QAM (Quadrature Amplitude Modulation) signal initial phase estimation method based on constellation diagram edge detection and fitting

Abstract

The invention relates to a QAM (Quadrature Amplitude Modulation) signal initial phase estimation method based on constellation diagram edge detection and fitting, which comprises the following steps of: firstly, calculating an initial phase of a QAM signal constellation point based on the quartic power of the constellation point to finish the capture of large phase offset; detecting the edge of the QAM constellation diagram to obtain points of the upper edge of the constellation diagram; fitting the upper edge, the lower edge, the left edge and the right edge of the QAM edge constellation diagram through an upper straight line, a lower straight line, a left straight line and a right straight line respectively; respectively calculating four initial phases according to the slopes of the four fitted straight lines, averaging the four initial phases to obtain an initial phase of primary iteration, and performing phase compensation on the QAM edge constellation diagram; and updating and fitting the four edges of the compensated edge constellation diagram again, calculating a new initial phase and compensating, and iterating the process until the four edges do not change any more or the maximum number of iterations is reached. The initial phase of the QAM signal constellation diagram is calculated through constellation diagram edge detection and fitting, the precision is high, the variance of the calculation result is small, and the influence of the occurrence probability of points on the signal constellation diagram is small.

Description

Technical Field

[0001] The invention belongs to the technical field of QAM signal initial phase estimation, and in particular relates to a QAM signal initial phase estimation method based on constellation diagram edge detection and fitting. Background Art

[0002] When demodulating QAM signals, it is necessary to estimate and compensate the initial phase of the signal constellation point, otherwise the signal demodulation will be wrong. In a system with a pilot, the pilot can be used to estimate the initial phase, and its performance is excellent; but in a system without a pilot, it can only rely on blind estimation methods for processing. A commonly used method is to calculate the fourth power of the constellation point, as described below.

[0003] Assume that the constellation diagram of the received signal is x(i), i=0, 1, ..., N-1, then the initial phase calculation formula is as follows:

[0004]

[0005] Here, angle(·) represents the angle of the complex number. This method is simple to calculate, and only needs to calculate the fourth power of the signal x(i). It can meet most demodulation application scenarios and has low requirements for EVM. However, when the requirements for EVM are very high, such as in the field of signal analysis, this method has large errors, especially when the points on the signal constellation diagram are not evenly distributed, the measurement error of this method is large.

[0006] In response to the above-mentioned problem of large measurement error, the current solution is to calculate the signal x(i) four times and then perform an initial phase estimate of the four points in the innermost circle. In theory, the four points in the innermost circle are similar to QPSK modulation, and the estimation result is more accurate. However, this method is easily affected by the probability of occurrence of constellation points on the constellation diagram. For higher-order modulation, sometimes the four points in the innermost circle cannot be found at all.

[0007] Therefore, the existing method of calculating the initial phase of the QAM signal by calculating the fourth power of the constellation point has large measurement errors in the field of signal analysis with high requirements on EVM, and the initial phase estimation of the four points in the innermost circle after calculating the fourth power of the constellation point is easily affected by the probability of occurrence of the constellation points on the constellation diagram. These are technical problems that need to be solved urgently. Summary of the invention

[0008] The purpose of the present invention is to provide a QAM signal initial phase estimation method based on constellation edge detection and fitting, so as to solve the technical problems that the initial phase of the QAM signal is calculated by calculating the fourth power of the constellation point, such as the signal analysis field with high requirements on EVM, the measurement error is large, and the initial phase estimation of the four points in the innermost circle after calculating the fourth power of the constellation point is easily affected by the probability of occurrence of the constellation point on the constellation diagram.

[0009] In order to solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0010] A QAM signal initial phase estimation method based on constellation diagram edge detection and fitting comprises the following steps:

[0011] S1: Calculate the initial phase of the QAM signal constellation point based on the fourth power of the constellation point. The specific formula is as follows

[0012]

[0013] Wherein, angle(·) represents the angle of the calculated complex number, x(i) is the point numbered i on the constellation diagram of the signal, i represents the number of the point on the constellation diagram of the signal, i=0, 1, ..., N-1, and the constellation diagram is compensated to complete the capture of large phase deviation;

[0014] S2: Detect the edge of the QAM constellation diagram and obtain the points of the constellation diagram on the edge, that is, the edge constellation diagram;

[0015] S3: The upper edge, lower edge, left edge and right edge of the QAM constellation diagram are fitted by four straight lines: upper, lower, left and right;

[0016] S4: Calculate four initial phases according to the slopes of the four fitted straight lines, and average the four initial phases to get the initial phase of one iteration.

[0017] S5: Phase compensation for the QAM edge constellation diagram, that is

[0018] S6: Update the top, bottom, left, and right edges, repeat step S3, and iterate until the four edges no longer change or the maximum number of iterations is reached. The sum of the initial phases of each iteration is the final initial phase.

[0019] Preferably, step S2 includes the following specific process:

[0020] S21: Establish an initial QAM constellation set {x(i), i=0, 1, ..., N-1};

[0021] S22: Establish a rectangular coordinate system with x(i) as the origin, determine the quadrants of the coordinate system where other x(j), j≠i, are located, and record x(j) as other points in the QAM constellation diagram set;

[0022] S23: When there are other x(j) in each quadrant of x(i), it means that x(i) cannot be on the edge, and x(i) is removed from the set;

[0023] S24: Return to step S22 and traverse each {x(i), i=0, 1, ..., N-1} in turn;

[0024] S25: The remaining constellation points are the edge constellation diagram.

[0025] Preferably, fitting the upper edge of the QAM constellation diagram in step S3 includes the following specific processes:

[0026] s311: determine the initial upper edge, which is all points in the edge constellation diagram that satisfy im(x(i))>thr, where im(·) represents the imaginary part and thr represents the threshold, and the value of thr is related to the modulation mode;

[0027] S312: A straight line is fitted to the points on the upper edge, that is, the fitting curve is y=a1x+b1, and the slope is a1. The inverse tangent value of a1 is calculated, and the inverse tangent value of a1 is the first initial phase.

[0028] Preferably, fitting the lower edge of the QAM constellation diagram in step S3 includes the following specific process:

[0029] S321: Determine an initial lower edge, which is all points in the edge constellation diagram that satisfy im(x(i))<-thr, where im(·) represents the imaginary part and thr represents the threshold;

[0030] S322: A straight line is fitted to the points on the lower edge, and the fitting curve is y=a2 x+b2, and the slope is a2. The inverse tangent value of a2 is calculated, and the inverse tangent value of a2 is the second initial phase.

[0031] Preferably, fitting the left edge of the QAM constellation diagram in step S3 includes the following specific process:

[0032] S331: Determine an initial left edge, which is all points in the edge constellation diagram that satisfy re(x(i))<-thr, where re(·) represents the real part and thr represents the threshold;

[0033] S332: A straight line is fitted to the points on the left edge, that is, the fitting curve is y=a3 x+b3, and the slope is a3. The inverse tangent value of a3 is calculated. The inverse tangent value of a3 That is the third initial phase.

[0034] Preferably, fitting the right edge of the QAM constellation diagram in step S3 includes the following specific process:

[0035] S341: Determine an initial right edge, which is all points in the edge constellation diagram that satisfy re(x(i))>thr, where re(·) represents the real part and thr represents the threshold;

[0036] S342: A straight line is fitted to the points on the right edge, that is, the fitting curve is y=a4x+b4, and the slope is a4. The inverse tangent value of a4 is calculated. That is the fourth initial phase.

[0037] Preferably, the process of calculating the first initial phase, the second initial phase, the third initial phase and the fourth initial phase is performed simultaneously. After calculating the four initial phases, the four phases are averaged to obtain an initial phase of one iteration. And perform initial phase compensation on the edge constellation diagram. Then update the upper edge, lower edge, left edge and right edge, re-fit and calculate the new initial phase and compensate. Iterate the above process. When the upper edge, lower edge, left edge and right edge are not updated or the set number of iterations is reached, it ends. The sum of the initial phases of each iteration is the final initial phase.

[0038] Preferably, in step S6:

[0039] The process of performing the first initial phase compensation on the edge constellation diagram, i.e. updating the upper edge, i.e. all points that still satisfy im(x(i))>thr, loops through step s312 and ends when the upper edge is not updated or the maximum number of iterations is reached;

[0040] The process of performing the second initial phase compensation on the edge constellation diagram, i.e. updating the lower edge, i.e. all points that still satisfy im(x(i))<-thr, loops through step S322, and ends when the lower edge is not updated or the maximum number of iterations is reached;

[0041] The process of performing the third initial phase compensation on the edge constellation diagram, i.e., updating the left edge, i.e., all points that still satisfy re(x(i))<-thr, looping through step S332, and ending when the left edge is not updated or the maximum number of iterations is reached;

[0042] The process of performing the fourth initial phase compensation on the edge constellation diagram, i.e. updating the right edge, i.e. all points that still satisfy re(x(i))>thr, loops through step S342 and ends when the right edge is not updated or the maximum number of iterations is reached.

[0043] The beneficial effects of the present invention include:

[0044] The QAM signal initial phase estimation method based on constellation edge detection and fitting provided by the present invention calculates the initial phase of the QAM signal constellation point based on the fourth power of the constellation point to complete the capture of large phase deviation. Then the edge of the QAM constellation is detected to obtain the point of the edge constellation; the upper edge, lower edge, left edge and right edge of the QAM edge constellation are fitted by four straight lines: upper, lower, left and right; four initial phases are calculated respectively according to the slopes of the four fitted straight lines, and the four initial phases are averaged to obtain the initial phase of one iteration. Perform phase compensation on the QAM edge constellation diagram. Update and fit the four edges of the compensated edge constellation diagram again, calculate the new initial phase and compensate, iterate the above process until the four edges no longer change or the maximum number of iterations is reached, and the sum of the initial phases of each iteration is the final initial phase. The initial phase of the QAM signal constellation diagram is calculated based on the idea of ​​constellation edge detection and fitting. It has high accuracy, small variance of the calculation result, and is less affected by the probability of occurrence of points on the signal constellation diagram. It is more suitable for scenarios such as spectrum analyzer signal analysis that require high EVM measurement accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is the constellation diagram after edge detection of 16QAM with an initial phase of 10° according to the present invention.

[0046] Figure 2 This is the constellation diagram after edge detection of 16QAM with an initial phase of 0° according to the present invention.

[0047] Figure 3 This is the constellation diagram after edge detection of 16QAM with an initial phase of -10° in the present invention.

[0048] Figure 4 It is a schematic diagram of 1000 initial phase estimation results and EVM values ​​of the present invention.

[0049] Figure 5 It is a mean comparison diagram of the estimation method for calculating the fourth power in the prior art and the method of the present invention.

[0050] Figure 6 It is a variance comparison diagram between the estimation method for calculating the fourth power in the prior art and the method of the present invention.

[0051] Figure 7 It is a comparison diagram of the mean values ​​of EVM between the estimation method of calculating the fourth power in the prior art and the method of the present invention.

[0052] Figure 8 It is a comparison diagram of EVM variance between the estimation method of calculating the fourth power in the prior art and the method of the present invention. DETAILED DESCRIPTION

[0053] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application usually described and shown in the drawings here can be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0054] The following is combined with Figure 1-Figure 8 The present invention is further described in detail:

[0055] Example 1

[0056] A QAM signal initial phase estimation method based on constellation diagram edge detection and fitting comprises the following steps:

[0057] S1: Calculate the initial phase of the QAM signal constellation point based on the fourth power of the constellation point. The specific formula is as follows

[0058]

[0059] Wherein, angle(·) represents the angle of the calculated complex number, x(i) is the point numbered i on the constellation diagram of the signal, i represents the number of the point on the constellation diagram of the signal, i=0, 1, ..., N-1. Compensate the constellation diagram to complete the capture of large phase deviation;

[0060] S2: Detect the edge of the QAM constellation diagram and obtain the points of the constellation diagram on the edge, that is, the edge constellation diagram;

[0061] S3: The upper edge, lower edge, left edge and right edge of the QAM constellation diagram are fitted by four straight lines: upper, lower, left and right;

[0062] S4: Calculate four initial phases according to the slopes of the four fitted straight lines, and average the four initial phases to get the initial phase of one iteration.

[0063] S5: Perform phase compensation on the QAM constellation diagram, that is,

[0064] S6: Update the top, bottom, left, and right edges, repeat step S3, and iterate until the four edges no longer change or the maximum number of iterations is reached. The sum of the initial phases of each iteration is the final initial phase.

[0065] The above-mentioned QAM signal initial phase estimation method based on constellation edge detection and fitting is based on the idea of ​​constellation edge detection and fitting, and proposes a constellation edge detection method, and calculates the initial phase of the QAM signal constellation diagram based on it. It has high accuracy, small variance of the calculation result, and is less affected by the probability of occurrence of points on the signal constellation diagram. It is more suitable for scenarios such as spectrum analyzer signal analysis that have high requirements for EVM measurement accuracy.

[0066] Example 2

[0067] Based on Example 1, step S2 includes the following specific process:

[0068] S21: Establish an initial QAM constellation set {x(i), i=0, 1, ..., N-1};

[0069] S22: Establish a rectangular coordinate system with x(i) as the origin, determine the quadrants of the coordinate system where other x(j), j≠i, are located, and record x(j) as other points in the QAM constellation diagram set;

[0070] S23: When there are other x(j) in each quadrant of x(i), it means that x(i) cannot be on the edge, and x(i) is removed from the set;

[0071] S24: Return to step S22 and traverse each {x(i), i=0, 1, ..., N-1} in turn;

[0072] S25: The remaining constellation points are the edge constellation diagram.

[0073] In this embodiment, since the edge of the QAM signal needs to be constrained by four straight line edges, after obtaining the edge of the constellation diagram, four straight lines need to be determined. The fitting of the upper edge of the QAM constellation diagram in step S3 includes the following specific processes:

[0074] s311: determine the initial upper edge, which is all points in the edge constellation diagram that satisfy im(x(i))>thr, where im(·) represents the imaginary part and thr represents the threshold, and the value of thr is related to the modulation mode;

[0075] S312: A straight line is fitted to the points on the upper edge, that is, the fitting curve is y=a1x+b1, and the slope is a1. The inverse tangent value of a1 is calculated, and the inverse tangent value of a1 is the first initial phase.

[0076] Among them, thr is related to the specific QAM modulation mode. See the threshold parameters of the initial edge in the following table. Different QAM modulation modes correspond to different thr values.

[0077] Modulation thr 8QAM 0.408248 16QAM 0.632456 32QAM 0.894427 64QAM 0.925820 128QAM 1.104315 256QAM 1.073751 512QAM 1.211060 1024QAM 1.148759

[0078] Example 3

[0079] Based on Embodiment 1 or Embodiment 2, fitting the lower edge of the QAM constellation diagram in step S3 includes the following specific process:

[0080] S321: Determine an initial lower edge, which is all points in the edge constellation diagram that satisfy im(x(i))<-thr, where im(·) represents the imaginary part and thr represents the threshold;

[0081] S322: A straight line is fitted to the points on the left edge, and the fitting curve is y=a2 x+b2. The slope is a2, and the inverse tangent value of a2 is calculated. The inverse tangent value of a2 is the second initial phase.

[0082] The fitting of the left edge of the QAM constellation diagram in step S3 includes the following specific processes:

[0083] S331: Determine an initial left edge, which is all points in the edge constellation diagram that satisfy re(x(i))<-thr, where re(·) represents the real part and thr represents the threshold;

[0084] S332: A straight line is fitted to the points on the left edge, that is, the fitting curve is y=a3 x+b3, and the slope is a3. The inverse tangent value of a3 is calculated. The inverse tangent value of a3 That is the third initial phase.

[0085] The fitting of the right edge of the QAM constellation diagram in step S3 includes the following specific processes:

[0086] S341: Determine an initial right edge, which is all points in the edge constellation diagram that satisfy re(x(i))>thr, where re(·) represents the real part and thr represents the threshold;

[0087] S342: A straight line is fitted to the points on the right edge, that is, the fitting curve is y=a4x+b4, and the slope is a4. The inverse tangent value of a4 is calculated. That is the fourth initial phase.

[0088] The above-mentioned process of calculating the first initial phase, the second initial phase, the third initial phase and the fourth initial phase is performed simultaneously. After calculating the four initial phases, the four initial phases are averaged to obtain an initial phase of one iteration. Perform initial phase compensation on the edge constellation diagram, update the upper edge, lower edge, left edge and right edge at the same time, and re-estimate the initial phase of the fitting. Iterate in sequence. When the upper edge, lower edge, left edge and right edge are not updated or the set number of iterations is reached, the process ends. The sum of the initial phases of each iteration is the final initial phase. Figure 1-Figure 3 ,The constellation diagram after edge detection of 16QAM with initial phases of 10°, 0°, and -10° respectively.

[0089] In this embodiment, in step S6:

[0090] The process of performing the first initial phase compensation on the edge constellation diagram, i.e. updating the upper edge, i.e. all points that still satisfy im(x(i))>thr, loops through step s312 and ends when the upper edge is not updated or the maximum number of iterations is reached.

[0091] The process of performing the second initial phase compensation on the edge constellation diagram, i.e. updating the lower edge, i.e. all points that still satisfy im(x(i))<-thr, loops through step S322 and ends when the lower edge is not updated or the maximum number of iterations is reached.

[0092] The process of performing the third initial phase compensation on the edge constellation diagram, that is, updating the left edge, that is, all points that still satisfy re(x(i))<-thr, loops through step S332 and ends when the left edge is not updated or the maximum number of iterations is reached.

[0093] The process of performing the fourth initial phase compensation on the edge constellation diagram, i.e. updating the right edge, i.e. all points that still satisfy re(x(i))>thr, loops through step S342 and ends when the right edge is not updated or the maximum number of iterations is reached.

[0094] See also Figure 4 , we can see the 1000 initial phase estimation results and the EVM values ​​after compensation under different SNRs. It is obvious that the initial phase estimation result error of the method adopted by the present invention is very small, and the larger the SNR, the smaller the variance. In addition, the EVM also meets the expected value, and the fluctuation is greatly reduced compared with the conventional method. The comparison results of the conventional method and the method of this paper under different SNRs can be found in Figures 5 to 8 As shown in the figure, Normal represents the conventional method and Edge represents the method in this paper. From the following results, we can see that the method in this paper significantly reduces the variance of the initial phase estimation, improves the accuracy of the initial phase measurement, has a good EVM performance, and the EVM measurement result has a small error and a small variance.

[0095] In summary, the QAM signal initial phase estimation method based on constellation edge detection and fitting provided by the present invention calculates the initial phase of the QAM signal constellation point based on the fourth power of the constellation point to complete the capture of the large initial phase. Then the edge of the QAM constellation is detected to obtain the points of the constellation on the edge, that is, the edge constellation; the upper edge, lower edge, left edge and right edge of the QAM edge constellation are fitted by four straight lines: upper, lower, left and right; four initial phases are calculated respectively according to the slopes of the four fitted straight lines, and the four initial phases are averaged to obtain the initial phase of one iteration. use Perform phase compensation on the QAM constellation diagram, then iteratively update the upper, lower, left, and right edges, and perform new edge fitting to obtain a new initial phase estimate. Iterate in sequence until each edge is no longer updated or the maximum number of iterations is reached. The sum of the initial phases of each iteration is The initial phase of the QAM signal constellation is calculated based on the idea of ​​edge detection and fitting of the constellation diagram. It has high accuracy, small variance of the calculation result, and is less affected by the probability of occurrence of points on the signal constellation diagram. It is more suitable for scenarios such as spectrum analyzer signal analysis that require high EVM measurement accuracy.

[0096] The above-mentioned embodiments only express the specific implementation method of the present application, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the protection scope of the present application. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the technical solution concept of the present application, and these all belong to the protection scope of the present application.

Claims

1. A QAM signal initial phase estimation method based on constellation edge detection and fitting, characterized in that: The following steps are involved: S1: Calculate the initial phase of the QAM signal constellation point based on the fourth power of the constellation point. The specific formula is as follows Wherein, angle(·) represents the angle of the calculated complex number, x(i) is the point numbered i on the constellation diagram of the signal, i represents the number of the point on the constellation diagram of the signal, i=0, 1, ..., N-1, and the constellation diagram is compensated to complete the capture of large phase deviation; S2: Detect the edge of the QAM constellation diagram and obtain the points of the constellation diagram on the edge, that is, the edge constellation diagram; S3: The upper edge, lower edge, left edge and right edge of the QAM edge constellation diagram are fitted by four straight lines: upper, lower, left and right; S4: Calculate four initial phases according to the slopes of the four fitted straight lines, and average the four initial phases to get the initial phase of one iteration. S5: Perform phase compensation on the QAM edge constellation diagram, that is, S6: Update the top, bottom, left, and right edges, repeat step S3, and iterate until the four edges no longer change or the maximum number of iterations is reached. The sum of the initial phases of each iteration is the final initial phase.

2. A QAM signal initial phase estimation method based on constellation edge detection and fitting according to claim 1, characterized in that: Step S2 includes the following specific processes: S21: Establish an initial QAM constellation set {x(i), i=0, 1, ..., N-1}; S22: Establish a rectangular coordinate system with x(i) as the origin, determine the quadrants of the coordinate system where other x(j), j≠i, are located, and record x(j) as other points in the QAM constellation diagram set; S23: When there are other x(j) in each quadrant of x(i), it means that x(i) cannot be on the edge, and x(i) is removed from the set; S24: Return to step S22 and traverse each {x(i), i=0, 1, ..., N-1} in turn; S25: The remaining constellation points are the points on the edge constellation diagram.

3. A QAM signal initial phase estimation method based on constellation edge detection and fitting according to claim 1, characterized in that: The fitting of the upper edge of the QAM constellation diagram in step S3 includes the following specific processes: S311: Determine an initial upper edge, which is all points in the edge constellation diagram that satisfy im(x(i))>thr, where im(·) represents the imaginary part, thr represents the threshold, and the value of thr is related to the modulation mode; S312: A straight line is fitted to the points on the upper edge, that is, the fitting curve is y=a1x+b1, and the slope is a1. The inverse tangent value of a1 is calculated, and the inverse tangent value of a1 is the first initial phase.

4. A QAM signal initial phase estimation method based on constellation edge detection and fitting according to claim 3, characterized in that: The fitting of the lower edge of the QAM constellation diagram in step S3 includes the following specific processes: S321: Determine an initial lower edge, which is all points in the edge constellation diagram that satisfy im(x(i))<-thr, where im(·) represents the imaginary part and thr represents the threshold; S322: A straight line is fitted to the points on the lower edge, and the fitting curve is y=a2x+b2, and the slope is a2. The inverse tangent value of a2 is calculated, and the inverse tangent value of a2 is the second initial phase.

5. The method for estimating the initial phase of a QAM signal based on edge detection and fitting of a constellation diagram according to claim 3, characterized in that: The fitting of the left edge of the QAM constellation diagram in step S3 includes the following specific processes: S331: Determine an initial left edge, which is all points in the edge constellation diagram that satisfy re(x(i))<-thr, where re(·) represents the real part and thr represents the threshold; S332: A straight line is fitted to the points on the left edge, that is, the fitting curve is y=a3 x+b3, and the slope is a3. The inverse tangent value of a3 is calculated. The inverse tangent value of a3 That is the third initial phase.

6. The method for estimating the initial phase of a QAM signal based on edge detection and fitting of a constellation diagram according to claim 3, characterized in that: The fitting of the right edge of the QAM constellation diagram in step S3 includes the following specific processes: S341: Determine an initial right edge, which is all points in the edge constellation diagram that satisfy re(x(i))>thr, where re(·) represents the real part and thr represents the threshold; S342: A straight line is fitted to the points on the left edge, that is, the fitting curve is y=a4x+b4, and the slope is a4. The inverse tangent value of a is calculated, and the inverse tangent value of a4 is calculated. That is the fourth initial phase.

7. A QAM signal initial phase estimation method based on constellation edge detection and fitting according to any one of claims 3 to 6, characterized in that: The process of calculating the first initial phase, the second initial phase, the third initial phase and the fourth initial phase is performed simultaneously. After calculating the four initial phases, the initial phase of one iteration is obtained by averaging. Perform initial phase compensation on the edge constellation diagram, then update the upper edge, lower edge, left edge and right edge, calculate the new initial phase and compensate. Iterate in sequence, and end when the upper edge, lower edge, left edge and right edge are not updated or the set number of iterations is reached.

8. The method for estimating the initial phase of a QAM signal based on edge detection and fitting of a constellation diagram according to claim 7, characterized in that: The process of performing the first initial phase compensation on the edge constellation diagram, i.e. updating the upper edge, i.e. all points that still satisfy im(x(i))>thr, loops through step S312, and ends when the upper edge is not updated or the maximum number of iterations is reached; The process of performing the second initial phase compensation on the edge constellation diagram, i.e. updating the lower edge, i.e. all points that still satisfy im(x(i))<-thr, loops through step S322, and ends when the lower edge is not updated or the maximum number of iterations is reached; The process of performing the third initial phase compensation on the edge constellation diagram, i.e., updating the left edge, i.e., all points that still satisfy re(x(i))<-thr, loops through step S332, and ends when the left edge is not updated or the maximum number of iterations is reached; The process of performing the fourth initial phase compensation on the edge constellation diagram, i.e. updating the right edge, i.e. all points that still satisfy re(x(i))>thr, loops through step S342 and ends when the right edge is not updated or the maximum number of iterations is reached.

Images