Detection method and system for specific frequency hopping signal

Abstract

The invention relates to a detection method and system for a specific frequency hopping signal, and the method aims at a narrowband signal, and comprises the following steps: 1, setting a parameter set, the parameters comprising a frame length, an FFT point, a filter frame length, a weak signal pulse threshold value, a middle signal pulse threshold value, a minimum signal pulse width, and a maximum signal pulse width; 2, converting the data file into a time-frequency two-dimensional logarithmic matrix for later signal detection; 3, searching all spectral lines which are greater than a pulse threshold value, and dynamically adapting to a reference noise level; step 4, screening based on pulse time difference; step 5, screening based on pulse frequency difference; step 6, screening based on pulse appearance time, pulse width and intensity; step 7, screening based on a pulse continuing relation; and step 8, detecting frequency hopping signal pulses in the data file based on hierarchical screening in the step 4 to the step 7, and outputting a pulse information table. According to the invention, specific frequency hopping signal detection processing can be efficiently and accurately carried out in a complex transmission environment.

Description

Technical Field

[0001] The present invention belongs to the field of communication technology, and in particular relates to a detection method and system for a specific frequency hopping signal. Background Art

[0002] In the process of mobile communication transmission, the transmission channel will be affected by the environment. In a complex environment, various interference signals may appear frequently. In order to resist the adverse effects of these interferences and ensure the accuracy and anti-interference of the communication process, frequency hopping communication technology, as a communication method with strong anti-interference ability and high spectrum utilization, has been widely used in military communications, mobile communications and other fields. With the continuous development of communication technology, the detection of frequency hopping signals faces more and more challenges, such as signal interference, noise influence, and rapid frequency switching. There are two main factors that make signal detection difficult. One is that the presence of fluctuating noise in the signal frequency band will weaken the signal strength of certain signal points and blur the difference between signal and noise; the other is that the strong noise in the local frequency band or time period may partially or completely suppress the signal point of the frequency band or time period, causing the characteristics of the signal point to be lost or incomplete. Therefore, it is of great significance to propose an efficient frequency hopping signal detection method. Summary of the invention

[0003] The purpose of the present invention is to provide a detection method and system for a specific frequency hopping signal, which is used to accurately detect the frequency set of the frequency hopping signal in a complex communication environment, and can efficiently and accurately detect and process a specific frequency hopping signal in a complex transmission environment.

[0004] The present invention provides a detection method for a specific frequency hopping signal. For a narrowband signal, the steps are as follows:

[0005] Step 1, setting a parameter set, wherein the parameters include frame length, FFT point, filter frame length, weak signal pulse threshold, medium signal pulse threshold, minimum signal pulse width, and maximum signal pulse width;

[0006] Step 2, converting the data file into a two-dimensional time-frequency logarithmic matrix for subsequent signal detection;

[0007] Step 3, search all spectral lines for those greater than the pulse threshold and dynamically adapt to the baseline noise level;

[0008] Step 4, screening based on pulse time difference;

[0009] Step 5, screening based on pulse frequency difference;

[0010] Step 6, screening based on pulse occurrence time, pulse width, and intensity;

[0011] Step 7, screening based on pulse connection relationship;

[0012] Step 8: Based on the hierarchical screening of steps 4 to 7, the frequency hopping signal pulses in the data file are detected and a pulse information table is output.

[0013] Furthermore, the step 1 comprises:

[0014] The frame length setting takes into account both the detection rate and the tolerance of the minimum signal pulse to noise interference. When the frame length is 0.5ms, the corresponding reference signal pulse width is 125 frames. Combined with the signal pulse change rate, as long as the number of detected signal pulse frames changes within the range of (0.85*125*N~1.15*125*N, N=1,...,10), it is regarded as a signal pulse to resist the influence of local noise;

[0015] The FFT point number setting takes into account the minimum frequency spacing between adjacent spectral lines, so that each signal point appears on at least one spectral line. If it appears on multiple spectral lines, the spectral line with the strongest SNR is selected as the signal carrier frequency to avoid the detected signal carrier frequency deviating too far from the actual carrier frequency, resulting in incomplete audio signals subsequently demodulated;

[0016] The weak signal pulse threshold is used to sort out possible signal frames in the frame sequence. All frames with a value greater than the threshold are regarded as frames that may carry signal pulses on the spectrum line, and vice versa.

[0017] The weak signal pulse threshold and the medium signal pulse threshold together constitute the ladder threshold of signal pulse detection. Through the combined effect of the two parameters, the immunity of the detection algorithm to noise is enhanced;

[0018] The filter frame length calculates the amplitude mean of the frame curve of each spectral line in units of the filter frame length to weaken the influence of the amplitude fluctuation of the frame curve on the pulse detection while maintaining a certain processing speed;

[0019] The minimum signal pulse width and the maximum signal pulse width work together to sort out pulses within a pulse width range determined by the two parameters from the entire pulse set.

[0020] Furthermore, the step 3 comprises:

[0021] Each pulse contains 7 parameters, and the recording order is: spectral line number, spectral line corresponding frequency, pulse start frame number, pulse end frame number, pulse width, pulse width and reference signal pulse width ratio, pulse SNR, the 7 parameters form a row of the pulse parameter matrix. If multiple pulses are found on a spectral line, multiple rows are recorded. The parameter matrix of L pulses is (L×7);

[0022] The reference noise level is established on each spectral line, indicating the change of the amplitude value of this spectral line over time; assuming that the amplitude value of the i-th spectral line is ai (n), n=1,2,…,N, the calculation method of the reference noise level is:

[0023]

[0024] Set B i The frames in are composed of frames that meet the formula conditions and are part of the original frame sequence. Frames with too large amplitude values ​​are removed, and most frames are kept to participate in the calculation of the reference noise level, so as to truly represent the actual noise level value of this spectrum line.

[0025] Let P represent the pulse matrix output in this step, then P={p i}, i = 1, 2, ..., L, L represents the total number of output pulses, corresponding to the total number of rows in the matrix, and each i corresponds to a row of the matrix;

[0026] Representation and parameter definition of the j-th row element: p j ={p j (1),p j (2),p j (3),p j (4),p j (5),p j (6),p j (7)};

[0027] g)p j (1): spectral line number;

[0028] h)p j (2): Frequency corresponding to the spectral line;

[0029] i)p j (3): Pulse start frame number;

[0030] j)p j (4): Pulse end frame number;

[0031] k)p j (5): Pulse width;

[0032] l)p j (6): Ratio of pulse width to reference signal pulse width;

[0033] p j (7): Pulse SNR.

[0034] Furthermore, the step 4 comprises:

[0035] Arrange the start frames of all pulses in the order of their appearance, and compare them in pairs. If the frame difference between the start frame and the end frame of two adjacent pulses is less than the maximum inter-frame distance parameter value, they are considered to be from the same pulse, otherwise they are two different pulses. Then, select the pulse with the largest SNR from the same pulse set as the representative of this pulse set and retain it. If it is not the same pulse, retain it directly. Specifically:

[0036] Define B as a set, which represents a set of pulses with approximately the same time period, and define the maximum inter-frame distance parameter as T2. Then the element b in the B set satisfies the following conditions:

[0037] B={b,||p i (3)-p j (3)|≤T2,i,j=1,2,..,L,i≠j}

[0038] The set B contains one or more pulses; for multiple pulses, P is selected from them. b (7) The largest pulse is retained as the pulse representing this set, and the remaining pulses are directly filtered out; let the pulse set after step 4 be M, then: |M|≤|L|.

[0039] Furthermore, the step 5 comprises:

[0040] Arrange all the pulses in the pulse set from small to large according to the spectral line number, and compare them in pairs. If the difference between the spectral line numbers of two adjacent pulses is less than the maximum spectral line distance parameter value, they are considered to be generated by the same pulse, otherwise they are two different pulses. Select the pulse with the largest SNR from the same pulse set as the representative of this pulse set and retain it, and directly retain the pulses that are not the same. Specifically, it includes:

[0041] Define C as a set, representing a set of pulses with approximately the same frequency, and define the maximum spectral line distance parameter as T3. Then the element c in the C set meets the following conditions:

[0042] C={c,||p i (1)-p j (1)|≤T3,i,j=1,2,..,M,i≠j}

[0043] The set C is one or more pulses; for multiple pulses, P is selected from them c (7) The largest pulse is retained as the pulse representing this set, and the remaining pulses are directly filtered out; let the pulse set after step 5 be N, then: |N|≤|M|.

[0044] Further, the step 6 comprises:

[0045] First, screening is performed based on the time when the pulse occurs. When screening pulses, all such pulses in the set are retained as signal pulses; second, screening is performed using the ratio of pulse width to reference signal pulse width; specifically, the following are included:

[0046] Define D as the pulse set after this step, define the signal pulse width change rate parameter as T4, and define the strong signal pulse threshold parameter as T5. Then D contains 3 subsets, D = D1, D2, D3, and the elements in each subset are:

[0047] D1={p i |p i (3)≤3,or p i (4)≤TotalFrames-3,i=1,2,..,N}

[0048] D2={p i ,|(1-T4)≤p i (6)≤(1.0+T4),i=1,2,..,N}

[0049] D3={p i ,|p i (7)≥T5,i=1,2,..,N}

[0050] In the formula, TotalFrames is the total number of frames corresponding to each data file; considering the influence of noise interference, the number 3 indicates that there is a certain degree of relaxation for the start frame and end frame of the data file.

[0051] Furthermore, the step 7 comprises:

[0052] The criteria for pulse screening based on pulse connection relationship include: (1) retaining pulses that have no time overlap and connection relationship with any pulse in the pulse set and retaining them as isolated signal pulses; (2) calculating the time connection relationship between each pulse and other pulses, and retaining all pulses whose time difference is less than the maximum connection distance parameter value as signal pulses, and filtering out pulses that do not meet criteria (1) and (2) as non-signal pulses; specifically, including:

[0053] Define F as the pulse set after step 7, define the maximum connection distance parameter as T6, then F contains two subsets, F = F1, F2, F1 is a set of signal pulses with a continuous relationship, and F2 is a set of isolated signal pulses;

[0054] Assume that the number of pulses in set D is All elements in D are arranged in the order of their appearance, and the absolute difference between the end time of each pulse in D and the start time of each pulse after this pulse is calculated in order to form a ; Set the undefined element values ​​in the matrix to infinity and construct a distance inverted triangular matrix:

[0055]

[0056] In the matrix: - indicates infinite value;

[0057] The calculation rules are as follows:

[0058] 1) Set the starting value i=1;

[0059] 2) If Arrange the values ​​of the elements in the i-th row from small to large;

[0060] 3) If the first value is less than the second value, and the first value is less than T6, then p i The continuation signal pulse is retained. After that, i jumps to the same i The pulse row with the smallest distance returns to step 2) and enters the next round of calculation; if the first value is less than the second value, and the first value is greater than T6, then p i The independent signal pulses are retained. At this time, set i+1 and return to step 2) to enter the next round of calculation;

[0061] 4) If the first value is equal to the second value, it means that there are two pulses and p i Equidistant, at this time, p i It is retained as a signal pulse. Define these two pulses as p j 、p k , determine p j 、p k Which pulse in the following pulse p l The pulses with a continuous relationship are selected as signal pulses and retained. Assuming that p k ,but At the same time, i jumps to p l row, return to step 2) and enter the next round of calculation; if there is no pulse row after that, select p j 、p k The pulse with a large width is retained as a signal pulse, and the calculation is terminated afterwards;

[0062] 5) If at this time or In any case, the i pulse is retained as a signal pulse.

[0063] The present invention also provides a detection system for a specific frequency hopping signal, comprising a detection module, wherein the detection module executes the detection method for the specific frequency hopping signal.

[0064] The present invention also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the detection method for a specific frequency hopping signal is implemented.

[0065] The present invention also provides an electronic device, comprising:

[0066] A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the detection method for a specific frequency hopping signal by executing the computer instructions.

[0067] By means of the above scheme, through the detection method and system for a specific frequency hopping signal, by dynamically adapting to the reference noise level and combining other parameters, the signal is hierarchically screened, and the relevant frequency set information of the frequency hopping signal can be correctly detected.

[0068] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention and implement it according to the contents of the specification, the following is a detailed description of the preferred embodiments of the present invention in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 The time-frequency two-dimensional feature space spectrum in one embodiment of the present invention

[0070] Figure 2 is a flow chart of a detection method for a specific frequency hopping signal according to the present invention;

[0071] Figure 3 The figure is a schematic diagram of an electronic device according to the present invention. DETAILED DESCRIPTION

[0072] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0073] Ginseng Figure 1 In the time-frequency two-dimensional feature space spectrum shown in the figure, each vertical line on the horizontal axis represents a frame, and each horizontal line on the vertical axis represents a spectrum line, and the spectrum lines are evenly spaced. The data between frames do not overlap. Each element in the matrix corresponds to a unique intersection point in the time-frequency two-dimensional space.

[0074] First, define the frame length, then calculate the time-frequency two-dimensional spectrogram corresponding to each data file of a certain duration in units of frames, generate a two-dimensional spectrogram matrix, and all subsequent detections are performed on this matrix. The specific frequency hopping signal is a narrowband signal, and the designed time-frequency two-dimensional spectrogram should make each frequency hopping signal point appear on one or two or more adjacent spectral lines. Although the characteristics of the noise cannot be predicted and the noise is not filtered, as long as the noise does not completely suppress the spectral lines in a certain period of time where the signal point exists, an appropriate algorithm can be designed to detect the signal point. For the same signal point repeatedly detected on multiple spectral lines, the spectral line with the strongest SNR can be selected as its carrier frequency. This is actually a verification mechanism and a reflection of the robustness of the detection method.

[0075] Ginseng Figure 2 As shown, the specific detection steps of a detection method for a specific frequency hopping signal are as follows:

[0076] 1. Parameter set settings:

[0077] The parameters set are closely related to the characteristics of the signal to be analyzed, and these parameter values ​​directly affect the detection performance. Users can appropriately adjust some parameters according to the changing signal characteristics to adapt to changes in the data acquisition environment and signal characteristics.

[0078] Here we first explain the selection basis and functions of several main parameters, and the auxiliary parameters are explained in the specific steps of their functions.

[0079] The frame length setting needs to take into account both the detection rate and the minimum signal pulse tolerance to noise interference. When the frame length is 0.5ms, the corresponding reference signal pulse width is 125 frames. Combined with the signal pulse change rate, as long as the detected signal pulse frame number changes within the range of (0.85*125*N~1.15*125*N, N=1,...,10), it is regarded as a signal pulse. Such a flexible setting can resist the influence of local noise to a certain extent.

[0080] The FFT point setting needs to consider the minimum frequency spacing between adjacent spectral lines. If it is too small, the detection calculation amount will increase, and if it is too large, the signal points between adjacent spectral lines may be lost. The FFT parameters given here can make each signal point appear on at least one spectral line. If it appears on multiple spectral lines, the algorithm can select the spectral line with the strongest SNR as the signal carrier frequency to avoid the detected signal carrier frequency deviating too far from the actual carrier frequency, resulting in incomplete audio signals from subsequent demodulation.

[0081] The weak signal pulse threshold is used to sort out possible signal frames in the frame sequence. All frames greater than the threshold are considered as frames that may carry signal pulses on the spectrum line, and vice versa. The weak signal pulse threshold is based on the baseline noise level of each spectrum line. In fact, the baseline noise level changes dynamically, which requires the algorithm to dynamically respond to this changing noise. The weak signal pulse threshold parameter, together with the medium signal pulse threshold, constitutes the ladder threshold for signal pulse detection. The frame that passes the first parameter is a "candidate frame", which is a detection based on each independent frame and is the minimum requirement for constituting a pulse frame. The pulse that passes the second parameter is the minimum amplitude requirement for the signal pulse, which is based on the average amplitude of multiple frames that constitute the pulse. Its physical meaning is to allow small fluctuations in the frame amplitude implicit in each pulse, which may be the result of noise. The combined effect of these two parameters can effectively enhance the immunity of the detection algorithm to noise.

[0082] 2. Calculate the Fourier transform two-dimensional logarithmic matrix.

[0083] The purpose of this step is to convert a data file of a certain length into a two-dimensional logarithmic matrix of time and frequency for subsequent signal detection.

[0084] 3. Search all spectral lines for values ​​greater than the pulse threshold.

[0085] This step uses five parameters: "filter frame length", "weak signal pulse threshold", "medium signal pulse threshold", "minimum signal pulse width" and "maximum signal pulse width". Among them, "filter frame length" calculates the amplitude mean of the frame curve of each spectral line in units of filter frame length, with the purpose of weakening the influence of frame curve amplitude fluctuation on pulse detection while maintaining a certain processing speed; the functions of "weak signal pulse threshold" and "medium signal pulse threshold" parameters have been mentioned before and will not be repeated here; the combined function of "minimum signal pulse width" and "maximum signal pulse width" is to sort out pulses with a pulse width range determined by these two parameters from the entire pulse set. Each pulse contains 7 parameters, and the recording order is: spectral line number, spectral line corresponding frequency, pulse start frame number, pulse end frame number, pulse width (measured in frames), pulse width to reference signal pulse width ratio, pulse SNR (difference with the reference noise level of the spectral line). These 7 parameters form a row of the pulse parameter matrix. If multiple pulses are found on a spectral line, multiple rows are recorded, and the parameter matrix of L pulses is (L×7).

[0086] What needs to be mentioned here is how to calculate the "reference noise level". The "reference noise level" is based on each spectral line and represents the change of the amplitude value of this spectral line over time. Assume that the amplitude value of the i-th spectral line is a i (n), n=1,2,…,N, the calculation method is:

[0087]

[0088] That is, set B i The frames in are composed of frames that meet the formula conditions and are part of the original frame sequence. Frames with excessively large amplitude values ​​are eliminated, while most frames are kept to participate in the calculation of the "reference noise level" in order to more realistically represent the actual noise level value of this spectrum line.

[0089] If P represents the pulse matrix output in this step, it can be expressed as: P = {p i}, i = 1, 2, …, L, L represents the total number of output pulses, corresponding to the total number of rows in the matrix, and each i corresponds to a row of the matrix.

[0090] Representation and parameter definition of the j-th row element: p j ={p j (1),p j (2),p j (3),p j (4),p j (5),p j (6),p j (7)}.

[0091] m)p j (1): Spectral line number

[0092] n)p j (2): Spectral line corresponding frequency

[0093] o)p j (3): Pulse start frame number

[0094] p)p j (4): Pulse end frame number

[0095] q)p j (5) Pulse width

[0096] r)p j (6): Ratio of pulse width to reference signal pulse width

[0097] s)p j (7) Pulse SNR

[0098] 4. Screening based on pulse time difference.

[0099] This step first arranges the start frames of all pulses in the order of the pulse occurrence, and then compares them in pairs. If the frame difference between the start frame and the end frame of two adjacent pulses is less than the parameter value of the "maximum frame distance", they are considered to be from the same pulse, otherwise they are two different pulses. Then, the pulse with the largest SNR is selected from the "same pulse set" as the representative of this pulse set and retained. Pulses that are not the same are directly retained.

[0100] The screening logic of this step is that at most one signal pulse can appear in a time period with almost the same starting point at the same time.

[0101] Mathematical description: Define B as a set, which represents a "pulse set with approximately the same time period", and define the "maximum inter-frame distance" parameter as T2, then the element b in the B set satisfies the following conditions:

[0102] B={b,||p i (3)-p j (3)|≤T2,i,j=1,2,..,L,i≠j}

[0103] There can be one or more pulses in set B. For multiple pulses, select P from them b (7) The largest pulse is retained as the pulse representing this set, and the rest of the pulses are directly filtered out. Let the pulse set after this step be M, then: |M|≤|L|.

[0104] 5. Screening based on pulse frequency difference.

[0105] In this step, all pulses in the pulse set are arranged in ascending order according to the spectral line number, and then compared in pairs. If the difference between the spectral line numbers of two adjacent pulses is less than the parameter value of "maximum spectral line distance", they are considered to be generated by the same pulse, otherwise they are two different pulses. Then, the pulse with the largest SNR is selected from the "same pulse set" as the representative of this pulse set and retained. Pulses that are not the same are directly retained.

[0106] The screening logic of this step is that at most one signal pulse can appear in any very narrow frequency band.

[0107] One more thing to add here is that this step requires a reasonable data file time length. The signal pulse can appear at the same frequency point again after a certain time length. This is a frequency hopping pattern design problem for both the sender and the receiver, provided that it is after a certain time window. In the existing data sample set, no case of two frequency hopping signal pulses appearing at the same frequency point in the same data file has been found.

[0108] Mathematical description: Define C as a set, representing "a set of pulses with approximately the same frequency", define the "maximum spectral line distance" parameter as T3, then the element c in the C set meets the following conditions:

[0109] C={c,||p i (1)-p j (1)|≤T3,i,j=1,2,..,M,i≠j}

[0110] Similarly, the set C can be one or more pulses. For multiple pulses, P is selected from them. c (7) The largest pulse is retained as the pulse representing this set, and the rest of the pulses are directly filtered out. Let the pulse set after this step be N, then: |N|≤|M|.

[0111] 6. Filter based on pulse occurrence time, pulse width and intensity.

[0112] This step first screens based on the time when the pulse appears. Due to the fixed data file size during signal acquisition, the signal pulses that appear at the head or tail of the data file may be split into two adjacent data files. Therefore, when screening pulses, all such pulses in the set are retained as signal pulses; secondly, the ratio of pulse width to reference signal pulse width is used for screening.

[0113] Generally speaking, the signal pulse width should be an integer multiple of the reference signal pulse width. Combined with the parameter "signal pulse width change rate" defined in the parameter set, this ratio is allowed to change within a certain range. All pulses that change within this range are retained as signal pulses; finally, for those pulses greater than the parameter value of "strong signal pulse threshold", they are retained as signal pulses. There are indeed some isolated pulses in the data sample set. Although such pulses do not meet the previous two rules, there seems to be no reason to directly lose them just because of their large SNR. It is better to keep them in the set and leave them for subsequent modules to screen.

[0114] Mathematical description: Define D as the pulse set after this step, define the "signal pulse width change rate" parameter as T4, and define the "strong signal pulse threshold" parameter as T5. Then D contains 3 subsets, D = D1, D2, D3, and the elements in each subset are:

[0115] D1={p i ,|p i (3)≤3,or p i (4)≤TotalFrames-3,i=1,2,..,N}

[0116] D2={p i,|(1-T4)≤p i (6)≤(1.0+T4),i=1,2,..,N}

[0117] D3={p i ,|pi(7))≥T5,i=1,2,..,N}

[0118] The "TotalFrames" in the expression refers to the total number of frames corresponding to each data file. The number "3" means that there is a certain degree of relaxation for the start frame and end frame of the data file, which also takes into account the influence of noise interference.

[0119] 7. Screening based on pulse connection relationship.

[0120] After the above three rounds of screening, it should be said that the signal pulse is in the remaining pulse set D. However, there may still be some non-signal pulses in D, such as pulses that partially overlap in time (the same starting frame but different ending frames, different starting frames but the same ending frames, different starting frames and different ending frames but overlapping frames in the middle). This type of pulse evades the various screening conditions set previously.

[0121] This step performs pulse screening based on the pulse continuity relationship. There are two specific screening criteria. The first is to retain those pulses that have no time overlap and continuity relationship with any pulse in the pulse set and retain them as isolated signal pulses. The second is to calculate the time continuity relationship between each pulse and other pulses. All pulses with a time difference less than the parameter value of the "maximum connection distance" will be retained as signal pulses, and pulses that do not meet these two criteria will be filtered out as non-signal pulses.

[0122] Mathematical description: Define F as the pulse set after the module is activated, and define the "maximum connection distance" parameter as T6. Then F contains two subsets, F=F1 and F2. F1 is a set of signal pulses with a connection relationship, and F2 is a set of isolated signal pulses.

[0123] Assume that the number of pulses in set D is After the above steps, all elements in D are arranged in the order of appearance. The absolute difference between the end time of each pulse in D and the start time of each pulse after this pulse is calculated in order to form a The values ​​of undefined elements in the matrix are set to infinity, thus constructing an inverted triangular distance matrix:

[0124]

[0125] The “-” in the matrix indicates infinite value.

[0126] Calculation rules:

[0127] 1) Set the starting value i=1;

[0128] 2) If Arrange the values ​​of the elements in the i-th row from small to large;

[0129] 3) If the first value is less than the second value, and the first value is less than T6, then p i The continuation signal pulse is retained. After that, i jumps to the same i The pulse row with the smallest distance returns to step 2 and enters the next round of calculation; if the first value is less than the second value, and the first value is greater than T6, then p i The independent signal pulses are retained. At this time, set i+1, return to step 2, and enter the next round of calculation;

[0130] 4) If the first value is equal to the second value, it means that there are two pulses and p i At this time, firstly set p i It is retained as a signal pulse. Define these two pulses as p j 、p k Then look at p j 、p k Which pulse in the following pulse p l The pulses with a continuous relationship are selected as signal pulses and retained. Here, it is assumed that p k ,but At the same time, i jumps to p l If there is no pulse row after that, select p j 、p k The pulse with a large width is retained as a signal pulse, and the calculation is terminated afterwards;

[0131] 5) If at this time or In any case, the i pulse is retained as a signal pulse.

[0132] As a result of the combined and sequential action of the above steps, the frequency hopping signal pulses in the data files collected in the complex electromagnetic environment can be detected and a pulse information table can be output.

[0133] The detection method for a specific frequency hopping signal can correctly detect the relevant frequency set information of the frequency hopping signal by dynamically adapting to the reference noise level and combining other parameters to perform hierarchical screening on the signal.

[0134] The effectiveness of the detection algorithm is closely related to the baseline noise level. The baseline noise level changes dynamically. Through the "baseline noise level" design, the algorithm can dynamically respond to this changing noise. Combined with other parameters, it can effectively enhance the algorithm's immunity to noise and compress the pulses to only dozens without losing signal pulses.

[0135] Through hierarchical screening and logic design, it can achieve the effect of "unraveling the mystery" layer by layer. By screening the signal pulses based on time difference, frequency difference, pulse width, intensity and connection relationship, the pulses screened out by each module are either non-signal pulses, or there are more representative signal pulses in the remaining set. The biggest advantage of the hierarchical screening module is that it ensures that the signal points in the pulse set will not be lost, while minimizing the number of non-signal points in the remaining pulse points, thereby improving the accuracy of detection.

[0136] This embodiment further provides a detection system for a specific frequency hopping signal, including a detection module, wherein the detection module executes the detection method based on the specific frequency hopping signal.

[0137] This embodiment further provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the detection method for a specific frequency hopping signal is implemented.

[0138] Ginseng Figure 3 As shown, this embodiment also provides an electronic device, including:

[0139] The memory 201 and the processor 202 are communicatively connected with each other, the memory 201 stores computer instructions, and the processor 202 executes the detection method for a specific frequency hopping signal by executing the computer instructions.

[0140] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the technical principles of the present invention, and these improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A detection method for a specific frequency hopping signal, characterized in that: For narrowband signals, the steps are as follows: Step 1, setting a parameter set, wherein the parameters include frame length, FFT point, filter frame length, weak signal pulse threshold, medium signal pulse threshold, minimum signal pulse width, and maximum signal pulse width; Step 2, converting the data file into a two-dimensional time-frequency logarithmic matrix for subsequent signal detection; Step 3, search all spectral lines for those with noise levels greater than the pulse threshold, and dynamically adapt to the baseline noise level; Step 4, screening based on pulse time difference; Step 5, screening based on pulse frequency difference; Step 6, screening based on pulse occurrence time, pulse width, and intensity; Step 7, screening based on pulse connection relationship; Step 8: Based on the hierarchical screening of steps 4 to 7, the frequency hopping signal pulses in the data file are detected and a pulse information table is output.

2. The detection method for a specific frequency hopping signal according to claim 1, characterized in that: The step 1 comprises: The frame length setting takes into account both the detection rate and the tolerance of the minimum signal pulse to noise interference. When the frame length is 0.5ms, the corresponding reference signal pulse width is 125 frames. Combined with the signal pulse change rate, as long as the number of detected signal pulse frames changes within the range of (0.85*125*N~1.15*125*N, N=1,...,10), it is regarded as a signal pulse to resist the influence of local noise; The FFT point number setting takes into account the minimum frequency spacing between adjacent spectral lines, so that each signal point appears on at least one spectral line. If it appears on multiple spectral lines, the spectral line with the strongest SNR is selected as the signal carrier frequency to avoid the detected signal carrier frequency deviating too far from the actual carrier frequency, resulting in incomplete audio signals subsequently demodulated; The weak signal pulse threshold is used to sort out possible signal frames in the frame sequence. All frames with a value greater than the threshold are regarded as frames that may carry signal pulses on the spectrum line, and vice versa. The weak signal pulse threshold and the medium signal pulse threshold together constitute the ladder threshold of signal pulse detection. Through the combined effect of the two parameters, the immunity of the detection algorithm to noise is enhanced; The filter frame length calculates the amplitude mean of the frame curve of each spectral line in units of the filter frame length to weaken the influence of the amplitude fluctuation of the frame curve on the pulse detection while maintaining a certain processing speed; The minimum signal pulse width and the maximum signal pulse width work together to sort out pulses within a pulse width range determined by the two parameters from the entire pulse set.

3. According to the detection method for a specific frequency hopping signal according to claim 2, the step 3 comprises: Each pulse contains 7 parameters, and the recording order is: spectral line number, spectral line corresponding frequency, pulse start frame number, pulse end frame number, pulse width, pulse width and reference signal pulse width ratio, pulse SNR, the 7 parameters form a row of the pulse parameter matrix. If multiple pulses are found on a spectral line, multiple rows are recorded. The parameter matrix of L pulses is (L×7); The reference noise level is established on each spectral line, indicating the change of the amplitude value of this spectral line over time; assuming that the amplitude value of the i-th spectral line is a i (n), n=1,2,…,N, the calculation method of the reference noise level is: Set B i The frames in are composed of frames that meet the formula conditions and are part of the original frame sequence. Frames with too large amplitude values ​​are removed, and most frames are kept to participate in the calculation of the reference noise level, so as to truly represent the actual noise level value of this spectrum line. Let P represent the pulse matrix output in this step, then P={p i }, i = 1, 2, ..., L, L represents the total number of output pulses, corresponding to the total number of rows in the matrix, and each i corresponds to a row of the matrix; Representation and parameter definition of the j-th row element: p j ={p j (1),p j (2),p j (3),p j (4),p j (5),p j (6),p j (7)}; a)p j (1): spectral line number; b)p j (2): Frequency corresponding to the spectral line; c)p j (3): Pulse start frame number; d)p j (4): Pulse end frame number; e)p j (5): Pulse width; f)p j (6): Ratio of pulse width to reference signal pulse width; p j (7): Pulse SNR.

4. The detection method for a specific frequency hopping signal according to claim 3, characterized in that: The step 4 comprises: Arrange the start frames of all pulses in the order of their appearance, and compare them in pairs. If the frame difference between the start frame and the end frame of two adjacent pulses is less than the maximum inter-frame distance parameter value, they are considered to be from the same pulse, otherwise they are two different pulses. Then, select the pulse with the largest SNR from the same pulse set as the representative of this pulse set and retain it. If it is not the same pulse, retain it directly. Specifically: Define B as a set, which represents a set of pulses with approximately the same time period, and define the maximum inter-frame distance parameter as T2. Then the element b in the B set satisfies the following conditions: B={b,||p i (3)-p j (3)|≤T2,i,j=1,2,..,L,i≠j} The set B contains one or more pulses; for multiple pulses, P is selected from them. b (7) The largest pulse is retained as the pulse representing this set, and the remaining pulses are directly filtered out; let the pulse set after step 4 be M, then: |M|≤|L|.

5. The detection method for a specific frequency hopping signal according to claim 4, characterized in that: The step 5 comprises: Arrange all the pulses in the pulse set from small to large according to the spectral line number, and compare them in pairs. If the difference between the spectral line numbers of two adjacent pulses is less than the maximum spectral line distance parameter value, they are considered to be generated by the same pulse, otherwise they are two different pulses. Select the pulse with the largest SNR from the same pulse set as the representative of this pulse set and retain it, and directly retain the pulses that are not the same. Specifically, it includes: Define C as a set, representing a set of pulses with approximately the same frequency, and define the maximum spectral line distance parameter as T3. Then the element c in the C set meets the following conditions: C={c,||p i (1)-p j (1)|≤T3,i,j=1,2,..,M,i≠j} The set C is one or more pulses; for multiple pulses, P is selected from them c (7) The largest pulse is retained as the pulse representing this set, and the remaining pulses are directly filtered out; let the pulse set after step 5 be N, then: |N|≤|M|.

6. The detection method for a specific frequency hopping signal according to claim 5, characterized in that: The step 6 comprises: First, screening is performed based on the time when the pulse occurs. When screening pulses, all such pulses in the set are retained as signal pulses; second, screening is performed using the ratio of pulse width to reference signal pulse width; specifically, the following are included: Define D as the pulse set after this step, define the signal pulse width change rate parameter as T4, and define the strong signal pulse threshold parameter as T5. Then D contains 3 subsets, D = D1, D2, D3, and the elements in each subset are: D1={p i ,|p i (3)≤3,or p i (4)≤TotalFrames-3,i=1,2,..,N} D2={p i ,|(1-T4)≤p i (6)≤(1.0+T4),i=1,2,..,N} D3={p i ,|p i (7)≥T5,i=1,2,..,N} In the formula, TotalFrames is the total number of frames corresponding to each data file; considering the influence of noise interference, the number 3 indicates that there is a certain degree of relaxation for the start frame and end frame of the data file.

7. The detection method for a specific frequency hopping signal according to claim 6, characterized in that: The step 7 comprises: The criteria for pulse screening based on pulse connection relationship include: (1) retaining pulses that have no time overlap and connection relationship with any pulse in the pulse set and retaining them as isolated signal pulses; (2) calculating the time connection relationship between each pulse and other pulses, and retaining all pulses whose time difference is less than the maximum connection distance parameter value as signal pulses, and filtering out pulses that do not meet criteria (1) and (2) as non-signal pulses; specifically, including: Define F as the pulse set after step 7, define the maximum connection distance parameter as T6, then F contains two subsets, F = F1, F2, F1 is a set of signal pulses with a continuous relationship, and F2 is a set of isolated signal pulses; Assume that the number of pulses in set D is All elements in D are arranged in the order of their appearance, and the absolute difference between the end time of each pulse in D and the start time of each pulse after this pulse is calculated in order to form a ; Set the undefined element values ​​in the matrix to infinity and construct a distance inverted triangular matrix: In the matrix: - indicates infinite value; The calculation rules are as follows: 1) Set the starting value i=1; 2) If Arrange the values ​​of the elements in the i-th row from small to large; 3) If the first value is less than the second value, and the first value is less than T6, then p i The continuous signal pulse is retained. After that, i jumps to the same i The pulse row with the smallest distance returns to step 2) and enters the next round of calculation; if the first value is less than the second value, and the first value is greater than T6, then p i The independent signal pulses are retained. At this time, set i+1 and return to step 2) to enter the next round of calculation; 4) If the first value is equal to the second value, it means that there are two pulses and p i Equidistant, at this time, p i It is retained as a signal pulse, Define these two pulses as p j 、p k , determine p j 、p k Which pulse in the following pulse p l The pulses with a continuous relationship are selected as signal pulses and retained. Assuming that p k ,but At the same time, i jumps to p l row, return to step 2) and enter the next round of calculation; if there is no pulse row after that, select p j 、p k The pulse with a large width is retained as a signal pulse, and the calculation is terminated afterwards; 5) If at this time or In any case, the i pulse is retained as a signal pulse.

8. A detection system for a specific frequency hopping signal, characterized in that: It comprises a detection module, and the detection module executes a detection method for a specific frequency hopping signal as described in any one of claims 1-7.

9. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, a detection method for a specific frequency hopping signal as described in any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the detection method for a specific frequency hopping signal as described in any one of claims 1 to 7 by executing the computer instructions.

Images