CN103647591A - Cooperative interference detection method based on support vector machine - Google Patents
Cooperative interference detection method based on support vector machine Download PDFInfo
- Publication number
- CN103647591A CN103647591A CN201310737317.9A CN201310737317A CN103647591A CN 103647591 A CN103647591 A CN 103647591A CN 201310737317 A CN201310737317 A CN 201310737317A CN 103647591 A CN103647591 A CN 103647591A
- Authority
- CN
- China
- Prior art keywords
- interference
- node
- characteristic parameter
- data fusion
- interference detection
- 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
- Noise Elimination (AREA)
- Monitoring And Testing Of Transmission In General (AREA)
Abstract
The invention discloses a cooperative interference detection method based on a support vector machine. The method comprises the following steps: firstly, respectively operating different categories of interference signals to obtain the characteristic parameters of the interference signals, and training the support vector machine through the obtained characteristic parameters to obtain a classification model of the support vector machine; secondarily, carrying out energy detection on the receiving signals of a plurality of nodes to obtain an interference detection result; transmitting the interference detection result and the characteristic parameters of the signals to a data fusion center, determining whether the interference signals exist according to the interference detection result by using the data fusion center, and if the interference signals exist, carrying out interference identification on the interference signals based on a support vector machine algorithm to determine the interference category. The cooperative interference detection method disclosed by the invention can be used for accurately identifying the characteristics of the interference signals and effectively resisting multi-path fading.
Description
Technical field
The present invention relates to interference signal characteristic parameter extraction, the interference identification method based on energy measuring, the classification of disturbance method based on SVMs and interfere information and merge, the multipoint cooperative that is specially adapted to wireless communication technology field disturbs identification.
Background technology
With respect to multipoint cooperative, disturb identification, the interference identification of single node exists blocks and the problem such as channel fading, cause and disturb recognition result all inaccurate: the node having disturbs formation undetected owing to blocking to cause to detect, and the empty inspection and undetected because the reasons such as channel fading form of some nodes.Especially in the middle of broadband wireless transmission, interference signal arrives receiver and has equally multipath fading.The interference signal that receiver receives has very large decline at frequency domain, and Interference Detection and the identification of depending merely on a node are accurately to obtain the feature of interference signal.
Summary of the invention
The technical problem of solution required for the present invention is to avoid the weak point in above-mentioned background technology and a kind of cooperation interference detection method based on SVMs is provided, thereby effectively resists multipath fading.
The technical solution used in the present invention is: a kind of cooperation interference detection method based on SVMs, based on a cooperative detection system with multinode, one of them node is as data fusion center, and all the other nodes, as cooperative node, comprise the following steps:
(1) extract the characteristic parameter of dissimilar interference signal, the characteristic parameter of interference signal is carrier wave factor coefficient, the smooth coefficient of average frequency spectrum, the white noise factor, frequency domain square coefficient of kurtosis and the frequency domain square coefficient of skewness, then the characteristic parameter of dissimilar interference signal is carried out to the computing of SVMs block algorithm, supported vector machine classification plane;
(2) frequency range that each node is all used the signal receiving is separately divided into a plurality of frequency sub-band, and the signal in each frequency sub-band is called subsignal;
(3) each node carries out energy measuring to subsignal separately respectively, and the energy detection results of subsignal separately being merged to the interference detection results that obtains each node, each cooperative node is transferred to data fusion center by interference detection results separately;
(4) each node all carries out the extraction of characteristic parameter to the signal receiving separately, and each cooperative node is transferred to data fusion center by the characteristic parameter extracting separately;
(5) data fusion center, according to the interference detection results at the interference detection results of each cooperative node and data fusion center self, adopts K order criterion to merge the interference detection results of each node, obtains total interference detection results; If total interference detection results shows, there is not interference signal, behind official hour interval, forward step (2) to, otherwise forward step (6) to;
(6) data fusion center is according to the characteristic parameter of each cooperative node transmission and the characteristic parameter at data fusion center self, adopt and select merging criterion to merge and obtain new characteristic parameter the characteristic parameter of each node, then utilize algorithm of support vector machine that new characteristic parameter is matched in SVMs classification plane, thus the classification of identification interference signal; After the interval time of regulation, forward step (2) to.
The present invention compared with prior art has the following advantages:
1,, by the SVMs classification and identification algorithm at data fusion center, effectively promote the recognition performance to interference signal;
2, each node does not need the interference signal receiving to be transferred to data fusion center, and the result only detecting to fusion center transmitting energy and the characteristic parameter of interference signal have reduced the expense detecting.
Accompanying drawing explanation
Fig. 1 is schematic flow sheet of the present invention;
Fig. 2 is that multi-node collaboration disturbs identification schematic diagram;
Fig. 3 is Multistage Support Vector Machine algorithm block diagram.
Embodiment
Below, in conjunction with Fig. 1 to Fig. 3, the invention will be further described:
A kind of cooperation interference detection method based on SVMs, in a cooperative detection system with multinode, one of them node is as data fusion center, each cooperative node will carry out the extraction of energy measuring and characteristic parameter to the interference signal receiving, and number is transferred to data fusion center, then at data fusion Centroid, carry out data fusion and disturb identification, schematic diagram as shown in Figure 2.SVMs as shown in Figure 3, traditional support vector machine method is based on two class problems, in a kind of multipoint cooperative interference identification method based on SVMs, adopt " one-to-many " method, multiclass problem is converted into two class problems, for K binary classifier of K class problem structure, in the i class that i SVMs used, training sample is as positive classification samples, and using other sample as negative sample, last output is that maximum class of binary classifier output.
Cooperation disturb identification be exactly in different spatial a plurality of nodes respectively signal is carried out to Interference Detection and identification, then the recognition result of a plurality of nodes is carried out to information fusion, thereby obtains disturbing accurately recognition result, for hiding and resisting to disturb, provide foundation.This is a kind of application scenario of collaboration diversity in fact, and the collaborative process of a plurality of nodes can reach the effect of space diversity, thereby effectively resists multipath fading.
SVMs has very outstanding advantage in theory, is widely used area of pattern recognition.Oneself successfully solves many pattern recognition problem, has good performance.SVMs can carry out data processing under limited data sample, obtains the optimal solution under Given information, and then algorithm is changed into a secondary optimization problem again, finally obtains theoretic globally optimal solution, thereby has avoided local extremum problem.Just, due to its good performance, existing oneself is widely used in the aspects such as word identification, signal identification, image recognition.
As shown in Figure 1, the present invention includes following steps:
(1) extract the characteristic parameter of dissimilar interference signal, the characteristic parameter of interference signal is carrier wave factor coefficient, the smooth coefficient of average frequency spectrum, the white noise factor, frequency domain square coefficient of kurtosis and the frequency domain square coefficient of skewness, then the characteristic parameter of dissimilar interference signal is carried out to the computing of SVMs block algorithm, supported vector machine classification plane;
The starting point of block algorithm is that the row and column that is zero corresponding to Lagrange multiplier in puncture table can not exert an influence to final result.The sample that all adds fixed qty in each step of algorithm retains the sample with non-zero Lagrange multiplier remaining in previous step simultaneously, and M the poorest sample that does not meet Karush-Kuhn-Tucker condition; If do not meet not enough M of the sample number of Karush-Kuhn-Tucker condition, the sample not satisfying condition all added.Each subproblem adopts the result of a subproblem as initial value.Finally all non-zero Lagrange multipliers are all found, thereby solved original quadratic programming problem.
(2) frequency range that each node is all used the signal receiving is separately divided into a plurality of frequency sub-band, and the signal in each frequency sub-band is called subsignal;
The available frequency band of supposing each node receive channel is 500MHz~600MHz, and the sensitivity of receiver is 20MHz, and each is examined available frequency band is divided into 5 sections of every section of 20MHz.
(3) each node carries out energy measuring to subsignal separately respectively, and the energy detection results of subsignal separately being merged to the interference detection results that obtains each node, each cooperative node is transferred to data fusion center by interference detection results separately;
Each node is respectively by 500MHz~520MHz, 520MHz~540MHz, and 540MHz~560MHz, 560MHz~580MHz, the energy detection results of five frequency ranges of 580MHz~600MHz is stitched together, and obtains the interference detection results of 500MHz~600MHz frequency range.
(4) each node all carries out the extraction of characteristic parameter to the signal receiving separately, and each cooperative node is transferred to data fusion center by the characteristic parameter extracting separately;
The characteristic parameter of interference signal can be carrier wave factor coefficient, the smooth coefficient of average frequency spectrum, the white noise factor, frequency domain square coefficient of kurtosis and the frequency domain square coefficient of skewness.
(5) data fusion center, according to the interference detection results at the interference detection results of each cooperative node and data fusion center self, adopts K order criterion to merge the interference detection results of each node, obtains total interference detection results; If total interference detection results shows, there is not interference signal, behind official hour interval, forward step (2) to, otherwise forward step (6) to;
K order criterion: suppose that the interference strength information that always has N node is gathered together, wherein have in K node recognition band frequency and have interference to think that this band frequency exists interference, the intensity of interference is the average of each node interference strength information; Wherein, N is greater than 1 natural number, and K is less than or equal to the natural number of node number N;
(6) data fusion center is according to the characteristic parameter of each cooperative node transmission and the characteristic parameter at data fusion center self, adopt and select merging criterion to merge and obtain new characteristic parameter the characteristic parameter of each node, then utilize algorithm of support vector machine that new characteristic parameter is matched in SVMs classification plane, thus the classification of identification interference signal; After the interval time of regulation, forward step (2) to.
Select merging criterion, therefore namedly conceive, from a plurality of receiving nodes, select exactly a best node of performance to carry out Interference Detection.The thought of svm classifier comes from statistical theory, and it is that interval between two class data samples is maximum as discriminant classification plane that SVM constructs a hyperplane during for the basic thought of classification problem.In the process of disturbing identification, by interference signal characteristic parameter, SVM is trained, svm classifier plane is divided, then the characteristic parameter immediately obtaining is matched to ready-portioned classification plane, thereby realize the classification to interference signal.
Claims (1)
1. the cooperation interference detection method based on SVMs, based on a cooperative detection system with multinode, one of them node is as data fusion center, and all the other nodes, as cooperative node, is characterized in that comprising the following steps:
(1) extract the characteristic parameter of dissimilar interference signal, the characteristic parameter of interference signal is carrier wave factor coefficient, the smooth coefficient of average frequency spectrum, the white noise factor, frequency domain square coefficient of kurtosis and the frequency domain square coefficient of skewness, then the characteristic parameter of dissimilar interference signal is carried out to the computing of SVMs block algorithm, supported vector machine classification plane;
(2) frequency range that each node is all used the signal receiving is separately divided into a plurality of frequency sub-band, and the signal in each frequency sub-band is called subsignal;
(3) each node carries out energy measuring to subsignal separately respectively, and the energy detection results of subsignal separately being merged to the interference detection results that obtains each node, each cooperative node is transferred to data fusion center by interference detection results separately;
(4) each node all carries out the extraction of characteristic parameter to the signal receiving separately, and each cooperative node is transferred to data fusion center by the characteristic parameter extracting separately;
(5) data fusion center, according to the interference detection results at the interference detection results of each cooperative node and data fusion center self, adopts K order criterion to merge the interference detection results of each node, obtains total interference detection results; If total interference detection results shows, there is not interference signal, behind official hour interval, forward step (2) to, otherwise forward step (6) to;
(6) data fusion center is according to the characteristic parameter of each cooperative node transmission and the characteristic parameter at data fusion center self, adopt and select merging criterion to merge and obtain new characteristic parameter the characteristic parameter of each node, then utilize algorithm of support vector machine that new characteristic parameter is matched in SVMs classification plane, thus the classification of identification interference signal; After the interval time of regulation, forward step (2) to.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310737317.9A CN103647591B (en) | 2013-12-27 | 2013-12-27 | A kind of multipoint cooperative interference detection method based on SVMs |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310737317.9A CN103647591B (en) | 2013-12-27 | 2013-12-27 | A kind of multipoint cooperative interference detection method based on SVMs |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103647591A true CN103647591A (en) | 2014-03-19 |
CN103647591B CN103647591B (en) | 2016-06-08 |
Family
ID=50252759
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310737317.9A Active CN103647591B (en) | 2013-12-27 | 2013-12-27 | A kind of multipoint cooperative interference detection method based on SVMs |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103647591B (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2018006798A1 (en) * | 2016-07-05 | 2018-01-11 | 深圳大学 | Wireless sensing-based temperature detection method and system |
CN109031403A (en) * | 2018-08-20 | 2018-12-18 | 吉林大学 | A kind of borehole strain data exception extraction method based on S-K feature |
CN113938228A (en) * | 2021-12-16 | 2022-01-14 | 天津七一二通信广播股份有限公司 | Network domain cooperation interference position detection method and system |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060072680A1 (en) * | 2002-06-20 | 2006-04-06 | Yingmin Wang | Method of space-time encoding and decoding for frequency selective fading channel |
CN101521896A (en) * | 2009-04-17 | 2009-09-02 | 东南大学 | Cooperative spectrum sensing method based on likelihood ratio in cognitive radio |
CN102571230A (en) * | 2011-12-22 | 2012-07-11 | 中国人民解放军总参谋部第六十三研究所 | Distributed collaborative signal identification method based on blind estimation of higher order statistics and signal to noise ratio |
CN102869064A (en) * | 2012-07-27 | 2013-01-09 | 南京邮电大学 | Cluster modulation identification method based on feature level and strategy level combined fusion |
CN103117820A (en) * | 2013-01-23 | 2013-05-22 | 南通大学 | Reliability-based weighted collaboration spectrum detection method |
CN103166723A (en) * | 2013-03-18 | 2013-06-19 | 广东工业大学 | Cognitive radio sensor network spectrum sensing method based on extensible correlation function |
-
2013
- 2013-12-27 CN CN201310737317.9A patent/CN103647591B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060072680A1 (en) * | 2002-06-20 | 2006-04-06 | Yingmin Wang | Method of space-time encoding and decoding for frequency selective fading channel |
CN101521896A (en) * | 2009-04-17 | 2009-09-02 | 东南大学 | Cooperative spectrum sensing method based on likelihood ratio in cognitive radio |
CN102571230A (en) * | 2011-12-22 | 2012-07-11 | 中国人民解放军总参谋部第六十三研究所 | Distributed collaborative signal identification method based on blind estimation of higher order statistics and signal to noise ratio |
CN102869064A (en) * | 2012-07-27 | 2013-01-09 | 南京邮电大学 | Cluster modulation identification method based on feature level and strategy level combined fusion |
CN103117820A (en) * | 2013-01-23 | 2013-05-22 | 南通大学 | Reliability-based weighted collaboration spectrum detection method |
CN103166723A (en) * | 2013-03-18 | 2013-06-19 | 广东工业大学 | Cognitive radio sensor network spectrum sensing method based on extensible correlation function |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2018006798A1 (en) * | 2016-07-05 | 2018-01-11 | 深圳大学 | Wireless sensing-based temperature detection method and system |
CN109031403A (en) * | 2018-08-20 | 2018-12-18 | 吉林大学 | A kind of borehole strain data exception extraction method based on S-K feature |
CN109031403B (en) * | 2018-08-20 | 2019-07-26 | 吉林大学 | A kind of borehole strain data exception extraction method based on S-K feature |
CN113938228A (en) * | 2021-12-16 | 2022-01-14 | 天津七一二通信广播股份有限公司 | Network domain cooperation interference position detection method and system |
CN113938228B (en) * | 2021-12-16 | 2022-03-01 | 天津七一二通信广播股份有限公司 | Network domain cooperation interference position detection method and system |
Also Published As
Publication number | Publication date |
---|---|
CN103647591B (en) | 2016-06-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10158437B2 (en) | Wireless analysis apparatus and wireless analysis method | |
CN105844216B (en) | Detection and matching mechanism for recognizing handwritten letters by WiFi signals | |
CN108600135B (en) | Method for identifying signal modulation mode | |
Pastor et al. | Robust estimation of noise standard deviation in presence of signals with unknown distributions and occurrences | |
Zhang et al. | A spectrum sensing method based on signal feature and clustering algorithm in cognitive wireless multimedia sensor networks | |
CN106483514B (en) | Airplane motion mode identification method based on EEMD and support vector machine | |
CN102760444A (en) | Support vector machine based classification method of base-band time-domain voice-frequency signal | |
Li et al. | A deep convolutional network for multitype signal detection and classification in spectrogram | |
US9473257B1 (en) | Radio communication system utilizing a radio signal classifier | |
CN105807264A (en) | Method for detecting radar pulse repetition frequency and estimating inceptive pulse arrival time | |
CN106170139B (en) | A kind of frequency spectrum detecting method and system | |
CN103647591A (en) | Cooperative interference detection method based on support vector machine | |
Wang et al. | A cooperative spectrum sensing method based on signal decomposition and K-medoids algorithm | |
CN105743756B (en) | Frame detection method based on adaboost algorithm in WiFi system | |
Miao et al. | Application of one-class classification in NLOS identification of UWB positioning | |
TWI775186B (en) | Rf fingerprint signal processing device and rf fingerprint signal processing method | |
Dang et al. | Air gesture recognition using WLAN physical layer information | |
Chen et al. | Unlocking signal processing with image detection: A frequency hopping detection scheme for complex EMI environments using STFT and CenterNet | |
US9332454B2 (en) | Method, system and program product for detecting, quantifying and localizing of wireless interferers | |
Li et al. | Recovery of collided RFID tags with frequency drift on physical layer | |
CN103076595B (en) | Abnormal type identifying method for multivariate discrete radar emitter signal | |
KR101427149B1 (en) | A method of recognizing complex PRI modulation type of radar signals based on 2-layer SVM using exponential moving average | |
CN109587136A (en) | Radio frequency fingerprint feature extraction and identification method based on double maximum values | |
Guo et al. | Multi modal human action recognition for video content matching | |
CN104392252A (en) | Method and device of identifying radiation sources |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | 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 |