CN101572574A - Smart antenna self-adapting interference suppression method based on least square-lowest mean square - Google Patents
Smart antenna self-adapting interference suppression method based on least square-lowest mean square Download PDFInfo
- Publication number
- CN101572574A CN101572574A CN 200910069090 CN200910069090A CN101572574A CN 101572574 A CN101572574 A CN 101572574A CN 200910069090 CN200910069090 CN 200910069090 CN 200910069090 A CN200910069090 A CN 200910069090A CN 101572574 A CN101572574 A CN 101572574A
- Authority
- CN
- China
- Prior art keywords
- algorithm
- mean square
- array
- training sequence
- interference suppression
- 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
Images
Landscapes
- Variable-Direction Aerials And Aerial Arrays (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention relates to a smart antenna self-adapting interference suppression method based on least square-lowest mean square, which is an interference suppression algorithm based on training sequences; and by combining the least square algorithm with the lowest mean square algorithm, the method increases the velocity of convergence of the least mean square algorithm. The method includes the following steps: radio-frequency signals received by array antenna are converted into IF signals by a down converter; the IF signals are processed by the operations of analogue-to-digital conversion and digital down-conversion to obtain zero IF digital signals; the zero IF digital signals obtained in the step 2; and the local training sequence are utilized for carrying out corresponding processing to obtain local reference signals by computation delay; the low snapshot least square algorithm is utilized for calculating the initial weight vector of an aerial array; the weight vector calculated in the step 4 is used as the initial weight vector of the least mean square algorithm, and the least mean square algorithm is utilized for updating the aerial array weight vector; and the array weight vector calculated in the step 5 is adopted for the interference suppression of user data. The method achieves the purposes of increasing the availability of frequency spectrum of the system and reducing the complexity of the system.
Description
Technical field
The present invention relates to a kind of smart antenna interference suppression algorithm.Particularly relate to a kind of smart antenna self-adapting interference suppression method based on least square-lowest mean square based on training sequence.
Background technology
At the beginning of the nineties in last century, array signal process technique is introduced in the mobile communication, has formed a new research focus-smart antenna very soon.Smart antenna is defined as having the antenna array of direction finding and wave beam formation ability, generally is divided into two big classes: multi-beam intelligent antenna and adaptive array smart antenna, be called for short multi-beam antenna and adaptive antenna.The sensing of each wave beam of multi-beam antenna is fixed, and it determines the direction of arrival of signal by detection technique, by regulating the weight coefficient of each array element, selects respective beam then, and is simple in structure.But along with moving of user, when its no longer fixed beam point to the center time, multi-beam antenna then can not obtain good reception.And adaptive antenna is discerned user's direction of arrival by array signal process technique, form main beam in that this side up then, and main beam constantly changes adjustment according to the change of signal direction of arrival, so that keep aiming at the desired signal direction, form zero at interference radiating way simultaneously and fall into, thereby be widely used in the mobile communication system.
At present, the smart antenna self-adapting interference mitigation technology roughly can be divided three classes: (1) based on the known beam-forming technology of useful signal direction vector, representational is that standard C apon wave beam forms (Standard Capon Beamformer is called for short SCB).(2) based on the known beam-forming technology of reference signal, this method minimizes the weight vectors of asking suitable by making the difference between array output and the reference signal.In general, reference signal is difficult to obtain.In communication system, in order to obtain this reference signal, the two is all known training sequence to transmitter and receiver with regard to must periodically sending.The transmitting training sequence will take the frequency resource of communication system preciousness.(3) blind adaptive beam-forming technology, this technology does not need the transmitting training signal, also need not know the prioris such as spatial autocorrelation matrix of array direction vector and interference and noise, but utilize the statistical property of signal or signal itself really qualitative property carry out wave beam and form.
Lowest mean square (Least Mean Squares based on training sequence, abbreviation LMS) the adaptive disturbance method is used widely in communication system, but because the LMS algorithm is slower to initial value sensitivity, convergence rate, and constringency performance is relevant with interference environment, therefore need just can reach desirable interference suppressioning effect than long training sequence, serious waste valuable frequency resource.Least square (Least Squares, abbreviation LS) is though optimal beam formation inhibition jamming performance is higher and disturb rejection and interference environment to have nothing to do, and algorithm operation quantity is big, is unfavorable for the real system realization.
Summary of the invention
Technical problem to be solved by this invention is, provide a kind of based on least square-lowest mean square (LeastSquares-Least Mean Squares, abbreviation LS-LMS) smart antenna self-adapting interference suppression method, this method goes out the weight vectors of array by little fast umber of beats LS algorithm computation, with the initial weight vectors of this weight vectors as the LMS algorithm, effectively improved LMS convergence of algorithm speed, reduce the length of training sequence, thereby reached the purpose that improves the system spectrum utilance, reduces system complexity.
The technical solution adopted in the present invention is: a kind of smart antenna self-adapting interference suppression method based on least square-lowest mean square, be based on the interference suppression algorithm of training sequence, by least-squares algorithm is combined with least mean square algorithm, improve the convergence rate of least mean square algorithm, include following steps:
(1) will become intermediate-freuqncy signal through low-converter by the radiofrequency signal of array antenna received;
(2) intermediate-freuqncy signal is carried out the A/D conversion, Digital Down Convert obtains the zero intermediate frequency digital signal;
(3) utilize step (2) gained zero intermediate frequency digital signal and local training sequence to carry out the relevant treatment computing relay and obtain local reference signal;
(4) utilize little fast umber of beats least-squares algorithm to calculate the initial weight vectors of aerial array;
(5) weight vectors that step (4) is calculated utilizes least mean square algorithm to carry out the renewal of antenna array weights vector as the initial weight vectors of least mean square algorithm;
(6) the array weight vector that adopts step (5) to calculate disturbs inhibition to user data.
Step (3) is described carries out the relevant treatment computing relay with digital intermediate frequency signal and local training sequence and obtains local reference signal, be that local training sequence is progressively postponed, training sequence after each delay is made related operation at a training sequence in the cycle with the array dateout respectively, the relatively output of each correlator then, therefrom select maximum output, the training sequence on the branch road corresponding with this maximum output is local reference signal.
The described initial weight vectors that utilizes the least-squares algorithm computing array of step (4), be by with the error sum of squares minimum as cost function, utilize little fast umber of beats to calculate weight vectors.
The described renewal that utilizes least mean square algorithm to carry out the antenna array weights vector of step (5), be to be that cost function, stochastic gradient algorithm are that optimization method carries out right value update with the mean square error with least mean square algorithm, the initial weight vectors of least mean square algorithm is provided by the weight vectors that step (4) calculates.
Smart antenna self-adapting interference suppression method based on least square-lowest mean square of the present invention, be to utilize the LS algorithm to estimate that array weight vector calculus amount is big at the long communication system of training sequence, the slow and a kind of disturbance restraining method that proposes of LMS algorithm the convergence speed.The present invention utilizes array weight vector that the LS algorithm computation the goes out initial weight vectors as the LMS algorithm, during LS algorithm computation weight vectors, utilization be that little fast umber of beats calculates, can reduce the operand of LS algorithm.The weight vectors that goes out with little fast umber of beats LS algorithm computation is as the initial weight vectors of LMS algorithm, accelerated LMS convergence of algorithm speed greatly, reduced the sensitiveness of LMS algorithm to interference environment, can reduce the length of training sequence, thereby reach the purpose that improves the system spectrum utilance, reduces system complexity.
Description of drawings
Fig. 1 is based on the smart antenna self-adapting interference suppression method flow chart of least square-lowest mean square;
The weight vectors that Fig. 2 is based on training sequence upgrades the array junctions composition;
Fig. 3 a is the array pattern of fast umber of beats=180 o'clock LMS algorithm;
Fig. 3 b is the array pattern of fast umber of beats=180 o'clock LS-LMS algorithm;
Fig. 3 c is the array pattern of fast umber of beats=100 o'clock LMS algorithm;
Fig. 3 d is the array pattern of fast umber of beats=100 o'clock LS-LMS algorithm;
Fig. 3 e is the array pattern of fast umber of beats=30 o'clock LMS algorithm;
Fig. 3 f is the array pattern of fast umber of beats=30 o'clock LS-LMS algorithm;
Fig. 4 a is that the initial weight vectors of LMS algorithm is [0,0,0,0]
T, dried LMS algorithm, LS-LMS algorithm output Signal to Interference plus Noise Ratio convergence curve figure when making an uproar than (Jamming-to-noise Ratio, be called for short JNR)=10dB;
Fig. 4 b is that the initial weight vectors of LMS algorithm is [0,0,0,0]
T, LMS algorithm, LS-LMS algorithm output Signal to Interference plus Noise Ratio convergence curve figure during JNR=15dB;
Fig. 4 c is that the initial weight vectors of LMS algorithm is [0,0,0,0]
T, LMS algorithm, LS-LMS algorithm output Signal to Interference plus Noise Ratio convergence curve figure during JNR=20dB;
Fig. 5 a is that the initial weight vectors of LMS algorithm is [0.5+0.5i, 0.5+0.5i, 0.5+0.5i, 0.5+0.5i]
T, LMS algorithm, LS-LMS algorithm output Signal to Interference plus Noise Ratio convergence curve figure during JNR=10dB;
Fig. 5 b is that the initial weight vectors of LMS algorithm is [0.5+0.5i, 0.5+0.5i, 0.5+0.5i, 0.5+0.5i]
T, LMS algorithm, LS-LMS algorithm output Signal to Interference plus Noise Ratio convergence curve figure during JNR=15dB;
Fig. 5 c is that the initial weight vectors of LMS algorithm is [0.5+0.5i, 0.5+0.5i, 0.5+0.5i, 0.5+0.5i]
T, LMS algorithm, LS-LMS algorithm output Signal to Interference plus Noise Ratio convergence curve figure during JNR=20dB;
Fig. 6 a is LS algorithm, LMS algorithm, LS-LMS algorithm operation quantity comparison diagram;
Fig. 6 b is that LS algorithm, LMS algorithm, LS-LMS algorithm are taken advantage of again and added number of times and change comparative graph with fast umber of beats;
Fig. 6 c is that LS algorithm, LMS algorithm, LS-LMS algorithm are taken advantage of again and added number of times and change comparative graph with array number.
Embodiment
Below in conjunction with the embodiment accompanying drawing smart antenna self-adapting interference suppression method based on least square-lowest mean square of the present invention is made a detailed description.
Of the present invention based on least square-lowest mean square (Least Squares-Least Mean Squares, abbreviation LS-LMS) smart antenna self-adapting interference suppression method, be based on the self-adapting interference suppression method of training sequence, by with LS (Least Squares, be called for short LS) algorithm and LMS (Least Mean Squares, abbreviation LMS) algorithm combines, improve LMS convergence of algorithm speed, promptly at receiving terminal known sequences signal of the periodic emission of transmitting terminal, receiver self produces the reference signal of this sequence signal as adaptive algorithm, and weight vector minimizes by cost function and obtains.As shown in Figure 1, include following steps:
The first step: will become intermediate-freuqncy signal through low-converter by the radiofrequency signal of array antenna received;
Second the step: with intermediate-freuqncy signal carry out the A/D conversion, Digital Down Convert obtains the zero intermediate frequency digital signal;
When having interference, array antenna received signals can be expressed as:
Wherein, l sampling of x (l) expression snap (l=0,1 ..., L-1), L represents hits, s
d(l) expression useful signal, s
k(l) (k=1 ..., K) k interference signal of expression, K represents the interference source number,
The steering vector of expression useful signal,
Represent k steering vector that disturbs, λ represents signal wavelength, e (l) represents the array received noise vector, and θ is the direction of arrival of signal, and M is an array number, d is an array element distance, adopt even linear array in the present embodiment, element number of array is 4, and spacing is 1/2 wavelength, received signal is a useful signal and an interference signal, and direction of arrival is respectively 30 ° and 0 °.
The 3rd step: utilize the second step gained zero intermediate frequency digital signal and local training sequence to carry out the relevant treatment computing relay and obtain local reference signal, in the adaptive algorithm based on training sequence, essential step is realize training sequence synchronous.Be that local training sequence is progressively postponed in the present embodiment, training sequence after each delay is made related operation at a training sequence in the cycle with the array dateout respectively, the relatively output of each correlator then, therefrom select maximum output, the training sequence on the branch road corresponding with this maximum output is local reference signal.
The 4th step: utilize the initial weight vectors of little fast umber of beats LS algorithm computation aerial array, be by with the error sum of squares minimum as cost function, utilize little fast umber of beats N (N<<L) calculate weight vectors;
Based on the adaptive array array structure of training sequence as shown in Figure 2.
Suppose the data vector x (n) of N snap, n=0,1 ..., N-1, then the cost function of LS algorithm is:
Wherein, d (n) is a n desired signal constantly,
Its gradient is:
Making it is the zero optimum weight vectors that obtains least square method:
w=(XX
H)
-1Xd
H (4)
Wherein: X=[x (0), x (1) ..., x (N-1)], d=[d (0), d (1) ..., d (N-1)],
In the present embodiment, fast umber of beats N=8 samples.
The 5th step: the weight vectors that the 4th step is calculated is as the initial weight vectors of LMS algorithm, utilize the LMS algorithm to carry out the renewal of antenna array weights vector, the LMS algorithm is that cost function, stochastic gradient algorithm are that optimization method carries out right value update with the mean square error, and the initial weight vectors of LMS algorithm is provided by the weight vectors that the 4th step calculates;
LMS algorithm criterion is that the mean-square value of evaluated error is minimized, and promptly cost function is:
J(w)=E{|e(l)|
2} (5)
E{} represents statistical average in the formula, and e (l) is an error, e (l)=w
HX (l)-d (l).
Then:
J(w)=E{e(l)e
*(l)}=E{|d(l)|
2}-2Re[w
Hr
xd]+w
HR
xxw (6)
Wherein, Re represents to get real part, R
Xx=E{x (l) x
H(l) } be the autocorrelation matrix of input vector, r
Xd=E{xl)) d
*(l) } be the cross-correlation matrix of input vector x (l) and desired signal d (l).
To formula (6) differentiate and to make it be zero to obtain:
Consider the stochastic gradient algorithm, the general formula that weight vector upgrades is:
Wherein,
μ is a step factor, can be got by formula (6):
Replace mathematic expectaion with instantaneous value in the LMS algorithm, promptly obtain the right value update formula:
The 6th step: the array weight vector that adopts the 5th step to calculate disturbs inhibition to user data.
Fig. 3 has provided at signal to noise ratio (Signal-to-noise Ratio, be called for short SNR)=condition of 10dB, SIR=-20dB under LMS carry out 100 Monte Carlos (Mont Carlo) with two kinds of algorithms of LS-LMS and test the array pattern that obtains, the initial value of LMS algorithm all is taken as [0.5+0.5i in the emulation experiment, 0.5+0.5i, 0.5+0.5i, 0.5+0.5i]
TLS-LMS utilizes 8 snaps to calculate initial weight vectors, the step-length of LMS and two kinds of algorithms of LS-LMS is 0.00005, Fig. 3 a, Fig. 3 b are respectively the array pattern of fast umber of beats=180 o'clock LMS, two kinds of algorithms of LM-LMS, Fig. 3 c, Fig. 3 d are respectively the array pattern of fast umber of beats=100 o'clock LMS, two kinds of algorithms of LM-LMS, and Fig. 3 e, Fig. 3 f are respectively the array pattern of fast umber of beats=30 o'clock LMS, two kinds of algorithms of LM-LMS.As seen from Figure 3, along with the minimizing LMS algorithm of fast umber of beats can not form zero point at interference radiating way, but new method is still aimed at useful signal with main beam, and zero falls into and aims at interference signal.
When Fig. 4 has provided SNR=10dB, LMS algorithm, LS-LMS algorithm output Signal to Interference plus Noise Ratio (Signal toInterference and Noise Ratio, be called for short SINR) convergence curve figure, wherein JNR is respectively 10dB, 15dB, 20dB among Fig. 4 a, Fig. 4 b, Fig. 4 c.The initial value of LMS algorithm is taken as [0,0,0,0]
T, utilize 8 snaps to calculate initial weight vectors among the LS-LMS, the step-length of two kinds of algorithms is 0.00005, carries out 100 Mont Carlo emulation experiments and obtains the average Signal to Interference plus Noise Ratio of Beam-former output with sampling snap transformation of variable curve as shown in Figure 4.Two kinds of algorithms of increase along with fast umber of beats all converge on theoretical value (the SCB method of correspondence when fast umber of beats is infinitely great) as can be seen, but the LMS algorithm is considerably slower than LS-LMS, and it is along with the increase LMS convergence of algorithm speed of JNR reduces, but little to the LS-LMS algorithm affects.Fig. 5 is the same with emulation experiment content and condition among Fig. 4, and just the initial value of LMS algorithm is taken as [0.5+0.5i, 0.5+0.5i, 0.5+0.5i, 0.5+0.5i]
TComparison diagram 4 can be found the LMS algorithm to initial weight vectors sensitivity with Fig. 5, and different initial weight vectors have considerable influence to the LMS algorithm the convergence speed, but little to the influence of LS-LMS algorithm the convergence speed.
Fig. 6 a has provided LS algorithm, LMS algorithm, LS-LMS algorithm operation quantity comparison diagram, wherein M is an element number of array, L is fast umber of beats, and N is the fast umber of beats that calculates initial weight vectors in the LS-LMS algorithm, is approximately the operand of matrix inversion among the figure with the complexity of matrix inversion operation.LS algorithm, LMS algorithm, LS-LMS algorithm were taken advantage of again and are added number of times with fast umber of beats variation comparative graph when Fig. 6 b provided M=4, N=8.LS algorithm, LMS algorithm, LS-LMS algorithm were taken advantage of again and are added number of times and change comparative graph with array number when Fig. 6 c provided L=120, N=8.Relatively can find sharply to increase by Fig. 6, but the increase of LS-LMS algorithm operation quantity is less relatively along with the increase LS algorithm operation quantity of fast umber of beats and array number.
Claims (4)
1. smart antenna self-adapting interference suppression method based on least square-lowest mean square, it is characterized in that, be based on the interference suppression algorithm of training sequence, by least-squares algorithm is combined with least mean square algorithm, improve the convergence rate of least mean square algorithm, include following steps:
(1) will become intermediate-freuqncy signal through low-converter by the radiofrequency signal of array antenna received;
(2) intermediate-freuqncy signal is carried out the A/D conversion, Digital Down Convert obtains the zero intermediate frequency digital signal;
(3) utilize step (2) gained zero intermediate frequency digital signal and local training sequence to carry out the relevant treatment computing relay and obtain local reference signal;
(4) utilize little fast umber of beats least-squares algorithm to calculate the initial weight vectors of aerial array;
(5) weight vectors that step (4) is calculated utilizes least mean square algorithm to carry out the renewal of antenna array weights vector as the initial weight vectors of least mean square algorithm;
(6) the array weight vector that adopts step (5) to calculate disturbs inhibition to user data.
2. the smart antenna self-adapting interference suppression method based on least square-lowest mean square according to claim 1, it is characterized in that, step (3) is described carries out the relevant treatment computing relay with digital intermediate frequency signal and local training sequence and obtains local reference signal, be that local training sequence is progressively postponed, training sequence after each delay is made related operation at a training sequence in the cycle with the array dateout respectively, the relatively output of each correlator then, therefrom select maximum output, the training sequence on the branch road corresponding with this maximum output is local reference signal.
3. the smart antenna self-adapting interference suppression method based on least square-lowest mean square according to claim 1, it is characterized in that, the described initial weight vectors that utilizes the least-squares algorithm computing array of step (4), be by with the error sum of squares minimum as cost function, utilize little fast umber of beats to calculate weight vectors.
4. the smart antenna self-adapting interference suppression method based on least square-lowest mean square according to claim 1, it is characterized in that, the described renewal that utilizes least mean square algorithm to carry out the antenna array weights vector of step (5), be to be that cost function, stochastic gradient algorithm are that optimization method carries out right value update with the mean square error with least mean square algorithm, the initial weight vectors of least mean square algorithm is provided by the weight vectors that step (4) calculates.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2009100690909A CN101572574B (en) | 2009-06-01 | 2009-06-01 | Smart antenna self-adapting interference suppression method based on least square-lowest mean square |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2009100690909A CN101572574B (en) | 2009-06-01 | 2009-06-01 | Smart antenna self-adapting interference suppression method based on least square-lowest mean square |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101572574A true CN101572574A (en) | 2009-11-04 |
CN101572574B CN101572574B (en) | 2012-10-31 |
Family
ID=41231820
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2009100690909A Expired - Fee Related CN101572574B (en) | 2009-06-01 | 2009-06-01 | Smart antenna self-adapting interference suppression method based on least square-lowest mean square |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101572574B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103095357A (en) * | 2013-01-18 | 2013-05-08 | 成都九华圆通科技发展有限公司 | Intelligence antenna self-adaptive beam forming method |
CN104679976A (en) * | 2014-10-31 | 2015-06-03 | 哈尔滨工业大学深圳研究生院 | Contractive linear and contractive generalized linear complex-valued least squares algorithm for signal processing |
CN107135022A (en) * | 2017-05-04 | 2017-09-05 | 成都航天通信设备有限责任公司 | A kind of quick multiple antennas weight merging approach |
CN107332601A (en) * | 2017-06-18 | 2017-11-07 | 南京理工大学 | A kind of self-adapting interference suppression method based on wave beam forming |
CN109217844A (en) * | 2018-10-30 | 2019-01-15 | 哈尔滨理工大学 | Hyperparameter optimization method based on the random Fourier's feature core LMS of pre-training |
CN109831236A (en) * | 2018-11-13 | 2019-05-31 | 电子科技大学 | A kind of beam selection method based on Monte Carlo tree search auxiliary |
CN112003645A (en) * | 2020-10-29 | 2020-11-27 | 南昌大学 | Synchronous reference signal extraction method and device |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4299323B2 (en) * | 2006-08-04 | 2009-07-22 | 株式会社日立国際電気 | Communications system |
CN101013916B (en) * | 2007-02-05 | 2010-05-19 | 华南理工大学 | Method and apparatus for implementing reception of diversity with intelligent antenna |
-
2009
- 2009-06-01 CN CN2009100690909A patent/CN101572574B/en not_active Expired - Fee Related
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103095357B (en) * | 2013-01-18 | 2015-11-18 | 成都九华圆通科技发展有限公司 | Smart antenna self-adapting Beamforming Method |
CN103095357A (en) * | 2013-01-18 | 2013-05-08 | 成都九华圆通科技发展有限公司 | Intelligence antenna self-adaptive beam forming method |
CN104679976A (en) * | 2014-10-31 | 2015-06-03 | 哈尔滨工业大学深圳研究生院 | Contractive linear and contractive generalized linear complex-valued least squares algorithm for signal processing |
CN104679976B (en) * | 2014-10-31 | 2017-11-28 | 哈尔滨工业大学深圳研究生院 | Contraction for signal transacting is linear and shrinks the multiple least-squares algorithm of generalized linear |
CN107135022B (en) * | 2017-05-04 | 2020-06-09 | 成都航天通信设备有限责任公司 | Rapid multi-antenna weighting combination method |
CN107135022A (en) * | 2017-05-04 | 2017-09-05 | 成都航天通信设备有限责任公司 | A kind of quick multiple antennas weight merging approach |
CN107332601B (en) * | 2017-06-18 | 2020-10-30 | 南京理工大学 | Adaptive interference suppression method based on beamforming |
CN107332601A (en) * | 2017-06-18 | 2017-11-07 | 南京理工大学 | A kind of self-adapting interference suppression method based on wave beam forming |
CN109217844A (en) * | 2018-10-30 | 2019-01-15 | 哈尔滨理工大学 | Hyperparameter optimization method based on the random Fourier's feature core LMS of pre-training |
CN109217844B (en) * | 2018-10-30 | 2022-02-25 | 哈尔滨理工大学 | Hyper-parameter optimization method based on pre-training random Fourier feature kernel LMS |
CN109831236A (en) * | 2018-11-13 | 2019-05-31 | 电子科技大学 | A kind of beam selection method based on Monte Carlo tree search auxiliary |
CN109831236B (en) * | 2018-11-13 | 2021-06-01 | 电子科技大学 | Beam selection method based on Monte Carlo tree search assistance |
CN112003645A (en) * | 2020-10-29 | 2020-11-27 | 南昌大学 | Synchronous reference signal extraction method and device |
Also Published As
Publication number | Publication date |
---|---|
CN101572574B (en) | 2012-10-31 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101572574A (en) | Smart antenna self-adapting interference suppression method based on least square-lowest mean square | |
JP6278961B2 (en) | Radar equipment | |
US10024958B2 (en) | Radar apparatus | |
JP7108929B2 (en) | Radar device and target determination method | |
CN109923433A (en) | Radar installations, signal processing apparatus and signal processing method | |
CN110988811B (en) | Grating lobe ambiguity resolution method applied to sparsely-arranged broadband phased array | |
CN101561490A (en) | Blind self-adaptive GPS interference suppression method based on codeword structure | |
JP6462365B2 (en) | Radar apparatus and radar signal processing method thereof | |
CN105158773A (en) | Multi-beam directional anti-interference processing method for satellite navigation reception system | |
CN104243001A (en) | Broadband beam switching system and method | |
CN110824414A (en) | Device and method for estimating angle of arrival | |
CN103941234B (en) | A kind of Beam-former method for designing for multipath signal | |
CN102664666A (en) | Efficient robust self-adapting beam forming method of broadband | |
CN103037494A (en) | Control method for multi-beam antenna power | |
CN105282761A (en) | Rapid LMS adaptive wave beam forming method | |
CN103701515A (en) | Digital multi-beam forming method | |
CN104868946B (en) | The disturbance restraining method of adaptive weighted Subarray mixing MIMO phased array systems | |
CN103248412B (en) | A kind of method of satellite navigation Multibeam synthesis | |
CN109669167B (en) | Airborne radar emission waveform selection method based on radio frequency stealth | |
Kamio et al. | Adaptive sidelobe control for clutter rejection of atmospheric radars | |
Anajemba et al. | Efficient switched digital beamforming radar system based on SIMO/MIMO receiver | |
CN101552630B (en) | Beam forming method based on aeronautical communication channel | |
Sharma et al. | SMI algorithm—Adaptive beamforming for radar systems | |
Jichuan et al. | The impact of step selection in NLMS algorithm on low velocity target detecting for passive radar | |
US20230033166A1 (en) | Communication apparatus and communication method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20121031 Termination date: 20180601 |