High-dimensional qam signal modulation system and modulation method based on clustering algorithm sequence extraction

Abstract

The application relates to a high-dimensional QAM signal modulation system and a modulation method based on a clustering algorithm sequence extraction, and the method comprises the following steps: firstly, generating an original high-dimensional constellation point set and an index bit stream; then, calculating the clustering number according to a target source entropy and a sequence length, and clustering the original constellation points by adopting a K-means++ algorithm, selecting the sequence closest to the center in each cluster as a representative sequence to form an optimized constellation point set; then, mapping the index bit stream to the index of the representative sequence, extracting the corresponding QAM symbol sequence according to the index; finally, modulating the sequence into a time domain signal and transmitting, and at the receiving end, calculating the Euclidean distance between the received signal and all the representative sequences, identifying the sequence according to the minimum distance criterion and restoring the index bit stream. The application can significantly improve the noise resistance and the spectrum efficiency, controls the system complexity, and has great practical value in adapting to high-speed and high-reliability communication scenes.

Description

Technical Field

[0001] This invention relates to the field of optical communication and wireless communication technologies, specifically to a high-dimensional QAM signal modulation system and modulation method based on sequence extraction using a clustering algorithm. Background Technology

[0002] The rapid development of fifth-generation and future communication technologies (5G / 6G), high-speed optical communication, and the Internet of Things (IoT) has placed increasingly stringent demands on the spectral efficiency and noise immunity of communication systems. Quadrature Amplitude Modulation (QAM), as a core modulation technology in modern communication systems, encompasses modulation formats ranging from low-order QPSK (which can be considered 4-QAM) to high-order 16-QAM, 64-QAM, and 256-QAM. Low-order QAM (such as QPSK) has the advantages of simple implementation and strong robustness, but its spectral efficiency is limited. While high-order QAM can significantly improve spectral efficiency, its noise immunity deteriorates sharply in high-code-rate transmission scenarios due to the reduced Euclidean distance between constellation points. Therefore, effectively balancing spectral efficiency and noise immunity among various QAM modulation schemes has become a key challenge in system design.

[0003] To overcome the aforementioned bottlenecks, advanced modulation techniques such as probabilistic shaping (PS) and geometric shaping (GS) have been extensively studied. However, both techniques have inherent limitations when applied to QAM of different orders. For low-order QAM (such as QPSK), probabilistic shaping approximates channel capacity by optimizing the symbol probability distribution, but its shaping gain is limited by the very small number of constellation points, resulting in low freedom of probability adjustment and limited performance improvement. Meanwhile, the distributed matching unit (DM) necessary for implementing PS introduces disproportionate complexity and cost to low-order modulation, leading to poor cost-effectiveness. For high-order QAM, while probabilistic shaping can play a greater role, the complexity of DM increases sharply with the modulation order, resulting in high hardware implementation costs. On the other hand, geometric shaping increases the minimum Euclidean distance by optimizing the constellation point positions, thereby improving noise margin. However, the constellation points of QPSK are uniformly distributed on the unit circle, forming an optimal orthogonal structure. Any modification will not only destroy its orthogonality but also significantly increase the accuracy requirements of the digital-to-analog / analog-to-digital converter (DAC / ADC), making it impractical in engineering. Therefore, GS is also unsuitable for QPSK. Existing research attempts to optimize constellation distribution in higher-dimensional spaces using multidimensional geometric shaping (MD-GS), but the complexity of its demapping algorithm increases exponentially with dimensionality, making it difficult to implement practically. More critically, existing time-domain multidimensional modulation methods based on random sequence extraction lack systematic optimization for high-dimensional constellation point distributions, resulting in limited and unstable performance gains. Therefore, existing technologies struggle to effectively probabilistically or geometrically shape QAM signals from QPSK to 256-QAM while maintaining low complexity, significantly limiting the application potential of QAM in next-generation high-performance communication systems.

[0004] Therefore, it is imperative to propose a high-dimensional QAM signal modulation system and method based on sequence extraction using clustering algorithms to overcome the aforementioned technical problems. This system should be able to achieve high-performance constellation distribution optimization with low complexity and be applicable to various QAM formats from low to high order, thereby improving the overall performance of the communication link.

[0005] It is understood that the above statements only provide background information related to the present invention and do not necessarily constitute prior art. Summary of the Invention

[0006] The purpose of this invention is to provide a high-dimensional QAM signal modulation system and method based on sequence extraction using a clustering algorithm. This system can achieve high-performance constellation distribution optimization with low complexity and is applicable to various QAM formats from low to high order, thereby improving the overall performance of the communication link.

[0007] To achieve the above objectives, this invention provides a high-dimensional QAM signal modulation method based on sequence extraction using a clustering algorithm, comprising the following steps: S1, generating an original high-dimensional constellation point set according to a preset sequence length S and a basic modulation order M; and generating a pseudo-random binary sequence as an index bit stream; wherein, the original high-dimensional constellation point set includes S1) A set of QAM symbol sequences; S2) Based on the target source entropy SE and sequence length S, calculate a preset number n, and use the K-means++ clustering algorithm to divide the original high-dimensional constellation point set into a preset number n clusters. In each cluster, select the sequence closest to the cluster center as the representative sequence to form an optimized high-dimensional constellation point set; S3) Map the index bitstream generated in S1 to the corresponding sequence index in the optimized high-dimensional constellation point set in S2; and extract the corresponding QAM symbol sequence from the optimized high-dimensional constellation point set according to the received sequence index; S4) Modulate the extracted QAM symbol sequence into a time-domain high-dimensional QAM signal and transmit it; and at the receiving end, convert the received signal into a baseband signal, calculate its Euclidean distance with all representative sequences in the optimized high-dimensional constellation point set, and identify the representative sequence with the smallest distance according to the minimum distance criterion, and restore its index to the original index bitstream.

[0008] Preferably, the preset quantity n is calculated based on the target source entropy SE and the sequence length S, using the following formula: .

[0009] Preferably, the objective function of the K-means++ clustering algorithm is to maximize the average Euclidean distance between clusters. Specifically: , In the formula, the preset quantity The total number of cluster centers. and These are the indices of the cluster centers. For the first Cluster center, For the first The center of a cluster.

[0010] Preferably, the specific steps of the K-means++ clustering algorithm in selecting a preset number of n cluster centers include: S21, randomly selecting the first cluster center from the original set of high-dimensional constellation points. S22. Calculate the distance from each point x in the cluster to the cluster center. The shortest distance D(x); S23, select the next cluster center according to the probability formula. S24. Repeat steps S22-S23 until the preset number of n cluster centers have been selected.

[0011] Preferably, in step S23, the probability formula is: In the formula, P(x) represents the probability that point x is selected as the next cluster center.

[0012] Preferably, before step S2, a standardization process is further included for the original high-dimensional constellation point set, and the standardization formula is: In the formula, Let be the original value of the i-th sample in the j-th dimension of the original sequence; i is the sample index; j is the dimension index. and These are the mean and standard deviation of the j-th dimension, respectively; This is the standardized data.

[0013] The present invention also provides a high-dimensional QAM signal modulation system based on K-means sequence extraction, for implementing the above-mentioned high-dimensional QAM signal modulation method based on K-means sequence extraction, comprising: an interconnected sequence construction system, a sequence extraction system, and a multi-dimensional modulation and demodulation system.

[0014] Preferably, the sequence construction system includes: a full sequence generation module, used to generate an original high-dimensional constellation point set based on a set sequence length S and a fundamental modulation order M, wherein the original high-dimensional constellation point set contains The system includes: a set of QAM symbol sequences; a pseudo-random binary sequence generation module for generating pseudo-random binary sequences as index bit streams; a sequence extraction system comprising: a K-means-based sequence extraction module connected to the full sequence generation module, which receives the original high-dimensional constellation point set and uses the K-means++ clustering algorithm to divide the original high-dimensional constellation point set into multiple clusters of a preset number n, and selects the sequence closest to the center of each cluster as a representative sequence to form an optimized high-dimensional constellation point set; and a multi-dimensional modulation and demodulation system comprising: a bit-to-sequence index mapping module connected to the pseudo-random binary sequence generation module. A column generation module is connected to the array and is used to map the index bit stream to the corresponding sequence index in the optimized high-dimensional constellation point set. A sequence symbol extraction generation module is connected to the K-means-based sequence extraction module and the bit-to-sequence index mapping module, respectively, and is used to extract the corresponding QAM symbol sequence from the optimized high-dimensional constellation point set according to the received sequence index. A signal transmission link is connected to the sequence symbol extraction generation module and is used to modulate the extracted QAM symbol sequence into a time-domain high-dimensional QAM signal and transmit it through the channel, and to demodulate the time-domain high-dimensional QAM signal at the receiving end to recover the sequence information.

[0015] Preferably, the signal transmission link includes: a sequence modulation module connected to the extracted sequence symbol generation module, used to modulate the extracted QAM symbol sequence to form a time-domain high-dimensional QAM signal; a transmitter connected to the sequence modulation module, used to convert the modulated time-domain high-dimensional QAM signal into a signal form suitable for transmission and transmit it; a channel connected to the transmitter, used to transmit the modulated time-domain high-dimensional QAM signal; a receiver connected to the channel, used to receive the signal transmitted through the channel and convert it into a baseband signal; and a sequence demodulation module connected to the receiver, used to perform sequence decision and bit recovery based on the minimum distance criterion by calculating the Euclidean distance between the baseband signal and each representative sequence in the high-dimensional constellation point set.

[0016] Preferably, the signal format suitable for transmission is a millimeter-wave signal format. In summary, compared with existing technologies, the high-dimensional QAM signal modulation system and method based on clustering algorithm sequence extraction provided by this invention achieves true polarization-time domain multidimensional modulation by extending the signal dimension in the time domain and optimizing the high-dimensional constellation using the K-means++ algorithm. Furthermore, it overcomes the dependence of traditional probabilistic shaping and geometric shaping techniques on existing digital signal processing architectures, requiring no modification to the traditional DSP core link; only sequence modulation and demodulation modules need to be added at the transmitter and receiver, respectively, greatly reducing the complexity and cost of system upgrades. Simultaneously, by maximizing the Euclidean distance between high-dimensional constellation points, this method significantly improves the signal's noise robustness. Moreover, this method has good compatibility with polarization multiplexing technology, and by flexibly configuring sequence length and source entropy, it can effectively balance system performance and decoding complexity, providing a highly promising high-performance solution for next-generation 5G / 6G wireless networks, optical communication systems, and other high-capacity communication applications. Attached Figure Description

[0017] Figure 1 This is a flowchart of the high-dimensional QAM signal modulation system based on sequence extraction using a clustering algorithm in this invention. Detailed Implementation

[0018] The present invention will be further described below with reference to the accompanying drawings and by providing a detailed description of a preferred embodiment.

[0019] It should be noted that the accompanying drawings are in a very simplified form and use non-precise proportions. They are only used to facilitate and clarify the purpose of illustrating the embodiments of the present invention, and are not intended to limit the implementation conditions of the present invention. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportional relationship, or adjustments to the size should still fall within the scope of the technical content disclosed in the present invention, provided that they do not affect the effects and objectives that the present invention can produce.

[0020] It should be noted that, in this invention, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only the expressly listed elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0021] like Figure 1As shown, the present invention provides a high-dimensional QAM signal modulation system based on clustering algorithm sequence extraction, including an interconnected sequence construction system, a sequence extraction system, and a multi-dimensional modulation and demodulation system.

[0022] The sequence construction system includes: The full sequence generation module 1 is used to generate the original high-dimensional constellation point set based on the set sequence length S and the basic modulation order M. The basic modulation order M is 4, 16, 64, or 256, corresponding to QPSK modulation, 16-QAM modulation, 64-QAM modulation, or 256-QAM modulation, respectively. The original high-dimensional constellation point set contains all possible QAM symbol sequences. Each QAM symbol can be represented as a point in a two-dimensional complex space; therefore, a QAM symbol sequence of length S can be considered as a point in a 2S-dimensional real space. Furthermore, for M-order modulation, each symbol has M possible values, so a sequence of length S has a total of... A number of possible combinations.

[0023] The pseudo-random binary sequence (PRBS) generation module 3 is used to generate pseudo-random binary sequences as index bit streams and output them to the multidimensional modulation and demodulation system to provide a basis for sequence mapping.

[0024] The sequence extraction system includes: The K-means-based sequence extraction module 2 is connected to the full sequence generation module 1. It is used to receive the original high-dimensional constellation point set and use the K-means++ clustering algorithm to divide the original high-dimensional constellation point set into multiple clusters of a preset number of n. In each cluster, the sequence closest to the center of the cluster is selected as the representative sequence to form an optimized high-dimensional constellation point set; where n is a positive integer.

[0025] The objective function of the K-means++ clustering algorithm is to maximize the average Euclidean distance between clusters. Specifically: , In the formula, the preset quantity The total number of cluster centers. and These are the indices of the cluster centers. For the first Cluster center, For the first The center of a cluster.

[0026] The preset quantity n is calculated based on the target source entropy SE and the sequence length S. The target source entropy SE is determined according to the actual situation, and the specific calculation formula is as follows: .

[0027] The multidimensional modulation and demodulation system includes: Bit-to-sequence index mapping module 4, connected to the PRBS generation module 3, receives the index bit stream and maps the index bit stream (each... (each bit is grouped together) and mapped to the corresponding sequence index in the optimized high-dimensional constellation point set.

[0028] The sequence symbol generation module 5 is connected to the K-means-based sequence extraction module 2 and the bit-to-sequence index mapping module 4, respectively, and is used to extract the corresponding QAM symbol sequence from the optimized high-dimensional constellation point set according to the received sequence index.

[0029] The signal transmission link is connected to the extracted sequence symbol generation module 5, and is used to modulate the extracted QAM symbol sequence into a time-domain high-dimensional QAM signal and transmit it through the channel, and to demodulate the time-domain high-dimensional QAM signal at the receiving end to recover the sequence information.

[0030] The signal transmission link includes: The sequence modulation module 6, connected to the extracted sequence symbol generation module 5, is used to modulate the extracted QAM symbol sequence to form a high-dimensional QAM signal in the time domain. The sequence modulation module 6 supports polarization multiplexing technology, which can assign different sequences or symbol sequences to different polarization states, thereby realizing polarization-time domain multidimensional modulation and further improving system capacity.

[0031] Transmitter 7, connected to the sequence modulation module 6, is used to convert the modulated time-domain high-dimensional QAM signal into a signal form suitable for transmission and transmit it; in a preferred embodiment of the present invention, transmitter 7 uses photon-assisted technology to upconvert the modulated time-domain high-dimensional QAM signal into a millimeter-wave signal and transmit it.

[0032] Channel 8 is a wireless transmission channel connected to the transmitter 7 and used to transmit modulated time-domain high-dimensional QAM signals; wherein, in a preferred embodiment of the present invention, the signal transmitted by channel 8 is a millimeter-wave signal, and the maximum reliable transmission distance of channel 8 is 4600m.

[0033] Receiver 9, connected to channel 8, is used to receive the modulated time-domain high-dimensional QAM signal transmitted through channel 8 and convert it into a baseband signal; in a preferred embodiment of the present invention, receiver 9 is used to receive millimeter-wave signals and perform down-conversion processing.

[0034] The sequence demodulation module 10, connected to the receiver 9, calculates the Euclidean distance between the baseband signal and each representative sequence in the high-dimensional constellation point set, and performs sequence decision and bit recovery based on the minimum distance criterion. That is, it selects the representative sequence with the smallest Euclidean distance as the decision result, and recovers the original index bit stream according to the index of the representative sequence in the high-dimensional constellation point set.

[0035] This invention also provides a modulation method for high-dimensional QAM signals based on sequence extraction using a clustering algorithm, implemented using the aforementioned modulation system, specifically comprising the following steps: S1. The full sequence generation module 1 generates an original high-dimensional constellation point set based on the preset sequence length S and the basic modulation order M; and the PRBS generation module 3 generates a pseudo-random binary sequence as an index bit stream; wherein, the original high-dimensional constellation point set includes A possible set of QAM symbol sequences; S2. K-means-based sequence extraction module 2 calculates a preset number n based on the target source entropy SE and sequence length S, and uses the K-means++ clustering algorithm to divide the original high-dimensional constellation point set into a preset number n clusters. In each cluster, the sequence closest to the cluster center is selected as the representative sequence to form an optimized high-dimensional constellation point set. S3, the bit-to-sequence index mapping module 4 maps the index bit stream generated in S1 to the corresponding sequence index in the optimized high-dimensional constellation point set in S2; the sequence symbol generation module 5 extracts the corresponding QAM symbol sequence from the optimized high-dimensional constellation point set according to the received sequence index; S4. The sequence modulation module 6 modulates the extracted QAM symbol sequence into a time-domain high-dimensional QAM signal and transmits it from channel 8 through transmitter 7. At the receiving end, receiver 9 converts the received signal into a baseband signal and calculates its Euclidean distance with all representative sequences in the optimized high-dimensional constellation point set through sequence demodulation module 10. Based on the minimum distance criterion, it identifies the representative sequence with the smallest distance and restores its index to the original index bit stream.

[0036] Furthermore, prior to step S2, a standardization process is included to standardize the original high-dimensional constellation point set. The specific standardization formula is as follows: In the formula, Let f be the original value of the e-th sample in the original sequence along the f-th dimension; e is the sample index; f is the dimension index. and are the mean and standard deviation of the f-th dimension, respectively; This is the standardized data.

[0037] Furthermore, in step S2, to ensure the dispersion of the K-means++ clustering algorithm when dividing into a preset number of clusters n, the following operations need to be performed: S21. Randomly select the first cluster center from the original set of high-dimensional constellation points. ; S22. Calculate the distance from each point x in the cluster to the cluster center. The shortest distance D(x); S23. Select the next cluster center according to the probability formula. ; S24. Repeat steps S22-S23 until the preset number of n cluster centers have been selected.

[0038] In step S3, the probability formula is: Where P(x) represents the probability that point x is selected as the next cluster center. This probability is proportional to the square of D(x), so that the farther away a point is from the selected center, the greater the probability of it being selected.

[0039] Specifically, in one embodiment, the sequence length is used. Target source entropy Taking a basic modulation order M=4 (i.e., QPSK) as an example; under this configuration, each QPSK symbol has 4 possible values ​​(0, 1, 2, 3), so the sequence of 5 symbols has a total of 4^5 = 1024 possible combinations; and at this time, according to the target source entropy SE and sequence length S It can be determined that the preset quantity n is 256.

[0040] In summary, this invention provides a high-dimensional QAM signal modulation system and method based on sequence extraction using a clustering algorithm. Compared to existing modulation and shaping techniques, this system and method achieve true polarization-time domain multidimensional modulation by extending the signal dimension in the time domain and optimizing the high-dimensional constellation using the K-means++ algorithm. Furthermore, it overcomes the dependence of traditional probabilistic shaping and geometric shaping techniques on existing digital signal processing architectures, requiring no modification to the traditional DSP core link. Only sequence modulation and demodulation modules need to be added at the transmitter and receiver, respectively, significantly reducing the complexity and cost of system upgrades. Simultaneously, by maximizing the Euclidean distance between high-dimensional constellation points, this method significantly improves the signal's noise robustness. Moreover, this method exhibits good compatibility with polarization multiplexing techniques, and by flexibly configuring sequence length and source entropy, it can effectively balance system performance and decoding complexity, providing a highly promising high-performance solution for next-generation 5G / 6G wireless networks, optical communication systems, and other high-capacity communication applications.

[0041] Although the present invention has been described in detail through the preferred embodiments above, it should be understood that the above description should not be considered as a limitation of the present invention. Various modifications and substitutions to the present invention will be apparent to those skilled in the art after reading the above description. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A high-dimensional QAM signal modulation method based on a sequence of clustering algorithm extraction, characterized in that, Includes the following steps: S1, generating a set of original high-dimensional constellation points according to a preset sequence length S and a basic modulation order M, and generating a pseudo-random binary sequence as an index bit stream; wherein the set of original high-dimensional constellation points comprises a set of QAM symbol sequences S2. Based on the target source entropy SE and the sequence length S, calculate the preset number n, and use the K-means++ clustering algorithm to divide the original high-dimensional constellation point set into the preset number n clusters. In each cluster, select the sequence closest to the cluster center as the representative sequence to form the optimized high-dimensional constellation point set. S3. Map the index bitstream generated in S1 to the corresponding sequence index in the optimized high-dimensional constellation point set in S2; and extract the corresponding QAM symbol sequence from the optimized high-dimensional constellation point set according to the received sequence index. S4. Modulate the extracted QAM symbol sequence into a time-domain high-dimensional QAM signal and transmit it; and at the receiving end, convert the received signal into a baseband signal, calculate its Euclidean distance with all representative sequences in the optimized high-dimensional constellation point set, identify the representative sequence with the smallest distance according to the minimum distance criterion, and restore its index to the original index bit stream.

2. The high-dimensional QAM signal modulation method based on a sequence of clustering algorithm extraction according to claim 1, characterized in that, The preset quantity n is calculated based on the target source entropy SE and the sequence length S, using the following formula: 。 3. The high-dimensional QAM signal modulation method based on a sequence of clustering algorithm extraction according to claim 2, characterized in that, The objective function of the K-means++ clustering algorithm is to maximize the average Euclidean distance between clusters Specifically, ， In the formula, the preset quantity The total number of cluster centers. and These are the indices of the cluster centers. For the first Cluster center, For the first The center of a cluster.

4. The high-dimensional QAM signal modulation method based on clustering algorithm sequence extraction as described in claim 3, characterized in that, The specific steps of the K-means++ clustering algorithm to select a preset number of n cluster centers include: S21, selecting a first cluster center randomly from the original high-dimensional constellation point set ; S22, compute the shortest distance D(x) of each point x in the cluster to the cluster center S22, compute the shortest distance D(x) of each point x in the cluster to the cluster center S23. Selecting the next cluster center according to the probability formula ; S24. Repeat steps S22-S23 until the preset number of n cluster centers have been selected.

5. The high-dimensional QAM signal modulation method based on the sequence extraction of clustering algorithm according to claim 4, characterized in that, In step S23, the probability formula is: In the formula, P(x) represents the probability that point x is selected as the next cluster center.

6. The method of claim 1, wherein the sequence of clustering algorithm based extraction of high dimensional QAM signal modulation is characterized by, Before step S2, the process also includes standardizing the original high-dimensional constellation point set. The standardization formula is as follows: In the formula, Let be the original value of the i-th sample in the j-th dimension of the original sequence; i is the sample index; j is the dimension index. and These are the mean and standard deviation of the j-th dimension, respectively; This is the standardized data.

7. The method of claim 1, wherein the sequence of clustering algorithm based extraction of high dimensional QAM signal modulation is characterized by, The basic modulation order M is 4, 16, 64 or 256, corresponding to QPSK modulation, 16-QAM modulation, 64-QAM modulation or 256-QAM modulation respectively.

8. A high-dimensional QAM signal modulation system based on sequence extraction using a clustering algorithm, used to implement the high-dimensional QAM signal modulation method based on sequence extraction using a clustering algorithm as described in any one of claims 1-7, characterized in that, This includes interconnected sequence construction systems, sequence extraction systems, and multidimensional modulation and demodulation systems; The sequence construction system includes: A full sequence generating module is configured to generate a set of original high-dimensional constellation points according to a set sequence length S and a basic modulation order M, and the set of original high-dimensional constellation points comprises A set of QAM symbol sequences A pseudo-random binary sequence generation module is used to generate pseudo-random binary sequences as index bit streams; The sequence extraction system includes: The K-means-based sequence extraction module is connected to the full sequence generation module. It receives the original high-dimensional constellation point set and uses the K-means++ clustering algorithm to divide the original high-dimensional constellation point set into multiple clusters with a preset number of n. In each cluster, the sequence closest to the center of the cluster is selected as the representative sequence to form an optimized high-dimensional constellation point set. The multidimensional modulation and demodulation system includes: The bit-to-sequence index mapping module, connected to the pseudo-random binary sequence generation module, is used to map the index bit stream to the corresponding sequence index in the optimized high-dimensional constellation point set. The sequence symbol generation module is connected to the K-means-based sequence extraction module and the bit-to-sequence index mapping module, respectively, and is used to extract the corresponding QAM symbol sequence from the optimized high-dimensional constellation point set according to the received sequence index. The signal transmission link is connected to the extracted sequence symbol generation module and is used to modulate the extracted QAM symbol sequence into a time-domain high-dimensional QAM signal and transmit it through the channel, and to demodulate the time-domain high-dimensional QAM signal at the receiving end to recover the sequence information.

9. The high-dimensional QAM signal modulation system based on sequence extraction of clustering algorithm according to claim 8, characterized in that, The signal transmission link includes: The sequence modulation module, connected to the extracted sequence symbol generation module, is used to modulate the extracted QAM symbol sequence to form a high-dimensional QAM signal in the time domain; A transmitter, connected to the sequence modulation module, is used to convert the modulated time-domain high-dimensional QAM signal into a signal form suitable for transmission and transmit it. The channel, connected to the transmitter, is used to transmit modulated time-domain high-dimensional QAM signals; A receiver, connected to the channel, is used to receive signals transmitted through the channel and convert them into baseband signals; The sequence demodulation module, connected to the receiver, is used to perform sequence decision and bit recovery based on the minimum distance criterion by calculating the Euclidean distance between the baseband signal and each representative sequence in the high-dimensional constellation point set.

10. The high-dimensional QAM signal modulation system based on sequence extraction of clustering algorithm as claimed in claim 8, wherein, The signal format suitable for transmission is millimeter wave signal format.

Images