CN115426002B - Frequency hopping signal detection and parameter estimation method and system based on time-frequency analysis - Google Patents
Frequency hopping signal detection and parameter estimation method and system based on time-frequency analysis Download PDFInfo
- Publication number
- CN115426002B CN115426002B CN202211046676.5A CN202211046676A CN115426002B CN 115426002 B CN115426002 B CN 115426002B CN 202211046676 A CN202211046676 A CN 202211046676A CN 115426002 B CN115426002 B CN 115426002B
- Authority
- CN
- China
- Prior art keywords
- frequency
- time
- connected domain
- frequency hopping
- clustering
- 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.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B1/00—Details of transmission systems, not covered by a single one of groups H04B3/00 - H04B13/00; Details of transmission systems not characterised by the medium used for transmission
- H04B1/69—Spread spectrum techniques
- H04B1/713—Spread spectrum techniques using frequency hopping
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B1/00—Details of transmission systems, not covered by a single one of groups H04B3/00 - H04B13/00; Details of transmission systems not characterised by the medium used for transmission
- H04B1/69—Spread spectrum techniques
- H04B1/713—Spread spectrum techniques using frequency hopping
- H04B1/715—Interference-related aspects
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a frequency hopping signal detection and parameter estimation method and a system based on time-frequency analysis, which belong to the technical field of electromagnetic spectrum monitoring, wherein the method comprises the steps of carrying out double threshold segmentation processing and connected region marking processing on a time-frequency image of a frequency hopping signal to obtain characteristic information of the connected region image; coarse clustering is carried out on the connected domain according to the length characteristics of the connected domain image; carrying out subdivision clustering treatment on the coarse clustering result; and finally, estimating any parameter or parameters such as the jump speed, amplitude, bandwidth, residence time, a frequency hopping frequency set and the like according to the connected domain image. According to the method, the coarse clustering result is subjected to subdivision clustering through the initial coordinate characteristics of the time-frequency image, so that the time complexity is greatly reduced compared with the conventional clustering method, and the clustering efficiency is high; meanwhile, the temporary hopping period is subjected to remainder processing based on the first difference value, so that the method is applicable to all clustering scenes, the detection efficiency is further improved, and the frequency hopping network stations with the same hopping speed and different starting time are selected.
Description
Technical Field
The invention relates to the technical field of electromagnetic spectrum monitoring, in particular to a frequency hopping signal detection and parameter estimation method and system based on time-frequency analysis.
Background
The frequency hopping technique uses a pseudo-random code sequence to perform frequency shift keying so that the carrier frequency is continuously hopped to spread the frequency spectrum. The frequency of the frequency hopping signal continuously hops randomly along with time in a wider frequency band, and the anti-interference capability and the anti-interception capability are outstanding. Since the mobile communication channel environment is bad, various kinds of interference may be insufficient, and in order to resist the occurrence of interference of certain frequencies, it is one of effective methods to employ a frequency hopping signal.
Meanwhile, the radio communication technology is rapidly developed, the electromagnetic environment is complex and changeable, and the monitoring difficulty of frequency hopping signals is increased. In order to better remove background noise, the existing frequency hopping monitoring method firstly removes the background noise through morphological filtering, then carries out connected region marking processing on signals in a binary time-frequency diagram, acquires parameter information of each time-frequency component, finally clusters according to each parameter by adopting a K-means method or an improved K-means method, judges a clustering result, and judges that frequency hopping signals exist when the clustering result exceeds a set threshold. The above method has the following problems:
1. the traditional K-means clustering method has the time complexity of O (N is equal to K is equal to T), wherein N represents the data length, K represents the cluster number, T represents the iteration number, and the signal types are more and continuous iteration convergence is needed, so that the calculation complexity is high, the convergence is slow, and the efficiency is low.
2. Whether the frequency hopping signal exists or not is judged through a single threshold, interference cannot be filtered well, and the threshold value is difficult to control, so that the detection accuracy is greatly reduced.
In summary, the invention of the frequency hopping signal detection and parameter estimation method based on time-frequency analysis, which can rapidly and accurately screen the frequency hopping signal to improve the detection speed, the interception probability and the recognition accuracy of the frequency hopping signal, is very necessary.
Disclosure of Invention
The invention aims to overcome the problems in the prior art and provides a frequency hopping signal detection and parameter estimation method and system based on time-frequency analysis.
The aim of the invention is realized by the following technical scheme: a real-time blind detection method of frequency hopping signals comprises the following steps:
performing double threshold segmentation processing and connected region marking processing on a time-frequency image of a frequency hopping signal to obtain feature information of the connected region image, wherein the feature information comprises length features of the connected region image; preferably, the feature information of the connected domain image further includes any one or more of a height feature, a centroid feature, an amplitude feature, and a coordinate feature.
Coarse clustering is carried out on the connected domain according to the length characteristics of the connected domain image, coarse clustering results of different hopping speeds are obtained, and sorting of the hopping network platforms of different hopping speeds is achieved.
Specifically, the segmentation processing is to separate a foreground target (feature information of the connected domain image) of the time-frequency image from background noise, so that the feature information of the connected domain image is effectively extracted, and further, the clustering of frequency hopping signals is realized. The connected domain marks mark the adjacent connected domains, and then each connected domain in the time-frequency image is extracted.
In an example, a morphological filtering step is further included before the connected domain labeling process to further suppress background noise. Of course, morphological filtering may be performed before the time-frequency image is subjected to the double threshold segmentation process. According to the invention, the double threshold is adopted to segment the time-frequency image, and compared with the single threshold image segmentation, the double threshold can better filter interference, so that the signal detection accuracy is improved.
More specifically, the invention performs coarse clustering based on the length characteristics of the connected domain image, and the time complexity is O (N) 1 )，N 1 The data length is represented, namely the complexity of the clustering time only depends on the number of data points, the whole process does not need iteration, coarse clustering results with different jump speeds can be obtained quickly, and jump is realizedAnd the real-time blind detection of the frequency signals greatly improves the clustering calculation efficiency.
In one example, the method of the present invention further comprises the steps of:
collecting and preprocessing frequency hopping signals; specifically, IQ signal (modulation signal with quadrature phase) data st of external electromagnetic signal are collected in continuous time, and step length of the data st is N 2 Is stored in a time-frequency matrix to obtain a time-frequency image X 2 ；
In an example, the dual threshold segmentation process includes the sub-steps of:
image segmentation processing is performed based on a manual threshold: performing double threshold segmentation processing on the time-frequency image according to the issued first upper threshold and the issued first lower threshold; and/or the number of the groups of groups,
image segmentation processing is carried out based on the self-adaptive threshold: calculating the energy distribution number of the time-frequency image, acquiring a denoising threshold value according to the energy distribution number, taking the denoising threshold value as a second lower threshold, enabling the maximum pixel value to be a second upper threshold, and carrying out double-threshold segmentation processing on the time-frequency image according to the second upper threshold and the second lower threshold.
Preferably, the image segmentation processing is realized based on the two modes, in an example, the current double threshold value is determined by judging the threshold type (manual threshold or self-adaptive threshold), if the current threshold type is the manual threshold, the double threshold value segmentation processing is directly carried out on the time-frequency image according to the issued first upper threshold and the issued first lower threshold; and if the current threshold type is the self-adaptive threshold, performing double-threshold segmentation processing on the time-frequency image through the second upper threshold and the second lower threshold. In a preferred example, the second threshold (corresponding to the second upper threshold and the second lower threshold) adapted to the current image is obtained through the adaptive threshold, and the second double threshold is further manually adjusted to obtain the optimal first threshold (corresponding to the first upper threshold and the first lower threshold), so that the image segmentation accuracy is improved, the detection effect is optimal, and the user experience and the signal detection accuracy are synchronously improved.
In an example, the first upper threshold, the first lower threshold may be determined based on historical data or technician experience. The calculation formula for performing double threshold segmentation processing on the time-frequency image according to the first upper threshold and the first lower threshold is as follows:
wherein B is 1 Representing a double threshold segmentation result; i.e 1 ,m 1 Respectively represent time-frequency images X 2 Length and width values of (a); threshold (threshold) H Representing an upper threshold, in this example a first upper threshold; threshold (threshold) L Representing a lower threshold, in this example a first lower threshold.
In one example, the second threshold is calculated as:
th(j)＝j*0.1*P j＝1,2,3,…,n
wherein P represents the total energy; n, M are sum symbol upper bounds respectively representing time-frequency image X 2 Length maximum and width maximum of (a); i. m is sum symbol lower bound, respectively representing time-frequency image X 2 A length start value and a width start value of (a); th (j) represents a denoising threshold; j represents the number of iterations; c (j) represents the energy distribution number. The energy distribution number is subjected to a second difference process, and the inflection point is obtained, wherein the coefficient k (j) corresponding to the inflection point is used as a noise coefficient, the value of the corresponding th (j) is used as a second threshold lower limit, and the second threshold upper limit is set to 255.
In an example, the coarse clustering of the connected domain according to the length feature of the connected domain image includes the following sub-steps:
S321: sequencing the length characteristics of the time-frequency images to obtain a length sequence and a corresponding connected domain label sequence;
s322: performing differential operation on the length sequence to obtain a differential length sequence;
s323: and comparing the differential length sequence with a first threshold value, and combining the connected domain labels corresponding to the differential length sequence to realize coarse clustering. Wherein if the absolute error of the length sequence estimation is e 1 The first threshold is 3e 1 。
More specifically, step S323 specifically includes the following sub-steps:
comparing whether the current differential length sequence is larger than a first threshold value, if so, storing the sequence number of the current differential length sequence into a label length array label len In, label len (n)—label len The tag sequences corresponding to (n+1) -1 are classified into the same cluster, and n represents the serial number of the tag length array;
repeating the steps S321-S323 until all the length features are polled to obtain a coarse clustering result, and sorting the frequency hopping signals with different frequency hopping speeds, namely sorting the frequency hopping network stations with different frequency hopping speeds.
In an example, the method further comprises a sub-clustering step comprising the sub-steps of:
according to the initial coordinate feature of the connected domain image, the coarse clustering result is subjected to subdivision clustering processing, and the initial coordinate feature represents the starting time of the frequency hopping signal, so that the frequency hopping network stations with the same frequency hopping speed and different starting time can be sorted according to the subdivision clustering result. In this example, the feature information of the connected domain image includes coordinate features including a start coordinate feature.
In an example, the subdivision clustering process includes the sub-steps of:
s331: sequencing the coarse clustering results according to the initial coordinate features of the time-frequency images to obtain a coordinate feature sequence and a corresponding connected domain label sequence;
s332: calculating the average value of the differential results of the initial coordinate features to obtain a temporary jump period;
s333: selecting any initial coordinate feature as a reference value, calculating a first difference value between the current initial coordinate feature and the reference value, and performing residual calculation on the first difference value and the temporary jump period to obtain a residual result;
s334: and comparing the residual result with a second threshold value, or comparing the second difference value of the temporary jump period and the residual result with the second threshold value, and combining the connected domain labels corresponding to the coordinate feature sequences to realize subdivision clustering. Wherein if the absolute error of the initial coordinate estimation is e 2 The second threshold is 3e 2 。
More specifically, step S334 specifically includes the following sub-steps:
s3341: comparing whether the residual result is smaller than or equal to a second threshold value or whether the second difference value between the temporary jump period and the residual result is smaller than or equal to the second threshold value, if so, storing the label corresponding to the initial coordinate feature into a new cluster group, otherwise, storing the label into a temporary initial coordinate array;
S3342: repeating the steps until all the initial coordinate features are polled;
s3343: assigning the temporary initial coordinate feature to the initial coordinate feature;
s3344: repeating steps S333 to S3343 until the temporary start coordinate feature is 0;
s3345: repeating the steps S331 to S3344 until the rough clustering result is polled, and obtaining the subdivision clustering result.
In an example, the invention further includes a result reporting step, configured to report real-time blind detection results of the frequency hopping signals, including coarse clustering results of the frequency hopping signals of different frequency hopping speeds (classification results of the frequency hopping network stations of different frequency hopping speeds) and/or fine clustering results of different time hopping speeds (classification results of the frequency hopping network stations of different time hopping speeds).
It should be further noted that the technical features corresponding to the examples of the real-time blind detection method may be combined with each other or replaced to form a new technical scheme.
The invention also comprises a method for estimating real-time parameters of the frequency hopping signals, the method is implemented based on the method for real-time blind detection of the frequency hopping signals formed by any one or more examples, at the moment, the characteristic information of the connected domain image comprises length characteristics, height characteristics, centroid characteristics, amplitude characteristics and coordinate characteristics, the coordinate characteristics comprise initial coordinate characteristics, and the method for estimating the parameters comprises the following steps:
Estimating the residence time of the frequency hopping signal according to the length characteristics of the connected domain image in each group of classification results; and/or the number of the groups of groups,
estimating the jumping period and the jumping speed of the frequency jumping signals according to the initial coordinate characteristics of the connected domain images in the classification results of each group; and/or the number of the groups of groups,
estimating the bandwidth of the frequency hopping signal according to the height characteristics of the connected domain image in each group of classification results; and/or the number of the groups of groups,
estimating the amplitude of the frequency hopping signal according to the amplitude characteristics of the connected domain image in each group of classification results; and/or the number of the groups of groups,
and estimating a frequency set of the frequency hopping signal according to the centroid characteristics of the connected domain image in each group of classification results.
It should be further noted that the technical features corresponding to the examples above may be combined with each other or replaced to form a new technical solution.
In one example, the estimated calculation formula for residence time is:
wherein T is r Representing the estimated residence time; length characteristic sequence l k Is (l) 1 ,l 2 ,l 3 ,…,l k ) K is a sequence number.
In an example, the estimated calculation formula of the jump period and the jump speed is:
V h ＝1/T h
wherein T is h To represent an estimated hop period; v (V) h Representing the estimated jump rate; the initial coordinate feature sequence is { (xs) 1 ,ys 1 ),(xs 2 ,ys 2 ),…,(xs k ,ys k ) And k is a sequence number.
In one example, the bandwidth estimate is calculated as:
wherein B represents the estimated bandwidth; f (f) s Representing the sampling rate; height characteristic sequence b k Is (b) 1 ,b 2 ,b 3 ,…,b k ) K is a sequence number.
In an example, in the amplitude estimation step, the amplitude of each hop signal, i.e. the value of the centroid, is at the corresponding position in the time-frequency matrix.
In one example, the predictive calculation formula for the frequency meter is:
f k ＝x k f s /N 2
wherein f k Representing the estimated frequency hopping frequency; f (f) s Representing the sampling rate; n (N) 2 Is the short-time Fourier transform step length; centroid feature sequence x k Is { (x) 1 ,y 1 ),(x 2 ,y 2 ),…,(x k ,y k ) And k is a sequence number.
In an example, the invention further includes a result reporting step, configured to report a real-time blind detection result and a parameter estimation result of the frequency hopping signal, where the real-time blind detection result includes coarse clustering results of the frequency hopping signals with different frequency hopping speeds and/or fine clustering results with the same frequency hopping speed but different time to take off; the parameter estimation result comprises the parameter estimation result of any parameter or a plurality of parameter combinations such as jump speed, amplitude, bandwidth, residence time, frequency jump frequency set and the like.
It should be further noted that the technical features corresponding to the examples of the real-time parameter estimation method may be combined with each other or replaced to form a new technical scheme.
The invention also comprises a frequency hopping signal real-time blind detection system which has the same technical conception as the frequency hopping signal real-time blind detection method, and the system comprises:
The image processing unit is used for carrying out double threshold segmentation processing and connected region marking processing on the time-frequency image of the frequency hopping signal to obtain characteristic information of the connected region image, wherein the characteristic information comprises length characteristics;
and the signal blind detection unit is used for carrying out coarse clustering on the connected domain according to the length characteristics of the connected domain image to obtain coarse clustering results with different jump speeds.
In an example, the system further comprises a subdivision and cluster unit, which is used for performing subdivision and cluster processing on the coarse cluster result according to the initial coordinate characteristics of the connected domain image, so as to realize sorting of the network platforms with the same jump speed and different jump time.
Preferably, the image processing unit, the signal blind detection unit and the subdivision clustering unit can execute the frequency hopping signal real-time blind detection method formed by any one or more of the above examples.
In an example, the system further includes a data acquisition and preprocessing unit, configured to acquire and preprocess the frequency hopping signal, so as to obtain a time-frequency image of the frequency hopping signal.
In an example, the system further includes a reporting unit, configured to report a real-time blind detection result of the frequency hopping signal, where the real-time blind detection result includes coarse clustering results of the frequency hopping signals with different frequency hopping speeds and/or fine clustering results of the frequency hopping signals with the same frequency hopping speed and different starting times.
In an example, the system further comprises a display unit for displaying the real-time blind detection result.
It should be further noted that the technical features corresponding to the examples of the real-time blind detection system may be combined with each other or replaced to form a new technical scheme.
The invention also comprises a frequency hopping signal real-time parameter estimation system which has the same technical conception as the frequency hopping signal real-time parameter estimation method, and comprises a unit structure in the real-time blind detection system formed by any one or more examples, wherein at the moment, the characteristic information of the connected domain image obtained based on the image processing unit comprises a length characteristic, a height characteristic, a mass center characteristic, an amplitude characteristic and a coordinate characteristic, the coordinate characteristic comprises an initial coordinate characteristic, and the parameter estimation system comprises:
the residence time estimation unit is used for estimating the residence time of the frequency hopping signal according to the length characteristics of the connected domain image in each group of classification results; and/or the number of the groups of groups,
the jump period and jump speed estimating unit is used for estimating the jump period and jump speed of the frequency-hopping signal according to the initial coordinate characteristics of the connected domain image in the classification results of each group; and/or the number of the groups of groups,
the bandwidth estimation unit is used for estimating the bandwidth of the frequency hopping signal according to the height characteristics of the connected domain image in each group of classification results; and/or the number of the groups of groups,
The amplitude estimation unit is used for estimating the amplitude of the frequency hopping signal according to the amplitude characteristics of the connected domain image in each group of classification results; and/or the number of the groups of groups,
and the frequency set estimation unit is used for estimating the frequency set of the frequency hopping signal according to the centroid characteristics of the connected domain image in each group of classification results.
Preferably, the system comprises a dwell time estimation unit, a hop period and hop speed estimation unit, a bandwidth estimation unit, an amplitude estimation unit and a frequency set estimation unit.
In an example, the system further comprises a reporting unit, configured to report a real-time blind detection result and a parameter estimation result of the frequency hopping signal, where the real-time blind detection result includes coarse clustering results of the frequency hopping signals with different frequency hopping speeds and/or sub-clustering results of the frequency hopping signals with the same frequency hopping speed and different time to take off; the parameter estimation result comprises the parameter estimation result of any parameter or a plurality of parameter combinations such as jump speed, amplitude, bandwidth, residence time, frequency jump frequency set and the like.
In an example, the system further comprises a display unit for displaying real-time blind detection results and/or parameter estimation results.
It should be further noted that the technical features corresponding to the examples of the above-mentioned frequency hopping signal real-time parameter estimation system may be combined with each other or replaced to form a new technical scheme.
The invention also includes a storage medium having stored thereon computer instructions that, when executed, perform the steps of the method for real-time blind detection of a frequency-hopped signal formed by any one or more of the examples set forth above, and/or perform the steps of the method for real-time parameter estimation of a frequency-hopped signal formed by any one or more of the examples set forth above.
The invention also comprises a terminal which comprises a memory and a processor, wherein the memory stores computer instructions capable of being operated on the processor, and the processor executes the steps of the frequency hopping signal real-time blind detection method formed by any one or more examples and/or executes the steps of the frequency hopping signal real-time parameter estimation method formed by any one or more examples.
Compared with the prior art, the invention has the beneficial effects that:
1. in an example, the time-frequency image is segmented by adopting the double threshold, so that interference can be filtered better, and the signal detection accuracy is improved; meanwhile, coarse clustering is performed based on the length characteristics of the connected domain images, so that the time complexity is greatly reduced, convergence iteration is not needed in the whole clustering process, coarse clustering results with different hopping speeds can be obtained quickly, namely, sorting of the frequency hopping network stations with different hopping speeds is realized, real-time blind detection of frequency hopping signals is realized, and meanwhile, the clustering calculation efficiency is greatly improved.
2. In an example, by combining a dual-threshold issuing mode (manual threshold calculation) with an adaptive calculation dual-threshold mode (automatic threshold calculation), the dual threshold to be issued can be further adjusted based on the dual-threshold obtained by the adaptive calculation, so as to obtain an optimal dual-threshold adapted to the current detection environment, thereby improving the image segmentation accuracy, optimizing the detection effect, and synchronously improving the user experience and the signal detection accuracy.
3. In an example, the invention realizes the sorting of the network stations with the same jump speed and different jump time based on the initial coordinate characteristics of the connected domain image, has small operand and is effective in real time, and further expands the signal blind detection function.
4. In an example, the invention carries out subdivision clustering treatment on the coarse clustering result through the initial coordinate characteristic of the time-frequency image, and compared with the existing clustering method, the time complexity is greatly reduced, and the clustering efficiency is high; meanwhile, the temporary hopping period is subjected to remainder processing based on the first difference value, so that the method is applicable to all clustering scenes, the detection efficiency is further improved, and the frequency hopping network stations with the same hopping speed and different starting time are selected.
5. In an example, the invention realizes accurate estimation of frequency hopping parameters based on the length features, the height features, the centroid features, the amplitude features and the coordinate features of the connected domain images in the classification results of each group, and provides a new idea of frequency hopping parameter estimation, including frequency hopping, amplitude, bandwidth, residence time, frequency set and the like.
Drawings
The following detailed description of the present invention is further detailed in conjunction with the accompanying drawings, which are provided to provide a further understanding of the present application, and in which like reference numerals are used to designate like or similar parts throughout the several views, and in which the illustrative examples and descriptions thereof are used to explain the present application and are not meant to be unduly limiting.
FIG. 1 is a flow chart of a method in an example of the invention;
FIG. 2 is a flow chart of a preferred method in an example of the invention;
FIG. 3 is a schematic diagram of a connected domain image with image binarization segmentation performed according to an example of the present invention;
FIG. 4 is a schematic diagram of a connected domain image with morphological filtering according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a 4-contiguous connected domain signature in an example of the present invention;
FIG. 6 is a schematic diagram of an 8-neighbor connected domain marker in an example of the present invention;
FIG. 7 is a schematic diagram of a connected domain image with connected domain labeling according to an example of the present invention;
FIG. 8 is a graph illustrating inter-frequency spacing performance in an example of the present invention;
FIG. 9 is a schematic diagram of residence time performance in an example of the present invention;
FIG. 10 is a schematic diagram illustrating a time-of-use performance analysis in an example of the present invention.
In the figure: 1-first connected domain, 2-second connected domain, 3-third connected domain, 4-fourth connected domain, 5-fifth connected domain, 6-sixth connected domain, 7-seventh connected domain, 8-eighth connected domain, 9-ninth connected domain, 10-tenth connected domain, 11-eleventh connected domain, 12-twelfth connected domain, 13-thirteenth connected domain, 14-fourteenth connected domain, 15-fifteenth connected domain, 16-sixteenth connected domain, 17-seventeenth connected domain, 18-eighteenth connected domain, 19-nineteenth connected domain, 20-twenty-fourth connected domain, 21-twenty-second connected domain, 22-twenty-second connected domain, 23-twenty-third connected domain, 24-twenty-fourth connected domain, 25-twenty-fifth connected domain, 26-twenty-sixth connected domain, 27-twenty-seventh connected domain, 28-twenty-eighth connected domain, 29-twenty-ninth connected domain, 30-thirty connected domain, 31-thirty connected domain, 32-thirty connected domain, 33-thirty connected domain, 33-twenty-fourth connected domain, 35-thirty connected domain, 35-thirty-eighth connected domain, 35-thirty connected domain, and thirty-six connected domain.
Detailed Description
The following description of the embodiments of the present invention will be made apparent and fully understood from the accompanying drawings, in which some, but not all embodiments of the invention are shown. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In the description of the present invention, it should be noted that directions or positional relationships indicated as being "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. are based on the directions or positional relationships described in the drawings are merely for convenience of describing the present invention and simplifying the description, and do not indicate or imply that the apparatus or elements to be referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention. Further, ordinal words (e.g., "first and second," "first through fourth," etc.) are used to distinguish between objects, and are not limited to this order, but rather are not to be construed to indicate or imply relative importance.
In the description of the present invention, it should be noted that, unless explicitly specified and limited otherwise, terms "mounted," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
In addition, the technical features of the different embodiments of the present invention described below may be combined with each other as long as they do not collide with each other.
In an example, as shown in fig. 1, a method for real-time blind detection of a frequency hopping signal specifically includes the following steps:
step 1: performing double threshold segmentation processing and connected region marking processing on a time-frequency image of a frequency hopping signal to obtain feature information of the connected region image, wherein the feature information comprises length features;
step 2: and performing coarse clustering on the connected domain according to the length characteristics of the connected domain image to obtain coarse clustering results with different jump speeds.
According to the invention, the double threshold is adopted to segment the time-frequency image, and compared with the single threshold image segmentation, the double threshold can better filter interference, so that the signal detection accuracy is improved. Meanwhile, coarse clustering is performed based on the length characteristics of the connected domain images, the time complexity is only dependent on the number of data points, iteration is not needed in the whole process, coarse clustering results with different hopping speeds can be obtained rapidly, real-time blind detection of frequency hopping signals is achieved, and meanwhile clustering calculation efficiency is improved greatly.
As a preferred embodiment, the method for real-time blind detection and parameter estimation of a frequency hopping signal is obtained by combining the method for real-time blind detection and parameter estimation of a frequency hopping signal, as shown in fig. 2, and specifically comprises the following steps:
S1: data acquisition and pretreatment: electromagnetic data are collected and preprocessed, and the method comprises the following substeps:
s11: collecting external electromagnetic signal IQ data st in a continuous 1s time;
specifically, a complex CNI signal environment simulator (CNI-SIM-1000) is adopted to transmit frequency hopping signals and interference signals, and a 20 MHz-8 GHz vertical polarization monitoring antenna is used to radiate the signals, so that the acquisition of the signals is realized; the method comprises the steps of performing IQ data acquisition on external electromagnetic signals by adopting a CS-8000AT digital broadband monitoring receiver of Chengdu star-world electronic technology limited company, wherein the sampling frequency is 204.8MHz; the data processing and interface display are carried out by adopting CSCCRMS6.0 software of Chengdu Zhongshi electronic technology Co. Meanwhile, due to limitation of the picture space and better observation effect, in this embodiment, the data display results in the 112ms window are shown in fig. 3 to 4 and fig. 7, and the rest are the data display results in the 1s window, which will not be described again.
S12: step length of data st is N 2 Is stored in a time-frequency matrix.
S2: image segmentation, morphological filtering and connected domain marking are carried out on the time-frequency matrix by adopting an image processing technology;
s21: determining threshold classes, based on upper threshold (threshold) of different threshold classes H ) And a lower threshold (threshold) L ) Image double-threshold segmentation is carried out, and the embodiment adopts an automatic threshold;
s211: based on an energy statistics algorithm, an adaptive threshold is calculated auto ；
S212: based on the obtained threshold auto The image binarization segmentation (double threshold segmentation processing is performed on the time-frequency image of the frequency hopping signal) is completed, as shown in fig. 3, and the white frame in fig. 3 is a connected domain.
S22: selecting structural elements to perform morphological filtering on the segmented image, wherein the obtained connected domain image is shown in fig. 4;
s23: for the filtered time-frequency matrix B 2 Performing connected domain marking, namely 4 adjacent connected domain marking as shown in fig. 5, wherein a current connected domain is represented by a black filled rectangular box; rectangular square filled by oblique linesThe boxes represent contiguous connected domains; as shown in fig. 6, which shows an 8-adjacent connected domain mark or the like, the 8-adjacent connected domain mark employed in the present embodiment is marked 41 connected domains in total, and as shown in fig. 7, includes a first connected domain 1, a second connected domain 2, a third connected domain 3, a fourth connected domain 4, a fifth connected domain 5, a sixth connected domain 6, a seventh connected domain 7, an eighth connected domain 8, a ninth connected domain 9, a tenth connected domain 10, an eleventh connected domain 11, a twelfth connected domain 12, a thirteenth connected domain 13, a fourteenth connected domain 14, a fifteenth connected domain 15, a sixteenth connected domain 16, a seventeenth connected domain 17, an eighteenth connected domain 18, a nineteenth connected domain 19, a twenty-fifth connected domain 20, a twenty-second connected domain 22, a twenty-third connected domain 23, a twenty-fourth connected domain 24, a twenty-fifth connected domain 25, a twenty-sixth connected domain 26, a twenty-seventh connected domain 27, an eighth connected domain 28, a twenty-ninth connected domain 29, a thirty-fifth connected domain 30, a thirty-eighth connected domain 31, a thirty-seventh connected domain 31, a thirty-eighth connected domain 33, a thirty-ninth connected domain 34, a thirty-seventh connected domain 35. It should be further noted that fig. 3-4 and fig. 7 are simulation diagrams, and are not used for limiting the protection scope of the technical scheme of the present application; in fig. 7, a part of the connected domains such as the fourth connected domain 4, the seventeenth connected domain 17, the eighteenth connected domain 18, the twenty-first connected domain 21, the twenty-second connected domain 22, and the like are not visually recognized.
S3: blind detection of frequency hopping signals is realized according to the length characteristics of the connected domain image, sorting of frequency hopping network stations with different frequency hopping speeds is realized according to the coordinate characteristics of the connected domain image, and the method specifically comprises the following sub-steps:
s31: and extracting feature information of the connected domain image, wherein the feature information comprises a length feature, a height feature, a centroid feature, an amplitude feature and a coordinate feature, and the coordinate feature comprises a starting coordinate feature.
S32: coarse clustering is carried out on the connected domain according to the length characteristics of the image, and the method specifically comprises the following substeps:
s321: sequencing according to the length characteristic length to form a length sequence Len and a tag sequence Label of the connected domain mark;
s322: the length sequence Len is subjected to differential operation, and in this embodiment, a differential length sequence Dlen with a length of 40 is obtained.
S323: selecting a first threshold value l When Dlen (k)>threshold l When the sequence number k of the differential length sequence Dlen is stored into a label length array label len In (a) and (b);
s324: label is put into len (n)—label len The tag sequences corresponding to (n+1) -1 are classified into the same cluster;
s325: step S324 is repeated until all labels label are clustered, and the coarse clustering result in this embodiment is shown in table 1:
TABLE 1 coarse clustering results Table
S33: based on the step S32, according to the initial coordinate feature S-corodinate of the image, the coarse clustering result is subdivided and clustered again, and the method specifically comprises the following substeps:
S331: sequencing the clustering results in the step S32 according to the initial coordinate features S-corodinate;
s332: the differential result of the initial coordinate feature s-corodinate is averaged to obtain a temporary jump period t c ；
S333: selecting a reference value basevalue, wherein the difference value between the initial coordinate array s-pivot (n) and the basevalue is tempvalue;
s334: selecting a second threshold value s If the residual result of the difference value tempvalue and the temporary jump period is less than or equal to threshold s Storing the label corresponding to the initial coordinate feature s-coordinate into a new clustering group; otherwise, storing the s-corodinate (n) into a temporary initial coordinate array s-corodinate-temp;
s335: repeating S334 until the initial coordinate feature polling is completed;
s336: assigning s-chord-temp to s-chord;
s337: repeating S333-S336 until the size of the S-cordinate-temp is 0;
s338: and repeating S331-S337 until the grouping and polling of the coarse clustering result of S32 are completed. In this example, a schematic diagram of the subdivision clusters is shown in table 2:
TABLE 2 subdivision and clustering result table
| Clustering grouping | Tag of connected domain |
| 1 | {6,9,12,15,19,24,27,30,33,36} |
| 2 | {7,10,13,16,20,25,28,31,34,37} |
| 3 | {5,8,11,14,23,26,29,32,35,38} |
S34: judging whether a frequency hopping signal exists or not according to a clustering result (a coarse clustering result and/or a subdivision clustering result);
s4: frequency hopping parameter estimation, including frequency hopping estimation, amplitude estimation, bandwidth estimation, dwell time estimation, frequency hopping set estimation, etc.; in this example, the final parameter estimation results are shown in tables 3 and 4:
Table 3 parameter estimation results table for the first time period
Table 4 parameter estimation results table for the second period
S5: reporting the result: and reporting the frequency hopping signal blind detection result and the parameter estimation result and displaying the result in real time.
S51: packing the number of the frequency hopping signals, relevant signal parameters, frequency hopping patterns and other information;
s52: reporting to a display control interface.
In this embodiment, the blind detection performance and the parameter estimation performance of the frequency hopping signal are analyzed, and first, a single signal is tested and verified, and performances such as a detectable frequency hopping signal inter-frequency interval, a residence time estimation error, a frequency set capturing probability and the like are specifically analyzed.
For the inter-frequency interval performance, signals of 1000hop/s are set, the frequency ranges are 7200MHz to 7260MHz, the frequency points of the frequency sets are respectively set to 64, 128, 256 and 300, and the frequencies are generated in an increasing mode in the set range, so that the inter-frequency intervals are respectively: 952.38kHz,472.44kHz,235.29kHz,200.67kHz. As shown in fig. 8, the abscissa in the figure is the inter-frequency spacing in kHz; the ordinate is the detection accuracy, and when the inter-frequency interval is greater than or equal to 9 pixels, the signal detection accuracy is greater than 95%.
For the residence time performance, the frequency range 7200 MHz-7260 MHz is set, the frequency point number of the frequency set is set to 256, and the hop period is respectively set to 0.1s, 0.01s, 5ms, 4ms, 2.5ms, 2ms and 1ms, so that the residence time is approximately equal to the hop period because the CNI-SIM-1000 has fast frequency cutting speed and negligible time. As shown in fig. 9, the abscissa in the figure is the number of tests; the ordinate is the estimated error in microseconds, and as can be seen from fig. 9, the difference between the estimated dwell time and the actual dwell time is less than one time pixel, which is better.
For performance analysis of frequency estimation errors, a single 1000hop/s signal is set, the frequency ranges are 7200MHz to 7260MHz, the frequency points of the frequency sets are respectively set to 16 and 32, and the obtained estimated frequency sets are shown in Table 5:
TABLE 5 frequency set look-up table
As can be seen from Table 5, the estimation error of the frequency is substantially 13kHz and below, and the estimation error is substantially within one pixel range, and the performance is good.
In terms of the capturing probability performance of the frequency set, a single signal of 1000hop/s is set, the frequency ranges of 7200MHz to 7260MHz, the frequency points of the frequency set are respectively set to 32, 64, 128 and 256, and the test results are shown in Table 6:
TABLE 6 frequency Point intercept test results Table
| Actual frequency point number | Capturing frequency points | Probability of frequency set |
| 32 | 32 | 100％ |
| 64 | 64 | 100％ |
| 128 | 128 | 100％ |
| 256 | 256 | 100％ |
As can be seen from table 6, the frequency set acquisition probability is as high as 100%.
The screening capability of the net platform of the method is tested. For sorting network platforms with different hopping speeds, signal hopping speeds of 1000hop/s, 200hop/s and 100hop/s are set, the frequency ranges are 7200 MHz-7260 MHz, the frequency points are 130, 58 and 30 respectively, and the three hopping frequencies are in a bandwidth range, and the frequency is staggered but not overlapped. As can be seen from table 3, the present embodiment can sort 3 or more synchronous frequency hopping network stations simultaneously.
For the same-hop asynchronous network table sorting, three hopping signals with the hopping speed of 400hop/s are set, the frequency sets of the three hopping signals are the same, the hopping patterns are different, and the starting time of the three hopping signals is respectively set to be 0s, 0.9ms and 1.8ms. As can be seen from table 4, this embodiment can accurately sort 3 or more asynchronous orthogonal frequency hopping network stations (frequency hopping network stations with the same hopping speed and different starting time) based on the asynchronous non-orthogonality.
The time performance analysis of the method proposed by the embodiment is as shown in fig. 10, wherein the abscissa in the figure is a data window, and the unit is ms; the ordinate is in ms when the algorithm is used. As can be seen from fig. 10, when the data window is smaller than 1s, the time used in the method is smaller than 1s, i.e. the electromagnetic signal within 1s can be detected and the parameters can be estimated in real time within 1 s; when the data window is 3s, the time used for the method is 1.8s, which is far lower than the time of the data window. In summary, the method provided by the embodiment can quickly, effectively and accurately detect the frequency hopping signal and perform parameter estimation on the frequency hopping signal.
Furthermore, the method for real-time blind detection and parameter estimation of the preferred frequency hopping signal is widely applied to CS-100A, CS-200A, CS-200B, CS-300A, CS-900A and other series monitoring direction-finding systems and CS-8000A, CS-8000AT, CS-8000B, CS-8000BT, CS-18GB, CS-26.5GB and other series digital broadband monitoring receiver equipment, and has good detection effect and parameter estimation effect.
Based on the same inventive concept of the frequency hopping signal real-time blind detection and parameter estimation method, a frequency hopping signal real-time blind detection system and a frequency hopping signal real-time parameter estimation system are combined to obtain the frequency hopping signal real-time blind detection and parameter estimation system, and the system comprises:
the data acquisition and preprocessing unit is used for acquiring and preprocessing the frequency hopping signals so as to obtain time-frequency images of the frequency hopping signals;
the image processing unit is used for carrying out double threshold segmentation processing and connected region marking processing on the time-frequency image of the frequency hopping signal to obtain characteristic information of the connected region image, wherein the characteristic information comprises length characteristics;
the signal blind detection unit is used for performing coarse clustering on the connected domain according to the length characteristics of the connected domain image to obtain coarse clustering results with different jump speeds;
the subdivision clustering unit is used for carrying out subdivision clustering treatment on the coarse clustering result according to the initial coordinate characteristics of the connected domain image, so as to realize the sorting of the network stations with the same jump speed and different jump time;
the residence time estimation unit is used for estimating the residence time of the frequency hopping signal according to the length characteristics of the connected domain image in each group of classification results;
the jump period and jump speed estimating unit is used for estimating the jump period and jump speed of the frequency-hopping signal according to the initial coordinate characteristics of the connected domain image in the classification results of each group;
The bandwidth estimation unit is used for estimating the bandwidth of the frequency hopping signal according to the height characteristics of the connected domain image in each group of classification results;
the amplitude estimation unit is used for estimating the amplitude of the frequency hopping signal according to the amplitude characteristics of the connected domain image in each group of classification results;
the frequency set estimation unit is used for estimating the frequency set of the frequency hopping signal according to the centroid characteristics of the connected domain image in each group of classification results;
the reporting unit is used for reporting real-time blind detection results and parameter estimation results of the frequency hopping signals;
and the display unit is used for displaying the real-time blind detection result and/or the parameter estimation result.
The application also comprises a storage medium, which has the same inventive concept as the method for real-time blind detection and parameter estimation of the frequency hopping signal, and computer instructions are stored on the storage medium, and the computer instructions execute the steps of the method for real-time blind detection and parameter estimation of the frequency hopping signal when running.
Based on such understanding, the technical solution of the present embodiment may be essentially or a part contributing to the prior art or a part of the technical solution may be embodied in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The application also comprises a terminal which has the same inventive concept as the method for real-time blind detection and parameter estimation of the frequency hopping signal, and comprises a memory and a processor, wherein the memory stores computer instructions which can be operated on the processor, and the processor executes the steps of the method for real-time blind detection and parameter estimation of the frequency hopping signal when the processor operates the computer instructions. The processor may be a single or multi-core central processing unit or a specific integrated circuit, or one or more integrated circuits configured to implement the invention.
The functional units in the embodiments provided in the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.
The foregoing detailed description of the invention is provided for illustration, and it is not to be construed that the detailed description of the invention is limited to only those illustration, but that several simple deductions and substitutions can be made by those skilled in the art without departing from the spirit of the invention, and are to be considered as falling within the scope of the invention.
Claims (10)
1. The frequency hopping signal detection and parameter estimation method based on time-frequency analysis is characterized by comprising the following steps of: which comprises the following steps:
Performing double threshold segmentation processing and connected region marking processing on a time-frequency image of a frequency hopping signal to obtain feature information of the connected region image, wherein the feature information comprises length features;
coarse clustering is carried out on the connected domain according to the length characteristics of the connected domain image, so that coarse clustering results with different jump speeds are obtained;
the method further comprises a sub-clustering step comprising the sub-steps of:
according to the initial coordinate characteristics of the connected domain images, performing subdivision clustering treatment on the coarse clustering result, and further realizing sorting of network stations with the same jump speed and different jump time;
the method further comprises a parameter estimation step, which specifically comprises the following sub-steps:
estimating the residence time of the frequency hopping signal according to the length characteristics of the connected domain image in each group of classification results; and/or the number of the groups of groups,
estimating the jumping period and the jumping speed of the frequency jumping signals according to the initial coordinate characteristics of the connected domain images in the classification results of each group; and/or the number of the groups of groups,
estimating the bandwidth of the frequency hopping signal according to the height characteristics of the connected domain image in each group of classification results; and/or the number of the groups of groups,
estimating the amplitude of the frequency hopping signal according to the amplitude characteristics of the connected domain image in each group of classification results; and/or the number of the groups of groups,
estimating a frequency set of the frequency hopping signal according to the centroid characteristics of the connected domain image in each group of classification results;
Finally, coarse clustering results of frequency hopping signals with different hopping speeds and/or sub-clustering results with the same hopping speed but different starting time are reported, and parameter estimation results of any parameter or combination of a plurality of parameters of the hopping speed, amplitude, bandwidth, residence time and frequency hopping frequency set are reported;
the subdivision clustering process comprises the following substeps:
sequencing the coarse clustering results according to the initial coordinate features of the time-frequency images to obtain a coordinate feature sequence and a corresponding connected domain label sequence;
calculating the average value of the differential results of the initial coordinate features to obtain a temporary jump period;
selecting any initial coordinate feature as a reference value, calculating a first difference value between the current initial coordinate feature and the reference value, and performing residual calculation on the first difference value and the temporary jump period to obtain a residual result;
and comparing the residual result with a second threshold value, or comparing the second difference value of the temporary jump period and the residual result with the second threshold value, and combining the connected domain labels corresponding to the coordinate feature sequences to realize subdivision clustering.
2. The method for detecting and estimating frequency hopping signals based on time-frequency analysis according to claim 1, wherein: the double threshold segmentation process comprises the sub-steps of:
Performing double threshold segmentation processing on the time-frequency image according to the issued first upper threshold and the issued first lower threshold; and/or the number of the groups of groups,
calculating the energy distribution number of the time-frequency image, acquiring a denoising threshold value according to the energy distribution number, taking the denoising threshold value as a second lower threshold, enabling the maximum pixel value to be a second upper threshold, and carrying out double-threshold segmentation processing on the time-frequency image according to the second upper threshold and the second lower threshold.
3. The method for detecting and estimating frequency hopping signals based on time-frequency analysis according to claim 2, wherein: the calculation formula for performing double threshold segmentation processing on the time-frequency image according to the issued first upper threshold and the issued first lower threshold is as follows:
wherein B is 1 Representing a double threshold segmentation result; i.e 1 ,m 1 Respectively represent time-frequency images X 2 Length and width values of (a); threshold (threshold) H Representing a first upper threshold; threshold (threshold) L Representing a first lower threshold.
4. The method for detecting and estimating frequency hopping signals based on time-frequency analysis according to claim 2, wherein: the calculation formula of the second lower threshold is:
th(j)＝j*0.1*P,j＝1,2,3,…,n
wherein P represents the total energy; n, M are sum symbol upper bounds respectively representing time-frequency image X 2 Length maximum and width maximum of (a); i. m is sum symbol lower bound, respectively representing time-frequency image X 2 A length start value and a width start value of (a); th (j) represents a denoising threshold; j represents the number of iterations; c (j) represents an energy distribution number; and carrying out secondary difference processing on the energy distribution number, obtaining an inflection point of the energy distribution number, taking a coefficient k (j) corresponding to the inflection point as a noise coefficient, and taking a value corresponding to th (j) as a second lower threshold.
5. The method for detecting and estimating frequency hopping signals based on time-frequency analysis according to claim 1, wherein: the estimated calculation formula of the residence time is as follows:
wherein T is r Representing the estimated residence time; length characteristic sequence l k Is (l) 1 ,l 2 ,l 3 ,…,l k ) K is a sequence number.
6. The method for detecting and estimating frequency hopping signals based on time-frequency analysis according to claim 1, wherein: the estimated calculation formula of the jump period and the jump speed is as follows:
V h ＝1/T h
wherein T is h To represent an estimated hop period; v (V) h Representing the estimated jump rate; the initial coordinate feature sequence is { (xs) 1 ,ys 1 ),(xs 2 ,ys 2 ),…,(xs k ,ys k ) And k is a sequence number.
7. The method for detecting and estimating frequency hopping signals based on time-frequency analysis according to claim 1, wherein: the estimated calculation formula of the bandwidth is as follows:
wherein B represents the estimated bandwidth; f (f) s Representing the sampling rate; height characteristic sequence b k Is (b) 1 ,b 2 ,b 3 ,…,b k ) K is a sequence number; n (N) 2 Representing the short time fourier transform step size.
8. The method for detecting and estimating frequency hopping signals based on time-frequency analysis according to claim 1, wherein: in the amplitude estimation step, the amplitude of each hop signal is the value of the corresponding position of the centroid in the time-frequency matrix.
9. The method for detecting and estimating frequency hopping signals based on time-frequency analysis according to claim 1, wherein: the estimated calculation formula of the frequency is as follows:
f k ＝x k f s /N 2
wherein f k Representing the estimated frequency hopping frequency; f (f) s Representing the sampling rate; n (N) 2 Is the short-time Fourier transform step length; centroid feature sequence x k Is { (x) 1 ,y 1 ),(x 2 ,y 2 ),…,(x k ,y k ) And k is a sequence number.
10. The frequency hopping signal detection and parameter estimation system based on time-frequency analysis is characterized in that: the system comprises:
the image processing unit is used for carrying out double threshold segmentation processing and connected region marking processing on the time-frequency image of the frequency hopping signal to obtain characteristic information of the connected region image, wherein the characteristic information comprises length characteristics;
the signal blind detection unit is used for performing coarse clustering on the connected domain according to the length characteristics of the connected domain image to obtain coarse clustering results with different jump speeds;
the subdivision clustering unit is used for carrying out subdivision clustering treatment on the coarse clustering result according to the initial coordinate characteristics of the connected domain image, so as to realize the sorting of the network stations with the same jump speed and different jump time; the subdivision clustering process comprises the following substeps:
Sequencing the coarse clustering results according to the initial coordinate features of the time-frequency images to obtain a coordinate feature sequence and a corresponding connected domain label sequence;
calculating the average value of the differential results of the initial coordinate features to obtain a temporary jump period;
selecting any initial coordinate feature as a reference value, calculating a first difference value between the current initial coordinate feature and the reference value, and performing residual calculation on the first difference value and the temporary jump period to obtain a residual result;
comparing the residual result with a second threshold value, or comparing the second difference value of the temporary jump period and the residual result with the second threshold value, and combining the connected domain labels corresponding to the coordinate feature sequences to realize subdivision clustering;
the residence time estimation unit is used for estimating the residence time of the frequency hopping signal according to the length characteristics of the connected domain image in each group of classification results; and/or the number of the groups of groups,
the jump period and jump speed estimating unit is used for estimating the jump period and jump speed of the frequency-hopping signal according to the initial coordinate characteristics of the connected domain image in the classification results of each group; and/or the number of the groups of groups,
the bandwidth estimation unit is used for estimating the bandwidth of the frequency hopping signal according to the height characteristics of the connected domain image in each group of classification results; and/or the number of the groups of groups,
the amplitude estimation unit is used for estimating the amplitude of the frequency hopping signal according to the amplitude characteristics of the connected domain image in each group of classification results;
And/or the number of the groups of groups,
the frequency set estimation unit is used for estimating the frequency set of the frequency hopping signal according to the centroid characteristics of the connected domain image in each group of classification results;
the reporting unit is used for reporting coarse clustering results of the frequency hopping signals with different hopping speeds and/or sub-divided clustering results with the same hopping speed but different starting time, and parameter estimation results of any parameter or combination of a plurality of parameters of the hopping speed, amplitude, bandwidth, residence time and frequency hopping frequency set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211046676.5A CN115426002B (en) | 2022-06-15 | 2022-06-15 | Frequency hopping signal detection and parameter estimation method and system based on time-frequency analysis |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210675321.6A CN114759951B (en) | 2022-06-15 | 2022-06-15 | Frequency hopping signal real-time blind detection method, parameter estimation method, system and terminal |
| CN202211046676.5A CN115426002B (en) | 2022-06-15 | 2022-06-15 | Frequency hopping signal detection and parameter estimation method and system based on time-frequency analysis |
Related Parent Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210675321.6A Division CN114759951B (en) | 2022-06-15 | 2022-06-15 | Frequency hopping signal real-time blind detection method, parameter estimation method, system and terminal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115426002A CN115426002A (en) | 2022-12-02 |
| CN115426002B true CN115426002B (en) | 2023-07-11 |
Family
ID=82336673
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211046676.5A Active CN115426002B (en) | 2022-06-15 | 2022-06-15 | Frequency hopping signal detection and parameter estimation method and system based on time-frequency analysis |
| CN202210675321.6A Active CN114759951B (en) | 2022-06-15 | 2022-06-15 | Frequency hopping signal real-time blind detection method, parameter estimation method, system and terminal |
Family Applications After (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210675321.6A Active CN114759951B (en) | 2022-06-15 | 2022-06-15 | Frequency hopping signal real-time blind detection method, parameter estimation method, system and terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN115426002B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115422985B (en) * | 2022-11-07 | 2023-07-21 | 湖南赛能环测科技有限公司 | Micro-signal rapid detection method under hybrid scene |
| CN115496114B (en) * | 2022-11-18 | 2023-04-07 | 成都戎星科技有限公司 | TDMA burst length estimation method based on K-means clustering |
| CN116232378B (en) * | 2023-02-01 | 2023-08-29 | 南京中新赛克科技有限责任公司 | Frequency hopping signal sorting method based on statistical analysis |
| CN116886124B (en) * | 2023-09-05 | 2023-11-14 | 成都九华圆通科技发展有限公司 | Frequency hopping signal tracking and suppressing method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102752015A (en) * | 2012-07-09 | 2012-10-24 | 中国人民解放军国防科学技术大学 | Method for detecting and processing narrow-band interference of DS (Direct Sequence)/FH (Frequency Hopping) hybrid spread spectrum communication system |
| CN103607249A (en) * | 2013-09-22 | 2014-02-26 | 郝建华 | Method and system for testing DS/FH (Direct Sequence/Frequency Hopping) mixed spread frequency signal parameters |
| CN107273860A (en) * | 2017-06-20 | 2017-10-20 | 电子科技大学 | Frequency Hopping Signal dynamic clustering extracting method based on connected component labeling |
| CN113824518A (en) * | 2021-09-16 | 2021-12-21 | 中国电子科技集团公司第二十研究所 | Non-cooperative signal detection method based on frequency spectrum |
Family Cites Families (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7035311B2 (en) * | 2001-12-05 | 2006-04-25 | Sicom Systems, Ltd. | Apparatus and method for a digital, wideband, intercept and analysis processor for frequency hopping signals |
| US7974299B1 (en) * | 2006-11-27 | 2011-07-05 | Marvell International Ltd. | Methods and apparatus for switching transmission channels |
| US8218487B2 (en) * | 2008-04-09 | 2012-07-10 | Texas Instruments Incorporated | System and method of adaptive frequency hopping with look ahead interference prediction |
| CN106228551B (en) * | 2016-07-19 | 2019-04-23 | 清华大学 | The adaptive decomposition method of filter is generated based on image Segmentation Technology |
| CN106685478B (en) * | 2016-12-19 | 2020-06-19 | 电子科技大学 | Frequency hopping signal parameter estimation method based on signal time-frequency image information extraction |
| JP6905432B2 (en) * | 2017-09-12 | 2021-07-21 | 株式会社東芝 | Wireless communication device and signal detection method |
| CN109446896B (en) * | 2018-09-19 | 2021-08-17 | 青岛国数信息科技有限公司 | Unmanned aerial vehicle signal identification detection algorithm based on image processing |
| CN109975770B (en) * | 2019-03-13 | 2021-01-22 | 中国电子科技集团公司第二十九研究所 | Separation method and device of time-frequency overlapped multi-component linear frequency modulation signals |
| CN110334591B (en) * | 2019-05-24 | 2022-07-15 | 西华大学 | Unmanned aerial vehicle frequency hopping signal detection and identification method based on cluster analysis |
| CN111428631B (en) * | 2020-03-23 | 2023-05-05 | 中南大学 | Visual identification and sorting method for unmanned aerial vehicle flight control signals |
| CN112994741B (en) * | 2021-05-11 | 2021-07-23 | 成都天锐星通科技有限公司 | Frequency hopping signal parameter measuring method and device and electronic equipment |
| CN114050951B (en) * | 2021-10-30 | 2023-05-02 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Blind estimation method for parameters of ultrashort wave frequency hopping signals |
| CN114580476A (en) * | 2022-03-02 | 2022-06-03 | 北京清瑞维航技术发展有限公司 | Unmanned aerial vehicle signal identification model construction method and corresponding identification method and system |
-
2022
- 2022-06-15 CN CN202211046676.5A patent/CN115426002B/en active Active
- 2022-06-15 CN CN202210675321.6A patent/CN114759951B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102752015A (en) * | 2012-07-09 | 2012-10-24 | 中国人民解放军国防科学技术大学 | Method for detecting and processing narrow-band interference of DS (Direct Sequence)/FH (Frequency Hopping) hybrid spread spectrum communication system |
| CN103607249A (en) * | 2013-09-22 | 2014-02-26 | 郝建华 | Method and system for testing DS/FH (Direct Sequence/Frequency Hopping) mixed spread frequency signal parameters |
| CN107273860A (en) * | 2017-06-20 | 2017-10-20 | 电子科技大学 | Frequency Hopping Signal dynamic clustering extracting method based on connected component labeling |
| CN113824518A (en) * | 2021-09-16 | 2021-12-21 | 中国电子科技集团公司第二十研究所 | Non-cooperative signal detection method based on frequency spectrum |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115426002A (en) | 2022-12-02 |
| CN114759951B (en) | 2022-09-02 |
| CN114759951A (en) | 2022-07-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115426002B (en) | Frequency hopping signal detection and parameter estimation method and system based on time-frequency analysis | |
| CN110927706A (en) | Convolutional neural network-based radar interference detection and identification method | |
| CN107911183B (en) | Unmanned aerial vehicle identification method and system based on real-time signal characteristic analysis | |
| Iorkyase et al. | Improving RF-based partial discharge localization via machine learning ensemble method | |
| CN110197209B (en) | Radiation source identification method based on multi-feature fusion | |
| Rock et al. | Deep interference mitigation and denoising of real-world fmcw radar signals | |
| CN109947238B (en) | Non-cooperative gesture recognition method based on WIFI | |
| CN111428631B (en) | Visual identification and sorting method for unmanned aerial vehicle flight control signals | |
| CN112560803A (en) | Radar signal modulation identification method based on time-frequency analysis and machine learning | |
| CN106199351A (en) | The sorting technique of local discharge signal and device | |
| CN109598271A (en) | A kind of character segmentation method and device | |
| US20200097852A1 (en) | Systems and methods for detecting and grouping anomalies in data | |
| CN108734228A (en) | The polarimetric SAR image random forest classification method of comprehensive multiple features | |
| CN112003803B (en) | Detection and reception equipment for VHF and UHF band aviation radio station signals | |
| CN113312996A (en) | Detection and identification method for aliasing short-wave communication signals | |
| CN110046639A (en) | A kind of Hyperspectral imaging noise label detection method based on super-pixel weight density | |
| Liu et al. | A novel signal separation algorithm for wideband spectrum sensing in cognitive networks | |
| JP2019193049A (en) | Spectrum analysis program, spectrum analysis method and spectrum analysis apparatus | |
| CN116342656B (en) | Space-time image speed measurement method and device based on self-adaptive edge detection | |
| CN113495193A (en) | Fault diagnosis method based on multi-wavelet transform and cluster analysis | |
| Wang et al. | A learning-based signal parameter extraction approach for multi-source frequency-hopping signal sorting | |
| CN110390253A (en) | The modulation mode of communication signal recognition methods extracted based on a variety of points of shape spectrum signatures | |
| Rock et al. | CNNs for interference mitigation and denoising in automotive radar using real-world data | |
| CN110048977A (en) | The recognition methods of short-wave signal system and device based on the detection of gray level co-occurrence matrixes textural characteristics | |
| CN114580476A (en) | Unmanned aerial vehicle signal identification model construction method and corresponding identification method and system |
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 |