather than
Whole embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art are not making creative work premise
Under every other embodiment obtained, shall fall within the protection scope of the present invention.
Embodiment one:
Referring to Fig. 1, Fig. 1 is a kind of reality of the frequency spectrum sensing method based on information geometry and spectral clustering in the embodiment of the present invention
Flow chart is applied, this method may comprise steps of:
S101: it obtains and treats the progress preset times sampling of perception primary user's signal respectively by each secondary user and be calculated
Each covariance matrix.
Multiple secondary users can be preset to detect the signal of primary user, perception is treated by each secondary user respectively
Primary user's signal carries out preset times and samples to obtain the corresponding sampled data of each secondary user, sampled data corresponding to each secondary user
It carries out that the corresponding covariance matrix of each secondary user is calculated, obtains each covariance matrix.For example, in cognition radio
The case where single user is to primary user detection in network, can be expressed as with the dualism hypothesis model in statistics:
Wherein, x (k) is the signal that time user arrives in k reception, and s (k) indicates the authorization user signal received, n
(k) noise received, H are indicated0The case where indicate perception is noise, H1The case where indicate perception is signal.Assuming that noise
N (k) is that independent same distribution, mean value 0, variance areWhite Gaussian noise, s (k) be primary user send signal.Then two
Under kind is assumed, data x obeys distribution N (0, R respectivelyn) and N (0, (Rs+Rn)), RnIndicate the covariance square of noise vector n (k)
Battle array, RsIndicate the covariance matrix of random signal s (k).
Assuming that there are M time users, and the signal of M user's acquisitions constitutes a vector matrix X in cognition network
=[x1,x2...,xM], wherein xi=[xi(1),xi(2)...,xi(N)]TIndicate that the signal sampling value of i-th user, N are
Sampling number.Therefore X is the matrix of N × M dimension:
Each column vector of matrix X is the corresponding covariance matrix R of each secondary user.
S102: each covariance matrix is mapped to matrix manifold using information geometry, obtains each covariance matrix in matrix
Corresponding coordinate points in manifold.
After the corresponding covariance matrix of each secondary user is calculated, information geometry can use by each covariance
Matrix is mapped to matrix manifold, obtains each covariance matrix corresponding coordinate points in matrix manifold, so that tradition is believed
Number test problems convert for the geometrical issues on statistical manifold.For any one user's covariance square being calculated
Battle array, general its obeys zero-mean gaussian distribution, so that the family of probability distribution parameterized by each covariance matrix is obtained, according to
Information geometry is theoretical, and under certain topological structure, family of probability distribution may be constructed a manifold that can be micro-, and be referred to as to count
Manifold, covariance matrix R are the coordinate of the manifold.Due to manifold S parameter R be covariance matrix, then again can S be referred to as matrix
Manifold.Two kinds of hypothesis distribution ps (x | H0) and p (x | H1) two points in manifold are respectively corresponded, and the two put corresponding coordinates
For RnAnd Rs+Rn。
S103: Riemann's mean value by obtaining in advance to noise covariance matrix training is obtained.
Wherein, noise covariance matrix noise signal of environment where primary user's signal to be perceived is constituted.
The noise signal of environment where primary user's signal to be perceived can be obtained in advance obtains being made of each noise signal
Multiple noise covariance matrixes are trained each noise covariance matrix, obtain Riemann's mean value, obtain Riemann's mean value.It is right
The process that each noise covariance matrix training obtains Riemann's mean value can indicate are as follows:
N number of noise covariance matrix is mapped to matrix manifold, obtains N number of signaling point, R in matrix manifoldk(k=1,2,
3,, N), for objective function:
So that point objective function J (R) corresponding when being minimized, as Riemann's mean value:
For example, for any two point R1And R2The case where,Equal to connection R1And R2Geodesic midpoint, Riemann
Mean value are as follows:
S104: the geodesic curve distance between each coordinate points and Riemann's mean value is calculated separately.
The corresponding coordinate points of each covariance matrix are being obtained, and training obtains the ring where primary user's signal to be perceived
After the Riemann's mean value for each noise covariance matrix that the noise signal in border is constituted, it is equal with Riemann that each coordinate points can be calculated separately
Geodesic curve distance (Geodesic Distance, GD) between value is used to gauge signal by using the size of geodesic curve distance
Between difference, geodesic curve illustrates that it is higher with noise signal similarity, primary user's signal to be perceived is likely to apart from smaller
Noise；Geodesic curve distance is bigger, illustrates that it is lower with the similarity of noise signal, primary user's signal to be perceived is likely to user
Signal.
S105: classifying to each geodesic curve distance using spectral clustering, obtains the frequency for treating perception primary user's signal
Compose sensing results.
After the geodesic curve distance being calculated between each coordinate points and Riemann's mean value, each geodesic curve distance can be made
For the input sample of spectral clustering, each geodesic curve distance is carried out using spectral clustering (spectral clustering)
Classification obtains the frequency spectrum perception result for treating perception primary user's signal.Spectral clustering can carry out first each geodesic curve distance
Dimensionality reduction clusters again, and compared with traditional K-Means algorithm, the calculation amount of cluster is much smaller, realizes simply, compared to existing needs
The mode of decision threshold is chosen, the embodiment of the present invention is larger by classifying using spectral clustering to each geodesic curve distance
Ground improves the detection efficiency of frequency spectrum perception, significantly improves the accuracy of frequency spectrum perception.
Using method provided by the embodiment of the present invention, obtain by each secondary user treat respectively perception primary user's signal into
Each covariance matrix that row preset times are sampled and are calculated；Each covariance matrix is mapped to matrix stream using information geometry
Shape obtains each covariance matrix corresponding coordinate points in matrix manifold；It obtains by advance to noise covariance matrix
Riemann's mean value that training obtains；Wherein, the noise signal structure of noise covariance matrix environment where primary user's signal to be perceived
At；Calculate separately the geodesic curve distance between each coordinate points and Riemann's mean value；Using spectral clustering to each geodesic curve distance into
Row classification obtains the frequency spectrum perception result for treating perception primary user's signal.By using spectral clustering to passing through information geometry
The geodesic curve distance being calculated is classified, and avoids the complicated calculations to decision threshold, and can calculate by spectral clustering
Method carries out first dimensionality reduction to each geodesic curve distance and clusters again, reduces the calculation amount of cluster, improves the detection efficiency of frequency spectrum perception,
Significantly improve the accuracy of frequency spectrum perception.
It should be noted that based on the above embodiment one, the embodiment of the invention also provides be correspondingly improved scheme.Rear
Involved in continuous embodiment with can mutually be referred between same steps or corresponding steps in above-described embodiment one, corresponding beneficial effect
Can also be cross-referenced, it is no longer repeated one by one in improvement embodiment below.
Embodiment two:
Referring to fig. 2, Fig. 2 is the another kind of the frequency spectrum sensing method based on information geometry and spectral clustering in the embodiment of the present invention
Implementation flow chart, this method may comprise steps of:
S201: it obtains and treats the progress preset times sampling of perception primary user's signal respectively by each secondary user and be calculated
Each covariance matrix.
S202: the corresponding probability-distribution function of each covariance matrix is calculated separately, probability-distribution function race is obtained.
After the corresponding covariance matrix of each secondary user is calculated, for any one use being calculated
Family covariance matrix, general its obey zero-mean gaussian distribution, can calculate separately the corresponding probability distribution of each covariance matrix
Function, then its distribution expression formula can indicate are as follows:
Obtain covariance matrix R ∈ Cn×nParametrization family of probability distribution S=p (x | R) | R ∈ Cn×n, wherein Cn×nFor n
The opener in × n-tuple space.
S203: probability-distribution function race is mapped as matrix manifold using information geometry, obtains each covariance matrix in square
Corresponding coordinate points in battle array manifold.
The corresponding probability-distribution function of each covariance matrix is being calculated separately, it, can be with after obtaining probability-distribution function race
Probability-distribution function race is mapped as matrix manifold using information geometry, it is right respectively in matrix manifold to obtain each covariance matrix
The coordinate points answered.Above-described embodiment is accepted, it, can be then according to information geometry theory, certain after obtaining family of probability distribution S
Topological structure under S may be constructed a manifold that can be micro-, and referred to as matrix manifold, obtain each covariance matrix R in matrix
Corresponding coordinate points in manifold.
S204: it obtains equal by the Riemann obtained in advance to the training of multiple noise covariance matrixes using gradient descent algorithm
Value.
Wherein, noise covariance matrix noise signal of environment where primary user's signal to be perceived is constituted.
For the accuracy for guaranteeing the Riemann's mean value acquired in advance, the noise signal obtained in advance can repeatedly be adopted
Sample obtains a large amount of sampled data, and sampled data is divided into multiple data groups, calculates corresponding noise to each data group
Covariance matrix, so that multiple noise covariance matrix R are obtained, multiple coordinate points in homography prevalence.It is a for N (N > 2)
The case where coordinate points, can use Riemann's mean value that gradient descent algorithm in advance obtains the training of multiple noise covariance matrixes,
The calculation expression of Riemann's mean value are as follows:
Wherein, τ is iteration step length, and i is iterative steps, and k is the ordinal number of coordinate points.
S205: the geodesic curve distance between each coordinate points and Riemann's mean value is calculated separately.
S206: the similarity matrix of each geodesic curve distance is calculated, and corresponding degree matrix is calculated according to similarity matrix.
After calculating separately to obtain each geodesic curve distance, the similarity matrix of each geodesic curve distance, and root can be calculated
Corresponding degree matrix is calculated according to similarity matrix.Since frequency spectrum perception result is divided into two classes, one kind is that sensing results are signal, separately
One kind is that sensing results are noise, so the data of the clustering cluster of each geodesic curve distance of output are 2.For by n geodesic curve
The sample set that distance is constitutedThe similar matrix W of n*n is calculated using following formula:
Wherein, sijFor the element of the i-th column jth row in similar matrix W, σ indicates width neighborhood.
And using face formula calculating degree matrix D:
D is diThe n*n diagonal matrix of composition.
S207: corresponding Laplacian Matrix is calculated in conjunction with similarity matrix and degree matrix.
After the similarity matrix and degree matrix that each geodesic curve distance is calculated, similarity matrix and degree can be combined
Matrix calculates corresponding Laplacian Matrix.Example in undertaking, is calculated by the following formula Laplacian Matrix:
L=D-W；
S208: the characteristic value of Laplacian Matrix is calculated, and each characteristic value is subjected to size sequence, obtains ranking results.
After calculating Laplacian Matrix, the characteristic value of Laplacian Matrix can be calculated, and by each characteristic value into
The sequence of row size, obtains ranking results.Example in undertaking calculates the characteristic value of L, characteristic value can be sorted from small to large, can also
To sort from large to small characteristic value, ranking results are obtained.
S209: the first two characteristic value is chosen in the small one end of characteristic value from ranking results, and calculates separately two characteristic values
Corresponding feature vector obtains the objective matrix being made of two feature vectors.
Wherein, two feature vectors are column vector.
Since the data of the clustering cluster of each geodesic curve distance are 2, one end that characteristic value is small from ranking results is chosen
The first two characteristic value, and calculate separately the corresponding feature vector of two characteristic values: u1, u2, and two feature vectors are column vector,
Obtain the objective matrix U={ u being made of two feature vectors1, u2, U ∈ Rn*2。
S210: choosing each row vector of objective matrix as target to be sorted, using default clustering algorithm by mesh to be sorted
Mark cluster cluster, obtains the frequency spectrum perception result for treating perception primary user's signal.
Two vectors of every a line of objective matrix are chosen as each new sample point, enable yr∈R2Be the r row of U to
It measures, wherein r=1,2 ..., n.New sample set is represented by Y={ y1, y2,…yn, it will be wait divide using default clustering algorithm
Class target is clustered into cluster, and the number of clustering cluster is specially 2, i.e. A1, A2, whether deposited in perception primary user's signal to obtain treating
In the frequency spectrum perception result of primary user's signal.First dimensionality reduction is carried out to each geodesic curve distance by spectral clustering to cluster again, it is larger
Ground reduces the dimension of clustering object, significantly reduces the calculation amount of cluster, improves the detection effect of frequency spectrum perception significantly
Rate.
It is to be utilized respectively the frequency spectrum sense based on information geometry and spectral clustering provided by the embodiment of the present invention referring to Fig. 3, Fig. 3
Perception method and existing frequency spectrum sensing method carry out the frequency spectrum detection performance map of frequency spectrum perception, and two perception users are in noise
In the case where being -12 than SNR, recipient's operating characteristic curve ROC of method and conventional method that the embodiment of the present invention proposes is bent
Line, PdIt is detection probability, PfaIt is false-alarm probability.IG-PC indicates that the information geometry that the embodiment of the present invention proposes and spectral clustering combine
Method, MSE indicate be based on maximum eigenvalue-energy frequency spectrum perception algorithm.By comparison as can be seen that false-alarm probability
When identical, the algorithm detection performance that the embodiment of the present invention proposes is higher.
Corresponding to above method embodiment, the embodiment of the invention also provides a kind of based on information geometry and spectral clustering
Frequency spectrum sensing device, the frequency spectrum sensing device described below based on information geometry and spectral clustering are based on information with above-described
Geometry can correspond to each other reference with the frequency spectrum sensing method of spectral clustering.
Referring to fig. 4, Fig. 4 is a kind of knot of the frequency spectrum sensing device based on information geometry and spectral clustering in the embodiment of the present invention
Structure block diagram, the apparatus may include:
Matrix obtains module 41, treats perception primary user's signal progress preset times respectively by each secondary user for obtaining
Each covariance matrix for sampling and being calculated；
Coordinate points obtain module 42, for each covariance matrix to be mapped to matrix manifold using information geometry, obtain each
Covariance matrix corresponding coordinate points in matrix manifold；
Riemann's mean value obtains module 43, trains obtained Riemann equal noise covariance matrix in advance for obtaining to pass through
Value；Wherein, noise covariance matrix noise signal of environment where primary user's signal to be perceived is constituted；
Geodesic curve distance calculation module 44, for calculating separately the geodesic curve distance between each coordinate points and Riemann's mean value；
As a result module 45 is obtained to obtain treating perception for classifying to each geodesic curve distance using spectral clustering
The frequency spectrum perception result of primary user's signal.
Using device provided by the embodiment of the present invention, obtain by each secondary user treat respectively perception primary user's signal into
Each covariance matrix that row preset times are sampled and are calculated；Each covariance matrix is mapped to matrix stream using information geometry
Shape obtains each covariance matrix corresponding coordinate points in matrix manifold；It obtains by advance to noise covariance matrix
Riemann's mean value that training obtains；Wherein, the noise signal structure of noise covariance matrix environment where primary user's signal to be perceived
At；Calculate separately the geodesic curve distance between each coordinate points and Riemann's mean value；Using spectral clustering to each geodesic curve distance into
Row classification obtains the frequency spectrum perception result for treating perception primary user's signal.By using spectral clustering to passing through information geometry
The geodesic curve distance being calculated is classified, and avoids the complicated calculations to decision threshold, and can calculate by spectral clustering
Method carries out first dimensionality reduction to each geodesic curve distance and clusters again, reduces the calculation amount of cluster, improves the detection efficiency of frequency spectrum perception,
Significantly improve the accuracy of frequency spectrum perception.
In a kind of specific embodiment of the invention, as a result obtaining module 45 includes distance classification submodule, distance point
Class submodule includes:
Matrix calculation unit, for calculating the similarity matrix of each geodesic curve distance, and according to similarity matrix calculating pair
The degree matrix answered；Corresponding Laplacian Matrix is calculated in conjunction with similarity matrix and degree matrix；
Ranking results obtaining unit carries out big float for calculating the characteristic value of Laplacian Matrix, and by each characteristic value
Sequence obtains ranking results；
Objective matrix obtaining unit, the first two characteristic value is chosen in one end small for the characteristic value from ranking results, and divides
Not Ji Suan the corresponding feature vector of two characteristic values, obtain the objective matrix being made of two feature vectors；Wherein, two features
Vector is column vector；
Object selection unit to be sorted, for choose objective matrix any row two vectors as target to be sorted；
Target classification unit is classified for treating class object using default clustering algorithm.
In a kind of specific embodiment of the invention, it includes matrix manifold mapping submodule that coordinate points, which obtain module 42,
Matrix manifold mapping submodule includes:
Family of functions's obtaining unit obtains probability point for calculating separately the corresponding probability-distribution function of each covariance matrix
Cloth family of functions；
Matrix manifold mapping unit, for probability-distribution function race to be mapped as matrix manifold using information geometry.
In a kind of specific embodiment of the invention, it is specially to obtain by utilizing gradient that Riemann's mean value, which obtains module 43,
The module for Riemann's mean value that descent algorithm in advance obtains the training of multiple noise covariance matrixes.
Corresponding to above method embodiment, referring to Fig. 5, Fig. 5 is provided by the present invention poly- based on information geometry and spectrum
The schematic diagram of the frequency spectrum perception equipment of class, the equipment may include:
Memory 51, for storing computer program；
Processor 52 can realize following steps when for executing the computer program of the above-mentioned storage of memory 51:
It obtains and treats the progress preset times sampling of perception primary user's signal respectively by each secondary user and be calculated each
Covariance matrix；Each covariance matrix is mapped to matrix manifold using information geometry, obtains each covariance matrix in matrix stream
Corresponding coordinate points in shape；Obtain Riemann's mean value by obtaining in advance to noise covariance matrix training；Wherein, noise
Covariance matrix noise signal of environment where primary user's signal to be perceived is constituted；Calculate separately each coordinate points and Riemann's mean value
Between geodesic curve distance；Classified using spectral clustering to each geodesic curve distance, obtains treating perception primary user's signal
Frequency spectrum perception result.
Above method embodiment is please referred to for the introduction of equipment provided by the invention, this will not be repeated here by the present invention.
It is computer-readable the present invention also provides a kind of computer readable storage medium corresponding to above method embodiment
It is stored with computer program on storage medium, can realize following steps when computer program is executed by processor:
It obtains and treats the progress preset times sampling of perception primary user's signal respectively by each secondary user and be calculated each
Covariance matrix；Each covariance matrix is mapped to matrix manifold using information geometry, obtains each covariance matrix in matrix stream
Corresponding coordinate points in shape；Obtain Riemann's mean value by obtaining in advance to noise covariance matrix training；Wherein, noise
Covariance matrix noise signal of environment where primary user's signal to be perceived is constituted；Calculate separately each coordinate points and Riemann's mean value
Between geodesic curve distance；Classified using spectral clustering to each geodesic curve distance, obtains treating perception primary user's signal
Frequency spectrum perception result.
The computer readable storage medium may include: USB flash disk, mobile hard disk, read-only memory (Read-Only
Memory, ROM), random access memory (Random Access Memory, RAM), magnetic or disk etc. is various to deposit
Store up the medium of program code.
Above method embodiment is please referred to for the introduction of computer readable storage medium provided by the invention, the present invention exists
This is not repeated them here.
Each embodiment in this specification is described in a progressive manner, the highlights of each of the examples are with it is other
The difference of embodiment, same or similar part may refer to each other between each embodiment.For being filled disclosed in embodiment
It sets, for equipment and computer readable storage medium, since it is corresponded to the methods disclosed in the examples, so the comparison of description
Simply, reference may be made to the description of the method.
Used herein a specific example illustrates the principle and implementation of the invention, and above embodiments are said
It is bright to be merely used to help understand technical solution of the present invention and its core concept.It should be pointed out that for the common of the art
, without departing from the principle of the present invention, can be with several improvements and modifications are made to the present invention for technical staff, these
Improvement and modification are also fallen within the protection scope of the claims of the present invention.
Claims (10)
1. a kind of frequency spectrum sensing method based on information geometry and spectral clustering characterized by comprising
It obtains and treats each association side that perception primary user's signal carries out preset times sampling and is calculated respectively by each secondary user
Poor matrix；
Each covariance matrix is mapped to matrix manifold using information geometry, obtains each covariance matrix in the square
Corresponding coordinate points in battle array manifold；
Obtain Riemann's mean value by obtaining in advance to noise covariance matrix training；Wherein, the noise covariance matrix by
The noise signal of environment is constituted where primary user's signal to be perceived；
Calculate separately the geodesic curve distance between each coordinate points and Riemann's mean value；
Classified using spectral clustering to each geodesic curve distance, obtains the frequency spectrum to primary user's signal to be perceived
Sensing results.
2. the frequency spectrum sensing method according to claim 1 based on information geometry and spectral clustering, which is characterized in that utilize spectrum
Clustering algorithm classifies to each geodesic curve distance, comprising:
The similarity matrix of each geodesic curve distance is calculated, and corresponding degree matrix is calculated according to the similarity matrix；
Corresponding Laplacian Matrix is calculated in conjunction with the similarity matrix and the degree matrix；
The characteristic value of the Laplacian Matrix is calculated, and each characteristic value is subjected to size sequence, obtains ranking results；
The first two characteristic value is chosen in the small one end of characteristic value from the ranking results, and calculates separately two characteristic values pair
The feature vector answered obtains the objective matrix being made of two described eigenvectors；Wherein, two described eigenvectors be column to
Amount；
Each row vector of the objective matrix is chosen as target to be sorted, using default clustering algorithm by the target to be sorted
Cluster cluster.
3. the frequency spectrum sensing method according to claim 1 or 2 based on information geometry and spectral clustering, which is characterized in that benefit
Each covariance matrix is mapped to matrix manifold with information geometry, comprising:
The corresponding probability-distribution function of each covariance matrix is calculated separately, probability-distribution function race is obtained；
The probability-distribution function race is mapped as matrix manifold using information geometry.
4. the frequency spectrum sensing method according to claim 3 based on information geometry and spectral clustering, which is characterized in that obtain logical
After the Riemann's mean value obtained in advance to noise covariance matrix training, comprising:
Obtain Riemann's mean value by obtaining in advance to the training of multiple noise covariance matrixes using gradient descent algorithm.
5. a kind of frequency spectrum sensing device based on information geometry and spectral clustering characterized by comprising
Matrix obtains module, treats the sampling of perception primary user's signal progress preset times respectively by each secondary user simultaneously for obtaining
Each covariance matrix being calculated；
Coordinate points obtain module and obtain each institute for each covariance matrix to be mapped to matrix manifold using information geometry
State covariance matrix corresponding coordinate points in the matrix manifold；
Riemann's mean value obtains module, for obtaining the Riemann's mean value for passing through and obtaining in advance to noise covariance matrix training；Wherein,
Noise covariance matrix noise signal of environment where primary user's signal to be perceived is constituted；
Geodesic curve distance calculation module, for calculate separately the geodesic curve between each coordinate points and Riemann's mean value away from
From；
As a result module is obtained, for being classified using spectral clustering to each geodesic curve distance, is obtained to described wait feel
Know the frequency spectrum perception result of primary user's signal.
6. the frequency spectrum sensing device according to claim 5 based on information geometry and spectral clustering, which is characterized in that the knot
It includes distance classification submodule that fruit, which obtains module, and the distance classification submodule includes:
Matrix calculation unit, for calculating the similarity matrix of each geodesic curve distance, and according to the similarity matrix meter
Calculate corresponding degree matrix；Corresponding Laplacian Matrix is calculated in conjunction with the similarity matrix and the degree matrix；
Ranking results obtaining unit carries out greatly for calculating the characteristic value of the Laplacian Matrix, and by each characteristic value
Small sequence obtains ranking results；
Objective matrix obtaining unit, the first two characteristic value is chosen in one end small for the characteristic value from the ranking results, and divides
Not Ji Suan the corresponding feature vector of two characteristic values, obtain the objective matrix being made of two described eigenvectors；Wherein,
Two described eigenvectors are column vector；
It is clustered into cluster unit, each row vector for choosing the objective matrix is calculated as target to be sorted using default cluster
The target to be sorted is clustered cluster by method.
7. the frequency spectrum sensing device according to claim 5 or 6 based on information geometry and spectral clustering, which is characterized in that institute
Stating coordinate points and obtaining module includes matrix manifold mapping submodule, and the matrix manifold mapping submodule includes:
Family of functions's obtaining unit obtains probability point for calculating separately the corresponding probability-distribution function of each covariance matrix
Cloth family of functions；
Matrix manifold mapping unit, for the probability-distribution function race to be mapped as matrix manifold using information geometry.
8. the frequency spectrum sensing device according to claim 7 based on information geometry and spectral clustering, which is characterized in that Li Manjun
It is specially the multitude obtained by being obtained in advance to the training of multiple noise covariance matrixes using gradient descent algorithm that value, which obtains module,
The module of graceful mean value.
9. a kind of frequency spectrum perception equipment based on information geometry and spectral clustering characterized by comprising
Memory, for storing computer program；
Processor, realized when for executing the computer program as described in any one of Claims 1-4 based on information geometry and
The step of frequency spectrum sensing method of spectral clustering.
10. a kind of computer readable storage medium, which is characterized in that be stored with computer on the computer readable storage medium
Program is realized as described in any one of Claims 1-4 when the computer program is executed by processor based on information geometry and spectrum
The step of frequency spectrum sensing method of cluster.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910564127.9A CN110288025A (en) | 2019-06-25 | 2019-06-25 | Frequency spectrum sensing method, device and equipment based on information geometry and spectral clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910564127.9A CN110288025A (en) | 2019-06-25 | 2019-06-25 | Frequency spectrum sensing method, device and equipment based on information geometry and spectral clustering |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110288025A true CN110288025A (en) | 2019-09-27 |
Family
ID=68007627
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910564127.9A Pending CN110288025A (en) | 2019-06-25 | 2019-06-25 | Frequency spectrum sensing method, device and equipment based on information geometry and spectral clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110288025A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110874615A (en) * | 2019-11-14 | 2020-03-10 | 深圳前海微众银行股份有限公司 | Feature clustering processing method, cluster server and readable storage medium |
| CN111506766A (en) * | 2020-04-20 | 2020-08-07 | 腾讯音乐娱乐科技（深圳）有限公司 | Audio frame clustering method, device and equipment |
| 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 |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080010245A1 (en) * | 2006-07-10 | 2008-01-10 | Jaehwan Kim | Method for clustering data based convex optimization |
| CN107071788A (en) * | 2017-04-21 | 2017-08-18 | 广东工业大学 | Frequency spectrum sensing method and device in a kind of cognition wireless network |
| CN109004997A (en) * | 2018-08-09 | 2018-12-14 | 广东工业大学 | A kind of frequency spectrum sensing method, device and computer readable storage medium |
-
2019
- 2019-06-25 CN CN201910564127.9A patent/CN110288025A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080010245A1 (en) * | 2006-07-10 | 2008-01-10 | Jaehwan Kim | Method for clustering data based convex optimization |
| CN107071788A (en) * | 2017-04-21 | 2017-08-18 | 广东工业大学 | Frequency spectrum sensing method and device in a kind of cognition wireless network |
| CN109004997A (en) * | 2018-08-09 | 2018-12-14 | 广东工业大学 | A kind of frequency spectrum sensing method, device and computer readable storage medium |
Non-Patent Citations (3)
| Title |
|---|
| WANG Y,CHEN Q,LI J: "A Spectrum Sensing Algorithm Based on Information Geometry and K-medoids Clustering", 《SPRINGER》 * |
| 杨清宇，孙凤伟，张曌，张迪，庄健: "利用测地线距离的改进谱聚类算法", 《西安交通大学学报》 * |
| 相洁，赵冬琴: "改进谱聚类算法在MCI患者检测中的应用研究", 《通信学报》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110874615A (en) * | 2019-11-14 | 2020-03-10 | 深圳前海微众银行股份有限公司 | Feature clustering processing method, cluster server and readable storage medium |
| CN110874615B (en) * | 2019-11-14 | 2023-09-26 | 深圳前海微众银行股份有限公司 | Feature clustering processing method, cluster server and readable storage medium |
| CN111506766A (en) * | 2020-04-20 | 2020-08-07 | 腾讯音乐娱乐科技（深圳）有限公司 | Audio frame clustering method, device and equipment |
| CN111506766B (en) * | 2020-04-20 | 2023-03-10 | 腾讯音乐娱乐科技（深圳）有限公司 | Audio frame clustering method, device and equipment |
| 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 |
|---|---|---|
| CN110288025A (en) | Frequency spectrum sensing method, device and equipment based on information geometry and spectral clustering | |
| Varish et al. | Image retrieval scheme using quantized bins of color image components and adaptive tetrolet transform | |
| CN104135327B (en) | frequency spectrum sensing method based on support vector machine | |
| TWI464604B (en) | Data clustering method and device, data processing apparatus and image processing apparatus | |
| CN106295498B (en) | Remote sensing image target area detection device and method | |
| CN104424235A (en) | Method and device for clustering user information | |
| CN110224771A (en) | Frequency spectrum sensing method and device based on BP neural network and information geometry | |
| CN109039503A (en) | A kind of frequency spectrum sensing method, device, equipment and computer readable storage medium | |
| Zhang et al. | WiFi-based cross-domain gesture recognition via modified prototypical networks | |
| CN104881682B (en) | A kind of image classification method based on guarantor office mapping and principal component analysis | |
| CN109004997A (en) | A kind of frequency spectrum sensing method, device and computer readable storage medium | |
| CN110097011A (en) | A kind of signal recognition method and device | |
| CN110991525A (en) | Accompanying pattern matching method based on operator track data | |
| CN107392887A (en) | A kind of heterogeneous method for detecting change of remote sensing image based on the conversion of homogeneity pixel | |
| CN107071788A (en) | Frequency spectrum sensing method and device in a kind of cognition wireless network | |
| CN106470435B (en) | The method and system of identification WiFi groups | |
| CN110503113A (en) | A kind of saliency object detection method restored based on low-rank matrix | |
| CN110069546A (en) | A kind of data classification method, device for classifying data and terminal device | |
| CN109246728B (en) | Method and device for identifying coverage abnormal cell | |
| CN108880717A (en) | A kind of frequency spectrum sensing method of the α divergence based on information geometry | |
| CN106373279B (en) | The communication means of intelligent IC card water meter for multilist centralized meter-reading system | |
| CN103530377A (en) | Scene information searching method based on binary feature codes | |
| CN102609718A (en) | Method for generating vision dictionary set by combining different clustering algorithms | |
| CN105472380A (en) | Compression domain significance detection algorithm based on ant colony algorithm | |
| CN106023184A (en) | Depth significance detection method based on anisotropy center-surround difference |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190927 |