Separation method of dense wireless signals
Through the grouping, spectrum processing and threshold adaptive determination of wireless signal data, dense wireless signal separation is achieved without prior information, solving the problems of poor separation effect and insufficient universality in the prior art, and improving separation Effect and applicability.
Patent Information
- Application Number
- CN202411211953.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-12-06
AI Technical Summary
The prior art is not effective in dense wireless signal separation, especially when the frequency domain is densely distributed, and requires prior information, and lacks universality and algorithm complexity.
By receiving wireless signal data, grouping, calculating the frequency spectrum and eliminating burst frequency points, counting the probability distribution to obtain the noise base, adaptively determining the threshold value for filtering and separation, and designing a bandpass filter for signal separation.
Achieve high-sensitivity signal separation without prior information, improves separation effect and versatility, reduces calculation complexity, and is suitable for most dense signal scenarios.
Smart Images

Figure CN119094052A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a method for separating dense wireless signals, and belongs to the technical field of communication signal processing. Background Art
[0002] With the development of space information networks, the frequency-using devices connected and the types of services supported are gradually enriched, which leads to very tight spectrum resources, so it is necessary to realize dynamic allocation of spectrum resources. In particular, in a dense signal environment, quickly detecting existing signals and separating them is a prerequisite for the above requirements. In response to this demand, the present invention designs a method for separating dense wireless signals. In the absence of prior information, it can perform high-sensitivity detection on sampled data that simultaneously contains multiple densely distributed wireless signals, and quickly separate each signal to obtain data containing only one signal. This method can provide a technical basis for applications such as radio management and spectrum monitoring.
[0003] However, there are still some shortcomings in the separation methods of dense wireless signals: 1. The separation effect is poor when the signals are densely distributed in the frequency domain; 2. Prior information is required when separating dense signals, but the receiver usually has limited prior information; 3. The separation of specific signals requires special methods designed according to their parameters, which is not very universal and the algorithm is complex. Summary of the invention
[0004] The technical problem solved by the present invention is: in view of the fact that there are many deficiencies in the separation means of dense wireless signals in the current prior art, a method for separating dense wireless signals is proposed.
[0005] The present invention solves the above technical problems by the following technical solutions:
[0006] A method for separating dense wireless signals, comprising:
[0007] receiving wireless signal data as sampled data and grouping them according to a preset continuous data volume;
[0008] Calculate the spectrum of each group of data after grouping and eliminate the burst frequency points;
[0009] Statistically calculate the probability distribution of the frequency spectrum of each group of data to obtain the noise floor of each group of data;
[0010] Calculate the data threshold value of each group of data according to the noise floor of each group of data;
[0011] The threshold values of each group of data are used to distinguish the signal type of the current group of data, and filtering and separation processing is performed according to the signal type.
[0012] The method to group the sampled data is:
[0013] The number of continuous data quantities is preset, and the continuous data satisfying the number of continuous data quantities in the sampled data are grouped into one group. After the continuous grouping, the data whose number of data quantities does not meet the condition is filled with 0 and then grouped into the last group.
[0014] The method for calculating the spectrum of each group of data after grouping and eliminating the burst frequency points is:
[0015] Perform Fourier transform processing on each group of sampling data to obtain the spectrum of each group of sampling data; obtain the modulus value of each frequency point of each group of spectrum and discard the phase value, and eliminate the sudden frequency points of the spectrum modulus value through low-pass filtering; save the data spectrum after eliminating the sudden frequency points.
[0016] The method to obtain the probability distribution of each set of data spectrum is:
[0017] The spectrum of each group of data is counted using the histogram method, and the maximum and minimum values of the corresponding spectrum values are searched;
[0018] Divide a preset number of intervals between the maximum and minimum values of the spectrum value, count the number of spectrum values in each interval, and save the statistical results;
[0019] For the statistical results, search for the interval with the largest number of spectrum values, determine the minimum and maximum values of the interval, and complete the probability distribution statistics.
[0020] The noise floor of each set of data is calculated as follows:
[0021]
[0022] In the formula, r i is the noise floor value of the i-th group of data, Y Amin is the minimum value of the interval A with the largest number of spectrum values in the i-th group of data, Y Amax It is the interval maximum value of the interval A with the largest number of spectrum values in the i-th group of data.
[0023] The calculation method of the data threshold value of each group of data is:
[0024] The corresponding data threshold is determined based on the noise floor calculated for each group of data. The threshold of the first group of data is based on the noise floor of the first group of data. The data threshold of the i-th group of data except the first group of data is th i The calculation method is:
[0025] th i =(1-p)th i-1 +pr i (i>1)
[0026] Where p is the adjustment factor;
[0027] After the data threshold corresponding to each group of data is calculated, it is saved until all data groups are traversed.
[0028] The method to distinguish the signal type of each group of data is:
[0029] Select any group of data, read the corresponding spectrum, and judge the value of each frequency point in the spectrum. If it is greater than the data threshold, there is a signal at the current frequency point, otherwise it is marked that there is no signal at the current frequency point;
[0030] If there is a signal at the current frequency point, all frequency points of the group data where the current frequency point is located are judged. When the continuous frequency points are greater than the data threshold value, the current data is wireless signal data, otherwise it is noise data.
[0031] When there is a signal at any frequency point in any group of data, the signal is taken as the first starting frequency point greater than the data threshold value. Starting from the starting frequency point, there are continuous frequency points that are greater than the data threshold value. The last frequency point greater than the data threshold value is determined as the ending frequency point. When the number of continuous frequency points meets the requirement for the number of wireless signal frequencies, it is determined that the signal corresponding to the current data is a wireless signal.
[0032] After determining that the group signal corresponding to each group of data threshold values is a wireless signal, determine the signal carrier frequency estimation value f c for:
[0033]
[0034] Determine the signal bandwidth estimate B as:
[0035] B=f1-f0
[0036] Determine the power estimate P e for:
[0037]
[0038] Where P(f) is the power spectrum of the sampled data, Δf is the frequency resolution of the power spectrum of the sampled data, and f1 is the sampling rate of the data.
[0039] According to the calculated parameter information of the wireless signal, and according to the starting frequency and ending frequency of each wireless signal, a bandpass filter is designed to perform filtering, and the wireless signal is subjected to signal target separation processing.
[0040] The advantages of the present invention compared with the prior art are:
[0041] (1) The present invention provides a method for separating dense wireless signals. By statistically analyzing the probability distribution of the wireless signal spectrum, a threshold is adaptively determined based on the result. The signal is filtered and separated according to the threshold and the signal parameters are estimated. The method can achieve signal separation in a dense electromagnetic spectrum environment without relying on prior information. It has no special requirements on the signal communication system and is suitable for signal separation in most cases.
[0042] (2) The present invention can adaptively determine the threshold for distinguishing noise from signals, thereby solving the problem of inaccurate noise and signal distinction when signals are densely distributed, improving the separation effect, and designing a specific low-pass filter and filtering the spectrum when calculating the probability distribution of the wireless signal spectrum, thereby eliminating the influence of burst noise on the statistical probability distribution, improving the accuracy of threshold selection without significantly increasing the computational complexity;
[0043] (3) The present invention can separate signals without prior information such as communication system parameters of the signals. The required prior information is small and it is suitable for dense signal separation scenarios in most cases. In addition to separating dense signals, the parameters of unknown signals can also be estimated. The complexity of the estimation algorithm is low but the accuracy meets the requirements. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] Figure 1 A flow chart of the dense signal separation method provided for the invention; DETAILED DESCRIPTION
[0045] A method for separating dense wireless signals. In the absence of prior information, the sampled data containing multiple densely distributed wireless signals is detected with high sensitivity, the probability distribution of the wireless signal spectrum is statistically analyzed, the threshold is adaptively determined based on the result, the signal is filtered and separated according to the threshold, the signal parameters are estimated, and each signal is quickly separated to obtain data containing only one signal. The wireless signal separation method provides a technical basis for applications such as radio management and spectrum monitoring.
[0046] The method for separating dense wireless signals includes the following specific steps:
[0047] receiving wireless signal data as sampled data and grouping them according to a preset continuous data volume;
[0048] Calculate the spectrum of each group of data after grouping and eliminate the burst frequency points;
[0049] Statistically calculate the probability distribution of the frequency spectrum of each group of data to obtain the noise floor of each group of data;
[0050] Calculate the data threshold value of each group of data according to the noise floor of each group of data;
[0051] The threshold values of each group of data are used to distinguish the signal type of the current group of data, and filtering and separation processing is performed according to the signal type.
[0052] The method to group the sampled data is:
[0053] The number of continuous data quantities is preset, and the continuous data satisfying the number of continuous data quantities in the sampled data are grouped into one group. After the continuous grouping, the data whose number of data quantities does not meet the condition is filled with 0 and then grouped into the last group.
[0054] The method for calculating the spectrum of each group of data after grouping and eliminating the burst frequency points is:
[0055] Perform Fourier transform processing on each group of sampling data to obtain the spectrum of each group of sampling data; obtain the modulus value of each frequency point of each group of spectrum and discard the phase value, and eliminate the sudden frequency points of the spectrum modulus value through low-pass filtering; save the data spectrum after eliminating the sudden frequency points.
[0056] The method to obtain the probability distribution of each set of data spectrum is:
[0057] The spectrum of each group of data is counted using the histogram method, and the maximum and minimum values of the corresponding spectrum values are searched;
[0058] Divide a preset number of intervals between the maximum and minimum values of the spectrum value, count the number of spectrum values in each interval, and save the statistical results;
[0059] For the statistical results, search for the interval with the largest number of spectrum values, determine the minimum and maximum values of the interval, and complete the probability distribution statistics.
[0060] The noise floor of each set of data is calculated as:
[0061]
[0062] In the formula, r i is the noise floor value of the i-th group of data, Y Amin is the minimum value of the interval A with the largest number of spectrum values in the i-th group of data, Y Amax It is the interval maximum value of the interval A with the largest number of spectrum values in the i-th group of data.
[0063] The calculation method of the data threshold value of each group of data is:
[0064] The corresponding data threshold is determined based on the noise floor calculated for each group of data. The threshold of the first group of data is based on the noise floor of the first group of data. The data threshold of the i-th group of data except the first group of data is th i The calculation method is:
[0065] th i=(1-p)th i-1 +pr i (i>1)
[0066] Where p is the adjustment factor;
[0067] After the data threshold corresponding to each group of data is calculated, it is saved until all data groups are traversed.
[0068] The method to distinguish the signal type of each group of data is:
[0069] Select any group of data, read the corresponding spectrum, and judge the value of each frequency point in the spectrum. If it is greater than the data threshold, there is a signal at the current frequency point, otherwise it is marked that there is no signal at the current frequency point;
[0070] If there is a signal at the current frequency point, all frequency points of the group data where the current frequency point is located are judged. When the continuous frequency points are greater than the data threshold value, the current data is wireless signal data, otherwise it is noise data.
[0071] When there is a signal at any frequency point in any group of data, the signal is taken as the first starting frequency point greater than the data threshold value. Starting from the starting frequency point, there are continuous frequency points that are greater than the data threshold value. The last frequency point greater than the data threshold value is determined as the ending frequency point. When the number of continuous frequency points meets the requirement for the number of wireless signal frequencies, it is determined that the signal corresponding to the current data is a wireless signal.
[0072] After determining that the group signal corresponding to each group of data threshold values is a wireless signal, determine the signal carrier frequency estimation value f c for:
[0073]
[0074] Determine the signal bandwidth estimate B as:
[0075] B=f1-f0
[0076] Determine the power estimate P e for:
[0077]
[0078] Where P(f) is the power spectrum of the sampled data, Δf is the frequency resolution of the power spectrum of the sampled data, and f1 is the sampling rate of the data.
[0079] According to the calculated parameter information of the wireless signal, and according to the starting frequency and ending frequency of each wireless signal, a bandpass filter is designed to perform filtering, and the wireless signal is subjected to signal target separation processing.
[0080] The following is further described in conjunction with the accompanying drawings and preferred embodiments of the specification:
[0081] In the current embodiment, the specific process of the separation method is as follows Figure 1 As shown:
[0082] Step 1: Divide the sampled data into multiple groups, calculate the spectrum of each group of data, and eliminate the burst frequency points;
[0083] The specific steps are:
[0084] Step 11, the sampled data is divided into multiple groups according to 4096 consecutive data as a group, and the data less than 4096 at the end are filled with 0 to make up 4096 data;
[0085] Step 12: For each set of data, perform a 4096-point fast Fourier transform (FFT) to obtain the spectrum of each set of data;
[0086] Step 13: For each set of spectrum, find the modulus value of each frequency point and discard the phase value;
[0087] Step 14: For the spectrum modulus value, use a low-pass filter to filter, eliminate the burst frequency points, and then save each group of filtered spectrum. The value of vector b is:
[0088] b=[-0.0001 0.0000 0.0000 0.0000 0.0001 0.0001 0.0001 0.00010.00010.0002 0.0002 0.0003 0.0003 0.0004 0.0004 0.0005 0.0006 0.0006 0.00070.0008 0.0009 0.0011 0.0012 0.0013 0.0015 0.0016 0.0018 0.0020 0.0022 0.00240.0026 0.0028 0.0031 0.0033 0.0036 0.0039 0.0042 0.0045 0.0048 0.0051 0.00550.0058 0.0062 0.0065 0.0069 0.0073 0.0077 0.0081 0.0085 0.0089 0.0093 0.00970.0101 0.0105 0.0109 0.0113 0.0117 0.0121 0.0125 0.0129 0.0132 0.0136 0.01390.0143 0.0146 0.0149 0.0152 0.0155 0.0157 0.0159 0.0162 0.0164 0.0165 0.01670.0168 0.0169 0.0170 0.0171 0.0171 0.0171 0.0171 0.0171 0.0170 0.0169 0.01680.0167 0.0165 0.0164 0.0162 0.0159 0.0157 0.0155 0.0152 0.0149 0.0146 0.01430.0139 0.0136 0.0132 0.0129 0.0125 0.0121 0.0117 0.0113 0.0109 0.0105 0.01010.0097 0.0093 0.0089 0.0085 0.0081 0.0077 0.0073 0.0069 0.0065 0.0062 0.00580.0055 0.0051 0.0048 0.0045 0.0042 0.0039 0.0036 0.0033 0.0031 0.0028 0.00260.0024 0.0022 0.0020 0.0018 0.0016 0.0015 0.0013 0.0012 0.0011 0.0009 0.00080.0007 0.0006 0.0006 0.0005 0.0004 0.0004 0.0003 0.0003 0.0002 0.0002 0.0001 0.0001 0.0001 0.0001 0.0000 0.0000 0.0000 -0.0001].
[0089] Step 2: Use the histogram method to calculate the probability distribution of the spectrum of each set of data and obtain the noise floor of each set of data;
[0090] The specific steps are:
[0091] Step 21: For the spectrum of the i-th group of data, search for the maximum and minimum values of the spectrum value;
[0092] Step 22: For the spectrum value of the i-th group of data, divide it into 100 intervals at equal intervals according to its maximum value and minimum value, count the number of spectrum values falling into each interval, and save the statistical results;
[0093] Step 23: For the statistical results, search for the interval A with the largest number of spectrum values, and determine the minimum value Y of interval A. Amin And the interval maximum value Y Amax ;
[0094] Step 24: Calculate the noise floor value r of the data set i , noise floor value r i Determined by the following formula:
[0095]
[0096] Step 3: Determine the threshold for distinguishing noise from signal through an iterative method;
[0097] The specific steps are:
[0098] Step 31: The threshold value th1 of the first set of data is determined by the following formula:
[0099] th1=r1
[0100] Step 32: Threshold value th of the i-th group of data (i>1) i Determined by the following formula:
[0101] th i =(1-p)th i-1 +pr i
[0102] Where p is the adjustment factor, p = 0.8;
[0103] Step 33: Save the threshold value th of the i-th group of data i ;
[0104] Step 34, go to step 32, until the threshold values of all groups of data are calculated;
[0105] Step 4: Distinguish noise from signal based on the threshold and estimate the frequency band of the signal;
[0106] The specific steps are:
[0107] Step 41: for the i-th group of data, read the spectrum saved in step 1;
[0108] Step 42: For each frequency point value of the i-th group of data spectrum, if it is greater than th i , then it is marked that there is a signal at this frequency point, otherwise it is marked that there is no signal at this frequency point;
[0109] Step 43: For the i-th group of data, if there is a signal, the starting frequency point f0 of the signal is the first one greater than the threshold value th i The end frequency f1 is the frequency point that is continuously greater than the threshold value th starting from the frequency point f0. i The last frequency point of the signal; if there are multiple signals, the starting frequency point and the ending frequency point of each signal are estimated according to the above method; otherwise, there is no signal in this group of data;
[0110] Step 5: For the detected signal, estimate its three signal parameters of carrier frequency, bandwidth and power, and perform filtering and separation according to requirements;
[0111] The specific steps are:
[0112] Step 51, calculate the power spectrum P(f) of the sampled data; the frequency resolution Δf of the power spectrum of the sampled data is calculated by the following formula:
[0113]
[0114] where f s is the sampling rate of the data;
[0115] Step 52: For the detected signal, its carrier frequency estimation value f c Calculated by the following formula:
[0116]
[0117] Step 53: For the detected signal, its bandwidth estimation value B is calculated by the following formula:
[0118] B=f1-f0
[0119] Step 54: For the detected signal, its power estimation value P e Calculated by the following formula:
[0120]
[0121] Step 55: For the detected signal, a bandpass filter is designed according to its starting frequency and ending frequency to filter and separate the target signal.
[0122] The current embodiment can adaptively determine the threshold for distinguishing noise and signals by counting the probability distribution of the wireless signal spectrum. When counting the probability distribution of the wireless signal spectrum, a specific low-pass filter is designed and the spectrum is filtered to eliminate the impact of burst noise on the statistical probability distribution. It is suitable for dense signal separation scenarios in most cases.
[0123] Although the present invention has been disclosed as above in the form of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications to the technical solution of the present invention by using the methods and technical contents disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall fall within the protection scope of the technical solution of the present invention.
[0124] The contents not described in detail in the specification of the present invention belong to the common knowledge of those skilled in the art.
Claims
1. A method for separating dense wireless signals, characterized in that include: receiving wireless signal data as sampled data and grouping them according to a preset continuous data volume; Calculate the spectrum of each group of data after grouping and eliminate the burst frequency points; Statistically calculate the probability distribution of the frequency spectrum of each group of data to obtain the noise floor of each group of data; Calculate the data threshold value of each group of data according to the noise floor of each group of data; The threshold values of each group of data are used to distinguish the signal type of the current group of data, and filtering and separation processing is performed according to the signal type.
2. The method for separating dense wireless signals according to claim 1, characterized in that: The method to group the sampled data is: The number of continuous data quantities is preset, and the continuous data satisfying the number of continuous data quantities in the sampled data are grouped into one group. After the continuous grouping, the data whose number of data quantities does not meet the condition is filled with 0 and then grouped into the last group.
3. The method for separating dense wireless signals according to claim 1, characterized in that: The method for calculating the spectrum of each group of data after grouping and eliminating the burst frequency points is: Perform Fourier transform processing on each group of sampling data to obtain the spectrum of each group of sampling data; obtain the modulus value of each frequency point of each group of spectrum and discard the phase value, and eliminate the sudden frequency points of the spectrum modulus value through low-pass filtering; save the data spectrum after eliminating the sudden frequency points.
4. The method for separating dense wireless signals according to claim 1, characterized in that: The method to obtain the probability distribution of each set of data spectrum is: The spectrum of each group of data is counted using the histogram method, and the maximum and minimum values of the corresponding spectrum values are searched; Divide a preset number of intervals between the maximum and minimum values of the spectrum value, count the number of spectrum values in each interval, and save the statistical results; For the statistical results, search for the interval with the largest number of spectrum values, determine the minimum and maximum values of the interval, and complete the probability distribution statistics.
5. The method for separating dense wireless signals according to claim 1, characterized in that: The noise floor of each set of data is calculated as follows: In the formula, r i is the noise floor value of the i-th group of data, Y Amin is the minimum value of the interval A with the largest number of spectrum values in the i-th group of data, Y Amax It is the interval maximum value of the interval A with the largest number of spectrum values in the i-th group of data.
6. The method for separating dense wireless signals according to claim 5, characterized in that: The calculation method of the data threshold value of each group of data is: The corresponding data threshold is determined based on the noise floor calculated for each group of data. The threshold of the first group of data is based on the noise floor of the first group of data. The data threshold of the i-th group of data except the first group of data is th i The calculation method is: th i =(1-p)th i-1 +pr i (i>1) Where p is the adjustment factor; After the data threshold corresponding to each group of data is calculated, it is saved until all data groups are traversed.
7. The method for separating dense wireless signals according to claim 1, characterized in that: The method to distinguish the signal type of each group of data is: Select any group of data, read the corresponding spectrum, and judge the value of each frequency point in the spectrum. If it is greater than the data threshold, there is a signal at the current frequency point, otherwise it is marked that there is no signal at the current frequency point; If there is a signal at the current frequency point, all frequency points of the group data where the current frequency point is located are judged. When the continuous frequency points are greater than the data threshold value, the current data is wireless signal data, otherwise it is noise data.
8. The method for separating dense wireless signals according to claim 7, characterized in that: When there is a signal at any frequency point in any group of data, the signal is taken as the first starting frequency point greater than the data threshold value. Starting from the starting frequency point, there are continuous frequency points that are greater than the data threshold value. The last frequency point greater than the data threshold value is determined as the ending frequency point. When the number of continuous frequency points meets the requirement for the number of wireless signal frequencies, it is determined that the signal corresponding to the current data is a wireless signal.
9. The method for separating dense wireless signals according to claim 8, characterized in that: After determining that the group signal corresponding to each group of data threshold values is a wireless signal, determine the signal carrier frequency estimation value f c for: Determine the signal bandwidth estimate B as: B=f1-f0 Determine the power estimate P e for: Where P(f) is the power spectrum of the sampled data, Δf is the frequency resolution of the power spectrum of the sampled data, and f1 is the sampling rate of the data.
10. The method for separating dense wireless signals according to claim 9, characterized in that: According to the calculated parameter information of the wireless signal, and according to the starting frequency and ending frequency of each wireless signal, a bandpass filter is designed to perform filtering, and the wireless signal is subjected to signal target separation processing.
