An analog joint coding modulation method based on helical amplitude phase keying

By constructing a spiral amplitude phase keying constellation, optimizing the constellation point distribution using probability and radial shaping parameters, and combining it with angular control, the problem of insufficient matching between coding mapping and modulation mapping is solved, source reconstruction distortion is reduced, and the accuracy of signal transmission is improved.

CN123074901APending Publication Date: 2026-11-10HUAQIAO UNIVERSITY +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611568654.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-12
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

In the prior art, when carrying continuous coded signals, the insufficient matching between the coding mapping and the modulation mapping of finite constellation mapping leads to large distortion of the source reconstruction at the receiver, and the selection of coding and modulation parameters fails to effectively reflect the impact of channel noise.

Method used

A helical amplitude phase keying-based analog joint coding modulation method is adopted. The constellation point radius is determined by probability shaping and radial shaping parameters, and a helical amplitude phase keying constellation is constructed by combining angular control parameters. The constellation point with the smallest Euclidean distance to the two-dimensional signal is selected as the modulation symbol, and the parameter combination is optimized under a given signal-to-noise ratio to reduce source reconstruction distortion.

Benefits of technology

It improves the adjustability of the spatial distribution of finite constellation points, enhances the matching relationship between the coded signal and constellation points, reduces quantization error, and optimizes the end-to-end distortion effect of source reconstruction.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123074901A_ABST
    Figure CN123074901A_ABST
Patent Text Reader

Abstract

This invention provides an analog joint coding and modulation method based on helical amplitude phase keying (HPS). Two Gaussian analog sources are coded using S-K encoding based on the double-helix spacing parameter in a candidate parameter set. After normalization and demultiplexing, a two-dimensional signal is obtained. The constellation point radius and angle are determined using probability shaping parameters, radial shaping parameters, and angular control parameters to construct a helical amplitude phase keying constellation. Modulation symbols are selected according to the minimum Euclidean distance criterion. The reconstructed source is obtained through channel transmission, constellation decision, inverse normalization, and S-K decoding. Under a given signal-to-noise ratio (SNR), the signal distortion ratios corresponding to each candidate parameter set are compared, and the parameter set with the largest SNR is selected for coding and modulation. This invention improves the matching relationship between coding and modulation by jointly adjusting the coding curve and constellation spatial distribution, thereby reducing end-to-end source reconstruction distortion.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and in particular to an analog joint coding modulation method based on helical amplitude phase keying. Background Technology

[0002] Analog joint source channel coding achieves source compression and channel transmission by mapping sampled values ​​of analog sources to channel symbols. The two-dimensional to one-dimensional Shannon-Kotelnikov mapping can convert two analog sources into a single continuously coded signal, suitable for communication scenarios requiring compressed transmission dimensions. To adapt the continuously coded signal to modulation equipment using finite constellations, the continuously coded signal can be normalized and combined in two dimensions. The combined signal is then mapped to a finite number of complex constellation points. After channel transmission, the receiving end performs constellation decision and source reconstruction.

[0003] In the aforementioned transmission process, the errors introduced by the finite constellation mapping, along with channel noise, jointly affect the estimated value of the coded signal obtained at the receiver. This estimated value requires a nonlinear inverse mapping to reconstruct the analog source. Therefore, the magnitude of the error in the modulation symbol domain and the final source reconstruction error are not simply linearly related. Even if a constellation can closely approximate the amplitude distribution of the coded signal, or has a large constellation spacing, it does not mean that the combination of that constellation and the adopted coding mapping will result in a smaller source reconstruction distortion.

[0004] Schemes employing fixed ring constellations and separately determining coding and modulation parameters typically rely on the compression characteristics of the coding mapping itself or the distribution characteristics of the constellation itself to determine the parameters, neglecting the end-to-end reconstruction error generated by continuous-value coded outputs after finite constellation quantization, noise interference, and inverse mapping. Because the non-uniform statistical characteristics of the coded output are not fully incorporated into the constellation geometry design, even if coding and modulation each achieve local optima, their performance gains are difficult to effectively reflect in the reconstruction accuracy at the receiver.

[0005] In view of the above, this application is hereby submitted. Summary of the Invention

[0006] This invention discloses an analog joint coding and modulation method based on helical amplitude phase keying, which aims to solve the problem of large source reconstruction distortion at the receiver due to insufficient matching between coding mapping and modulation mapping when a finite constellation carries continuous coded signals.

[0007] This invention provides an analog joint coding modulation method based on helical amplitude phase keying, comprising: Based on the double helix spacing parameter in the candidate parameter group, the two Gaussian analog sources are SK encoded to obtain a continuous encoded signal. The continuous encoded signal is then normalized and combined to obtain a two-dimensional signal. Based on the candidate parameter set, the cumulative probability is obtained by power transformation of the normalized constellation point index using the probability shaping parameter. The cumulative probability is then subtracted from 1 to obtain the probability difference. The natural logarithm of the probability difference is taken and its opposite is obtained to obtain the radial transformation value. The radial transformation value is then power-operated using the radial shaping parameter as the exponent to obtain the radius. The angle is determined based on the angular control parameter and the constellation point index. A helical amplitude phase keying constellation is constructed according to the radius and angle. The constellation point with the smallest Euclidean distance to the two-dimensional signal is selected from this constellation as the modulation symbol. The modulation symbols are transmitted through the channel, and the received signals are decided based on the spiral amplitude phase keying constellation. The decision results are denormalized and then SK decoding is performed to obtain the reconstructed source. Given a signal-to-noise ratio, encoding, modulation, and source reconstruction are performed for each candidate parameter group. The signal distortion ratio is calculated based on the reconstructed source and the two Gaussian analog sources. The candidate parameter group with the largest signal distortion ratio is selected to configure SK encoding and spiral amplitude phase keying modulation.

[0008] Beneficial effects: The constellation point radius is determined by probability shaping parameters and radial shaping parameters, and the constellation point angle is determined by angular control parameters. Radial distribution adjustment and angular arrangement adjustment are used together to construct the helical amplitude phase keying constellation, improving the adjustability of the spatial distribution of finite constellation points. On this basis, the constellation point with the smallest Euclidean distance to the two-dimensional signal is selected as the modulation symbol, improving the matching relationship between the constellation point distribution and the continuous coded signal, and reducing the quantization error caused by the continuous coded signal being mapped to finite constellation points.

[0009] By performing coding, modulation, and source reconstruction on different candidate parameter groups under a given signal-to-noise ratio, and using the signal distortion ratio between the original source and the reconstructed source as a unified evaluation index, the selection of the double helix spacing parameter and the constellation shaping parameter is correlated. This allows the parameter selection to reflect the reconstruction effect after the combined effects of coding mapping, finite constellation mapping, and channel noise, thereby obtaining a coding and modulation parameter combination that fits the given signal-to-noise ratio and reducing end-to-end source reconstruction distortion. Attached Figure Description

[0010] Figure 1 This is a flowchart illustrating an analog joint coding modulation method based on helical amplitude phase keying provided in the first embodiment of the present invention; Figure 2 This is a schematic diagram of the spiral amplitude phase keying constellation distribution provided in an embodiment of the present invention; Figure 3 This is a comparison chart of the signal distortion ratios of the spiral amplitude-phase keying scheme and the Gaussian amplitude-phase keying scheme provided in the embodiments of the present invention; Figure 4This is a schematic diagram of a module of an analog joint coding modulation device based on helical amplitude phase keying provided in the second embodiment of the present invention. Detailed Implementation

[0011] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] To better understand the technical solution of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0013] Please see Figure 1 The first embodiment of this invention provides an analog joint coding modulation method based on helical amplitude phase keying (HAP), applied to the communication transmission of analog sources. The transmitting end converts two analog sources into continuously coded signals, which are then normalized and combined in two dimensions to form the input of a constellation mapping. The receiving end recovers the estimated value of the continuously coded signal based on the constellation decision result and reconstructs the two sources through the corresponding inverse transform. During the parameter configuration stage, the above transmission processing is performed on each candidate parameter group, and the candidate parameter groups are evaluated based on the source reconstruction results. After determining the adopted parameter group, the transmitting end and receiving end are configured according to this parameter group to complete the analog joint coding modulation under the corresponding channel conditions.

[0014] S1. Perform SK encoding on the two Gaussian analog sources according to the double helix spacing parameter in the candidate parameter group to obtain a continuous encoded signal. After normalizing the continuous encoded signal, the signals are combined separately to obtain a two-dimensional signal. Obtain two independent Gaussian simulation sources with the same probability distribution. Each source contains... There are 1 sampling point, among which, This represents the total number of sampling points for each signal source. In this embodiment, each signal source follows a Gaussian distribution with a mean of zero and a variance of one, and an even number of sampling points constitute a sequence to be processed. According to the sampling order of the two signal sources, the sampling values ​​at corresponding positions are paired to form a two-dimensional signal source point. ,in, The sampled values ​​from the first signal source. These are the sampled values ​​at the corresponding positions in the second source. Each pair of sampled values ​​serves as input for one SK encoding operation, and all sampled values ​​are processed according to the same pairing relationship.

[0015] The SK encoding adopts a two-dimensional to one-dimensional Sha. The on-Kotelnikov mapping transforms a pair of original source sample values ​​into a single continuous coded value. The transmitter performs the mapping according to the original order of the two-dimensional source points, ensuring a one-to-one correspondence between the generated continuous coded values ​​and the input sample pairs. During parameter evaluation, the original source sequence and sampling positions are preserved, and the reconstructed results are subsequently compared with the original sources at these positions. The source pairing order remains unchanged throughout the processing of different candidate parameter groups, ensuring that the same two-dimensional source points are used as input even when the coding curve parameters change.

[0016] The current candidate parameter set is represented as:

[0017] in, Candidate parameter group; This is the double helix spacing parameter, used to determine the spacing between the helix arms of the SK encoding curve; These are radial shaping parameters used to determine the growth pattern of the radius of subsequent constellation points; This is a probability integer parameter used to determine the cumulative probability corresponding to subsequent constellation point indices; These are angular control parameters used to determine the angles of subsequent constellation points. The encoding process reads these parameters. The other three parameters, together with this parameter, constitute the current system configuration, which will be used to build the corresponding constellation later.

[0018] The double Archimedean spiral curve is used as the SK encoding curve, and its parametric equation is as follows:

[0019] in, The SK encoding curve is defined by the signed polar angle parameter. Determined two-dimensional coordinates; and These are the first and second coordinate components of the two-dimensional coordinate system, respectively. , indicating the spacing between adjacent spiral arms; The range of values ​​is ; Pi; For a sign function, its independent variable takes the values ​​1, 0, and 1 respectively when it is positive, zero, and negative. 1.

[0020] Given the double-helix spacing parameter and the signed polar angle parameter, the curve position can be determined according to the parametric equation. The signed polar angle parameter participates in the calculation of the two coordinate components, with positive and negative polar angles corresponding to different branches of the double Archimedean spiral curve. When the double-helix spacing parameter changes, the distribution of the curve in the two-dimensional source space changes accordingly. Therefore, the same two-dimensional source point can correspond to different curve positions under different candidate parameter sets. Each candidate parameter set uses its own encoded curve to process the complete source sequence, avoiding the mixing of different curve parameters within the same sequence.

[0021] For the current two-dimensional source point, its position on the SK coding curve is determined according to the minimum Euclidean distance criterion, and the signed polar angle parameter corresponding to this position is used as the spiral mapping parameter:

[0022] in, These are the spiral mapping parameters corresponding to the current two-dimensional source point; Represents the Euclidean norm; This represents the polar angle parameter that minimizes the Euclidean norm. During calculation, the two-dimensional source point is used as a fixed input, and the signed polar angle parameter is used as the variable to be determined. The curve coordinates are calculated based on the parametric equation, and the curve position with the minimum distance to the input source point is determined by numerical minimization. For multiple positions with the same minimum distance, one position is selected according to a pre-fixed selection order, thus obtaining a definite mapping result for the current source point.

[0023] The curve mapping represents the two-dimensional source point with its corresponding position on the curve, and the resulting spiral mapping parameters serve as input for subsequent nonlinear transformations. A distance may exist between the curve position and the original source point; this distance reflects the approximation error generated by the curve mapping. The source point recovered at the receiver using the same curve is an estimate of the original source point, and the impact of this error is retained in subsequent end-to-end evaluations. The minimum Euclidean distance selection aims to make the selected position as close as possible to the input source point under the current curve conditions; the result still needs to be evaluated together with subsequent constellation mapping and channel transmission.

[0024] Perform a nonlinear transformation on the obtained spiral mapping parameters to obtain continuous encoded values:

[0025] in, This represents the continuous encoded value corresponding to the current two-dimensional source point; It is a nonlinear transformation; These are nonlinear distributed control parameters, and ; This represents the absolute value. This embodiment will... The value is fixed at 1.5 and remains unchanged during the processing of different candidate parameter groups. The transformation preserves the sign of the spiral mapping parameter through a sign function and adjusts its magnitude through exponentiation. When the spiral mapping parameter is zero, the corresponding continuous encoded value is zero.

[0026] Under positive nonlinear distribution control parameters, the aforementioned nonlinear transformation has a definite inverse transformation relationship, allowing the receiver to recover the spiral mapping parameter estimates from the continuous coded signal estimates using the same parameters. The inverse transformation is performed on the relationship between the curve parameters and the coded values, without eliminating the approximation errors generated by the previous curve mapping. After fixing the nonlinear distribution control parameters, the coding curve is configured by changing the double-helix spacing parameter, giving a clear source to the coding differences in the candidate parameter evaluation. Subsequent processing is then based on the actually generated continuous coded signal.

[0027] Process all two-dimensional source points sequentially, and arrange the resulting continuous encoded values ​​in their original order to form a channel containing... A signal sequence of consecutive encoded values. Normalization is performed based on the average power of this consecutive encoded signal:

[0028] in, The normalized signal value; Represents the mathematical expectation; The average power of the continuously encoded signal; This is the normalization factor. For a finite-length signal sequence, the average power estimate is obtained by averaging the squares of each consecutive coded value, and the normalization factor is determined accordingly. The average power of the currently processed sequence should be greater than zero, and the same normalization factor is used for each consecutive coded value to maintain the relative amplitude relationship between them.

[0029] The normalization directly affects the SK-coded output, adjusting the current sequence to a uniform average power scale. The normalization factor corresponds to the coded sequence generated by the current candidate parameter group and serves as the matching parameter used during inverse normalization at the receiver. When the double-helix spacing parameter changes and generates a new continuous coded signal, the normalization factor is redefined based on the newly generated signal. The receiver uses the same factor when reconstructing the sequence, ensuring consistency in amplitude scale between transmission and reception processing.

[0030] The normalized signal sequence is split into two signals of equal length, and then the signal values ​​at corresponding positions are paired to form a two-dimensional signal. .in, and These are the corresponding signal values ​​in the two split signals, which respectively constitute the first coordinate component and the second coordinate component of the two-dimensional signal. In this embodiment, according to the original order of the normalized signal, the first of two adjacent signal values ​​is assigned to the first path, and the second is assigned to the second path. Then, the signal values ​​at corresponding positions in the two signals are paired up.

[0031] Since this embodiment uses an even number of sampling points, all normalized signal values ​​can be combined according to adjacent pairing. Each signal value participates in only one set of two-dimensional signals, and the number of generated two-dimensional signal points is half the number of the original consecutive coded values. After the receiving end separates the two components of each set of estimated two-dimensional signal values, the original order of the normalized signals can be restored by alternating the order of the first and second components. This correspondence remains fixed during the transmission and reception of the current signal sequence.

[0032] Both components of the two-dimensional signal originate from a continuously coded signal, and each component already contains a mapping result of a pair of original source sample values. Demultiplexing is used to organize the one-dimensional coded value sequence into a two-dimensional constellation mapping input without changing the correspondence between each coded value and its original sample pair. Further details will follow... The signal coordinates in the two-dimensional plane are compared with the constellation point coordinates constructed by the current candidate parameter group, and the resulting two-dimensional signal directly participates in the selection of modulation symbols.

[0033] S2, based on the candidate parameter group, the cumulative probability is obtained by performing a power transformation on the normalized constellation point index using the probability shaping parameter. The cumulative probability is then subtracted from 1 to obtain the probability difference. The natural logarithm of the probability difference is taken and its opposite is taken to obtain the radial transformation value. The radial transformation value is then exponentially operated on using the radial shaping parameter as the exponent to obtain the radius. The angle is determined based on the angular control parameter and the constellation point index. A helical amplitude phase keying constellation is constructed according to the radius and angle. The constellation point with the smallest Euclidean distance to the two-dimensional signal is selected from this constellation as the modulation symbol. In this embodiment, the two-dimensional signal output from the aforementioned encoding process is acquired, and the current candidate parameter group corresponding to this signal is read. A helical amplitude phase keying constellation is constructed using the probability shaping parameters, radial shaping parameters, and angular control parameters within this group. For each pair of paired signal values ​​in the two-dimensional signal, the Euclidean distance between it and each constellation point in the constructed constellation is calculated. The constellation point with the smallest distance is selected as the modulation symbol, and a sequence of modulation symbols to be transmitted is formed according to the original order of the paired signal values. The current two-dimensional signal and the current constellation jointly determine the mapping result for this operation.

[0034] The total number of constellation points is determined based on the target modulation order. ,in, The integer is greater than 1. The constellation points are numbered according to their generation order, and the constellation point index is denoted as . The value is At the same modulation order, each candidate parameter group uses the same number of constellation points, and the positions of the constellation points are adjusted by changing the parameters. The constellation point index is used to associate the cumulative probability, radius, and angle, and to ensure that the three calculation results corresponding to the same index belong to the same constellation point.

[0035] For the For each constellation point, subtract half from its index and divide by the total number of constellation points to obtain the normalized constellation point index. Then, perform a power transformation with the probability integer parameter in the current candidate parameter group as the exponent to calculate the corresponding cumulative probability:

[0036] in, For the first The cumulative probability corresponding to each constellation point Let be the probability integer parameter, and The normalized constellation point index takes values ​​between zero and one. After a power transformation with a positive exponent, the cumulative probability still lies between zero and one. By subtracting half of the index, the cumulative probabilities corresponding to the first and last constellation points avoid the endpoints, providing effective input for subsequent negative natural logarithm calculations.

[0037] The cumulative probability is used as an intermediate quantity to determine the radial position of the constellation. For a given total number of constellation points, as the probability shaping parameter changes, the cumulative probability corresponding to each index changes accordingly, and the radius of the constellation point calculated subsequently changes accordingly. The constellation generation process does not need to sum the cumulative probabilities of each constellation point and use them as the transmission probability allocation; the actual frequency at which each constellation point is selected is determined by the distribution of the input two-dimensional signal and the distance mapping result. The parameter participates in signal mapping by adjusting the spatial position of the constellation points.

[0038] After obtaining the cumulative probability, calculate the negative natural logarithm of the difference between the cumulative probability and the cumulative probability, and then perform a power transformation with the radial shaping parameter in the current candidate parameter group as the exponent to determine the constellation point radius:

[0039] in, For the first The radius of each constellation point For radial shaping parameters, and , This represents the natural logarithm. Since the cumulative probability is between zero and one, the difference between one and the cumulative probability is positive, and its negative natural logarithm is also positive. Therefore, the calculated radius is a positive real number. For the current finite number of constellation points, each index produces a finite and definite radius value.

[0040] The probabilistic shaping parameter and the radial shaping parameter work together in a series of interconnected operations to affect the constellation point radius. The probabilistic shaping parameter first changes the position of the cumulative probability, and the radial shaping parameter then alters the mapping relationship between the negative natural logarithm result and the radius. With these two positive parameters fixed, as the constellation point index increases, the cumulative probability and its corresponding radius increase sequentially, forming radial positions arranged from the inside out according to the index. After the parameters are changed, the relative distribution of constellation points in different radial regions is also adjusted accordingly.

[0041] The radial shaping parameter's effect is manifested through exponentiation; its influence on the specific radius is related to the magnitude of the negative natural logarithm. The radial distribution of the current constellation can be determined by calculating point-by-point according to the formula. Using both parameters together for radius generation allows for the construction of all constellation points through a defined set of analytical relationships, avoiding the need to independently determine the radius for each constellation point. The generated radius is then combined with the angle corresponding to the same index to form two-dimensional coordinates for actual distance comparison.

[0042] While determining the radius, calculate the constellation point angle based on the angular control parameters in the current candidate parameter group:

[0043] in, For the first The angle of each constellation point This represents the angular increment between adjacent constellation points arranged by index. The angle of the first constellation point is zero, and subsequent constellation points are incremented by this angular increment. The angle calculation follows the constellation point index with the same radius, ensuring a clear pairing between the radial and angular positions of each constellation point.

[0044] The angles, accumulated by index, can exceed a complete circle, and their directions in the two-dimensional plane are determined by a complex exponential relationship. When the directions corresponding to different indices coincide, constellation points can still have different positions due to different radii. When the angular direction control parameters take different values, the angular direction distribution corresponding to each index changes, and together with the radial position, it affects the constellation point spacing. All candidate parameters are calculated according to the same angle generation method, and the specific distribution effect is determined by subsequent source reconstruction evaluation.

[0045] Using the radius corresponding to the same index as the polar radius and the angle as the polar angle, generate the complex coordinates of the constellation points:

[0046] in, For the first Complex coordinates of constellation points, It is a natural constant. The unit is the imaginary number. The current spiral amplitude phase-keying constellation is composed of all constellation points, and the set of constellation points is denoted as . ,in, This represents the set of constellation points corresponding to the current candidate parameter group. The real and imaginary parts of each complex coordinate are respectively used as the first and second coordinate components in the two-dimensional plane.

[0047] In the coordinate generation process, the probability shaping parameter and the radial shaping parameter determine the radial position, while the angular control parameter determines the angular position. These three parameters, using the coordinates of the same constellation point, participate in subsequent distance calculations. Changing the radius affects the amplitude difference between the constellation point and the input signal, while changing the angle affects their directional difference; both types of changes ultimately manifest as distance changes in a two-dimensional plane. The output of constellation construction is a complete set of constellation points, and the modulation processing directly uses the coordinates from this set.

[0048] like Figure 2 As shown, each discrete point in the helical amplitude phase keying constellation represents a constellation point. The horizontal axis represents the in-phase component of the constellation point, and the vertical axis represents the orthogonal component. The position of each constellation point in the two-dimensional plane is determined by the corresponding radius and angle. After the constellation is constructed, the two-dimensional signal obtained in step S1 and the constellation points are placed in the same in-phase-orthogonal coordinate plane. The Euclidean distance between the two-dimensional signal and each constellation point is calculated, and the constellation point with the smallest Euclidean distance is selected as the corresponding modulation symbol.

[0049] Taking a set of candidate parameters with a target modulation order of 64 as an example, we take... , , , , The double-helix spacing parameter is used to generate the current two-dimensional signal, while the other three parameters are used to construct the corresponding constellation. For the first constellation point, the calculation is as follows: , , The corresponding two-dimensional coordinates are approximately For the second constellation point, the calculation is as follows: , , The corresponding two-dimensional coordinates are approximately .

[0050] The values ​​above are approximate results for display. Actual calculations retain the precision of intermediate values, generating the remaining constellation points sequentially using the same formula, resulting in a candidate constellation containing 64 points. The angle increment in this example is... Radius is used to distinguish constellation positions even if subsequent indices correspond to the same direction. This set of parameters serves as candidate configurations in the search space, and its performance is evaluated based on the signal distortion ratio generated during the subsequent complete transmission process.

[0051] After the constellation is constructed, the paired signal values ​​in the two-dimensional signal are read sequentially. This is used as the signal coordinates on a two-dimensional plane. Simultaneously, the real and imaginary parts of each constellation point in the current constellation point set are read and used as the two coordinate components of the constellation point. The Euclidean distance between the current signal coordinates and the coordinates of all constellation points is calculated, and the constellation point with the smallest distance is selected. Each set of input signals participates in the comparison with complete two-dimensional coordinates, and the two components jointly determine the final mapped position.

[0052] In practice, the differences between the signal coordinates and the constellation point coordinates in the two components are calculated separately. These two differences are squared and then summed, and the squared distances are directly compared. The square root operation does not change the order of non-negative distances, therefore the comparison is consistent with the minimum Euclidean distance criterion. The distance comparison iterates through the current constellation point set, selecting the radius not only based on the input amplitude but also incorporating the influence of angular position on the distance in the same calculation, ensuring that both the generated radius and angle actually participate in the modulation symbol selection.

[0053] The complex coordinates of the selected constellation points are used as the modulation symbols corresponding to the current paired signal values, denoted as... ,in, This represents the modulation symbol to be transmitted. When multiple identical minimum distances exist, one constellation point is selected according to a pre-fixed constellation point index order. The same selection rule is applied throughout the current sequence to ensure that each input signal point corresponds to a specific output symbol. All output symbols are arranged according to the original order of the paired input signal values, forming the modulation symbol sequence to be transmitted.

[0054] During the modulation of a complete sequence, the current constellation point set remains unchanged. When the input signal point changes, the distance is recalculated within this set, and the corresponding symbol is selected. The constellation position is not changed based on the selection result of the previous signal point. Each modulation symbol is a constellation point in the current set. After the receiver constructs this set with the same parameters, it can determine the received signal to the same candidate coordinates, thus providing a unified reference position between modulation and demodulation.

[0055] There is usually a distance between the selected constellation points and the original two-dimensional input signal. This distance corresponds to the quantization deviation caused by mapping a continuous signal to a finite number of constellation points. The minimum Euclidean distance selection determines the output with the smallest distance under the current constellation conditions. Even if the subsequent receiver correctly determines the transmitted constellation point, the two recovered components may still deviate from the original normalized signal values. Therefore, the selection of constellation parameters continues to consider the channel transmission and the source reconstruction results after SK decoding, evaluating the impact of this deviation on the final source recovery.

[0056] When switching candidate parameter groups, the double-helix spacing parameter is used to generate the corresponding two-dimensional signal, and the constellation is reconstructed using the three constellation parameters, followed by distance mapping. Each modulation symbol sequence corresponds to its input two-dimensional signal, constellation point set, and candidate parameter group. The current output is used for subsequent channel transmission, and the current constellation point set is used for reception decision, enabling constellation construction, symbol selection, and source reconstruction to jointly participate in the end-to-end evaluation of candidate parameters.

[0057] S3, the modulation symbol is transmitted through the channel, the received signal is decided based on the spiral amplitude phase keying constellation, the decision result is denormalized and then SK decoding is performed to obtain the reconstructed source; In this embodiment, the aforementioned modulated symbol sequence is transmitted through a channel, and at the receiving end, a decision is made based on the same spiral amplitude phase keying constellation as at the transmitting end. The decision result is separated into two signal estimates. The normalized signal estimate sequence is recovered according to the branching combination order at the transmitting end. After inverse normalization, a continuously coded signal estimate sequence is obtained. Then, nonlinear inverse transform and curve coordinate recovery are performed to obtain two reconstructed signal sources. The entire processing maintains the correspondence between the received symbols, the continuously coded signal estimates, and the signal source sampling positions.

[0058] The channel employs additive white Gaussian noise, with its input being the modulation symbols generated by the transmitter and its output being the received symbols perturbed by noise. The channel model is represented as follows:

[0059] in, The received symbol in the received signal. For the corresponding modulation symbol, The mean is zero and the variance is Circularly symmetric complex Gaussian white noise, The complex Gaussian white noise The variance of the transmitted symbol is calculated. The real and imaginary parts of the transmitted symbol together represent the position of the selected constellation point in the two-dimensional plane. After noise superposition, the received coordinates deviate from the position of the transmitted constellation point, and the receiver uses this to make a candidate constellation point decision.

[0060] During parameter evaluation, the noise conditions of the channel model are set according to a given signal-to-noise ratio (SNR). The complete modulation symbol sequence generated by the current candidate parameter group is input into the model to obtain the corresponding received sequence. In actual communication processing, the transmitted and received symbols under the corresponding conditions are described using the same additive noise relationship. The candidate parameter evaluation adopts a consistent SNR definition, and the channel conditions correspond to the average power of the current transmitted symbol sequence, so that the subsequent reconstruction results have clear evaluation conditions.

[0061] The receiver uses the same total number of constellation points, probability shaping parameters, radial shaping parameters, and angular control parameters as the current candidate parameter set to generate a set of constellation points for decision-making. The same parameters generate the same constellation point coordinates using the same cumulative probability, radius, and angle formulas. Therefore, the receiver can use the constellation position actually used by the transmitter as a decision reference. All received symbols in the current sequence are processed using the same set, avoiding changes in symbol correspondence caused by switching candidate constellations within a sequence.

[0062] For each received symbol, calculate its Euclidean distance to all candidate constellation points in the current constellation point set, and select the constellation point with the smallest distance according to the maximum likelihood decision criterion to obtain the estimated value of the transmitted modulation symbol:

[0063] in, The complex coordinates of the constellation points corresponding to the judgment result. This is the set of constellation points for the current spiral amplitude phase-keying constellation. For the first Complex coordinates of candidate constellation points Index the constellations; Represents the modulus of a complex number; This represents a candidate constellation point that minimizes the squared distance. The decision result belongs to the current constellation point set and corresponds to the currently received symbol.

[0064] Under the additive white Gaussian noise model, the noise distributions corresponding to each candidate transmitted symbol have the same variance. The smaller the distance between the received coordinates and the candidate coordinates, the greater the corresponding conditional likelihood of reception. Therefore, this embodiment uses distance minimization to achieve maximum likelihood decision. This decision takes the coordinates of the current received symbol and the candidate constellation points as input. The finite constellation mapping deviation of the input signal and the impact of subsequent nonlinear decoding on the error are still evaluated through the final reconstructed source.

[0065] In the specific distance calculation, the difference between the real part of the received symbol and the real part of the candidate constellation point, as well as the difference between their imaginary parts, are calculated separately. The two differences are squared and summed, and the squared distances are compared. After comparing all candidate constellation points, the current decision result is determined. If multiple candidate constellation points achieve the same minimum distance, one constellation point is selected according to the index order consistent with the preset decision rule, so that each received symbol obtains a definite output.

[0066] All decision results are arranged in the original order of the received symbols to form a modulation symbol estimation sequence. When the noise disturbance is small, the received symbol can still use the transmitted constellation point as the nearest candidate point; when the disturbance changes the nearest candidate point, the decision result may fall into other constellation positions. Each decision result continues to participate in subsequent component separation, denormalization and decoding. The changes in the decisions are passed to the reconstructed source through the complete processing chain, rather than using whether the symbol is correctly judged as the evaluation criterion for parameter quality.

[0067] The real and imaginary parts of each decision result are separated sequentially. The real part is used as the estimate for the first component, and the imaginary part is used as the estimate for the second component. The transmitting end splits the normalized signal value into two paths and pairs them. The receiving end splits the constellation coordinates into two estimated components according to the same correspondence. At this point, each component corresponds to an estimate of a normalized coded value. It is still necessary to restore the sequence order and amplitude scale, and then use SK decoding to recover the sampled values ​​of the original two signal sources.

[0068] In the implementation where adjacent signal values ​​are sequentially assigned to the first and second paths at the transmitting end, the receiving end alternates between real and imaginary components. First, the real and imaginary parts of the first decision result are arranged, then the real and imaginary parts of the second decision result are arranged, and so on until the end of the sequence, thus restoring the normalized signal order before the transmitting end's demultiplexing. Each decision result generates two normalized signal estimates, preserving both the pairing relationship and the original time order.

[0069] The restored normalized signal estimate is denoted as... Inverse normalization is performed using the normalization factor used during normalization at the sending end:

[0070] in, For the estimated value of the continuously coded signal, These are the continuous encoded values ​​before normalization at the sending end. Represents the mathematical expectation. This is the average power of the continuously encoded signal used in the current sequence at the transmitting end, and it is greater than zero. The current received sequence uses the same average power value as the transmitted sequence, so that the normalized signal estimate is restored to the amplitude scale corresponding to the nonlinear coded output.

[0071] When the transmitter estimates the average power using a finite-length coded sequence, the receiver uses the estimation result corresponding to that sequence, and the normalization factor remains consistent in both the transmitter and receiver configurations. Within the current sequence, all normalized signal estimates are proportionally recovered using the same factor. Since the amplitude of the received symbol is already affected by constellation mapping and channel noise, the inverse normalization uses the factor configured at the transmitter and does not change it based on the current decision result, thus avoiding additional scale bias affecting subsequent nonlinear inverse transforms.

[0072] During the candidate parameter evaluation process, the normalization factor determined during encoding at the transmitting end is used in conjunction with the corresponding receiving reconstruction process. After switching candidate parameter groups, the receiving end synchronously adopts the factor corresponding to that group of encoded sequences and the constellation constructed by that group of parameters. After denormalization is completed, a continuous coded signal estimation sequence is formed according to the order of the recovered continuous coded values, and each estimation value is then independently entered into the corresponding SK decoding process.

[0073] SK decoding first performs the inverse nonlinear transform. The transmitter preserves the sign of the spiral mapping parameters using a sign function and raises the absolute value to the power of the nonlinear distribution control parameter. Therefore, the receiver obtains the absolute value of the estimated continuous coded signal, raises it to the power of the reciprocal of the distribution control parameter, and then assigns the original sign to the estimated value, thus obtaining the estimated spiral mapping parameters. When the estimated value of the continuous coded signal is zero, the obtained estimated spiral mapping parameters are also zero.

[0074] Let the above estimated values ​​of the spiral mapping parameters be denoted as... ,in, The signed polar angle parameter estimate is obtained by nonlinear inverse transformation of the current continuously coded signal estimate. The receiver uses the same nonlinear distributed control parameters as the transmitter. In this embodiment, the value is fixed at 1.5. This inverse transform preserves the sign correspondence, enabling the positive and negative code estimates to be recovered to their respective curve branches. The inverse transform result is then used as the input for curve coordinate calculation.

[0075] Substitute the estimated values ​​of the spiral mapping parameters into the parametric equations of the aforementioned double Archimedean spiral curve, and use the double spiral spacing parameter from the current candidate parameter set. Calculate the two coordinate components of the corresponding curve position. Use the first coordinate component as the reconstructed sample value of the first source, denoted as... The second coordinate component is used as the reconstructed sample value of the second source, denoted as... .in, and These are the estimation results for two sampled values ​​in the original sample pair, respectively.

[0076] Each continuously coded signal estimate, after the above processing, generates a pair of reconstructed sample values. All first coordinate components are arranged sequentially to form the first reconstructed signal source, and all second coordinate components are arranged sequentially to form the second reconstructed signal source. According to the current demultiplexing method, each received symbol recovers two continuously coded signal estimates, and each continuously coded signal estimate recovers a pair of source sample values, ultimately resulting in two reconstructed sequences corresponding to the original input length.

[0077] The reconstructed sampled values ​​lie on the curve determined by the current double-helix spacing parameter, and may deviate from the original two-dimensional source points. This deviation includes the combined effects of curve mapping approximation, finite constellation quantization, and channel noise after decision and inverse transformation. Even if the decision result is consistent with the transmitted constellation point, it is still necessary to calculate the error based on the original and reconstructed sampled values ​​in order to evaluate the recovery effect of the entire analog joint coding and modulation process.

[0078] The receiver's symbol decision, demultiplexing, inverse normalization, and curve decoding sequentially use the output of the previous process. The two reconstructed signal sources retain the original sampling order and the correspondence between the current candidate parameter groups. During the parameter evaluation phase, the reconstructed result is compared with the saved original signal sources. When using selected parameters for transmission, the receiver completes the reconstruction based on the same constellation and decoding configuration. Therefore, the signal source estimation results generated in the current transmission process can be directly used for subsequent distortion calculation and parameter selection.

[0079] S4. Under a given signal-to-noise ratio, encoding, modulation, and source reconstruction are performed for each candidate parameter group. The signal distortion ratio is calculated based on the reconstructed source and the two Gaussian analog sources. The candidate parameter group with the largest signal distortion ratio is selected to configure SK encoding and spiral amplitude phase keying modulation.

[0080] During the parameter configuration phase, for a given signal-to-noise ratio (SNR), different candidate parameter sets are used to complete coding, constellation construction, modulation, channel transmission, and source reconstruction. The difference between the original source and the reconstructed source is used to evaluate the corresponding parameter set. After evaluating the candidate parameters within the preset joint search space, the candidate parameter set with the largest signal-to-distortion ratio is selected for configuring SK coding and spiral amplitude phase-keying modulation under the corresponding channel conditions. The parameter search organizes the complete processing of each candidate configuration, and once selected, signal transmission is performed according to that configuration.

[0081] The evaluation input uses the aforementioned two independent Gaussian analog sources, each following a Gaussian distribution with a mean of zero and a variance of one, including... Each sampling point is used. When comparing different candidate parameter groups under the same given signal-to-noise ratio, the original source sequence, the total number of constellation points, the total number of sampling points, and the order of branch combinations are kept consistent. The original source sampling values ​​are saved, and after reception and reconstruction, they are compared at the same position to ensure that the evaluation results reflect the differences in processing the same input by each candidate parameter.

[0082] The joint search targets the double helix spacing parameter. Radial shaping parameters Probability Integral Parameters and angular control parameters The preset value ranges of each parameter are discretized, and then the discretized values ​​are combined into candidate parameter groups. Each candidate parameter group contains all four parameters, and the coding curve and modulation constellation are configured accordingly so that the current reconstruction result can be attributed to a specific coding and modulation combination.

[0083] The double-helix spacing parameter ranges from 0.1 to 6.0, with a discretization step size of 0.1; the radial shaping parameter ranges from 0.1 to 2.5, with a discretization step size of 0.1; the probabilistic shaping parameter ranges from 0.1 to 1.5, with a discretization step size of 0.1; and the angular control parameter ranges from... to The discretization step size is 0.002 radians. Each parameter is generated with a fixed step size starting from the corresponding lower limit, and the value does not exceed the upper limit. When the last value does not coincide with the upper limit, the value generated with the step size and not exceeding the upper limit is retained, and other values ​​are added without changing the step size.

[0084] Let the joint search space consisting of all candidate parameter sets be denoted as . Each element is The system iterates through the discrete values ​​of each parameter in a preset order, ensuring that all candidate parameter groups in the search space participate in the evaluation. (Nonlinear distributed control parameters) The value is fixed at 1.5 and remains unchanged throughout the evaluation of all candidate parameters. The double-helix spacing parameter participates in the joint search, while the fixed nonlinear distribution control parameter is used to maintain the same nonlinear transformation form.

[0085] For the current candidate parameter set, the coding curve is first constructed using the double helix spacing parameter. The helix mapping parameters are then calculated pairwise for each pair of original signals to generate a continuous coded signal. The average power of the current coded signal is then normalized, and a two-dimensional signal is generated in a predetermined order. Subsequently, the cumulative probability, radius, and angle are calculated based on the three constellation parameters in the current candidate parameter set to construct the corresponding constellation. A minimum Euclidean distance mapping is then performed on the current two-dimensional signal to obtain the modulation symbol sequence corresponding to this parameter set.

[0086] The current modulated symbol sequence is input into a channel with a given signal-to-noise ratio. The receiver uses the same constellation to perform decision-making, reconstructs the sequence according to the same decoupling relationship, and decodes using the current normalization factor and the corresponding coding parameters. Each candidate parameter group obtains its own two reconstructed signal sources through complete processing, without replacing the current reconstruction result with the decision results of other candidate parameter groups. Since different double-helix spacings may change the coded signal and its normalization scale, the current normalization factor is always matched with the current coded sequence.

[0087] The given signal-to-noise ratio (SNR) is determined using a uniform caliber, which is the ratio of the average power of the modulation symbol to the variance of the complex noise. This ratio is maintained consistent across different candidate parameter groups. When evaluating using an additive white Gaussian noise model, the corresponding noise variance is determined based on the average power of the current modulation symbol sequence. Changes in constellation parameters may alter the average power of the modulation symbols; therefore, noise conditions are set using the same SNR caliber to ensure consistent channel quality evaluation across all parameter groups.

[0088] After obtaining the two reconstructed signal sources, align them according to their original sampling positions. Then, align the first original signal source... Each sampled value is denoted as The corresponding reconstructed value is denoted as The second original source Each sampled value is denoted as The corresponding reconstructed value is denoted as .in, This is the sampling point index, with a value of The correspondence is based on the sequence positions preserved during the encoding, decoding, and decoding processes.

[0089] Calculate the difference between the original sampled value and the corresponding reconstructed sampled value, sum the squares of the two errors, and then average them over all sampled values ​​to obtain the end-to-end mean square error:

[0090] in, This represents the end-to-end mean square error of the current candidate parameter set. The denominator is... The total number of sampled values ​​from the two signal sources used in the evaluation is represented by the sum of the squared errors of the two sources, which are then averaged to ensure that the reconstruction biases of both sources are reflected in the evaluation result. The squaring operation ensures that both positive and negative errors are accumulated as non-negative values, preventing them from canceling each other out. Both signal sources use the same variance and the same number of samples, and are evaluated uniformly according to this expression.

[0091] The mean square error (MSE) takes the original source and the final reconstructed source as inputs, reflecting the combined effects of curve mapping, finite constellation mapping, and channel transmission into a single value. Before evaluation, the corresponding order and amplitude scale of the two sources are maintained, and the reconstruction results are not rearranged or independently normalized to change their error magnitude. Therefore, the MSE corresponds to the reconstruction deviation output by the actual coding and modulation process, providing a consistent data basis for comparing different candidate parameters.

[0092] Calculate the signal distortion ratio based on the end-to-end mean square error:

[0093] in, The signal-to-distortion ratio of the current candidate parameter group, in decibels; This represents the logarithm to base 10. In this embodiment, the original source has a mean of zero and a variance of one, and the average power of a single source is one; therefore, the above expression is used. When the mean square error is greater than zero, the smaller the error, the greater the signal distortion ratio. For the theoretically zero mean square error case, which corresponds to the limit of no reconstruction distortion, it is treated as a case superior to the positive error result in parameter comparison.

[0094] The current signal distortion ratio is associated with and recorded with the candidate parameter group and the given signal-to-noise ratio. Then, the next candidate parameter group is processed according to the search order. The evaluation values ​​of different candidate parameter groups all come from their own complete transmission and reconstruction processes. The constellation construction results or local symbol distances are only used as intermediate processing results. Even if a constellation has a large inter-point distance, its signal distortion ratio is determined based on the final reconstructed source, so that the coding curve and constellation distribution are jointly selected through the same evaluation index.

[0095] In the specific comparison, the maximum signal-to-distortion ratio and corresponding parameter are recorded in the evaluated candidate parameter groups. After evaluating each new candidate parameter group, the new result is compared with the recorded maximum value; if the new value is larger, the maximum value and corresponding parameter are updated; otherwise, the original result is retained. If multiple candidate parameter groups obtain the same maximum value, one of them is retained according to a pre-fixed search order. This comparison method can be completed during the traversal process, without waiting for all reconstructed sequences to be saved simultaneously before making a judgment.

[0096] After evaluating all candidate parameter sets within the joint search space, the optimal parameter set is represented as:

[0097] in, This represents the signal distortion ratio of the corresponding candidate parameter group at the current given signal-to-noise ratio; This represents the candidate parameter group that maximizes the evaluation value. , , and These are the selected double-helix spacing parameter, radial shaping parameter, probabilistic shaping parameter, and angular control parameter, respectively. The optimal parameter set corresponds to the joint search space determined by the preset value range, discretization step size, and current evaluation conditions.

[0098] After determining the optimal parameter set, the double-helix spacing parameter is used to configure the SK coding curve, and the radial shaping parameter, probabilistic shaping parameter, and angular control parameter are used to construct the constellation for modulation. Subsequent sources to be transmitted generate continuous coded signals according to the selected coding curve. The normalization factor is determined based on the current sequence to be transmitted, completing the two-dimensional combination and constellation mapping. The receiver uses the same constellation, corresponding normalization factor, and the same decoding parameters to recover the source, thus implementing the selected parameters into the specific configuration for transmission and reception processing.

[0099] For different signal-to-noise ratios (SNRs) to be evaluated, the above parameter evaluations are performed under the corresponding channel conditions, and the optimal parameter set corresponding to each SNR is recorded. The evaluation results for different SNRs are compared and used separately, and the parameter configurations are applicable to their corresponding channel conditions. Under each given SNR, the double-helix spacing parameter and the three constellation parameters are jointly determined through the same end-to-end reconstruction evaluation, and the resulting configuration is used to complete the analog joint coding modulation and source recovery under the corresponding conditions.

[0100] like Figure 3 As shown, the horizontal axis represents the channel signal-to-noise ratio (SNR), and the vertical axis represents the signal distortion ratio of the source reconstruction, both expressed in decibels (dB). The GAPSK curve in the figure represents the comparative scheme using a Gaussian amplitude-phase-keying (APSK) constellation, while the Spiral APSK curve represents the scheme using the spiral amplitude-phase-keying (SPSK) constellation described in this embodiment. Under the test conditions shown, the signal distortion ratio of both schemes increases overall with increasing SNR, gradually stabilizing in the higher SNR range. Specifically, in the high SNR range shown in the figure, the signal distortion ratio of the spiral SPSK scheme is higher than that of the Gaussian SPSK comparative scheme. For the same source variance, a higher signal distortion ratio corresponds to a lower mean square error, indicating that the scheme described in this embodiment can improve the source reconstruction effect under the corresponding test conditions.

[0101] Please see Figure 4 The second embodiment of the present invention provides an analog joint coding modulation device based on helical amplitude phase keying, comprising: The encoding processing unit 201 is used to perform SK encoding on two Gaussian analog sources according to the double helix spacing parameter in the candidate parameter group to obtain a continuous encoded signal, and then normalize and combine the continuous encoded signal to obtain a two-dimensional signal. The constellation modulation unit 202 is used to obtain a cumulative probability by performing a power transformation on the normalized constellation point index using a probability shaping parameter according to the candidate parameter group, to obtain a radius by performing a power transformation on the negative natural logarithm of the difference between the cumulative probability and the radial shaping parameter, to determine an angle according to the angular control parameter and the constellation point index, to construct a spiral amplitude phase keying constellation according to the radius and angle, and to select the constellation point with the smallest Euclidean distance to the two-dimensional signal from the constellation as the modulation symbol. The transmission reconstruction unit 203 is used to transmit the modulation symbol through the channel, make a decision on the received signal based on the spiral amplitude phase keying constellation, and perform SK decoding after inverse normalization of the decision result to obtain the reconstructed signal source. The parameter optimization unit 204 is used to complete encoding, modulation and source reconstruction for each candidate parameter group under a given signal-to-noise ratio, calculate the signal distortion ratio based on the reconstructed source and the two Gaussian analog sources, and select the candidate parameter group with the largest signal distortion ratio to configure SK encoding and spiral amplitude phase keying modulation.

[0102] The third embodiment of the present invention provides an analog joint coding modulation device based on helical amplitude phase keying, including a memory and a processor. The memory stores a computer program, which can be executed by the processor to implement an analog joint coding modulation method based on helical amplitude phase keying as described in any of the above embodiments.

[0103] The fourth embodiment of the present invention provides a computer-readable storage medium storing a computer program, which can be executed by a processor of the device in which the computer-readable storage medium is located, to implement an analog joint coding modulation method based on helical amplitude phase keying as described in any of the above claims.

[0104] Exemplarily, the computer program described in the third and fourth embodiments of the present invention can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in implementing an analog joint coding modulation device based on helical amplitude phase keying. For example, the apparatus described in the second embodiment of the present invention.

[0105] The processor referred to can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the aforementioned analog joint coding modulation method based on helical amplitude-phase keying (HAPK), connecting various parts of the entire HAPK-based analog joint coding modulation method through various interfaces and lines.

[0106] The memory can be used to store the computer program and / or modules. The processor, by running or executing the computer program and / or modules stored in the memory, and by calling the data stored in the memory, implements various functions of an analog joint coding modulation method based on helical amplitude phase keying. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, text conversion function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, text message data, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0107] If the implemented module is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0108] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0109] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An analog joint coding modulation method based on helical amplitude phase keying, characterized in that, include: Based on the double helix spacing parameter in the candidate parameter group, the two Gaussian analog sources are SK encoded to obtain a continuous encoded signal. The continuous encoded signal is then normalized and combined to obtain a two-dimensional signal. Based on the candidate parameter set, the cumulative probability is obtained by performing a power transformation on the normalized constellation point index using the probability shaping parameter. The cumulative probability is then subtracted from 1 to obtain the probability difference. The natural logarithm of the probability difference is then taken and its opposite is taken to obtain the radial transformation value. Using the radial shaping parameter as the exponent, the radial transformation value is exponentially operated to obtain the radius. The angle is determined according to the angular control parameter and the constellation point index. A helical amplitude phase keying constellation is constructed according to the radius and the angle. The constellation point with the smallest Euclidean distance to the two-dimensional signal is selected from the constellation as the modulation symbol. The modulation symbols are transmitted through the channel, and the received signals are decided based on the spiral amplitude phase keying constellation. The decision results are denormalized and then SK decoding is performed to obtain the reconstructed source. Given a signal-to-noise ratio, encoding, modulation, and source reconstruction are performed for each candidate parameter group. The signal distortion ratio is calculated based on the reconstructed source and the two Gaussian analog sources. The candidate parameter group with the largest signal distortion ratio is selected to configure SK encoding and spiral amplitude phase keying modulation.

2. The analog joint coding modulation method according to claim 1, characterized in that, The step of performing SK encoding on the two Gaussian analog sources based on the double helix spacing parameter in the candidate parameter group to obtain a continuous coded signal is as follows: A double Archimedean spiral curve is used as the SK encoding curve, and the parametric equation of the SK encoding curve is determined based on the double spiral spacing parameter: The corresponding sampled values ​​from the two Gaussian analog sources are used to form a two-dimensional source point. The spiral mapping parameters corresponding to the two-dimensional source point on the SK coding curve are determined according to the minimum Euclidean distance criterion. The continuous coded signal is obtained by performing a nonlinear transformation on the spiral mapping parameters: in, The double helix spacing parameter; These are signed polar angle parameters; These are the two-dimensional coordinates on the SK encoding curve; and These are the first and second coordinate components of the two-dimensional coordinate system, respectively. It is a symbolic function; For the two-dimensional source point, and These are the corresponding sampled values ​​from the two Gaussian analog signal sources; The spiral mapping parameters corresponding to the two-dimensional source points; Represents the Euclidean norm; This represents the polar angle parameter that minimizes the Euclidean norm. The encoded value of the continuous encoded signal; For the nonlinear transformation; These are nonlinear distributed control parameters, and ; It represents the absolute value.

3. The analog joint coding modulation method according to claim 2, characterized in that, The normalization and demultiplexing of the continuous coded signal to obtain a two-dimensional signal is specifically as follows: Based on the average power of the continuously encoded signal, the continuously encoded signal is normalized to obtain a normalized signal: The normalized signal sequence is split into two signals of equal length, and corresponding signal values ​​from the two signals are paired to form the two-dimensional signal. ; in, The signal value of the normalized signal; Represents the mathematical expectation; The average power of the continuously encoded signal; and The corresponding signal values ​​in the two split signals constitute the first coordinate component and the second coordinate component of the two-dimensional signal, respectively.

4. The analog joint coding modulation method according to claim 1, characterized in that, The cumulative probability is obtained by performing a power transformation on the normalized constellation point index using a probability shaping parameter. The cumulative probability is then subtracted from 1 to obtain the probability difference. The natural logarithm of this probability difference is taken, and its negative value is obtained to obtain the radial transformation value. Finally, the radial transformation value is exponentially raised to the radial shaping parameter to obtain the radius. Specifically: The total number of constellation points is determined based on the target modulation order. Each constellation point is numbered sequentially. The constellation point index is then subtracted by half and divided by the total number of constellation points to obtain the normalized constellation point index. Using the probability shaping parameter as the exponent, a power transformation is performed on the normalized constellation point index to obtain the cumulative probability of the corresponding constellation point: Calculate the negative natural logarithm of the difference between the cumulative probabilities and the negative natural logarithm, and then perform a power transformation on the negative natural logarithm using the radial shaping parameter as the exponent to obtain the radius of the corresponding constellation point: in, The total number of points in the constellation is an integer greater than 1; For the constellation point index, ; For the first The cumulative probability corresponding to each constellation point; Let be the probability shaping parameter, and ; For the first The radius of each constellation point; The radial shaping parameter is, and .

5. The analog joint coding modulation method according to claim 4, characterized in that, The process of determining the angle based on the angular control parameters and constellation point index, and constructing a helical amplitude phase keying constellation according to the radius and angle, specifically involves: Subtract one from the constellation point index and multiply it by the angular control parameter to determine the angle of the corresponding constellation point: Using the radius and angle corresponding to the same constellation point index as the polar radius and polar angle of that constellation point, respectively, the complex coordinates of that constellation point are determined: The spiral amplitude phase-keying constellation is composed of all constellation points, and its constellation point set is as follows: ; in, The angular control parameter represents the angular increment between adjacent constellation points arranged by constellation point index; For the first The angle of each constellation point; For the first Complex coordinates of constellation points; It is a natural constant; It is the imaginary unit.

6. The analog joint coding modulation method according to claim 5, characterized in that, The step of selecting the constellation point with the smallest Euclidean distance from the two-dimensional signal as the modulation symbol from the constellation specifically involves: For each pair of paired signal values ​​in the two-dimensional signal, the pair of paired signal values ​​are used as coordinates in the two-dimensional plane, and the real and imaginary parts of the complex coordinates of each constellation point in the spiral amplitude phase keying constellation are used as the first coordinate component and the second coordinate component of the constellation point in the two-dimensional plane, respectively. Calculate the Euclidean distance between the coordinates corresponding to the paired signal values ​​and the coordinates corresponding to each constellation point, select the constellation point with the smallest Euclidean distance, and determine the complex coordinates of the selected constellation point as the modulation symbol corresponding to the paired signal values. The modulation symbols are arranged in the original order of each pair of paired signal values ​​to form a sequence of modulation symbols to be transmitted.

7. The analog joint coding modulation method according to claim 3, characterized in that, The process involves transmitting the modulation symbols through the channel, making decisions on the received signals based on the spiral amplitude phase keying constellation, inversely normalizing the decision results, and then performing SK decoding to obtain the reconstructed source. Specifically: The modulation symbol is transmitted through an additive white Gaussian noise channel to obtain the received signal: At the receiving end, the same spiral amplitude phase keying constellation as at the transmitting end is used. The Euclidean distance between the received signal and each candidate constellation point is calculated. The candidate constellation point with the smallest Euclidean distance is selected according to the maximum likelihood decision criterion, and the decision result is obtained. Separate the real and imaginary parts of the decision result to obtain the estimated values ​​of the two signals. Then, according to the order in which the normalized signal is split at the transmitting end, restore the estimated values ​​of the two signals to the normalized signal estimation sequence. Using the average power used during normalization at the transmitting end, the normalized signal estimation sequence is inversely normalized to obtain the continuously coded signal estimation sequence: Perform the inverse transformation of the nonlinear transformation on the estimated sequence of the continuous coded signal to obtain the estimated value of the spiral mapping parameter. Substitute the estimated value of the spiral mapping parameter into the parameter equation of the SK coding curve with the same double spiral spacing parameter as the transmitting end to obtain the reconstructed signal source. in, The received symbol is the received signal; The corresponding modulation symbol; The mean is zero and the variance is Circularly symmetric complex Gaussian white noise; This represents the power of complex noise. This refers to the set of constellation points of the spiral amplitude phase-keying constellation; For the first Complex coordinates of candidate constellation points; Index the constellations; The complex coordinates of the constellation point corresponding to the judgment result; Represent candidate constellation points that minimize the squared distance; Represents the modulus of a complex number; This is the estimated value of the normalized signal after restoration; This is the estimated value of the corresponding continuous coded signal.

8. The analog joint coding modulation method according to claim 1, characterized in that, The calculation of the signal distortion ratio based on the reconstructed source and the two Gaussian analog sources specifically involves: The sampled values ​​in the reconstructed source are compared with the corresponding sampled values ​​in the two Gaussian analog sources. The squares of the sampling errors for each source are calculated, and the average of the squares of all sampling errors is used to obtain the end-to-end mean square error. The signal distortion ratio is calculated based on the end-to-end mean square error: in, This represents the total number of sampling points for each Gaussian analog source. For sampling point index; and These are the first two Gaussian analog signal sources, respectively. Each sample value; and These are the corresponding reconstructed sample values; The end-to-end mean square error; The signal distortion ratio is given.

9. The analog joint coding modulation method according to claim 1, characterized in that, The process of encoding, modulation, and source reconstruction for each candidate parameter group under a given signal-to-noise ratio is as follows: The preset value ranges of the double helix spacing parameter, radial shaping parameter, probabilistic shaping parameter, and angular control parameter are discretized respectively. The discretized values ​​of each parameter are combined to form a joint search space. Each candidate parameter group in the joint search space is represented as follows: For each candidate parameter group, the double helix spacing parameter is used to perform SK encoding on the two Gaussian analog sources. The resulting continuous encoded signals are normalized and then combined to obtain the two-dimensional signal corresponding to the candidate parameter group. By using the probability shaping parameters, radial shaping parameters, and angular control parameters in the candidate parameter set, the radius and angle of each constellation point are determined, and the spiral amplitude phase keying constellation corresponding to the candidate parameter set is constructed. The constellation is then used to modulate the two-dimensional signal corresponding to the candidate parameter set. The modulation symbol is transmitted at the given signal-to-noise ratio. At the receiving end, the spiral amplitude phase keying constellation corresponding to the candidate parameter group is used for decision-making. The normalization factor corresponding to the candidate parameter group is used for inverse normalization. The double spiral spacing parameter in the candidate parameter group is used for SK decoding to obtain the reconstructed signal source corresponding to the candidate parameter group.

10. The analog joint coding modulation method according to claim 9, characterized in that, The selection of the candidate parameter group with the largest signal distortion ratio to configure SK encoding and spiral amplitude phase keying modulation is specifically as follows: Calculate and compare the signal distortion ratios corresponding to each candidate parameter group, and determine the optimal parameter group according to the following formula: The SK coding is configured using the double-helix spacing parameter in the optimal parameter set, and the helical amplitude-phase keying constellation for modulation is constructed using the radial shaping parameter, probability shaping parameter and angular control parameter in the optimal parameter set to complete the analog joint coding modulation under the given signal-to-noise ratio; in, The candidate parameter group; , , and These are the double helix spacing parameter, radial shaping parameter, probability shaping parameter, and angular control parameter, respectively. The joint search space consists of all candidate parameter sets; The signal distortion ratio of the corresponding candidate parameter group at the given signal-to-noise ratio; This represents a group of candidate parameters that maximize the signal distortion ratio. , , and These are the corresponding parameters in the optimal parameter set.