A method and device for calculating symbol rate for frequency modulation mode

Abstract

The invention relates to a symbol rate calculation method for frequency modulation mode. According to the self-characteristics of signals in radio phase amplitude modulation mode, the density clustering and statistical methods are used to calculate the symbol rate. Specifically, it includes obtaining the range of the overall data distribution, dividing intervals and performing extreme value analysis, sequence segmentation processing, continuity processing, construction of segmentation point sequences, construction of interval sequences, clustering analysis, and point calculation for individual symbols. It also discloses a symbol rate calculation device for frequency modulation mode and a storage medium. The beneficial effects achieved by the present invention are: under the condition of containing noise, the symbol rate can still be correctly calculated, and the longer the duration of the signal, the higher the accuracy of symbol rate calculation, which has strong robustness.

Description

Technical Field

[0001] The present invention relates to the field of communication technologies, and particularly to a symbol rate calculation method, apparatus, and storage medium for a frequency modulation method. Background Art

[0002] Since the invention of radio communication, it has played an important role in many fields and is now inseparable from human society. More and more scientific and technological personnel are engaged in the research and optimization of radio communication technologies. Due to the characteristics of radio itself, during the transmission process, it will be affected by factors such as weather, obstacles, and electromagnetic fields, resulting in signal attenuation or interference during transmission. These factors will have a serious impact on radio signal analysis. When demodulating a radio signal, it is necessary to calculate its symbol rate, and the calculation accuracy requirement is relatively high. And noise will directly affect the calculation accuracy of the symbol rate. Therefore, in actual calculation scenarios, it is generally required that the radio signal has a high signal-to-noise ratio.

[0003] In recent years, machine learning technologies have been applied to more and more industries and achieved good results; how to combine machine learning with radio signal technologies has become a trend. In radio signal technologies, the calculation of the symbol rate plays a key role in demodulation. How to improve the accuracy of the symbol rate and how to accurately calculate the number of symbols under noisy conditions have become a topic studied by more and more people. Summary of the Invention

[0004] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a symbol rate calculation method, apparatus, and storage medium for a frequency modulation method, which can still correctly calculate the symbol rate under noisy conditions, and the longer the duration of the signal, the higher the accuracy of the symbol rate calculation, and it has strong robustness.

[0005] The purpose of the present invention is achieved through the following technical solutions: A symbol rate calculation method for a frequency modulation method, including:

[0006] a. Perform phase adjustment on the forward difference sequence of the input signal with respect to the phase, and obtain the range of the overall data distribution;

[0007] b. Divide the range of the overall data distribution into M intervals evenly, and count the number of signals for the phase adjustment sequence in different intervals;

[0008] Perform extreme value analysis on each interval, and obtain the positions of the maximum value and the minimum value;

[0009] c. According to the position of the minimum value, further divide the entire sequence into N regions;

[0010] Construct N all-zero sequences of the same length as the original sequence; the construction method is that each all-zero sequence is mapped to the segmentation intervals, and if there is data at the position of the interval, the position is set to 1;

[0011] d. Conduct continuity analysis on all the all-zero sequences to obtain the starting position information of the signals in each interval;

[0012] Conduct repeatability analysis on the starting positions of the signals in all intervals to obtain a suspected symbol jump position sequence, sort this sequence, and calculate the interval sequence;

[0013] e. Obtain the reference interval by conducting reference interval analysis on the interval sequence;

[0014] f. Calculate the number of sampling points occupied by a single symbol based on the suspected symbol jump positions and the reference interval;

[0015] g. The ratio of the original frequency sampling rate to the number of sampling points occupied by a single symbol is the symbol rate of this phase amplitude modulation method; the original frequency sampling rate refers to the frequency sampling rate of the input signal in a.

[0016] Furthermore, in a, it includes:

[0017] a1. Conduct phase analysis and calculation on the input signal IQ data to obtain a phase data sequence; it should be noted that the signal IQ data means that the collected signal is provided to the user in the IQ form of the signal, and what the user gets is a series of IQ sample points of the signal;

[0018] a2. Conduct forward difference processing on the signal phase sequence to obtain a signal forward difference sequence; the minimum value and the maximum value of the signal forward difference sequence are -π and π respectively.

[0019] a3. According to the distribution of the signal difference sequence, conduct overall phase adjustment on the sequence and obtain the range of the overall data distribution.

[0020] It should be noted that in a2 of this solution, the specific steps of the forward difference of the signal are very traditional and conventional operations for those skilled in the art, and will not be elaborated in detail here.

[0021] Further preferably, in a3:

[0022] a31. Judge whether the proportion of the number of forward phase difference sequences between -3 and 3 in the whole is greater than 0.9. If it is greater than 0.9, end and return the result; and judge whether the number of iterations is greater than 6 times. If it is iterated six times, end and return the adjustment result with the largest proportion;

[0023] a32. Move the whole data down by 0.5 unit. If the data is less than -π, it needs to be processed by adding π, and then go to a31.

[0024] Explanation of the understanding of a3:

[0025] Since the minimum value of the entire forward difference sequence is -π and the maximum value is π, then within the range from -π to π, -3 to 3 is selected as the range for overall phase adjustment;

[0026] Since the minimum value of the forward difference sequence is -π and the maximum value is π; then when the data exceeds π, the data will jump to -π and surge from the bottom up. Therefore, it is necessary to adjust the overall data, that is, move the overall data downward. Then the data originally at -π will jump to a position close to the lower side of π, so that the overall data distribution is very concentrated and there will be no jump from π to -π;

[0027] During the process of moving the overall data downward, a specific threshold is set. In this solution, it is set to 0.5, that is, move down by 0.5. When the forward difference sequence of the signal within the range of -3 to 3 occupies more than 0.9 (more than 90%) of the entire forward difference sequence of the signal, that is, there is no need to continue moving downward;

[0028] If after moving down 6 times (0.5 x 6 = 3), it still cannot be satisfied that the forward difference sequence of the signal within the range of -3 to 3 occupies more than 0.9 (more than 90%) of the entire forward difference sequence of the signal; then there is a problem with the selected -3 to 3, and it is necessary to expand the interval within the range of -3 to 3 until the requirements are met.

[0029] Furthermore, in this solution, b is mainly for segmentation, and c is mainly for clustering. The signal parameters are calculated based on the overall distribution law of the signal. Even if there is noise in the signal, the accuracy of the signal can be guaranteed.

[0030] In b, according to the effect diagram of the phase forward difference sequence adjustment, it is divided into M intervals along the up and down directions. Then, the number of corresponding data of the phase forward difference sequence in each interval is counted, and the statistical situation is made into a statistical effect diagram of the number of phase forward difference intervals;

[0031] In addition, in b, within the M divided intervals, the maximum value and the minimum value of each interval are found; according to the height where the minimum value is located, then on the entire effect diagram of the phase forward difference sequence adjustment, it is divided into N intervals along the up and down directions.

[0032] Regarding the physical meaning understanding of dividing N intervals by the minimum value in b. For example, for 8PSK (octal phase shift keying), there are 8 data distribution regions, and then the isolation points (i.e., the minimum values) in each data distribution region are found.

[0033] Further, in c, for each of the N intervals, a new sequence is newly constructed, and the length of the new sequence is the same as the length of the corresponding original interval sequence. The purpose is to find the signal transitions.

[0034] In c, N all-zero sequences of the same length as the original sequence are constructed, and the construction method is as follows:

[0035] c1. Map the segmented intervals;

[0036] c2. If there is no data, it is represented by 0, indicating the absence of a signal; if there is data, it is represented by 1, indicating the presence of a signal (a continuous signal is represented by the same data unit, that is, the same continuous signal is represented by a horizontal line segment represented by a 1; the length of the horizontal line segment represents the duration of the signal).

[0037] The understanding of c is as follows: In the constructed all-zero sequence, the horizontal line segment itself represents a segment of the signal (if there are slight fluctuations in the signal itself, such as 1.1 and 1.05 near 1, they are all represented by 1); the length of the horizontal line segment represents the time period during which the signal exists; and the starting point of the horizontal line segment is the signal transition point.

[0038] Further, in d, the analysis of the continuity of the all-zero sequence refers to finding continuous signals, that is, those with a certain length in the horizontal line segment; obtaining the starting position information of the signals within each interval. The starting position of the signal is the suspected symbol transition position. (Because of noise and interference, all the symbols obtained here are suspected symbols)

[0039] In this solution, in d, the starting positions of the signals within all intervals are analyzed repeatedly to obtain a sequence of suspected symbol transition positions, which is understood as:

[0040] Among all the suspected symbols (in the horizontal line segment), for example, there are suspected symbols with lengths of 6, 12, 12, 24, 6, 7, 14... Since 12 is a multiple of 6, and 28 and 14 are multiples of 7, then the suspected symbols with lengths of 6 and 7 are the reference suspected symbols; then the reference suspected symbol with a length of 6 appears 5 times repeatedly (6, 12, 12, 24, 6), and the reference suspected symbol with a length of 7 appears 2 times. Therefore, 6 appears most frequently.

[0041] In d, sorting the sequence of suspected symbol transition positions and calculating the interval sequence means:

[0042] Sort the suspected symbols according to their positions, calculate the distance in the length direction between the positions of two adjacent suspected symbols, and form an interval sequence with all the calculated distances.

[0043] Further, in e:

[0044] e1. First, perform minimum length cleaning on the interval sequence, and remove all interval values with a length less than 4 from the sequence;

[0045] e2. Statistically analyze the cleaned interval sequence, and put the intervals that appear more than N times into the suspected interval set;

[0046] e3. Perform similar multiple interval merging and statistics on the values in the suspected interval set. When the following formula is satisfied, the occurrence count of the suspected interval is incremented by one.

[0047]

[0048] Among them, gap is the suspected interval value, x is the value in the cleaned interval sequence, and round represents rounding to the nearest integer;

[0049] Finally, find the most frequently occurring suspected interval as the reference interval, and assign gap 1 to the reference interval;

[0050] e4. Screen out the intervals whose basic intervals satisfy the above formula to form a symbol oversampling multiple interval sequence.

[0051] Furthermore, in f: According to the symbol oversampling multiple interval and the reference interval, calculate the number of sampling points occupied by a single symbol as follows. The calculation formula is as follows:

[0052]

[0053] Among them, g n is the interval value in the calculation interval sequence (here, the calculation interval sequence refers to the symbol oversampling multiple interval sequence; for example, if the reference interval is 6, the obtained calculation interval sequence is in the form of 6, 12, 12, 6, 6, 24, 35,), m is the number of elements in the interval set, gap 1 is the reference interval, and f(g n ) is the multiple of the calculation interval occupying the reference interval gap 1 , that is, the symbol oversampling multiple.

[0054] It should be noted that e1 to e3 are actually the content in d; e4 is the real content in e; but for the sake of the coherence of expression, e1 to e3 originally belonging to d are placed in e.

[0055] A symbol rate calculation device for a frequency modulation method, comprising:

[0056] A memory for storing a computer program;

[0057] A processor for implementing steps of a symbol rate calculation method for a frequency modulation method when executing a computer program.

[0058] A computer-readable storage medium including a program, where the program can be executed by a processor to support the completion of a symbol rate calculation method for a frequency modulation method.

[0059] The present invention has the following advantages:

[0060] (1) The symbol rate calculation method of the present invention has high accuracy;

[0061] Traditional frequency modulation methods also analyze and process the phase forward difference sequence of the signal, but there are noise points in the sequence, which often leads to the inability to guarantee the accuracy of symbol rate calculation;

[0062] According to the signal characteristics of the frequency modulation method, the present application calculates the signal parameters through the overall distribution law of the signal by means of segmentation and clustering. Even when there is noise in the signal, the accuracy of the signal can be guaranteed;

[0063] (2) The symbol rate calculation method of the present invention has high precision;

[0064] When calculating the symbol rate by traditional frequency modulation methods, the problem of reduced accuracy will occur due to the introduction of noise;

[0065] In the process of processing data in the present invention, the number of correct data is increased through repetitive analysis in d. Then, after analyzing and processing the data, the noise information does not participate in the calculation during the final analysis, ensuring the final calculation accuracy of the symbol rate;

[0066] (3) In the process of signal analysis and processing of the present invention, that is, e, f, g essentially use the prior knowledge of signal data and the method of clustering; the achieved effect is that even if there is noise and partial interference in the signal, the symbol rate can still be correctly calculated, and the longer the duration of the signal, the higher the accuracy rate of symbol rate calculation, with strong robustness. Description of the Drawings

[0067] Figure 1 It is a schematic diagram of the overall process of the present invention;

[0068] Figure 2 It is a schematic diagram of the overall process of an embodiment of the present invention;

[0069] Figure 3 It is an effect diagram of the phase forward difference of the signal in an embodiment of the present invention;

[0070] Figure 4 It is an effect diagram of the adjustment of the phase forward difference sequence in an embodiment of the present invention;

[0071] Figure 5 The statistical effect diagram of the number of phase forward difference intervals in the embodiment of the present invention;

[0072] Figure 6 The signal segmentation effect diagram of the embodiment of the present invention;

[0073] Figure 7 The signal duration effect diagram of the interval continuity analysis in the embodiment;

[0074] Figure 8 The schematic diagram of the basic interval analysis process in the embodiment. Detailed implementation manners

[0075] The present invention will be further described below with reference to the accompanying drawings, but the protection scope of the present invention is not limited to the following.

[0076] A symbol rate calculation method for a frequency modulation method, the overall process of which is as Figure 1 , Figure 2 shown, including:

[0077] a. Perform phase adjustment on the forward difference sequence of the input signal with respect to the phase, and obtain the range of the overall data distribution;

[0078] Specifically,

[0079] a1. Perform phase analysis and calculation on the IQ data of the input signal to obtain a phase data sequence;

[0080] a2. Perform forward difference processing on the signal phase sequence to obtain a signal forward difference sequence, and the effect is as Figure 3 shown; the minimum value and the maximum value of the signal forward difference sequence are -π and π respectively.

[0081] a3. According to the distribution of the signal forward difference sequence, perform overall phase adjustment on the sequence and obtain the range of the overall data distribution, as Figure 4 shown;

[0082] For a3, it specifically includes:

[0083] a31. Judge whether the proportion of the number of phase difference sequences between -3 and 3 in the whole is greater than 0.9. If it is greater than 0.9, end and return the result; and judge whether the number of iterations is greater than 6 times. If it is iterated six times, end and return the adjustment result with the largest proportion;

[0084] a32. Move the whole data down by 0.5 unit. If the data is less than -π, it needs to be processed by adding π, and then go to a31;

[0085] It should be noted that the signal IQ data refers to providing the acquired signal to the user in the IQ form of the signal, and what the user obtains is a series of IQ sample points of the signal.

[0086] It should be noted that the minimum and maximum values of the signal forward difference sequence are -π and π respectively, which means that when the signal phase sequence is forward differenced, the minimum and maximum values of the difference are -π and π;

[0087] It should be noted that, as Figure 3 and Figure 4 shown: In a3, from -π to π, -3 to 3 is selected as the range of overall phase adjustment; when adjusting, it is set that the overall data moves 0.5; at most two times downward, the signal forward difference sequence can occupy more than 90% of the entire signal forward difference sequence; therefore, only through 1 to 2 times, the overall data is moved downward to the Figure 4 position in

[0088] b. Divide the overall distribution range into M intervals evenly, and count the number of signal pairs in different intervals of the phase adjustment sequence; the statistical distribution effect is as Figure 5 shown;

[0089] For each interval, perform extreme value analysis to obtain all the maximum value positions and minimum value positions that conform to the distribution in the sequence;

[0090] c. First, according to the position of the minimum value, divide the entire sequence into N regions again, as Figure 6 shown;

[0091] Then construct N all-zero sequences of the same length as the original sequence;

[0092] Specifically, the construction method is: c1. Map the divided intervals; c2. If there is no data, it is represented by 0, that is, it indicates that there is no signal; if there is data, it is represented by 1, that is, it indicates that there is a signal, and continuous signals represent the same data unit;

[0093] It should be noted that: after the entire sequence is divided into intervals again, as Figure 6 shown (the effect diagram after dividing the signal differential phase data); when constructing the all-zero sequence, as Figure 6 becomes Figure 7 shown, if there is data at the corresponding position in this interval, it is represented by 1, if not, it is represented by 0.

[0094] It should be noted that for Figure 7In the description, in the constructed all-zero sequence, the horizontal line segment is essentially a suspected symbol, and the suspected symbol represents a continuous signal (if there are slight fluctuations in the signal itself, it is still represented by 1. For example, if the values of the signal are 1, 1.1, and 1.05, they are all represented by 1); the starting point of the suspected symbol is the signal transition point (in Figure 7 , it is the starting point of the horizontal line segment, and the length of the horizontal line segment represents the time when the signal exists);

[0095] d. Conduct a continuity analysis on all the constructed all-zero sequences to obtain the starting position information of the signals in each interval, that is, the starting point of each suspected symbol (horizontal line segment) in Figure 7 ;

[0096] Conduct a repeatability analysis on the starting positions of the signals in all intervals to obtain a sequence of suspected symbol transition positions, sort this sequence, and calculate the interval sequence;

[0097] The above sorting of the sequence of suspected symbol transition positions and calculating the interval sequence can be understood as

[0098] sorting the suspected symbols according to their positions, calculating the distance in the length direction between the positions of two adjacent suspected symbols, and forming an interval sequence with all the calculated distances;

[0099] e. Obtain the reference interval through a reference interval analysis of the interval sequence;

[0100] For, specifically, as Figure 8 shown,

[0101] e1. First, perform a minimum length cleaning on the interval sequence, and remove all interval values with a length less than 4 from the sequence;

[0102] e2. Conduct a statistics on the cleaned interval sequence, and put the intervals that appear more than N times into the suspected interval set;

[0103] e3. Conduct a similar multiple interval merging and statistics on the values in the suspected interval set. When the following formula is satisfied, the occurrence count of the suspected interval is incremented by one,

[0104]

[0105] where gap is the suspected interval value, x is the value in the cleaned interval sequence, and round represents rounding to the nearest integer;

[0106] Finally, find the suspected interval that appears most frequently as the reference interval, and assign gap 1 to the reference interval;

[0107] e4. Then, select the intervals that satisfy the above formula for the basic interval to form the symbol oversampling multiple intervals.

[0108] f. Calculate the number of sampling points occupied by a single symbol based on the symbol oversampling multiple intervals and the reference interval.

[0109] Specifically,

[0110] The calculation formula for the number of sampling points occupied by a single symbol is as follows:

[0111]

[0112] where g n is the interval value in the calculation interval set, m is the number of elements in the interval set, gap 1 is the reference interval, f(g n ) is the multiple of the basic interval gap 1 occupied by the calculation interval, that is, the symbol oversampling multiple;

[0113] g. The ratio of the original frequency sampling rate to the number of sampling points occupied by a single symbol is the symbol rate of this phase-amplitude modulation method; the original frequency sampling rate refers to the frequency sampling rate of the input signal in a.

[0114] In the above embodiments, a is mainly for segmentation, c is mainly for clustering, and the signal parameters are calculated based on the overall distribution law of the signal. Even if there is noise in the signal, the accuracy of the signal can be guaranteed.

[0115] Optionally, the present application also provides a symbol rate calculation device for the phase-amplitude modulation method.

[0116] Specifically, it includes a memory for storing computer programs; and a processor for implementing the steps of a symbol rate calculation method for the phase-amplitude modulation method when executing the computer programs.

[0117] Optionally, the present application also provides a computer-readable storage medium.

[0118] Specifically, it includes a program, and the program can be executed by a processor to support the completion of a symbol rate calculation method for the phase-amplitude modulation method.

[0119] The above embodiments only represent relatively preferred implementation manners, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the inventive concept, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention.

Claims

1. A method for calculating symbol rate for frequency modulation mode, characterized in that: It includes: a. Perform phase adjustment on the forward difference sequence of the input signal with respect to phase to obtain the range of the overall data distribution; it includes: a1. Perform phase analysis and calculation on the IQ data of the input signal to obtain a phase data sequence; a2. Perform forward difference processing on the signal phase data sequence to obtain a signal forward difference sequence; the minimum value and the maximum value of the signal forward difference sequence are -π and π respectively; a3. According to the distribution of the signal forward difference sequence, perform overall phase adjustment on the sequence and obtain the range of the overall data distribution; specifically: a31. Judge whether the proportion of the number of the phase difference sequence within -3 to 3 in the whole is greater than 0.

9. If it is greater than 0.9, end and return the result; and judge whether the number of iterations is greater than 6 times. If it is iterated six times, end and return the adjustment result with the largest proportion; a32. Move the whole data down by 0.5 unit. If the data is less than -π, it needs to be processed by adding π, and then go to a31 b. According to the phase forward difference sequence adjustment effect diagram: divide the range of the overall data distribution into M intervals along the up and down direction, then count the number of corresponding data of the phase forward difference sequence in each interval, and make a statistical effect diagram of the number of phase forward difference intervals; In addition, within the M divided intervals, find the maximum value and the minimum value of each interval; c. According to the height where the minimum value is located, then divide the whole phase forward difference sequence adjustment effect diagram into N intervals along the up and down direction; Construct N all-zero sequences with the same length as the original sequence; the construction method is: c1. Map the divided intervals; c2. If there is no data, it is represented by 0, that is, it indicates that there is no signal; if there is data, it is represented by 1, that is, it indicates that there is a signal, and continuous signals represent the same data unit; a continuous signal is represented by the same data unit, that is, the same continuous signal is represented by a horizontal line segment represented by 1; the length of the horizontal line segment represents the duration of the signal; in the constructed all-zero sequence, the horizontal line segment is actually a suspected symbol, and the suspected symbol represents continuous signals; the starting point of the suspected symbol is the signal jump point; d. Perform continuity analysis on all the all-zero sequences to obtain the starting position information of the signals in each interval; Perform repeatability analysis on the starting positions of the signals in all intervals to obtain a suspected symbol jump position sequence, sort this sequence, and calculate an interval sequence; Sorting the suspected symbol jump position sequence and calculating the interval sequence means: sort the suspected symbols according to their positions, calculate the distance in the length direction between the positions of two adjacent suspected symbols, and form an interval sequence with all the distances; e. By performing baseline interval analysis on the interval sequence, a baseline interval is obtained. Specifically, e1. First, perform minimum length cleaning on the interval sequence, and remove all interval values with lengths less than 4 from the sequence; e2. Statistically analyze the cleaned interval sequence, and put the intervals that appear more than N times into the suspected interval set; e3. Perform similar multiple interval merging and statistics on the values in the suspected interval set. When the following formula is satisfied, the occurrence count of the suspected interval is incremented by one, Among them, gap is the suspected interval value, x is the value in the cleaned interval sequence, and round represents rounding to the nearest integer; finally, the most frequently occurring suspected interval is found as the reference interval, and gap 1 is assigned as the reference interval; e4. Then, the intervals that satisfy the above formula with the basic interval are screened out to form a symbol oversampling multiple interval sequence; f. Calculate the number of sampling points occupied by a single symbol based on the suspected symbol jump position and the baseline interval; g. The ratio of the original frequency sampling rate to the number of sampling points occupied by a single symbol is the symbol rate of this frequency modulation method; the original frequency sampling rate refers to the frequency sampling rate of the input signal in a.

2. A method for calculating the symbol rate of a frequency modulation method according to claim 1, characterized in that: in the above f: The calculation formula for the number of sampling points occupied by a single symbol is as follows, where g n is the interval value in the calculation interval sequence, m is the number of elements in the interval set, gap 1 is the reference interval, and f(g n ) is the multiple of the reference interval gap 1 occupied by the calculation interval.

3. A device for calculating the symbol rate of a frequency modulation method, characterized in that: comprising: a memory for storing a computer program; a processor for implementing the steps of the method for calculating the symbol rate of a frequency modulation method according to any one of claims 1 to 2 when executing the computer program.

4. A computer-readable storage medium, characterized in that: including a program, and the program can be executed by a processor to support the completion of the method for calculating the symbol rate of a frequency modulation method according to any one of claims 1 to 2.

Images