Wireless signal sorting method, device and storage medium

Abstract

This invention provides a wireless signal sorting method, apparatus, and storage medium, relating to the field of wireless communication technology. The method includes: coarsely classifying multiple signal blocks to be processed to obtain first-class and second-class signal blocks; clustering the first-class signal blocks based on signal characteristics to obtain multiple third-class signal blocks; identifying third-class signal blocks whose number exceeds a first preset threshold as fixed-frequency signal blocks, and identifying the remaining third-class signal blocks as second-class signal blocks; determining associated signal blocks in the target signal block set based on time-domain feature parameters; extracting parameter information from the associated signal blocks, and determining the final signal block category based on the parameter information. This invention does not rely on prior information, but only utilizes the signal characteristics of the signal blocks themselves to sort fixed-frequency and frequency-hopping signals, thereby meeting the signal sorting requirements in completely blind scenarios and effectively improving the adaptability of signal sorting in complex and unknown electromagnetic environments.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and in particular to a wireless signal sorting method, apparatus and storage medium. Background Technology

[0002] In wireless communication systems, signal sorting refers to the process of obtaining the characteristic parameters of a mixed signal containing multiple signals through a series of processing steps, in order to separate signals belonging to different types.

[0003] Existing wireless signal sorting methods typically rely on prior information, generating local synchronization correlation signals based on N different synchronization sequences. They then extract received signal blocks of appropriate length according to each signal protocol. The received signal is then subjected to sliding correlation with each of the N local synchronization correlation signals to find the maximum value. Finally, the maximum value among the N maximum values ​​is identified, and the corresponding local synchronization signal is determined to be the signal type of that signal block. This process is repeated for all signal blocks in the received signal, and signals of the same type are sorted into a group. However, existing technologies rely on prior signal information, such as frame format, synchronization sequence, and modulation scheme, making it difficult to meet the signal sorting requirements in completely blind scenarios. Summary of the Invention

[0004] This invention provides a wireless signal sorting method, apparatus, and storage medium to solve the technical problem that existing technologies rely on prior signal information, such as frame format, synchronization sequence, and modulation method, which makes it difficult to meet the signal sorting requirements in completely blind scenarios.

[0005] This invention provides a wireless signal sorting method, comprising: Multiple signal blocks to be processed are coarsely classified to obtain first-class signal blocks and second-class signal blocks; Based on signal characteristics, the first type of signal blocks are clustered to obtain multiple third type signal blocks; The third type of signal block with a number of signal blocks exceeding the first preset threshold is defined as a fixed-frequency signal block, and the third type of signal block with a number of signal blocks not exceeding the first preset threshold is defined as a second type of signal block; The steps for determining associated signal blocks are as follows: determine the time-domain characteristic parameters of the target signal block set in the second type of signal blocks, and determine the associated signal blocks in the target signal block set based on the time-domain characteristic parameters; Extract the parameter information from the associated signal block, and determine the final signal block category based on the parameter information.

[0006] According to a wireless signal sorting method provided by the present invention, the step of coarsely classifying multiple signal blocks to be processed to obtain a first type of signal block and a second type of signal block includes: Signal blocks whose duration exceeds a preset threshold time are identified as first-class signal blocks; Signal blocks whose duration is less than or equal to a preset threshold time are identified as second-class signal blocks.

[0007] According to a wireless signal sorting method provided by the present invention, the signal features include at least one of frequency, bandwidth, and modulation scheme. The method involves clustering the first type of signal blocks based on these signal features to obtain multiple third type signal blocks, including: A feature vector is generated based on the signal characteristics of the signal blocks in the first type of signal block; The feature vector of the first signal block is used as the initial class center; Starting from the second signal block, calculate the Euclidean distance between the signal features in the current signal block and the signal features in the initial class center. Convert the Euclidean distance into a confidence score. If the confidence score is greater than a preset confidence threshold, classify the current signal block into the current category and update the initial class center to obtain a new class center. If the confidence score is less than or equal to the preset confidence threshold, classify the signal block into a new category and use the feature vector of the signal block as the initial class center of the new category. By traversing all signal blocks of the first type of signal block, multiple signal blocks of the third type are obtained.

[0008] According to a wireless signal sorting method provided by the present invention, before determining the time-domain feature parameters of the target signal block set in the second type of signal block, and determining the associated signal blocks in the target signal block set based on the time-domain feature parameters, the method further includes: Based on signal characteristics, the second type of signal blocks are clustered to obtain multiple fourth type signal blocks; The fourth type of signal blocks, whose number of signal blocks does not exceed the second preset threshold, are removed to obtain the target signal block set.

[0009] According to a wireless signal sorting method provided by the present invention, the time-domain feature parameters include signal period, and the step of determining the time-domain feature parameters of the target signal block set in the second type of signal block includes: The signal start times of the target signal block set are sorted in ascending order to obtain a signal start time sequence. Based on the signal start time sequence, calculate the k-th level time difference, where the k-th level time difference is the difference between the i-th signal start time and the ik-th signal start time; i = k+1, ..., N; N is the number of signal start times in the signal start sequence; Count the number of time differences belonging to different numerical intervals; Select the preset number range of values ​​with the largest number of time differences, and use the time differences corresponding to the preset number range as candidate time differences; The signal period is determined based on the preset detection threshold and the difference between the candidate times.

[0010] According to a wireless signal sorting method provided by the present invention, the step of determining the signal period based on a preset detection threshold and the candidate time difference includes: If the number of samples corresponding to the candidate time difference and twice the number of samples corresponding to the candidate time difference both exceed a preset detection threshold, the candidate time difference is taken as the signal period.

[0011] According to a wireless signal sorting method provided by the present invention, the time-domain feature parameters include signal period, and the step of determining the associated signal blocks in the target signal block set based on the time-domain feature parameters includes: According to the signal period, a related subsequence search is performed in the signal start time sequence. If a related subsequence is found, the related subsequence is deleted from the signal start time sequence, and the related signal block determination step is repeated. If the search stopping condition is met, the signal block corresponding to the associated subsequence is determined as the associated signal block.

[0012] According to a wireless signal sorting method provided by the present invention, the parameter information includes signal frequency, and the step of determining the final signal block category based on the parameter information includes: Based on the signal frequency, calculate the frequency variance in the associated signal block. If the frequency variance is less than a preset frequency variance threshold, determine the associated signal block as a fixed-frequency signal block. If the frequency variance is greater than or equal to a preset frequency variance threshold, the associated signal block is determined to be a frequency hopping signal.

[0013] The present invention also provides a wireless signal sorting device, comprising: The first classification module is used to perform coarse classification on multiple signal blocks to be processed, and obtain the first type of signal blocks and the second type of signal blocks. The clustering module is used to cluster the first type of signal blocks based on signal features to obtain multiple third type signal blocks; The second classification module is used to identify third-class signal blocks whose number of signal blocks exceeds the first preset threshold as fixed-frequency signal blocks, and to identify third-class signal blocks whose number of signal blocks does not exceed the first preset threshold as second-class signal blocks; The associated signal block determination module is used to determine the time-domain feature parameters of the target signal block set in the second type of signal blocks, and determine the associated signal blocks in the target signal block set based on the time-domain feature parameters; The third classification module is used to extract parameter information from the associated signal blocks and determine the final signal block category based on the parameter information.

[0014] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the wireless signal sorting method as described above.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the wireless signal sorting method as described above.

[0016] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the wireless signal sorting method as described above.

[0017] This invention distinguishes between first-class and second-class signal blocks by coarsely classifying multiple signal blocks to be processed based on their duration. Then, unsupervised clustering is performed on the signal features of the first-class signal blocks to obtain multiple third-class signal blocks. The third-class signal blocks with more than a first preset threshold are directly identified as fixed-frequency signal blocks, while the third-class signal blocks with fewer than the first preset threshold are classified as second-class signal blocks. Furthermore, the associated signal blocks among the second-class signal blocks are determined based on their time-domain feature parameters. The entire sorting process does not rely on prior information such as frame format, synchronization sequence, or modulation method. It can sort fixed-frequency signals and frequency-hopping signals using only the signal features of the signal blocks themselves, thereby meeting the signal sorting requirements in completely blind scenarios and effectively improving the adaptability of signal sorting in complex and unknown electromagnetic environments.

[0018] Furthermore, in this embodiment of the invention, after sorting the signal start times, multi-level time differences are calculated, and interval statistics are performed on the differences to suppress measurement noise. Then, several differences that occur most frequently are selected as candidate periods. Finally, the true signal period is determined by combining them with a preset threshold. This method can determine the signal period from complex signals with aliased or missing pulses without relying on prior information. It effectively avoids false periods caused by harmonic interference and random noise, and provides highly reliable period parameters for subsequent time correlation and fixed-hop signal sorting. This can effectively improve the accuracy and reliability of wireless signal sorting. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is one of the flowcharts illustrating the wireless signal sorting method provided by the present invention.

[0021] Figure 2 This is the second flowchart of the wireless signal sorting method provided by the present invention.

[0022] Figure 3 This is a schematic diagram of the structure of the wireless signal sorting device provided by the present invention.

[0023] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0025] Figure 1 This is one of the flowcharts illustrating a wireless signal sorting method provided by the present invention, such as... Figure 1 As shown, the method includes the following: This invention provides a wireless signal sorting method applicable to wireless signal sorting in completely blind scenarios. For example, in non-cooperative communication reconnaissance scenarios in complex electromagnetic environments, the receiving device cannot predict signal protocol parameters, rendering traditional correlation matching methods ineffective. The wireless signal sorting method provided by this invention first coarsely classifies signal blocks to be processed according to their duration. Then, it performs unsupervised feature clustering on long-duration signal blocks to directly output fixed-frequency signals. For short-duration signal blocks, it uses multi-level difference period estimation and sequence search to find time-related signal blocks. Finally, it completes the fixed-frequency / frequency-hopping decision based on frequency variance. This method can automatically and in real-time separate different types of signal sources without relying on any prior knowledge, thereby significantly improving the autonomous adaptability of signal reconnaissance and spectrum monitoring systems.

[0026] S1. Perform coarse classification on multiple signal blocks to be processed to obtain the first type of signal blocks and the second type of signal blocks; In this embodiment of the invention, the signal block to be processed is a signal block that needs to be sorted. It can be a continuous signal segment extracted from the received raw wireless signal according to certain rules (such as energy detection, frame synchronization header detection). Each signal block corresponds to a potential independent signal unit, which may be a pulse, a frequency hopping segment, or a burst data frame, etc.

[0027] In this embodiment of the invention, the receiver can scan the spectrum or use an energy detection algorithm. When the energy at a certain frequency exceeds the background noise threshold, signal recording begins until the energy falls back below the threshold, thus capturing a complete signal block. Simultaneously, the start and end times of this signal block are recorded, and the duration is calculated as end time - start time.

[0028] In this embodiment of the invention, the first type of signal can be an initial fixed-frequency signal, and the second type of signal can be an initial non-fixed-frequency signal, that is, a signal that has not been completely sorted and needs to be further sorted to obtain the final signal sorting result.

[0029] S2. Cluster the first type of signal blocks based on signal features to obtain multiple third type signal blocks; In this embodiment of the invention, the number of categories of the third type of signal block is obtained after clustering processing, and there can be multiple categories, such as 10, 20, etc.

[0030] S3. The third type of signal block with a number of signal blocks exceeding the first preset threshold is identified as a fixed-frequency signal block, and the third type of signal block with a number of signal blocks not exceeding the first preset threshold is identified as a second type of signal block; In this embodiment of the invention, the first preset threshold number can be determined by statistically analyzing the sample number distribution of all third-class signal blocks using a histogram, and selecting a natural dividing point (such as the valley of a bimodal distribution) as the threshold. The first preset threshold number is used to further determine which categories of signal blocks in the third-class signal blocks are fixed-frequency signal blocks and which are not, thereby further processing the second-class signal blocks that are not fixed-frequency signal blocks to obtain a more refined sorting result.

[0031] S4. Associated signal block determination step: Determine the time-domain characteristic parameters of the target signal block set in the second type of signal blocks, and determine the associated signal blocks in the target signal block set based on the time-domain characteristic parameters; In this embodiment of the invention, the target signal block set can be a subset of signal blocks from the second type of signal blocks. The target signal blocks can be determined according to actual needs. For example, the second type of signal blocks can be clustered to obtain several signal block categories. Signal blocks from these categories that meet preset conditions can be selected as the target signal block set. For instance, signal blocks from categories whose number of signal blocks exceeds a preset threshold can be selected as the target signal block set. The time-domain feature parameter can be the signal period. Based on the signal period within the signal blocks, associated signal blocks in the target signal block set can be accurately determined, thereby providing data support for subsequent signal sorting.

[0032] S5. Extract the parameter information from the associated signal block and determine the final signal block category based on the parameter information.

[0033] In this embodiment of the invention, the parameter information includes the signal frequency. The final signal block category can be determined based on the parameter information. For example, if the frequency variance is less than a preset frequency variance threshold, the associated signal block is determined as a fixed-frequency signal block; if the frequency variance is greater than or equal to the preset frequency variance threshold, the associated signal block is determined as a frequency-hopping signal.

[0034] This invention distinguishes between first-class and second-class signal blocks by coarsely classifying multiple signal blocks to be processed based on their duration. Then, unsupervised clustering is performed on the signal features of the first-class signal blocks to obtain multiple third-class signal blocks. The third-class signal blocks with more than a first preset threshold are directly identified as fixed-frequency signal blocks, while the third-class signal blocks with fewer than the first preset threshold are classified as second-class signal blocks. Furthermore, the associated signal blocks among the second-class signal blocks are determined based on their time-domain feature parameters. The entire sorting process does not rely on prior information such as frame format, synchronization sequence, or modulation method. It can sort fixed-frequency signals and frequency-hopping signals using only the signal features of the signal blocks themselves, thereby meeting the signal sorting requirements in completely blind scenarios and effectively improving the adaptability of signal sorting in complex and unknown electromagnetic environments.

[0035] In one embodiment, step S1, coarsely classifying multiple signal blocks to be processed to obtain a first type of signal block and a second type of signal block, includes: S11. The signal blocks to be processed with a duration greater than a preset threshold time are identified as the first type of signal blocks; S12. The signal blocks to be processed with a duration less than or equal to a preset threshold time are identified as the second type of signal blocks.

[0036] In this embodiment of the invention, the duration is the length of time from the start to the end of a signal block, typically expressed in microseconds, milliseconds, or the number of sampling points. The preset threshold time is a pre-defined time threshold value used to distinguish between long-duration and short-duration signal blocks. This threshold can be obtained based on prior knowledge (such as the known continuous transmission of fixed-frequency signals and the shorter dwell time of frequency-hopping signals) or through statistical analysis and training.

[0037] In this embodiment of the invention, the signal blocks with longer durations typically correspond to fixed-frequency signals (such as continuous waves, communication signals with fixed frequencies for long periods), long-pulse radar signals, or non-frequency-hopping continuous data transmission; the second type of signal blocks are signal blocks with shorter durations, typically corresponding to each frequency segment of a frequency-hopping signal, short-pulse radar signals, burst control signals, etc.

[0038] This invention compares the signal duration with a preset threshold time to initially classify different types of signal blocks, and performs corresponding processing on different types of signal blocks, which can effectively improve the accuracy of wireless signal sorting.

[0039] In one embodiment, before or after coarsely classifying the signal blocks awaiting processing, data preprocessing can be performed on the feature quantities of each signal block. Taking frequency as an example, the average and standard deviation of the frequency features of all current signal blocks are calculated, and the preprocessed data is obtained by the following formula: Where N is the total number of signal blocks, x i The frequency characteristic measurement value of the i-th signal block. The mean, Standard deviation The frequency characteristic measurement value of the signal block to be processed. This is the result after normalization.

[0040] In one embodiment, the signal features include at least one of frequency, bandwidth, and modulation scheme. Step S2 involves clustering the first type of signal blocks based on the signal features to obtain multiple third type signal blocks, including: S21. Generate a feature vector based on the signal characteristics of the signal blocks in the first type of signal blocks; In this embodiment of the invention, the values ​​of multiple signal features can be arranged in a fixed order to form a feature vector, which is the basic processing unit of the clustering algorithm. For example, signal block A has a frequency of 100.1MHz, a bandwidth of 25kHz, and a modulation method of 4QPSK (Quadrature Phase Shift Keying) → feature vector [100.1,25,4].

[0041] S22. Use the feature vector of the first signal block as the initial class center; S23. Starting from the second signal block, calculate the Euclidean distance between the signal features in the current signal block and the signal features in the initial class center, convert the Euclidean distance into confidence. If the confidence is greater than a preset confidence threshold, classify the current signal block into the current category and update the initial class center to obtain a new class center. If the confidence is less than or equal to the preset confidence threshold, classify the signal block into a new category and use the feature vector of the signal block as the initial class center of the new category. In this embodiment of the invention, converting Euclidean distance into confidence level can be achieved by using a preset function to map the Euclidean distance to the interval (0,1), such that the smaller the distance, the higher the confidence level. The preset function includes a negative exponential function, an inverse proportional function, and a piecewise linear function.

[0042] In this embodiment of the invention, the confidence level of the current signal block is compared with a preset confidence threshold. If the confidence level is greater than the preset confidence threshold, the current signal block is classified into the current category. The current category is a category in the cluster. Through this clustering method, multiple categories can be obtained after clustering. These category signal blocks are called third category signal blocks. That is, the third category signal blocks include multiple category signal blocks obtained by clustering. The multiple category signal blocks are distinguished by the confidence threshold in the feature space. The size of the class (number of signal blocks) may vary greatly. For example, some classes may contain only one signal block (isolated point), while some classes may contain hundreds or thousands of signal blocks.

[0043] S24. Traverse all signal blocks of the first type of signal block to obtain multiple third type signal blocks.

[0044] The embodiments of the present invention perform clustering by directly utilizing the characteristics of the signal itself, such as frequency, bandwidth, and modulation method, without relying on prior information such as the signal frame format and synchronization sequence, thus making it applicable to signal sorting scenarios in a completely blind environment.

[0045] In one embodiment, before step S4, which determines the time-domain feature parameters of the target signal block set in the second type of signal blocks, and determines the associated signal blocks in the target signal block set based on the time-domain feature parameters, the method further includes: S401. Cluster the second type of signal blocks based on signal characteristics to obtain multiple fourth type signal blocks; In this embodiment of the invention, based on the same technical concept as the clustering in step S2, the second type of signal blocks are clustered based on signal features to obtain multiple fourth type of signal blocks.

[0046] S402. Remove the fourth type of signal blocks whose number of signal blocks does not exceed the second preset threshold number to obtain the target signal block set.

[0047] In this embodiment of the invention, the fourth type of signal block to be removed refers to interference or noise signals relative to the target signal block set. For multiple types of fourth type signal blocks, each type includes at least one signal block. By counting the number of signal blocks in each type of fourth type signal block, fourth type signal blocks whose number of signal blocks does not exceed a second preset threshold are removed, while fourth type signal blocks whose number of signal blocks exceeds the second preset threshold are retained. These retained fourth type signal blocks are used as updated second type signal blocks, and the associated signal block determination step can be performed based on the updated second type signal blocks.

[0048] It's important to note that subsequent steps in determining associated signal blocks require period estimation based on the signal's initial time, such as using a Cumulative Difference Histogram (CDIF) and sequence search. Period estimation relies on a sufficient number of sample points (e.g., at least 3-5 pulses are needed for effective period detection). If a class contains only one or two signal blocks, a meaningful difference histogram cannot be calculated, nor can the existence of a fixed period be verified. Forcing period estimation with very few samples can easily lead to false periods or misassociations, contaminating the sorting results. In real-world receiving environments, isolated signal blocks often exist due to random noise floor increases, transient interference, and false detections. These blocks lack any regularity and do not belong to any genuine communication or radar signal source. These isolated subclasses are often few in number and can be efficiently eliminated using a simple size threshold, avoiding wasting resources on invalid data in subsequent complex algorithms. The second type of signal block may contain a large number of subclasses. Retaining all of them would significantly increase the number of period estimations and sequence searches (each subclass requires an attempt to estimate the period), and most searches would fail, consuming computational resources. By pre-eliminating minor categories and retaining only the fourth category of signal blocks, which are more numerous (i.e., appear frequently), resources can be concentrated on processing truly meaningful signal sources.

[0049] Clustering algorithms may split signal blocks that should belong to the same signal source into multiple tiny clusters due to noise or slight feature jitter. If these tiny clusters are directly retained, it will be difficult to merge them later, leading to signal source fragmentation. By removing these tiny clusters, it is possible to force those scattered signal blocks to not participate in subsequent sorting.

[0050] In one embodiment, the time-domain feature parameters include the signal period. Step S4, determining the time-domain feature parameters of the target signal block set in the second type of signal block, includes: S411. Sort the signal start times of the target signal block set in ascending order to obtain a signal start time sequence; In this embodiment of the invention, the signal start time ST (Start Time) is the timestamp at the beginning of each signal block, typically expressed in absolute time (ms or sampling point number). For example, the start time of receiving a pulse.

[0051] In an embodiment of the present invention, one form of expression for the signal start time sequence can be: Where N is the number of signal start times in the signal start time sequence.

[0052] S412. Based on the signal start time sequence, calculate the k-th level time difference, where the k-th level time difference is the difference between the i-th signal start time and the ik-th signal start time; i=k+1,…N; N is the number of signal start times in the signal start sequence; In this embodiment of the invention, k can be initialized to 1, and the time difference at the k-th stage can be... The expression is as follows: .

[0053] S413. Count the number of time differences belonging to different numerical intervals; By classifying continuous time differences into discrete intervals and counting them, this invention can effectively suppress measurement errors and noise fluctuations, making the periodic peaks more stably identified.

[0054] S414. Select the preset number range of values ​​with the largest number of time differences, and use the time difference value corresponding to the preset number range as the candidate time difference value. In this embodiment of the invention, the difference between several intervals with the largest number of intervals is selected as candidate periods, which effectively narrows the range of periods to be verified, thereby effectively improving the efficiency of the algorithm and avoiding the omission of multiple possible candidate periods.

[0055] S415. Determine the signal period based on the preset detection threshold and the candidate time difference.

[0056] In this embodiment of the invention, it is assumed that there are 6 signal blocks in a certain class, and their start times (in ms) are sorted as ST[100, 110, 120, 135, 145, 155]. The first-level time difference... The expression is: .

[0057] Divide the difference into intervals according to a certain resolution, for example, with 1ms as an interval, and define the interval as left closed and right open [m,m+1). Then the possible integer intervals are: [10~11), [11~12), ..., [15~16), etc.

[0058] scanning The differences between the various starting times: 10 → belongs to the interval [10~11) → count +1; 10 → belongs to the interval [10~11) → count +1; 15 → belongs to the interval [15~16) → count +1; 10 → belongs to the interval [10~11) → count +1; 10 → belongs to the interval [10~11) → count +1; In summary, the number of time differences between [10~11) appears 4 times (number of time differences between numerical intervals), and the number of time differences between [15~16) appears once (number of time differences between numerical intervals). These number of time differences between numerical intervals can be called the sample number of the starting time difference.

[0059] In this embodiment of the invention, the number of samples determined by the starting time difference of level k can be added to the number of samples in the same numerical interval of level (k-1) (k=2,3,4,N) to obtain the total number of samples in that numerical interval. For example, when there is only one level of starting time difference, the number of time differences in the above numerical interval is the statistical value of this level; as another example, the result obtained by dividing the starting time difference of level 2 into intervals and statistically analyzing it is: 20 appears twice in the interval [20, 21); 25 → the interval [25, 26) appears twice.

[0060] The number of time differences in different numerical intervals after summing the starting time difference of Level 2 and the starting time difference of Level 1 is as follows: The interval [10,11) is 4+0=4 times; the interval [15,16) is 1+0=1 time; the interval [20,21) is 0+2=4 times; and the interval [25,26) is 0+2=2 times.

[0061] In this embodiment of the invention, the preset number of numerical intervals with the largest number of time differences are selected. Among the above four numerical intervals, the preset number of numerical intervals are selected based on the statistical results of each numerical interval, i.e., the number of time differences. For example, if there are four numerical intervals with time differences of 4, 1, 4, and 2 respectively, several numerical intervals with the highest time differences can be selected. For example, if four numerical intervals are selected, all numerical intervals are selected; if two numerical intervals are selected, that is, the numerical intervals corresponding to the number of time differences of 4 and 4 are selected, i.e., [10,11) and [20,21), and then the starting time differences corresponding to [10,11) and [20,21) are determined as candidate time differences, i.e., corresponding to 10 and 20 respectively.

[0062] This invention, through sorting the signal start times and calculating multi-level time differences, performs interval statistics on the differences to suppress measurement noise, selects the most frequent differences as candidate periods, and finally determines the true signal period by combining them with a preset threshold. This allows the determination of the signal period from complex signals with aliased or missing pulses without relying on prior information, effectively avoiding false periods caused by harmonic interference and random noise. It provides highly reliable period parameters for subsequent time correlation and fixed-hop signal sorting, thereby effectively improving the accuracy and reliability of wireless signal sorting.

[0063] In one embodiment, step S415, determining the signal period based on a preset detection threshold and the candidate time difference, includes: If the number of samples corresponding to the candidate time difference and twice the number of samples corresponding to the candidate time difference both exceed a preset detection threshold, the candidate time difference is taken as the signal period.

[0064] In this embodiment of the invention, the expression for the preset detection threshold is as follows: in, To detect threshold, For periodic variables, Represents the current total number of hops. Represents a series, Represents the total number of sampling points. It is a constant.

[0065] In this embodiment of the invention, assuming the detection threshold is 2, the candidate time difference d1 is 10, and the twice-candidate time difference d2 is 20, wherein, for the above two candidate time differences: d1 = 10, sample number = 4 (>2), 2d1 = 20, sample number = 4 (>2), both exceed the preset detection threshold, so d1 is taken as the signal period; When d2 is 20, the number of samples is 4 (>2), and when 2d2 = 40, the number of samples is 0 (<2). Since there are cases where the preset detection threshold is not exceeded, d2 is not used as the signal period.

[0066] In this embodiment of the invention, it is determined whether the number of samples corresponding to the candidate time difference and twice the difference both exceed the preset detection threshold, and harmonic verification is performed to avoid misjudging integer multiples such as twice the period as the real signal period, thereby effectively improving the accuracy of period estimation and the robustness against pulse loss and noise interference.

[0067] In one embodiment, the time-domain feature parameter includes the signal period, and step S4, determining the associated signal block in the target signal block set based on the time-domain feature parameter, includes: S421. Search for associated subsequences in the signal start time sequence according to the signal period. If an associated subsequence is found, delete the associated subsequence from the signal start time sequence and repeat the associated signal block determination step. In this embodiment of the invention, a sequence search method can be used to search for and obtain the associated signal block, specifically: For example, consider an original ST sequence [0,10,20,25,35,45,60,70,80]. Calculate the adjacent differences: [10,10,5,10,10,15,10,10]. The difference histogram shows 10 appears 6 times, 5 appears once, and 15 appears once. Assuming a threshold of 3, the difference 10 exceeds the threshold, becoming the first candidate period T=10ms. Searching using T=10ms: starting from 0, find 0+10=10 → find 10 (the second pulse); 10+10=20 → find 20 (the third pulse); 20+10=30 → no 30 (actually 25, outside the tolerance), the search pauses. Retrying from 25: 25+10=35 → find 35; 35+10=45 → find 45; 45+10=55 → none; stop. Starting from 60: 60 + 10 = 70 → 70; 70 + 10 = 80 → 80; resulting in the subsequence [60, 70, 80]. Ultimately, we obtain multiple subsequences: [0, 10, 20], [25, 35, 45], [60, 70, 80]. Each subsequence has a length ≥ 3, indicating a successful search. After deleting these pulses from the original sequence, there are no ST sequences, at which point the algorithm terminates.

[0068] In this embodiment of the invention, if several STs remain after deleting the associated subsequence from the signal start time sequence, the associated signal block determination step is repeated, including taking the remaining ST sequence as a new signal start time sequence, re-determining the signal period, and determining the associated signal block in the remaining ST sequence according to the signal period.

[0069] In this invention, if the search fails, the next ST difference value exceeding the detection threshold in the current level difference histogram is taken as the period for searching the associated subsequence.

[0070] S422. If the search stopping condition is met, the signal block corresponding to the associated subsequence is determined as the associated signal block.

[0071] In this embodiment of the invention, the search stopping condition can be: all candidate cycles fail to be searched or the number of remaining ST sequences is less than the minimum number of pulses.

[0072] This invention, based on a signal periodic correlation subsequence search and iterative deletion mechanism, can efficiently separate complete pulse subsequences from multiple overlapping signal sources with the same period. Even with pulse loss or measurement errors, correlated signal blocks can still be successfully extracted. Simultaneously, by repeatedly executing period estimation and sequence search steps, it ensures that hidden periodic signals in the remaining ST sequences can also be extracted one by one, avoiding omissions. When a single candidate period search fails, the next candidate is automatically tried, significantly improving the robustness and anti-interference capability of period detection. Furthermore, without prior information, it can accurately and completely determine temporally correlated signal blocks from complex overlapping second-type signal blocks, providing a reliable data foundation for subsequent fixed-frequency / frequency-hopping discrimination.

[0073] In one embodiment, the parameter information includes the signal frequency. Step S5, determining the final signal block category based on the parameter information, includes: S51. Calculate the frequency variance in the associated signal block based on the signal frequency. If the frequency variance is less than a preset frequency variance threshold, determine the associated signal block as a fixed-frequency signal block. In this embodiment of the invention, the preset frequency variance threshold is a pre-set threshold value used to distinguish between fixed frequency and frequency hopping. This threshold can be obtained through theoretical calculation (e.g., considering the variance of receiver frequency measurement error) or statistical learning of typical signals.

[0074] S52. If the frequency variance is greater than or equal to a preset frequency variance threshold, the associated signal block is determined to be a frequency hopping signal.

[0075] The embodiments of the present invention can distinguish between fixed frequency and frequency hopping by utilizing frequency statistical characteristics, and are suitable for wireless signal sorting in complex electromagnetic environments.

[0076] Please see Figure 2 This is a second schematic diagram of a wireless signal sorting method provided in an embodiment of the present invention. Figure 2 As shown, the input signal block to be processed is used to determine the first type of signal block and the second type of signal block through coarse classification. For the first type of signal block, the fixed frequency signal block is determined by feature clustering. For the second type of signal block, the denoised second type of signal block is determined by feature clustering. Then, through signal period estimation, determination of associated signal blocks and frequency variance calculation, the fixed frequency signal block and the frequency hopping signal block are finally determined.

[0077] Implementing the embodiments of the present invention has the following beneficial effects: This invention distinguishes between first-class and second-class signal blocks by coarsely classifying multiple signal blocks to be processed based on their duration. Then, unsupervised clustering is performed on the signal features of the first-class signal blocks to obtain multiple third-class signal blocks. The third-class signal blocks with more than a first preset threshold are directly identified as fixed-frequency signal blocks, while the third-class signal blocks with fewer than the first preset threshold are classified as second-class signal blocks. Furthermore, the associated signal blocks among the second-class signal blocks are determined based on their time-domain feature parameters. The entire sorting process does not rely on prior information such as frame format, synchronization sequence, or modulation method. It can sort fixed-frequency signals and frequency-hopping signals using only the signal features of the signal blocks themselves, thereby meeting the signal sorting requirements in completely blind scenarios and effectively improving the adaptability of signal sorting in complex and unknown electromagnetic environments.

[0078] Furthermore, in this embodiment of the invention, after sorting the signal start times, multi-level time differences are calculated, and interval statistics are performed on the differences to suppress measurement noise. Then, several differences that occur most frequently are selected as candidate periods. Finally, the true signal period is determined by combining them with a preset threshold. This method can determine the signal period from complex signals with aliased or missing pulses without relying on prior information. It effectively avoids false periods caused by harmonic interference and random noise, and provides highly reliable period parameters for subsequent time correlation and fixed-hop signal sorting. This can effectively improve the accuracy and reliability of wireless signal sorting.

[0079] The wireless signal sorting device provided by the present invention is described below. The wireless signal sorting device described below and the wireless signal sorting method described above can be referred to in correspondence.

[0080] Please see Figure 3 This invention provides a wireless signal sorting device, comprising: The first classification module 310 is used to perform coarse classification on multiple signal blocks to be processed to obtain first-class signal blocks and second-class signal blocks. Clustering module 320 is used to perform clustering processing on the first type of signal blocks based on signal features to obtain multiple third type of signal blocks; The second classification module 330 is used to identify third-class signal blocks whose number of signal blocks exceeds the first preset threshold as fixed-frequency signal blocks, and to identify third-class signal blocks whose number of signal blocks does not exceed the first preset threshold as second-class signal blocks; The associated signal block determination module 340 is used to determine the time-domain feature parameters of the target signal block set in the second type of signal blocks, and determine the associated signal blocks in the target signal block set based on the time-domain feature parameters; The third classification module 350 is used to extract parameter information from the associated signal block and determine the final signal block category based on the parameter information.

[0081] In one embodiment, the coarse classification of multiple signal blocks to be processed to obtain a first type of signal block and a second type of signal block includes: Signal blocks whose duration exceeds a preset threshold time are identified as first-class signal blocks; Signal blocks whose duration is less than or equal to a preset threshold time are identified as second-class signal blocks.

[0082] In one embodiment, the signal features include at least one of frequency, bandwidth, and modulation scheme, and the clustering of the first type of signal blocks based on the signal features to obtain multiple third type of signal blocks includes: A feature vector is generated based on the signal characteristics of the signal blocks in the first type of signal block; The feature vector of the first signal block is used as the initial class center; Starting from the second signal block, calculate the Euclidean distance between the signal features in the current signal block and the signal features in the initial class center. Convert the Euclidean distance into a confidence score. If the confidence score is greater than a preset confidence threshold, classify the current signal block into the current category and update the initial class center to obtain a new class center. If the confidence score is less than or equal to the preset confidence threshold, classify the signal block into a new category and use the feature vector of the signal block as the initial class center of the new category. By traversing all signal blocks of the first type of signal block, multiple signal blocks of the third type are obtained.

[0083] In one embodiment, before determining the time-domain feature parameters of the target signal block set in the second type of signal blocks, and determining the associated signal blocks in the target signal block set based on the time-domain feature parameters, the method further includes: Based on signal characteristics, the second type of signal blocks are clustered to obtain multiple fourth type signal blocks; The fourth type of signal blocks, whose number of signal blocks does not exceed the second preset threshold, are removed to obtain the target signal block set.

[0084] In one embodiment, the time-domain feature parameters include the signal period, and determining the time-domain feature parameters of the target signal block set includes: The signal start times of the target signal block set are sorted in ascending order to obtain a signal start time sequence. Based on the signal start time sequence, calculate the k-th level time difference, where the k-th level time difference is the difference between the i-th signal start time and the ik-th signal start time; i = k+1, ..., N; N is the number of signal start times in the signal start sequence; Count the number of time differences belonging to different numerical intervals; Select the preset number range of values ​​with the largest number of time differences, and use the time differences corresponding to the preset number range as candidate time differences; The signal period is determined based on the preset detection threshold and the difference between the candidate times.

[0085] In one embodiment, determining the signal period based on a preset detection threshold and the candidate time difference includes: If the number of samples corresponding to the candidate time difference and twice the number of samples corresponding to the candidate time difference both exceed a preset detection threshold, the candidate time difference is taken as the signal period.

[0086] In one embodiment, the time-domain feature parameter includes the signal period, and determining the associated signal block in the target signal block set based on the time-domain feature parameter includes: According to the signal period, a related subsequence search is performed in the signal start time sequence. If a related subsequence is found, the related subsequence is deleted from the signal start time sequence, and the related signal block determination step is repeated. If the search stopping condition is met, the signal block corresponding to the associated subsequence is determined as the associated signal block.

[0087] In one embodiment, the parameter information includes a signal frequency, and determining the final signal block category based on the parameter information includes: Based on the signal frequency, calculate the frequency variance in the associated signal block. If the frequency variance is less than a preset frequency variance threshold, determine the associated signal block as a fixed-frequency signal block. If the frequency variance is greater than or equal to a preset frequency variance threshold, the associated signal block is determined to be a frequency hopping signal.

[0088] Implementing the embodiments of the present invention has the following beneficial effects: This invention distinguishes between first-class and second-class signal blocks by coarsely classifying multiple signal blocks to be processed based on their duration. Then, unsupervised clustering is performed on the signal features of the first-class signal blocks to obtain multiple third-class signal blocks. The third-class signal blocks with more than a first preset threshold are directly identified as fixed-frequency signal blocks, while the third-class signal blocks with fewer than the first preset threshold are classified as second-class signal blocks. Furthermore, the associated signal blocks among the second-class signal blocks are determined based on their time-domain feature parameters. The entire sorting process does not rely on prior information such as frame format, synchronization sequence, or modulation method. It can sort fixed-frequency signals and frequency-hopping signals using only the signal features of the signal blocks themselves, thereby meeting the signal sorting requirements in completely blind scenarios and effectively improving the adaptability of signal sorting in complex and unknown electromagnetic environments.

[0089] Furthermore, in this embodiment of the invention, after sorting the signal start times, multi-level time differences are calculated, and interval statistics are performed on the differences to suppress measurement noise. Then, several differences that occur most frequently are selected as candidate periods. Finally, the true signal period is determined by combining them with a preset threshold. This method can determine the signal period from complex signals with aliased or missing pulses without relying on prior information. It effectively avoids false periods caused by harmonic interference and random noise, and provides highly reliable period parameters for subsequent time correlation and fixed-hop signal sorting. This can effectively improve the accuracy and reliability of wireless signal sorting.

[0090] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device may include: a processor 410, a communications interface 420, a memory 430, and a communication bus 440, wherein the processor 410, the communications interface 420, and the memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute a wireless signal sorting method, including: Multiple signal blocks to be processed are coarsely classified to obtain first-class signal blocks and second-class signal blocks; Based on signal characteristics, the first type of signal blocks are clustered to obtain multiple third type signal blocks; The third type of signal block with a number of signal blocks exceeding the first preset threshold is defined as a fixed-frequency signal block, and the third type of signal block with a number of signal blocks not exceeding the first preset threshold is defined as a second type of signal block; The steps for determining associated signal blocks are as follows: determine the time-domain characteristic parameters of the target signal block set in the second type of signal blocks, and determine the associated signal blocks in the target signal block set based on the time-domain characteristic parameters; Extract the parameter information from the associated signal block, and determine the final signal block category based on the parameter information.

[0091] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0092] On the other hand, the present invention also provides a computer program product, the computer program product comprising a computer program, the computer program being able to be stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer being able to execute the wireless signal sorting methods provided by the above methods, including: Multiple signal blocks to be processed are coarsely classified to obtain first-class signal blocks and second-class signal blocks; Based on signal characteristics, the first type of signal blocks are clustered to obtain multiple third type signal blocks; The third type of signal block with a number of signal blocks exceeding the first preset threshold is defined as a fixed-frequency signal block, and the third type of signal block with a number of signal blocks not exceeding the first preset threshold is defined as a second type of signal block; The steps for determining associated signal blocks are as follows: determine the time-domain characteristic parameters of the target signal block set in the second type of signal blocks, and determine the associated signal blocks in the target signal block set based on the time-domain characteristic parameters; Extract the parameter information from the associated signal block, and determine the final signal block category based on the parameter information.

[0093] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the wireless signal sorting method provided by the methods described above, including: Multiple signal blocks to be processed are coarsely classified to obtain first-class signal blocks and second-class signal blocks; Based on signal characteristics, the first type of signal blocks are clustered to obtain multiple third type signal blocks; The third type of signal block with a number of signal blocks exceeding the first preset threshold is defined as a fixed-frequency signal block, and the third type of signal block with a number of signal blocks not exceeding the first preset threshold is defined as a second type of signal block; The steps for determining associated signal blocks are as follows: determine the time-domain characteristic parameters of the target signal block set in the second type of signal blocks, and determine the associated signal blocks in the target signal block set based on the time-domain characteristic parameters; Extract the parameter information from the associated signal block, and determine the final signal block category based on the parameter information.

[0094] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0095] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A wireless signal sorting method, characterized in that, include: Multiple signal blocks to be processed are coarsely classified to obtain first-class signal blocks and second-class signal blocks; Based on signal characteristics, the first type of signal blocks are clustered to obtain multiple third type signal blocks; The third type of signal block with a number of signal blocks exceeding the first preset threshold is defined as a fixed-frequency signal block, and the third type of signal block with a number of signal blocks not exceeding the first preset threshold is defined as a second type of signal block; The steps for determining associated signal blocks are as follows: determine the time-domain characteristic parameters of the target signal block set in the second type of signal blocks, and determine the associated signal blocks in the target signal block set based on the time-domain characteristic parameters; Extract the parameter information from the associated signal block, and determine the final signal block category based on the parameter information.

2. The wireless signal sorting method as described in claim 1, characterized in that, The coarse classification of multiple signal blocks to be processed to obtain a first type of signal block and a second type of signal block includes: Signal blocks whose duration exceeds a preset threshold time are identified as first-class signal blocks; Signal blocks whose duration is less than or equal to a preset threshold time are identified as second-class signal blocks.

3. The wireless signal sorting method as described in claim 1, characterized in that, The signal features include at least one of frequency, bandwidth, and modulation scheme. The clustering of the first type of signal blocks based on these signal features to obtain multiple third type signal blocks includes: A feature vector is generated based on the signal characteristics of the signal blocks in the first type of signal block; The feature vector of the first signal block is used as the initial class center; Starting from the second signal block, calculate the Euclidean distance between the signal features in the current signal block and the signal features in the initial class center. Convert the Euclidean distance into a confidence score. If the confidence score is greater than a preset confidence threshold, classify the current signal block into the current category and update the initial class center to obtain a new class center. If the confidence score is less than or equal to the preset confidence threshold, classify the signal block into a new category and use the feature vector of the signal block as the initial class center of the new category. By traversing all signal blocks of the first type of signal block, multiple signal blocks of the third type are obtained.

4. The wireless signal sorting method as described in claim 1, characterized in that, Before determining the time-domain feature parameters of the target signal block set in the second type of signal block, and determining the associated signal blocks in the target signal block set based on the time-domain feature parameters, the method further includes: Based on signal characteristics, the second type of signal blocks are clustered to obtain multiple fourth type signal blocks; The fourth type of signal blocks, whose number of signal blocks does not exceed the second preset threshold, are removed to obtain the target signal block set.

5. The wireless signal sorting method as described in claim 1, characterized in that, The time-domain feature parameters include the signal period, and the time-domain feature parameters for determining the target signal block set in the second type of signal block include: The signal start times of the target signal block set are sorted in ascending order to obtain a signal start time sequence. Based on the signal start time sequence, calculate the k-th level time difference, where the k-th level time difference is the difference between the i-th signal start time and the ik-th signal start time; i = k+1, ..., N; N is the number of signal start times in the signal start sequence; Count the number of time differences belonging to different numerical intervals; Select the preset number range of values ​​with the largest number of time differences, and use the time differences corresponding to the preset number range as candidate time differences; The signal period is determined based on the preset detection threshold and the difference between the candidate times.

6. The wireless signal sorting method as described in claim 5, characterized in that, The step of determining the signal period based on the preset detection threshold and the candidate time difference includes: If the number of samples corresponding to the candidate time difference and twice the number of samples corresponding to the candidate time difference both exceed a preset detection threshold, the candidate time difference is taken as the signal period.

7. The wireless signal sorting method as described in claim 5, characterized in that, The time-domain feature parameters include the signal period, and determining the associated signal blocks in the target signal block set based on the time-domain feature parameters includes: According to the signal period, a related subsequence search is performed in the signal start time sequence. If a related subsequence is found, the related subsequence is deleted from the signal start time sequence, and the related signal block determination step is repeated. If the search stopping condition is met, the signal block corresponding to the associated subsequence is determined as the associated signal block.

8. The wireless signal sorting method as described in claim 1, characterized in that, The parameter information includes the signal frequency, and determining the final signal block category based on the parameter information includes: Based on the signal frequency, calculate the frequency variance in the associated signal block. If the frequency variance is less than a preset frequency variance threshold, determine the associated signal block as a fixed-frequency signal block. If the frequency variance is greater than or equal to a preset frequency variance threshold, the associated signal block is determined to be a frequency hopping signal.

9. A wireless signal sorting device, characterized in that, include: The first classification module is used to perform coarse classification on multiple signal blocks to be processed, and obtain the first type of signal blocks and the second type of signal blocks. The clustering module is used to cluster the first type of signal blocks based on signal features to obtain multiple third type signal blocks; The second classification module is used to identify third-class signal blocks whose number of signal blocks exceeds the first preset threshold as fixed-frequency signal blocks, and to identify third-class signal blocks whose number of signal blocks does not exceed the first preset threshold as second-class signal blocks; The associated signal block determination module is used to determine the time-domain feature parameters of the target signal block set in the second type of signal blocks, and determine the associated signal blocks in the target signal block set based on the time-domain feature parameters; The third classification module is used to extract parameter information from the associated signal blocks and determine the final signal block category based on the parameter information.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the wireless signal sorting method as described in any one of claims 1 to 8.

Images