A radio signal feature automatic extraction system and extraction method thereof

Abstract

The present invention discloses a radio signal feature automatic extraction system and an extraction method thereof, which relates to the technical field of radio signal feature extraction, including calculating the sample signal quality coefficient of each group of sample radio signal data and the quality threshold of the radio signal, screening the main influencing features of the signal quality and their feature weights; when the quality coefficient of the signal to be tested is greater than the quality threshold of the radio signal, sending out a radio signal interference analysis warning, calculating the interference degree of each main influencing feature of the signal quality, and obtaining the interference weight of each main influencing feature of the signal quality; after receiving the radio signal interference analysis warning, correcting the main influencing features of the signal quality in turn, judging whether the quality coefficient of the signal to be tested of the corrected radio signal data to be tested is less than the quality threshold of the radio signal, until the quality coefficient of the signal to be tested is less than the quality threshold of the radio signal. The overall quality of the signal can be gradually improved, and unnecessary correction operations can be avoided.

Description

Technical Field

[0001] The invention relates to the technical field of radio signal feature extraction, and in particular to a method for automatically extracting radio signal features. Background Art

[0002] With the rapid development of wireless communication technology, radio signals play an increasingly important role in daily life, industrial production, military applications and other fields. However, radio signals are easily affected by various interference factors during transmission, such as multipath effects, noise interference, signal attenuation, etc. These factors will lead to a decrease in signal quality, thereby affecting the reliability and stability of communication. In recent years, with the rapid development of artificial intelligence technologies such as machine learning and deep learning, new ideas and methods have been provided for the automatic extraction of radio signal features. These technologies can automatically learn the feature representation of signals through training models, thereby realizing automatic classification, recognition and extraction of signals. Compared with traditional methods, the automatic extraction method of radio signal features based on artificial intelligence has higher accuracy and faster processing speed, and can better adapt to complex and changing wireless communication environments.

[0003] In the Chinese invention application with application publication number CN113225145A, a method for visualizing multi-dimensional time-varying features of radio signals is disclosed, including obtaining radio signal data; classifying the obtained radio signals; calculating the average center frequency, average bandwidth, average signal strength and average carrier-to-noise ratio of each type of radio signal; dividing the time period into time slices, and calculating the average signal strength and average carrier-to-noise ratio of each time slice; drawing a frequency-time summary diagram of radio signal data and performing visual encoding; and drawing a signal flow diagram of radio signal data. By visualizing the changes of multi-dimensional features of radio signals over time and presenting the features of radio signals in a radio signal river diagram, users can obtain all the features of the required radio signals.

[0004] In the above invention application, the complete radio signal graph is divided into several visual coding units, and the "Morse code" signal summary visualization visual coding design allows users to have an overall understanding of the appearance and disappearance of the radio signal in a short time. Since this method mainly focuses on the presence or absence of the signal (i.e., appearance and disappearance) and does not care about the signal quality, it may ignore important information about the signal quality, such as signal strength, signal-to-noise ratio, phase stability, etc., and thus cannot fully understand the actual situation of the signal, resulting in misjudgment of the communication quality, which in turn affects the accuracy of the decision.

[0005] To this end, the present invention provides a radio signal feature automatic extraction system and an extraction method thereof. Summary of the invention

[0006] In view of the shortcomings of the prior art, the present invention provides a radio signal feature automatic extraction system and method thereof. The present invention selects the main influencing features of signal quality and their feature weights. , calculate the interference degree of each signal quality main influencing feature , obtain the interference weight of each signal quality main influencing feature , which helps to accurately identify the key factors that cause signal quality degradation, optimize resource allocation according to the weight, and invest more resources in monitoring, analyzing and processing features with higher interference weights, thereby improving the anti-interference ability and stability of the overall system. After receiving the radio signal interference analysis warning, correct the main influencing features of signal quality in turn, and determine the signal quality coefficient of the corrected radio signal data to be tested Is it less than the quality threshold Z1 of the radio signal until the signal quality coefficient to be tested When the quality threshold Z1 of the radio signal is less than the quality threshold, the overall quality of the signal can be gradually improved, and unnecessary correction operations can be avoided. If the signal quality has reached the standard, the correction process can be stopped, thereby saving time and resources, effectively reducing the impact of interference on the signal, and improving the stability and reliability of the signal, thereby solving the technical problems recorded in the background technology.

[0007] To achieve the above object, the present invention is implemented by the following technical scheme: a method for automatically extracting radio signal features, comprising the following steps:

[0008] Construct a standard database and sample data set for radio signals, and calculate the sample signal quality coefficient for each set of sample radio signal data and the quality threshold Z1 of the radio signal, and calculate each radio signal feature and signal quality coefficient The absolute value of the Pearson correlation coefficient is used to screen the main influencing features of signal quality and their feature weights ;

[0009] When the signal quality factor When the signal quality threshold Z1 is greater than the radio signal, a radio signal interference analysis warning is issued to extract the main influencing features of the signal quality of the radio signal to be tested. , calculate the interference degree of each signal quality main influencing feature , obtain the interference weight of each signal quality main influencing feature ;

[0010] After receiving the radio signal interference analysis warning, correct the main influencing features of the signal quality in turn, and determine the signal quality coefficient of the corrected radio signal data to be tested.

[0011] Is it less than the quality threshold Z1 of the radio signal until the signal quality coefficient to be tested

[0012] is less than the quality threshold Z1 of the radio signal.

[0013] Furthermore, a sample data set including radio signal characteristics and signal quality indicators is collected, and a sample signal quality coefficient of each set of sample radio signal data is calculated. :

[0014]

[0015] in, a Indicates the sequential number of each group of sample radio signal data in the sample data set, is the signal-to-noise ratio of the sample radio signal, represents the bit error rate of the sample radio signal, representing the signal-to-noise ratio of a standard radio signal in a standard database of radio signals, representing the bit error rate of a standard radio signal in a standard database of radio signals, is a positive number used to prevent division by zero errors, It is the midpoint offset of the Sigmoid function in the signal-to-noise ratio part, which is used to adjust the center position of the influence of the signal-to-noise ratio on the quality coefficient. It is the midpoint offset of the Sigmoid function in the bit error rate part, which is used to adjust the center position of the influence of the bit error rate on the quality coefficient.

[0016] Furthermore, the sample signal quality coefficient mean of the sample radio signal data is used and standard deviation To calculate the quality threshold Z1 of the radio signal, it is as follows:

[0017]

[0018] where 1.645 is obtained from the standard normal distribution table and corresponds to a 2.5% tail probability for a two-sided test.

[0019] Furthermore, a sample data set including radio signal characteristics and sample signal quality coefficients is included. Import SPSS analysis software to calculate the characteristics and signal quality coefficient of each radio signal The absolute value of the Pearson correlation coefficient is used to sort the top 10 features from large to small and select them as the main influencing features of signal quality. The absolute value of the Pearson correlation coefficient corresponding to the main influencing features of signal quality is recorded as the feature weight. .

[0020] The Pearson Correlation Coefficient is the most commonly used correlation measurement method. It is used to measure the linear relationship between two continuous variables. Its value range is between -1 and 1, where 1 indicates a perfect positive correlation, -1 indicates a perfect negative correlation, and 0 indicates no correlation.

[0021] Furthermore, a receiving antenna is used to receive the radio signal to be tested, and a digital signal processing technique is used to analyze the signal-to-noise ratio of the radio signal to be tested. and bit error rate , calculate the signal quality coefficient of the radio signal data to be tested :

[0022]

[0023] in, i Indicates the reception time of the radio signal data, i =1, 2, …, n。

[0024] Furthermore, when the signal quality factor to be measured is When the signal quality threshold Z1 is greater than the radio signal, a radio signal interference analysis warning is issued to extract the main influencing features of the signal quality of the radio signal to be tested. , combined with feature weights and a standard database of radio signals to calculate the degree of interference for each of the main influencing features of signal quality :

[0025]

[0026] in,

[0027] k

[0028] The number of the main influencing feature of signal quality,

[0029] k=

[0030] 1, 2, ..., 10,

[0031] Indicates the standard value of the main influencing feature of signal quality.

[0032] Furthermore, the interference degree of each main influencing feature of the signal quality of all radio signals to be tested is counted The mean of the two is recorded as the interference weight of each main influencing feature of signal quality. .

[0033] Furthermore, after receiving the radio signal interference analysis warning, the interference weight of each main influencing feature of signal quality is Sort from large to small, start correcting the main influencing characteristics of signal quality from the first place, and perform digital filtering, equalization or interpolation algorithm on the radio signal to be tested according to the correction operation of the main influencing characteristics of signal quality, so as to correct the main influencing characteristics of signal quality to the standard value.

[0034] For example, signal strength correction uses automatic gain control (AGC), frequency offset correction uses automatic frequency control (AFC) and frequency synchronization, and phase error correction uses phase synchronization, phase estimation and compensation.

[0035] Digital filtering is the process of processing a signal through a specific algorithm to remove unwanted frequency components or noise while retaining or enhancing the desired signal characteristics. Common digital filters include low-pass filters, high-pass filters, band-pass filters, and band-stop filters.

[0036] Equalization is a signal processing technique used to compensate for distortion caused by channel characteristics (such as multipath effects, uneven frequency response, etc.) during signal transmission. An equalizer is usually an adjustable filter whose parameters are adjusted according to the channel characteristics.

[0037] Interpolation is a method of estimating the value of a continuous function between discrete data points. In signal processing, interpolation is often used in scenarios such as resampling, signal recovery, and error correction.

[0038] Further, after the correction of the main influencing feature of the first signal quality is completed, the signal quality coefficient of the corrected radio signal data to be tested is determined. Is it less than the quality threshold Z1 of the radio signal? If it is less than the quality threshold Z1 of the radio signal, the interference correction process is stopped; if it is not less than the quality threshold Z1 of the radio signal, the next correction signal quality main influencing feature is corrected in sequence until the signal quality coefficient to be measured is is less than the quality threshold Z1 of the radio signal.

[0039] A radio signal feature automatic extraction system, comprising:

[0040] The main influencing feature screening module builds a standard database and sample data set for radio signals and calculates the sample signal quality coefficient of each set of sample radio signal data and the quality threshold Z1 of the radio signal, and calculate each radio signal feature and signal quality coefficient The absolute value of the Pearson correlation coefficient is used to screen the main influencing features of signal quality and their feature weights ;

[0041] Interference weight analysis module, when the signal quality coefficient to be tested When the signal quality threshold Z1 is greater than the radio signal, a radio signal interference analysis warning is issued to extract the main influencing features of the signal quality of the radio signal to be tested. , calculate the interference degree of each signal quality main influencing feature , obtain the interference weight of each signal quality main influencing feature ;

[0042] The interference correction module, after receiving the radio signal interference analysis warning, corrects the main influencing features of the signal quality in turn, and determines the quality coefficient of the measured signal of the corrected radio signal data to be tested

[0043] Is it less than the quality threshold Z1 of the radio signal until the signal quality coefficient to be tested

[0044] is less than the quality threshold Z1 of the radio signal.

[0045] The present invention provides a radio signal feature automatic extraction system and extraction method thereof, which have the following beneficial effects:

[0046] 1. Build a standard database and sample data set for radio signals, and calculate the sample signal quality coefficient of each set of sample radio signal data and the quality threshold Z1 of the radio signal, and calculate each radio signal feature and signal quality coefficient The absolute value of the Pearson correlation coefficient is used to screen the main influencing features of signal quality and their feature weights By quantitatively evaluating signal quality and screening the main influencing features and their weights, interference features can be identified more accurately, providing a basis for subsequent optimization of interference features and rapid response to signal quality issues.

[0047] 2. When the quality coefficient of the signal to be measured When the signal quality threshold Z1 is greater than the radio signal, a radio signal interference analysis warning is issued to extract the main influencing features of the signal quality of the radio signal to be tested. , calculate the interference degree of each signal quality main influencing feature , obtain the interference weight of each main influencing feature of signal quality It helps to accurately identify the key factors that lead to signal quality degradation and optimize resource allocation according to the weight. For features with higher interference weights, more resources can be invested in monitoring, analysis and processing, thereby improving the overall system's anti-interference ability and stability.

[0048] 3. After receiving the radio signal interference analysis warning, correct the main influencing features of the signal quality in turn, and determine the signal quality coefficient of the corrected radio signal data to be tested

[0049] Is it less than the quality threshold Z1 of the radio signal until the signal quality coefficient to be tested

[0050] When the quality threshold Z1 of the radio signal is less than the threshold value, the overall quality of the signal can be gradually improved to avoid unnecessary correction operations. If the signal quality has reached the standard, the correction process can be stopped, thus saving time and resources, effectively reducing the impact of interference on the signal, and improving the stability and reliability of the signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A schematic diagram of a flow chart of a method for automatically extracting radio signal features according to the present invention;

[0052] Figure 2 The present invention is a schematic diagram of the structure of a radio signal feature automatic extraction system. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] See also Figure 1 The present invention provides a method for automatically extracting radio signal features, comprising the following steps:

[0055] Step 1: Build a standard database and sample data set for radio signals, and calculate the sample signal quality coefficient for each set of sample radio signal data and the quality threshold Z1 of the radio signal, and calculate each radio signal feature and signal quality coefficient The absolute value of the Pearson correlation coefficient is used to screen the main influencing features of signal quality and their feature weights .

[0056] The step 1 includes the following contents:

[0057] Step 101: extract the features of radio signals under normal transmission conditions, including time domain features, frequency domain features, high-level features, space and time distribution features, and signal quality features, and construct a standard database of radio signals.

[0058] Time domain feature extraction includes calculating the signal's mean, variance, root mean square, peak factor, kurtosis coefficient and other time domain statistics; frequency domain feature extraction includes using frequency domain analysis methods such as Fourier transform to extract the signal's frequency distribution, bandwidth, centroid frequency, mean square frequency and other frequency domain features; advanced feature extraction includes using modern signal processing methods such as wavelet transform, empirical mode decomposition (EMD), blind signal processing (BSP) to further extract the signal's high-order features; spatial and temporal distribution features include extracting its position distribution, energy distribution and other features; signal quality features include signal strength, signal-to-noise ratio, and bit error rate.

[0059] Step 102: collect a sample data set containing radio signal characteristics and signal quality indicators through experimental measurement, simulation or recording of an actual communication system, and calculate a sample signal quality coefficient for each set of sample radio signal data. :

[0060]

[0061] in,

[0062] a

[0063] Indicates the sequential number of each group of sample radio signal data in the sample data set,

[0064] is the signal-to-noise ratio of the sample radio signal,

[0065] represents the bit error rate of the sample radio signal,

[0066] representing the signal-to-noise ratio of a standard radio signal in a standard database of radio signals,

[0067] representing the bit error rate of a standard radio signal in a standard database of radio signals,

[0068] is a positive number used to prevent division by zero errors,

[0069] It is the midpoint offset of the Sigmoid function in the signal-to-noise ratio part, which is used to adjust the center position of the influence of the signal-to-noise ratio on the quality coefficient.

[0070] It is the midpoint offset of the Sigmoid function in the bit error rate part, which is used to adjust the center position of the influence of the bit error rate on the quality coefficient.

[0071] Step 103: Using the sample signal quality coefficient mean of the sample radio signal data and standard deviation To calculate the quality threshold Z1 of the radio signal, it is as follows:

[0072]

[0073] where 1.645 is obtained from the standard normal distribution table and corresponds to a 2.5% tail probability for a two-sided test.

[0074] Step 104: The sample data set containing the radio signal characteristics and the sample signal quality coefficient Import SPSS analysis software to calculate the characteristics and signal quality coefficient of each radio signal The absolute value of the Pearson correlation coefficient is used to sort the top 10 features from large to small and select them as the main influencing features of signal quality. The absolute value of the Pearson correlation coefficient corresponding to the main influencing features of signal quality is recorded as the feature weight. .

[0075] The Pearson Correlation Coefficient is the most commonly used correlation measurement method. It is used to measure the linear relationship between two continuous variables. Its value range is between -1 and 1, where 1 indicates a perfect positive correlation, -1 indicates a perfect negative correlation, and 0 indicates no correlation.

[0076] When using, combine the contents in steps 101 to 104:

[0077] Construct a standard database and sample data set for radio signals, and calculate the sample signal quality coefficient for each set of sample radio signal data

[0078] and the quality threshold Z1 of the radio signal, and calculate each radio signal feature and signal quality coefficient

[0079] The absolute value of the Pearson correlation coefficient is used to screen the main influencing features of signal quality and their feature weights

[0080] By quantitatively evaluating signal quality and screening the main influencing features and their weights, interference features can be identified more accurately, providing a basis for subsequent optimization of interference features and rapid response to signal quality issues.

[0081] Step 2: When the quality coefficient of the signal to be tested is When the signal quality threshold Z1 is greater than the radio signal, a radio signal interference analysis warning is issued to extract the main influencing features of the signal quality of the radio signal to be tested. , calculate the interference degree of each signal quality main influencing feature , obtain the interference weight of each signal quality main influencing feature .

[0082] The step 2 includes the following contents:

[0083] Step 201: Use a receiving antenna to receive the radio signal to be tested, and use digital signal processing technology to analyze the signal-to-noise ratio of the radio signal to be tested. and bit error rate , calculate the signal quality coefficient of the radio signal data to be tested :

[0084]

[0085] in, i Indicates the reception time of the radio signal data, i =1, 2, …, n。

[0086] Step 202: When the quality coefficient of the signal to be measured is When the signal quality threshold Z1 is greater than the radio signal, a radio signal interference analysis warning is issued to extract the main influencing features of the signal quality of the radio signal to be tested. , combined with feature weights and a standard database of radio signals to calculate the degree of interference for each of the main influencing features of signal quality :

[0087]

[0088] in, k The number of the main influencing feature of signal quality, k= 1, 2, ..., 10, Indicates the standard value of the main influencing feature of signal quality.

[0089] Step 203: Count the interference degree of each main influencing feature of the signal quality of all radio signals to be tested The mean of the two is recorded as the interference weight of each main influencing feature of signal quality. .

[0090] When using, combine the contents in steps 201 to 203:

[0091] When the signal quality factor

[0092] When the signal quality threshold Z1 is greater than the radio signal, a radio signal interference analysis warning is issued to extract the main influencing features of the signal quality of the radio signal to be tested.

[0093] , calculate the interference degree of each signal quality main influencing feature

[0094] , obtain the interference weight of each main influencing feature of signal quality

[0095] It helps to accurately identify the key factors that lead to signal quality degradation and optimize resource allocation according to the weight. For features with higher interference weights, more resources can be invested in monitoring, analysis and processing, thereby improving the overall system's anti-interference ability and stability.

[0096] Step 3: After receiving the radio signal interference analysis warning, correct the main influencing features of the signal quality in turn, and determine the signal quality coefficient of the corrected radio signal data to be tested. Is it less than the quality threshold Z1 of the radio signal until the signal quality coefficient to be tested is less than the quality threshold Z1 of the radio signal.

[0097] The step three includes the following contents:

[0098] Step 301: After receiving the radio signal interference analysis warning, the interference weight of each signal quality main influencing feature is calculated. Sort from large to small, start correcting the main influencing characteristics of signal quality from the first place, and perform digital filtering, equalization or interpolation algorithm on the radio signal to be tested according to the correction operation of the main influencing characteristics of signal quality, so as to correct the main influencing characteristics of signal quality to the standard value.

[0099] For example, signal strength correction uses automatic gain control (AGC), frequency offset correction uses automatic frequency control (AFC) and frequency synchronization, and phase error correction uses phase synchronization, phase estimation and compensation.

[0100] Digital filtering is the process of processing a signal through a specific algorithm to remove unwanted frequency components or noise while retaining or enhancing the desired signal characteristics. Common digital filters include low-pass filters, high-pass filters, band-pass filters, and band-stop filters.

[0101] Equalization is a signal processing technique used to compensate for distortion caused by channel characteristics (such as multipath effects, uneven frequency response, etc.) during signal transmission. An equalizer is usually an adjustable filter whose parameters are adjusted according to the channel characteristics.

[0102] Interpolation is a method of estimating the value of a continuous function between discrete data points. In signal processing, interpolation is often used in scenarios such as resampling, signal recovery, and error correction.

[0103] Step 302: After the correction of the main influencing feature of the first signal quality is completed, the signal quality coefficient of the corrected radio signal data to be tested is determined. Is it less than the quality threshold Z1 of the radio signal? If it is less than the quality threshold Z1 of the radio signal, the interference correction process is stopped; if it is not less than the quality threshold Z1 of the radio signal, the next correction signal quality main influencing feature is corrected in sequence until the signal quality coefficient to be measured is is less than the quality threshold Z1 of the radio signal.

[0104] When used, combine the contents in steps 301 and 302:

[0105] After receiving the radio signal interference analysis warning, correct the main influencing features of the signal quality in turn, and determine the signal quality coefficient of the corrected radio signal data to be tested.

[0106] Is it less than the quality threshold Z1 of the radio signal until the signal quality coefficient to be tested

[0107] When the quality threshold Z1 of the radio signal is less than the threshold value, the overall quality of the signal can be gradually improved to avoid unnecessary correction operations. If the signal quality has reached the standard, the correction process can be stopped, thus saving time and resources, effectively reducing the impact of interference on the signal, and improving the stability and reliability of the signal.

[0108] See also Figure 2 The present invention provides a radio signal feature automatic extraction system, comprising:

[0109] The main influencing feature screening module builds a standard database and sample data set for radio signals and calculates the sample signal quality coefficient of each set of sample radio signal data and the quality threshold Z1 of the radio signal, and calculate each radio signal feature and signal quality coefficient The absolute value of the Pearson correlation coefficient is used to screen the main influencing features of signal quality and their feature weights ;

[0110] Interference weight analysis module, when the signal quality coefficient to be tested When the signal quality threshold Z1 is greater than the radio signal, a radio signal interference analysis warning is issued to extract the main influencing features of the signal quality of the radio signal to be tested. , calculate the interference degree of each signal quality main influencing feature , obtain the interference weight of each signal quality main influencing feature ;

[0111] The interference correction module, after receiving the radio signal interference analysis warning, corrects the main influencing features of the signal quality in turn, and determines the quality coefficient of the measured signal of the corrected radio signal data to be tested Is it less than the quality threshold Z1 of the radio signal until the signal quality coefficient to be tested is less than the quality threshold Z1 of the radio signal.

[0112] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product. A person of ordinary skill in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0113] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0114] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for automatically extracting radio signal features, characterized in that: The steps include: Construct a standard database and sample data set for radio signals, and calculate the sample signal quality coefficient for each set of sample radio signal data and the quality threshold Z1 of the radio signal, and calculate each radio signal feature and signal quality coefficient The absolute value of the Pearson correlation coefficient is used to screen the main influencing features of signal quality and their feature weights ; When the signal quality factor When the signal quality threshold Z1 is greater than the radio signal, a radio signal interference analysis warning is issued to extract the main influencing features of the signal quality of the radio signal to be tested. , calculate the interference degree of each signal quality main influencing feature , obtain the interference weight of each signal quality main influencing feature ; After receiving the radio signal interference analysis warning, correct the main influencing features of the signal quality in turn, and determine the quality coefficient of the measured signal of the corrected radio signal data to be tested. Is it less than the quality threshold Z1 of the radio signal until the signal quality coefficient to be tested is less than the quality threshold Z1 of the radio signal.

2. The method for automatically extracting radio signal features according to claim 1, characterized in that: Collect sample data sets containing radio signal characteristics and signal quality indicators, and calculate the sample signal quality coefficient of each set of sample radio signal data : in, a Indicates the sequential number of each group of sample radio signal data in the sample data set, is the signal-to-noise ratio of the sample radio signal, represents the bit error rate of the sample radio signal, representing the signal-to-noise ratio of a standard radio signal in a standard database of radio signals, representing the bit error rate of a standard radio signal in a standard database of radio signals, is a positive number used to prevent division by zero errors, It is the midpoint offset of the Sigmoid function in the signal-to-noise ratio part, which is used to adjust the center position of the influence of the signal-to-noise ratio on the quality coefficient. It is the midpoint offset of the Sigmoid function in the bit error rate part, which is used to adjust the center position of the influence of the bit error rate on the quality coefficient.

3. The method for automatically extracting radio signal features according to claim 2, characterized in that: Using the mean of sample signal quality coefficients of sample radio signal data and standard deviation To calculate the quality threshold Z1 of the radio signal, it is as follows: where 1.645 is obtained from the standard normal distribution table and corresponds to a 2.5% tail probability for a two-sided test.

4. The method for automatically extracting radio signal features according to claim 3, characterized in that: A sample dataset containing radio signal characteristics and sample signal quality coefficients Import SPSS analysis software to calculate the characteristics and signal quality coefficient of each radio signal The absolute value of the Pearson correlation coefficient is used to sort the top 10 features from large to small and select them as the main influencing features of signal quality. The absolute value of the Pearson correlation coefficient corresponding to the main influencing features of signal quality is recorded as the feature weight. .

5. The method for automatically extracting radio signal features according to claim 1, characterized in that: Use the receiving antenna to receive the radio signal to be tested, and use digital signal processing technology to analyze the signal-to-noise ratio of the radio signal to be tested and bit error rate , calculate the signal quality coefficient of the radio signal data to be tested : in, i Indicates the reception time of the radio signal data, i =1, 2, …, n。 6. The method for automatically extracting radio signal features according to claim 1, characterized in that: When the signal quality factor When the signal quality threshold Z1 is greater than the radio signal, a radio signal interference analysis warning is issued to extract the main influencing features of the signal quality of the radio signal to be tested. , combined with feature weights and a standard database of radio signals to calculate the degree of interference for each of the main influencing features of signal quality : in, k The number of the main influencing feature of signal quality, k= 1, 2, ..., 10, Indicates the standard value of the main influencing feature of signal quality.

7. The method for automatically extracting radio signal features according to claim 1, characterized in that: Count the interference degree of each main influencing feature of signal quality of all radio signals under test The mean of the two is recorded as the interference weight of each main influencing feature of signal quality. .

8. The method for automatically extracting radio signal features according to claim 7, characterized in that: After receiving the radio signal interference analysis warning, the interference weight of each main influencing feature of signal quality is calculated. Sort from largest to smallest, start correcting the main influencing feature of signal quality from the first one, and correct the main influencing feature of signal quality to the standard value.

9. The method for automatically extracting radio signal features according to claim 8, characterized in that: After the correction of the main influencing feature of the first signal quality is completed, the quality coefficient of the measured signal of the corrected radio signal data to be measured is determined. Is it less than the quality threshold Z1 of the radio signal? If it is less than the quality threshold Z1 of the radio signal, the interference correction process is stopped; if it is not less than the quality threshold Z1 of the radio signal, the next correction signal quality main influencing feature is corrected in sequence until the signal quality coefficient to be measured is is less than the quality threshold Z1 of the radio signal.

10. A radio signal feature automatic extraction system, used to implement the method according to any one of claims 1 to 9, characterized in that: include: The main influencing feature screening module builds a standard database and sample data set for radio signals and calculates the sample signal quality coefficient of each set of sample radio signal data and the quality threshold Z1 of the radio signal, and calculate each radio signal feature and signal quality coefficient The absolute value of the Pearson correlation coefficient is used to screen the main influencing features of signal quality and their feature weights ; Interference weight analysis module, when the signal quality coefficient to be tested When the signal quality threshold Z1 is greater than the radio signal, a radio signal interference analysis warning is issued to extract the main influencing features of the signal quality of the radio signal to be tested. , calculate the interference degree of each signal quality main influencing feature , obtain the interference weight of each signal quality main influencing feature ; The interference correction module, after receiving the radio signal interference analysis warning, corrects the main influencing features of the signal quality in turn, and determines the quality coefficient of the measured signal of the corrected radio signal data to be tested Is it less than the quality threshold Z1 of the radio signal until the signal quality coefficient to be tested is less than the quality threshold Z1 of the radio signal.

Images