CN110348402A - A kind of signal detecting method of the expectation likelihood of binding characteristic frequency - Google Patents
A kind of signal detecting method of the expectation likelihood of binding characteristic frequency Download PDFInfo
- Publication number
- CN110348402A CN110348402A CN201910637642.5A CN201910637642A CN110348402A CN 110348402 A CN110348402 A CN 110348402A CN 201910637642 A CN201910637642 A CN 201910637642A CN 110348402 A CN110348402 A CN 110348402A
- Authority
- CN
- China
- Prior art keywords
- characteristic frequency
- signal
- binding characteristic
- follows
- likelihood
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
Landscapes
- Radar Systems Or Details Thereof (AREA)
Abstract
The present invention provides a kind of signal detecting method of the expectation likelihood of binding characteristic frequency, belongs to signal detection technique field.The present invention establishes data covariance matrix by the sampled data of signal first, and obtains the Fourier transform W of the feature vector of the data covariance matrixm(θk)；Then by gained Wm(θk) frequency spectrum peak point where frequency be defined as characteristic frequency, the expectation likelihood statistic of calculations incorporated characteristic frequency calculates feature vector detection threshold；Recycle the expectation likelihood statistic thresholding of numerical calculations binding characteristic frequency；Testing result is finally obtained by obtained new binary hypothesis test formula.The present invention solves the prior art in the case where low signal-to-noise ratio, the low problem of detection probability.The signal detection that the present invention can be used in the case of low signal-to-noise ratio.
Description
Technical field
The present invention relates to a kind of signal detecting methods for it is expected likelihood, belong to signal detection technique field.
Background technique
It is expected that likelihood (EL:Expected Likelihood) technology is a kind of parameter estimation techniques, can be used for Gauss with it is non-
The parameter Estimation based on array signal processing under Gaussian Background.Document [Yuri I.Abramovich, Olivier
Besson.Regularized Covariance Matrix Estimation in Complex Elliptically
Symmetric Distributions Using the Expected Likelihood Approach—Part 1:The
Over-Sampled Case.IEEE Transactions on Signal Processing.2013,61 (23): 5807~
5818.] and document [Yuri I.Abramovich, Olivier Besson.Regularized Covariance Matrix
Estimation in Complex Elliptically Symmetric Distributions Using the Expected
Likelihood Approach—Part 2:The Under-Sampled Case.IEEE Transactions on
Signal Processing.2013,61 (23): 5819~5829.] respectively under the conditions of sample is more and less when make
Regularized covariance Matrix Estimation is carried out to multiple ellipsometry distributed data with desired likelihood method.Document [Yuri
I.Abramovich, Olivier Besson, Ben A.Johnson.Conditional Expected Likelihood
Technique for Compound Gaussian and Gaussian Distributed Noise Mixtures.IEEE
62 (24): Transactions on Signal Processing.2016 6640~6649.] proposes a kind of conditional expectation seemingly
Right technology is for the array signal parameter Estimation under Gaussian mixtures noise conditions.Document [Yuri I.Abramovich,
Olivier Besson.On the Expected Likelihood Approach for Assessment of
Regularization Covariance Matrix.IEEE Signal Processing Letters.2015,22 (06):
777~781.] estimation of regularized covariance matrix is evaluated and is optimized using desired likelihood method.Document [E.Tom
Northardt, Igal Bilik, Yuri I.Abramovich.Spatial Compressive Sensing for
Direction-of-Arrival Estimation With Bias Mitigation Via Expected
Likelihood.IEEE Transactions on Signal Processing.2013,61 (05): 1183~1195.] makes
Deviation correction is carried out with desired likelihood technology, and a kind of space compression cognitive method for DOA estimation is proposed based on this technology.
Document [Bosung Kang, Vishal Monga, Muralidhar Rangaswamy, Yuri Abramovich, Expected
Likelihood Approach for Determining Constraints in Covariance Estimation.IEEE
Transactions on Aerospace and Electronic Systems, 2016,52 (5): 2139~2156.] by the phase
Hope likelihood method for the definition of the constraint condition in the processing of radar space-time adaptive when covariance matrix.When desired likelihood
When technology is applied to target echo detection, it is advantageous that the target direction letter that array signal data can be made full use of to be included
Breath, to be conducive to improve echo signal ability of discovery.Document [E.Tom Northardt, Igal Bilik, Yuri
Abramovich.Bearings-only constant velocity target maneuver detection via
expected likelihood.IEEE Transactions on Aerospace and Electronic
Systems.2014,50 (04): the frame 2974~2988.] based on DOA algorithm for estimating is proposed desired likelihood technical application
In moving object detection.This method does not need to make statistical hypothesis to direction estimation, and moving object detection performance can be made with letter
It makes an uproar than the raising with data sample number and improves.Document [Olivier Besson, Yuri Abramovich.Sensitivity
Analysis of Likelihood Ratio Test in K Distributed and/or Gaussian Noise.IEEE
Signal Processing Letters.2015,22 (12): 2329~2333.] under the conditions of K partition noise or Gaussian noise
The robustness of detector is analyzed.
From document [Yuguan Hou, Application Research of Spatial Spectrum Estimation
Technology in Radar Signal Processing.2008.], it is known where the maximum spectrum point of each feature vector
Frequency include target direction information.[Shuning Zhang, Wei Xie, Hang Zhu, Huichang Zhao, Combined
Eigenvector Analysis and Independent Component Analysis For Multi-Component
Periodic Interferences Suppression In PRCPM-PD Detection System.IEEE
Access.2017,5:12552-12562.] propose it is a kind of dry based on feature vector analysis and independent component analysis (ICA)
Suppressing method is disturbed, the Generalized Periodic feature for receiving signal is utilized in it.
But in the case where low signal-to-noise ratio (SNR), the signal detection performance of EL method may be decreased or deficiency, that is,
When low low signal-to-noise ratio, detection probability is low.
Summary of the invention
The present invention is to solve the prior art in the case where low signal-to-noise ratio, and the low problem of detection probability provides a kind of knot
Close the signal detecting method of the expectation likelihood of characteristic frequency.
A kind of signal detecting method of the expectation likelihood of binding characteristic frequency of the present invention, it is real by the following technical programs
It is existing:
Step 1: establishing data covariance matrix by the sampled data of signal, and obtain the spy of the data covariance matrix
Levy the Fourier transform W of vectorm(θk)；
Step 2: by gained Wm(θk) frequency spectrum peak point where frequency be defined as characteristic frequency；
Step 3: the expectation likelihood statistic l of binding characteristic frequency is calculatedMEL；
Step 4: calculating feature vector detection threshold；
Step 5: the l obtained in conjunction with step 3MEL, counted using the expectation likelihood of numerical calculations binding characteristic frequency
Measure thresholding GMEL；
Step 6: obtaining testing result by following binary hypothesis test formula:
Wherein, H0Indicate no signal, H1Indicate signal,For m-th of feature vector detection threshold.
Present invention feature the most prominent and significant beneficial effect are:
The signal detecting method of the expectation likelihood of a kind of binding characteristic frequency according to the present invention, by data covariance square
The peak point of the Fourier transformation of the feature vector of battle array is defined as characteristic frequency, and it includes target direction information.Then, based on existing
Some EL statistics derive a kind of new EL statistic --- expectation likelihood statistic of binding characteristic frequency, and obtain one
The new binary hypothesis test of kind.Compared with traditional EL detection technique, the present invention has better detection performance, emulation experiment
In, under the same terms, the lowest signal-to-noise of (detection performance is excellent) is about than traditional side EL when the method for the present invention detection probability is close to 1
The low 5dB of method.
Detailed description of the invention
Fig. 1 is flow chart of the present invention；
Fig. 2 is the frequency distribution histogram of 1,000,000 Monte Carlo Experiment testing results in embodiment；
Fig. 3 is in embodiment in sampling number L=64, false-alarm probability PFA=10-4, the method for the present invention detection probability with
Signal-to-noise ratio change curve；
Fig. 4 is in embodiment in sampling number L=64, false-alarm probability PFA=10-5, conventional EL method detection probability with
Signal-to-noise ratio change curve；
It in sampling number is L=128, false-alarm probability PFA=10 that Fig. 5, which is in embodiment,-4, the method for the present invention and tradition EL
Method detection probability changes contrast curve chart with signal-to-noise ratio；
It in sampling number is L=128, false-alarm probability PFA=10 that Fig. 6, which is in embodiment,-5, the method for the present invention and tradition EL
Method detection probability changes contrast curve chart with signal-to-noise ratio；
It in sampling number is L=128, false-alarm probability PFA=10 that Fig. 7, which is in embodiment,-3, 10-4, 10-5, 10-6When, this
Inventive method detection probability is with false-alarm probability result of variations curve graph.
Specific embodiment
Specific embodiment 1: being illustrated in conjunction with Fig. 1 to present embodiment, a kind of combination that present embodiment provides is special
The signal detecting method of the expectation likelihood of frequency is levied, specifically includes the following steps:
Step 1: establishing data covariance matrix by the sampled data of signal, and obtain the spy of the data covariance matrix
Levy the Fourier transform W of vectorm(θk)；
Step 2: by gained Wm(θk) frequency spectrum peak point where frequency be defined as characteristic frequency；
Step 3: the expectation likelihood statistic l of binding characteristic frequency is calculatedMEL；
Step 4: calculating feature vector detection threshold；
Step 5: the l obtained in conjunction with step 3MEL, counted using the expectation likelihood of numerical calculations binding characteristic frequency
Measure thresholding GMEL；
Step 6: obtaining testing result by following binary hypothesis test formula:
Wherein, H0Indicate no signal, H1Indicate signal,For m-th of feature vector detection threshold.
Traditional expectation likelihood EL mode, the test problems to be solved can be expressed as binary hypothesis test:
Characteristic vector em(θk) function that signal guide vector sum receives the sum of signal noise can be expressed as.Again because making an uproar
Sound nkGaussian distributed, so em(θk) obey zero-mean, varianceGaussian Profile, so transformed Characteristic Vectors
Measure Wm(θk) meet relationshipIndicate zero-mean,The Gauss of variance point
Cloth.
It enablesIt indicates the thresholding that is detected to feature vector Fourier transformation, judges whether to detect target, it can be with
Pass through the Fourier transformation W of judging characteristic vectorm(θk) it whether is more than thresholding, whenWhen be detected by mesh
Mark.Have because but having W when target is not present in signalm(θk) the case where being more than thresholding referred to as false-alarm；To sum up, test problems are obtained
New binary hypothesis test formula (1).The method of the present invention is carried out based on traditional expectation likelihood EL method, and on its basis
It improves, therefore the method for the present invention can also be known as being based on the signal detecting method of improved expectation likelihood (MEL), phase of the present invention
Than conventional EL method, there is better detection performance at low signal-to-noise ratio (SNR).
Specific embodiment 2: the present embodiment is different from the first embodiment in that, the association of data described in step 1
Variance matrix are as follows:
Wherein,For data covariance matrix, L is sampling number；Subscript H indicates transposition, YkIt is by the L in k-th of period
The vector that a mutually independent array sampled point is expressed as,I=1,2 ..., L；K-th of period
The i-th moment array output vectorAre as follows:
Be i-th moment in k-th of period N × 1 tie up zero-mean, variance beGaussian reflectivity mirrors vector, Jk
It is target number, target angle vector is θk=[θk(0),…,θk(Jk)],It is the target amplitude vector that Jk × 1 is tieed up,A(θk) it is N × JkDirection matrix, N are element number of array.
Other steps and parameter are same as the specific embodiment one.
Specific embodiment 3: present embodiment obtains the spy unlike specific embodiment two, in step 1
Levy the Fourier transform W of vectorm(θk) detailed process includes:
Enable em(θk) it is data covariance matrixFeature vector, m=1 ..., Jk；Then em(θk) can indicate are as follows:
Wherein, nkIt indicates to receive the corresponding zero-mean of signal, variance σ2Gaussian Profile noise；It is signal guide vector representation parameter；a(θk(1))、a(θk(2))...a(θk(Jk)) be respectively as follows:
Then, to feature vector em(θk) carry out Fourier transformation obtain Wm(θk), matrix Wm(θk) in first of element
Are as follows:
Wherein, l=0 ..., N-1；em(θk, n) and indicate em(θk) in nth elements, n=0 ..., N-1；E is nature
Constant, j are imaginary unit.
Other steps and parameter are identical with embodiment two.
Specific embodiment 4: present embodiment is unlike specific embodiment two or three, to characteristic vector em(θk)
It carries out Fourier transformation and obtains Wm(θk), each feature vector has its corresponding characteristic frequency.The length of feature vector is equal to battle array
First number N, digitization frequencies resolving power are 1/N.Directly make Discrete Fourier Transform discrete frequency point be also it is N number of, generally not
Be feature vector frequency spectrum peak value where, be not easy to compare, judgement get up deviation it is also larger.Therefore it needs to increase sequence by zero padding
Column length is come where peaking (to be denoted as ωp)。
Biggish characteristic value will likely be worth for clarification of objective, herein, the differentiation of large and small characteristic value can utilize tradition
Method (such as information theory criterion).If the number of big characteristic value is Jk, for some big characteristic value, characteristic frequency peak value coverage area
It is represented by [ωk′-α△ω,ωk′+ α △ ω], (k '=1,2 ..., Jk) wherein, △ ω=2 π/N is digitization frequencies resolution
Interval, α ∈ [0,2], usual value 1.
The W when meeting following conditionm(θk, l) and there is maximum value:
Therefore, characteristic frequency described in step 2 are as follows:
Wherein, d is array element spacing, and λ is wavelength.
If linear array, element number of array N, array element spacing d, wavelength X, target angle vector θkThe normalization electric field E at place
(sinθk) it is in φ0There are peak swing, peak swing in place are as follows:
So:
ωp=φ0 (11)
It include target bearing information i.e. in characteristic frequency.
Other steps and parameter are identical as specific embodiment two or three.
Specific embodiment 5: present embodiment unlike specific embodiment four, combines spy described in step 3
Levy the expectation likelihood statistic l of frequencyMELSpecifically:
By document [E.Tom Northardt, Igal Bilik, Yuri Abramovich.Bearings-only
constant velocity target maneuver detection via expected likelihood.IEEE
Transactions on Aerospace and Electronic Systems.2014,50 (04): 2974~2988.], by L
A mutually independent array sampled point is expressed as vectorK' accumulation period EL statistic can be obtained to simplify
Log-likelihood expression:
Formula (12) is improved, the expectation likelihood statistic l of binding characteristic frequency is obtainedMEL:
Wherein, k=1 ..., K '；The number of cycles of K ' expression accumulation；En,kFor noise feature vector；For rectangular projection；I is unit matrix,For N × JkDirection
The estimated value of matrix,For θkEstimated value.
Other steps and parameter are identical as specific embodiment four.
Specific embodiment 6: present embodiment is unlike specific embodiment five, feature described in step 4 to
Measure the calculating of detection threshold are as follows:
According to document [M.A.Richards, Fundamentals of Radar Signal Processing.New
York, NY, USA:McGraw-Hill, 2005.], it is detected using m-th of feature vector Fourier transform pairs echo signal
False-alarm probability PF(m) it indicates are as follows:
Wherein, m=1 ..., N, σ2It (m) is the variance of the corresponding noise of m-th of feature vector；
Then m-th of feature vector detection threshold are as follows:
Other steps and parameter are identical as specific embodiment five.
Specific embodiment 7: present embodiment unlike specific embodiment six, combines spy described in step 5
Levy the expectation likelihood statistic thresholding G of frequencyMELSpecific calculating process are as follows:
Wherein, p (lMEL, x) and it is probability density function (PDF) value being calculated according to numerical value, h is upper limit of integral, and x is
Intensity value, PFA are false-alarm probability.
Other steps and parameter are identical as specific embodiment six.
Specific embodiment 8: present embodiment is unlike specific embodiment seven, the false-alarm probability PFA are as follows:
When target is not present in signal, but there is Wm(θk) the case where being more than thresholding referred to as false-alarm.When without target, m-th of feature
Probability more than thresholding is indicated P (A by vectorm), the probability for being less than thresholding indicates
False-alarm probability can then be indicated are as follows:
PFA=P (A1∪A2∪...∪AN) (17)
I.e. formula (17) can also indicate are as follows:
Therefore, false-alarm probability PFA are as follows:
PF(m) false-alarm probability for indicating m feature vector, being shown by formula can be by calculating PF(m) it is general to obtain false-alarm
Rate.
Other steps and parameter are identical as specific embodiment seven.
Embodiment
Beneficial effects of the present invention are verified using following embodiment:
1, l is calculated using numerical approachMELThreshold value:
Consider an even linear array being made of 16 array elements, target number Jkλ/2=1, array element spacing d=.Sampled point
Number L=1.After doing 1,000,000 Monte Carlo Experiments, the frequency disribution of the testing result detected using the method for the present invention is obtained
Histogram such as Fig. 2.
Pass through the expectation likelihood statistic l of binding characteristic frequencyMEL；The integral calculation l of probability distributionMELThreshold value, with full
The desired PFA of foot.
Here p (lMEL, x) and it is calculated by the numerical value of the histogram in Fig. 2.
2, MEL and EL detection performance:
Consider an even linear array being made of 16 array elements, target number Jkλ/2=1, array element spacing d=.
(1) sampling number L=64 is respectively PFA=10 in false-alarm probability-4And PFA=10-5When obtain the method for the present invention
It is as shown in Figure 3, Figure 4 with signal-to-noise ratio result of variations with conventional EL method detection probability；Wherein, Pd(MEL) inspection of the method for the present invention
Survey probability, PdIt (EL) is the detection probability of conventional EL method；By Fig. 3, Fig. 4 it is found that being respectively PFA=10 in false-alarm probability-4With
PFA=10-5When, the method for the present invention testing result is more preferable than conventional EL method detection performance.
(2) sampling number L=128 is respectively PFA=10 in false-alarm probability-4And PFA=10-5When obtain the method for the present invention
It is as shown in Figure 5, Figure 6 with signal-to-noise ratio result of variations with conventional EL method detection probability；It is found that as sampling number increases, this hair
Bright method and conventional EL method detection performance are all significantly improved.
(3) sampling number L=128 is detected using the method for the present invention, in false-alarm probability PFA=10-3, 10-4, 10-5,
10-6When obtain detection probability with false-alarm probability result of variations as shown in fig. 7, it is found that as false-alarm probability increases, inspection of the invention
Performance is surveyed to be significantly improved.
The present invention can also have other various embodiments, without deviating from the spirit and substance of the present invention, this field
Technical staff makes various corresponding changes and modifications in accordance with the present invention, but these corresponding changes and modifications all should belong to
The protection scope of the appended claims of the present invention.
Claims (8)
1. a kind of signal detecting method of the expectation likelihood of binding characteristic frequency, which is characterized in that specifically includes the following steps:
Step 1: establish data covariance matrix by the sampled data of signal, and obtain the feature of the data covariance matrix to
The Fourier transform W of amountm(θk)；
Step 2: by gained Wm(θk) frequency spectrum peak point where frequency be defined as characteristic frequency；
Step 3: the expectation likelihood statistic l of binding characteristic frequency is calculatedMEL；
Step 4: calculating feature vector detection threshold；
Step 5: the l obtained in conjunction with step 3MEL, utilize the expectation likelihood statistic door of numerical calculations binding characteristic frequency
Limit GMEL；
Step 6: obtaining testing result by following binary hypothesis test formula:
Wherein, H0Indicate no signal, H1Indicate signal,For m-th of feature vector detection threshold.
2. the signal detecting method of the expectation likelihood of a kind of binding characteristic frequency according to claim 1, which is characterized in that step
Data covariance matrix described in rapid one are as follows:
Wherein,For data covariance matrix, L is sampling number；Subscript H indicates transposition, YkIt is to be adopted by L of k-th of period
The vector that sampling point is expressed as,The array at i-th moment in k-th of period exports
VectorAre as follows:
It is that zero-mean, the variance at i-th moment in k-th of period isGaussian reflectivity mirrors vector, JkIt is target number, mesh
Mark angle vector is θk=[θk(0) ..., θk(Jk)],It is JkThe target amplitude vector of × 1 dimension,A
(θk) it is N × JkDirection matrix, N are element number of array.
3. the signal detecting method of the expectation likelihood of a kind of binding characteristic frequency according to claim 2, which is characterized in that step
The Fourier transform W of described eigenvector is obtained in rapid onem(θk) detailed process includes:
Enable em(θk) it is data covariance matrixFeature vector, m=1 ..., Jk；Then em(θk) can indicate are as follows:
Wherein, nkIt indicates to receive the corresponding zero-mean of signal, variance σ2Gaussian Profile noise；γ′M, 2、η′m?
For signal guide vector representation parameter；a(θk(1))、a(θk(2))…a(θk(Jk)) be respectively as follows:
To feature vector em(θk) carry out Fourier transformation obtain Wm(θk), Wm(θk) in first of element are as follows:
Wherein, l=0 ..., N-1；em(θk, n) and indicate em(θk) in nth elements, n=0 ..., N-1；E is natural constant,
J is imaginary unit.
4. a kind of signal detecting method of the expectation likelihood of binding characteristic frequency, feature according to Claims 2 or 3 exist
In characteristic frequency described in step 2 are as follows:
Wherein, d is array element spacing, and λ is wavelength.
5. the signal detecting method of the expectation likelihood of a kind of binding characteristic frequency according to claim 4, which is characterized in that step
The expectation likelihood statistic l of binding characteristic frequency described in rapid threeMELSpecifically:
Wherein, k=1 ..., K '；The number of cycles of K ' expression accumulation；EN, kFor noise feature vector；For rectangular projection；I is unit matrix,For N × JkDirection
The estimated value of matrix,For θkEstimated value.
6. the signal detecting method of the expectation likelihood of a kind of binding characteristic frequency according to claim 5, which is characterized in that step
The calculating of feature vector detection threshold described in rapid four are as follows:
Utilize the false-alarm probability P of m-th of feature vectorF(m) it indicates are as follows:
Wherein, m=1 ..., N, σ2It (m) is the variance of the corresponding noise of m-th of feature vector；
Then m-th of feature vector detection threshold are as follows:
。
7. the signal detecting method of the expectation likelihood of a kind of binding characteristic frequency according to claim 6, which is characterized in that step
The expectation likelihood statistic thresholding G of binding characteristic frequency described in rapid fiveMELSpecific calculating process are as follows:
Wherein, p (lMEL, x) and it is probability density function (PDF) value being calculated according to numerical value, h is upper limit of integral, and x is intensity
Value, PFA is false-alarm probability.
8. the signal detecting method of the expectation likelihood of a kind of binding characteristic frequency according to claim 7, which is characterized in that institute
State false-alarm probability PFA are as follows:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910637642.5A CN110348402B (en) | 2019-07-15 | 2019-07-15 | Signal detection method combining expected likelihood of characteristic frequency |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910637642.5A CN110348402B (en) | 2019-07-15 | 2019-07-15 | Signal detection method combining expected likelihood of characteristic frequency |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110348402A true CN110348402A (en) | 2019-10-18 |
| CN110348402B CN110348402B (en) | 2021-05-28 |
Family
ID=68176356
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910637642.5A Active CN110348402B (en) | 2019-07-15 | 2019-07-15 | Signal detection method combining expected likelihood of characteristic frequency |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110348402B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7616395B2 (en) * | 2004-08-20 | 2009-11-10 | Panasonic Corporation | Information reproduction apparatus |
| CN102523054A (en) * | 2011-12-07 | 2012-06-27 | 清华大学 | Multiple Input Multiple Output (MIMO) detecting method |
| CN103227761A (en) * | 2013-05-24 | 2013-07-31 | 北京邮电大学 | Estimation method of multi-path non-Gaussian noise channel based on empirical likelihood method |
| CN104166128A (en) * | 2014-08-06 | 2014-11-26 | 电子科技大学 | Multi-pass SAR coherent change detection method based on general likelihood ratio |
| CN109245814A (en) * | 2018-09-13 | 2019-01-18 | 哈尔滨工业大学 | Adaptive beamformer method based on maximum likelihood resampling |
| CN109492190A (en) * | 2018-08-12 | 2019-03-19 | 中国科学院大学 | A kind of subglacial layer position detecting method based on branch's formula gauss hybrid models |
-
2019
- 2019-07-15 CN CN201910637642.5A patent/CN110348402B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7616395B2 (en) * | 2004-08-20 | 2009-11-10 | Panasonic Corporation | Information reproduction apparatus |
| CN102523054A (en) * | 2011-12-07 | 2012-06-27 | 清华大学 | Multiple Input Multiple Output (MIMO) detecting method |
| CN103227761A (en) * | 2013-05-24 | 2013-07-31 | 北京邮电大学 | Estimation method of multi-path non-Gaussian noise channel based on empirical likelihood method |
| CN104166128A (en) * | 2014-08-06 | 2014-11-26 | 电子科技大学 | Multi-pass SAR coherent change detection method based on general likelihood ratio |
| CN109492190A (en) * | 2018-08-12 | 2019-03-19 | 中国科学院大学 | A kind of subglacial layer position detecting method based on branch's formula gauss hybrid models |
| CN109245814A (en) * | 2018-09-13 | 2019-01-18 | 哈尔滨工业大学 | Adaptive beamformer method based on maximum likelihood resampling |
Non-Patent Citations (5)
| Title |
|---|
| PEI-JUNG CHUNG ET.AL: "Estimation of seismic-wave parameters and signal detection using maximum-likelihood methods", 《COMPUTERS & GEOSCIENCES》 * |
| YUGUAN HOU ET.AL: "A modified KCCA for clutter separation in airborne MIMO SAR", 《2016 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM (IGARSS)》 * |
| 张娟: "MIMO雷达信号检测与参数估计算法研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
| 王隆: "MIMO通信系统中信号检测技术的研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 申京: "MIMO-OFDM系统中信道估计及信号检测算法的研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110348402B (en) | 2021-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6021376B2 (en) | A method for detecting a target in a radar signal using spatio-temporal adaptation processing in a complex Gaussian distribution environment, and a system for detecting a target in a radar signal using spatio-temporal adaptation processing in a complex Gaussian distribution environment | |
| Hao et al. | Persymmetric Rao and Wald tests for partially homogeneous environment | |
| JP2012220492A5 (en) | ||
| CN115032623B (en) | Method and system for detecting double-parameter weighted extended target during subspace signal mismatch | |
| Sim et al. | Enhanced DOA estimation using linearly predicted array expansion for automotive radar systems | |
| Ai et al. | Robust adaptive beamforming with subspace projection and covariance matrix reconstruction | |
| Ghasemi et al. | Joint scattering center enumeration and parameter estimation in GTD model | |
| Hari et al. | Narrowband signal detection techniques in shallow ocean by acoustic vector sensor array | |
| Zhou et al. | Spatial information and angular resolution of sensor array | |
| Bai et al. | Association of DOA estimation from two ULAs | |
| CN110727915A (en) | Robust self-adaptive beam forming method based on data correlation constraint | |
| Meng et al. | Robust beamforming for non-circular signals in uniform linear arrays with unknown mutual coupling | |
| Huang et al. | Low-complexity MDL method for accurate source enumeration | |
| CN110348402A (en) | A kind of signal detecting method of the expectation likelihood of binding characteristic frequency | |
| De Maio | Maximum likelihood estimation of structured persymmetric covariance matrices | |
| Monga et al. | Rank constrained ML estimation of structured covariance matrices with applications in radar target detection | |
| US9660715B2 (en) | Method for the detection of an electromagnetic signal by an antenna array, and device implementing said method | |
| Combernoux et al. | Performance of the low-rank adaptive normalized matched filter test under a large dimension regime | |
| Penghui et al. | Radar HRRP target recognition based on linear dynamic model | |
| Aouada et al. | Application of the bootstrap to source detection in nonuniform noise | |
| Kalson | Adaptive array CFAR detection | |
| Vasyiyshyn | Antenna array signal processing with high-resolution by modified beamspace music algorithm | |
| Aktas et al. | Acoustic direction finding in highly reverberant environment with single acoustic vector sensor | |
| Breloy et al. | CFAR property and robustness of the lowrank adaptive normalized matched filters detectors in low rank compound gaussian context | |
| JP2004257968A (en) | Angle measuring instrument for measuring position of movable body |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |