An adaptive blind demodulation method, system, terminal and medium based on a satellite communication system

Abstract

This invention discloses an adaptive blind demodulation method, system, terminal, and medium based on a satellite communication system. The method includes: performing parameter estimation, frequency offset compensation, and timing synchronization processing on the original baseband sampled data sequence to obtain a single symbol rate data symbol sequence; performing hierarchical phase-locked loop carrier coarse synchronization iterative processing on the single symbol rate data symbol sequence to obtain a target convergent constellation diagram symbol sequence; extracting high-order cumulant features based on the target convergent constellation diagram symbol sequence and performing modulation scheme identification and confidence judgment, iterating until a preset judgment condition is met to obtain a target modulation identification result; and performing adaptive carrier fine phase-locking and blind equalization joint processing on the target modulation scheme to obtain a demodulated symbol sequence. This invention solves the coupling problem between carrier synchronization and modulation identification by combining hierarchical carrier synchronization, high-order cumulant feature identification, and adaptive fine phase-locking blind equalization, thereby improving demodulation reliability.

Description

Technical Field

[0001] This invention relates to the field of digital communication technology, and in particular to an adaptive blind demodulation method, system, terminal, and medium based on a satellite communication system. Background Technology

[0002] With high-Earth orbit (HEO) and low-Earth orbit (LEO) satellite cooperative communication becoming the mainstream trend for full-domain coverage and the increasing urgency of non-cooperative reception, overcoming the limitations of blind demodulation with unknown signal parameters to achieve efficient information recovery in complex channels has become a key means to enhance satellite communication reconnaissance and countermeasure capabilities. Taking the complex scenario of LEO satellite high-speed motion causing Doppler frequency offset, HEO satellite long time delay attenuation, and the coexistence of multiple modulation schemes (such as APSK and QAM) as an example, resolving the contradiction between the lack of prior signal parameters and channel dynamic distortion under non-cooperative conditions is crucial to ensuring reliable interception and demodulation of communication links. However, current blind receiver solutions generally face common technical challenges when dealing with the significant differences between HEO and LEO satellite channels and the diversity of modern communication schemes, such as "cyclic dependency" deadlock in carrier synchronization and modulation identification, poor adaptability of multi-ring constellation signals, and deterioration of synchronization performance under nonlinear channels. The limitations of existing technologies are thus revealed. Specifically, at present, achieving blind demodulation of signals often relies on carrier synchronization based on phase-locked loops or modulation identification technology based on constellation diagram characteristics. While the former can achieve coherent demodulation, its phase detector design requires prior knowledge of the modulation method, resulting in its inability to start when faced with unknown modulation signals. While the latter can identify signal types, it is highly dependent on a clear and stable constellation diagram after carrier synchronization, creating a "chicken or egg" coupling dilemma. Furthermore, for APSK multi-loop modulation signals widely used in satellite communication, traditional synchronization algorithms based on single-loop assumptions (such as PSK or QAM) are difficult to adapt to their amplitude layering characteristics. In addition, the signal distortion introduced by the nonlinearity of the power amplifier makes it difficult for existing blind receiver systems to meet the practical application requirements in terms of acquisition accuracy, convergence speed, and versatility.

[0003] Therefore, existing technologies still need to be improved and enhanced. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an adaptive blind demodulation method, system, terminal and medium based on satellite communication system, which addresses the above-mentioned defects of the prior art. It aims to solve the cyclic coupling problem caused by the interdependence of carrier synchronization and modulation identification in traditional blind demodulation technology, as well as the technical bottleneck that it is difficult to automatically acquire, synchronize and demodulate unknown modulation signals in the absence of prior modulation information.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: In a first aspect, the present invention provides an adaptive blind demodulation method based on a satellite communication system, wherein the method includes: The original baseband sampling data sequence is acquired, and based on the original baseband sampling data sequence, parameter estimation, frequency offset compensation and timing synchronization processing are performed to obtain a single symbol rate data symbol sequence. Based on the single symbol rate data symbol sequence, perform hierarchical phase-locked loop carrier coarse synchronization iterative processing to obtain the target convergent constellation diagram symbol sequence; Based on the target convergent constellation diagram symbol sequence, high-order cumulant features are extracted and modulation mode identification and confidence judgment are performed. The process is iterated until the preset judgment conditions are met to obtain the target modulation identification result. Based on the target modulation identification results, a carrier phase-locked loop and blind equalization process with adaptive modulation mode are performed to obtain the demodulated symbol sequence.

[0006] In one implementation, the step of performing parameter estimation, frequency offset compensation, and timing synchronization processing based on the original baseband sampled data sequence to obtain a single symbol rate data symbol sequence includes: Spectral analysis is performed on the original baseband sampled data sequence to obtain the signal bandwidth and signal center frequency, and coarse frequency offset is obtained based on the signal center frequency; Based on the coarse frequency offset, the original baseband sampled data sequence is compensated for the frequency offset by digital spectrum shifting to obtain the coarse compensated signal sequence; Based on the signal bandwidth, a symbol rate estimate is calculated, and the coarsely compensated signal sequence is downsampled according to the symbol rate estimate to obtain an integer multiple symbol rate downsampled signal sequence and then matched filtering is performed. Timing error detection and interpolation adjustment are performed on the filtered sampled signal sequence at integer multiple symbol rates to obtain a single symbol rate data symbol sequence.

[0007] In one implementation, the step of performing hierarchical phase-locked loop carrier coarse synchronization iterative processing based on the single symbol rate data symbol sequence to obtain the target convergent constellation diagram symbol sequence includes: The first stage of the hierarchical phase-locked loop is initialized, the integral term of the loop filter and the cumulative phase of the numerically controlled oscillator are cleared, and the coefficients of the second-order loop filter specific to the coarse phase-locking stage are configured. The single-symbol-rate data symbol sequence is continuously iterated. For each current single-symbol-rate data symbol in the sequence, phase rotation compensation is performed on the current single-symbol-rate data symbol using the current accumulated phase of the numerically controlled oscillator. A phase detection operation is performed on the compensated symbol to obtain the phase detection error. The phase detection error is input into a second-order loop filter to update the integral term of the loop filter and obtain the phase adjustment amount. The accumulated phase of the numerically controlled oscillator is updated using the phase adjustment amount. The sliding window is used to calculate the moving average variance of the phase detection error and compare it with a preset steady-state judgment threshold. When the preset convergence condition is met, the symbols output at this time and subsequently after phase rotation compensation are taken as the target convergent constellation diagram symbol sequence.

[0008] In one implementation, the step of extracting high-order cumulant features based on the target convergent constellation diagram symbol sequence and performing modulation scheme identification and confidence judgment, iterating until a preset judgment condition is met, to obtain the target modulation identification result, includes: In each iteration, based on the target convergent constellation diagram symbol sequence, a preset number of target convergent constellation diagram symbol sequences are collected as the current identification samples, and the cumulative quantity features of the current identification samples are calculated; based on the cumulative quantity features, signal class differentiation and modulation mode subdivision identification are performed to obtain the current modulation mode, and a confidence decision is performed on the current modulation mode until the preset confidence decision conditions are met, and the target modulation identification result is obtained.

[0009] In one implementation, the step of performing signal category differentiation and modulation mode subdivision identification based on the accumulated quantity characteristics to obtain the current modulation mode includes: Based on the accumulated feature, a normalized fourth-order accumulated value is extracted. Based on the normalized fourth-order accumulated value, the signal corresponding to the identification sample is classified into major categories to obtain the major category classification result. If the major category classification result is PSK, calculate the fourth phase moment of the identified sample, and distinguish the QPSK and 8PSK types of the signal corresponding to the identified sample based on the fourth phase moment to obtain the subdivision modulation result; If the major category classification result is QAM, based on the normalized fourth-order cumulant and sixth-order cumulant in the cumulant calculation result, the signal corresponding to the identification sample is classified into 16QAM and 32QAM types to obtain the subdivision modulation result. If the major category classification result is APSK, the signal corresponding to the identification sample is classified into 16APSK and 32APSK types based on the sixth-order cumulant in the cumulant calculation result to obtain the subdivision modulation result; The current modulation scheme is obtained based on the subdivision modulation results.

[0010] In one implementation, the step of performing joint processing of carrier fine phase-locked loop and blind equalization with modulation mode adaptation based on the target modulation identification result to obtain the demodulated symbol sequence includes: Based on the target modulation recognition result, the system adaptively switches to a fine phase detector that perfectly matches the target modulation recognition result, and adjusts the coefficients of the second-order loop filter to values ​​specific to the fine phase-locked stage. During the switching process, the cumulative phase of the numerically controlled oscillator and the integral term of the loop filter remain unchanged. Based on the switched fine phase detector and the adjusted second-order loop filter, a carrier fine phase-locked iteration operation is performed on the single symbol rate data symbol sequence to obtain a stable constellation diagram symbol sequence. Based on the target modulation recognition result, a unique constant modulus value corresponding to the target modulation recognition result is matched and selected. The unique constant modulus algorithm corresponding to the unique constant modulus value is used to perform blind equalization processing on the stable constellation diagram symbol sequence to obtain the equalized symbol sequence. Based on the decision-guided phase-locked loop, phase compensation processing is performed on the equalized symbol sequence. The phase compensation error is smoothed and the phase compensation value is updated through a first-order loop filter to correct phase ambiguity and obtain a demodulated symbol sequence with accurate phase.

[0011] In one implementation, after obtaining the demodulated symbol sequence, the method further includes: The signal-to-noise ratio (SNR) of the demodulated symbol sequence is estimated using the error vector magnitude method or the M2M4 blind estimation algorithm to obtain the SNR estimate.

[0012] Secondly, embodiments of the present invention also provide an adaptive blind demodulation system based on a satellite communication system, wherein the system includes: The single symbol rate data symbol sequence acquisition module is used to perform parameter estimation, frequency offset compensation and timing synchronization processing based on the original baseband sampled data sequence to obtain the single symbol rate data symbol sequence. The target convergence constellation diagram symbol sequence acquisition module is used to perform hierarchical phase-locked loop carrier coarse synchronization iterative processing based on the single symbol rate data symbol sequence to obtain the target convergence constellation diagram symbol sequence; The target modulation recognition result acquisition module is used to extract high-order cumulative features based on the target convergent constellation diagram symbol sequence and perform modulation mode recognition and confidence judgment, iterating until the preset judgment conditions are met to obtain the target modulation recognition result. The demodulation symbol sequence acquisition module is used to perform joint processing of carrier fine phase-locking and blind equalization with modulation mode adaptation based on the target modulation identification result to obtain the demodulation symbol sequence.

[0013] Thirdly, embodiments of the present invention also provide a terminal, wherein the terminal includes a memory, a processor, and an adaptive blind demodulation program based on a satellite communication system stored in the memory and executable on the processor. When the processor executes the adaptive blind demodulation program based on a satellite communication system, it implements the steps of the adaptive blind demodulation method based on a satellite communication system as described in any of the above schemes.

[0014] Fourthly, embodiments of the present invention also provide a computer-readable storage medium, wherein the computer-readable storage medium stores an adaptive blind demodulation program based on a satellite communication system, and when the adaptive blind demodulation program based on a satellite communication system is executed by a processor, it implements the steps of the adaptive blind demodulation method based on a satellite communication system as described in any of the above schemes.

[0015] Beneficial Effects: This invention provides an adaptive blind demodulation method based on a satellite communication system. Compared with existing technologies, this invention first performs parameter estimation, frequency offset compensation, and timing synchronization processing on the original baseband sampled data sequence to obtain a single symbol rate data symbol sequence. This effectively eliminates the impact of sampling deviation, carrier frequency offset, and timing error on subsequent demodulation, improving signal synchronization accuracy and data reliability. Next, based on the single symbol rate data symbol sequence, iterative processing of hierarchical phase-locked loop carrier coarse synchronization is performed to obtain the target convergent constellation diagram symbol sequence. This achieves rapid coarse adjustment of the carrier phase and constellation diagram convergence, significantly reducing demodulation distortion caused by phase ambiguity and noise interference. Then, based on the target convergent constellation diagram symbol sequence, high-order cumulant features are extracted, and modulation mode identification and confidence judgment are performed. This process iterates until preset judgment conditions are met to obtain the target modulation identification result. Even in complex channel environments, the accuracy and robustness of modulation identification are guaranteed, avoiding the deterioration of demodulation performance due to misidentification. Finally, based on the target modulation identification results, a combined process of carrier fine phase-locking and blind equalization with modulation mode adaptation is performed to obtain the demodulated symbol sequence. This achieves fine carrier phase correction and channel distortion equalization compensation, further improving the signal-to-noise ratio and decision correctness of the demodulated symbols. This invention, through its overall architecture of step-by-step synchronization, hierarchical carrier locking, adaptive modulation identification, and fine demodulation combined processing, achieves efficient blind demodulation without prior information in complex time-varying channels, low signal-to-noise ratio, and multi-modulation system coexistence scenarios in satellite communication. This effectively improves demodulation accuracy, convergence speed, and system adaptability, enhancing the stability and transmission reliability of satellite communication links. Attached Figure Description

[0016] Figure 1 A flowchart illustrating a specific implementation of the adaptive blind demodulation method based on a satellite communication system provided in this embodiment of the invention.

[0017] Figure 2This is a flowchart illustrating the processing procedure of a preferred embodiment of the adaptive blind demodulation method based on a satellite communication system provided in this invention.

[0018] Figure 3 The flowchart illustrates the adaptive carrier synchronization process in a preferred embodiment of the adaptive blind demodulation method based on a satellite communication system provided in this invention.

[0019] Figure 4 This is a flowchart illustrating the implementation steps of a preferred embodiment of the adaptive blind demodulation method based on a satellite communication system provided in this invention.

[0020] Figure 5 This is a block diagram of the adaptive blind demodulation system based on a satellite communication system provided in an embodiment of the present invention.

[0021] Figure 6 This is a block diagram illustrating the internal structure of a terminal provided in an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and effects of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0023] With high-Earth orbit (HEO) and low-Earth orbit (LEO) satellite cooperative communication becoming the mainstream trend for full-domain coverage and the increasing urgency of non-cooperative reception, overcoming the limitations of blind demodulation with unknown signal parameters to achieve efficient information recovery in complex channels has become a key means to enhance satellite communication reconnaissance and countermeasure capabilities. Taking the complex scenario of LEO satellite high-speed motion causing Doppler frequency offset, HEO satellite long time delay attenuation, and the coexistence of multiple modulation schemes (such as APSK and QAM) as an example, resolving the contradiction between the lack of prior signal parameters and channel dynamic distortion under non-cooperative conditions is crucial to ensuring reliable interception and demodulation of communication links. However, current blind receiver solutions generally face common technical challenges when dealing with the significant differences between HEO and LEO satellite channels and the diversity of modern communication schemes, such as "cyclic dependency" deadlock in carrier synchronization and modulation identification, poor adaptability of multi-ring constellation signals, and deterioration of synchronization performance under nonlinear channels. The limitations of existing technologies are thus revealed. Specifically, at present, achieving blind demodulation of signals often relies on carrier synchronization based on phase-locked loops or modulation identification technology based on constellation diagram characteristics. While the former can achieve coherent demodulation, its phase detector design requires prior knowledge of the modulation method, resulting in its inability to start when faced with unknown modulation signals. While the latter can identify signal types, it is highly dependent on a clear and stable constellation diagram after carrier synchronization, creating a "chicken or egg" coupling dilemma. Furthermore, for APSK multi-loop modulation signals widely used in satellite communication, traditional synchronization algorithms based on single-loop assumptions (such as PSK or QAM) are difficult to adapt to their amplitude layering characteristics. In addition, the signal distortion introduced by the nonlinearity of the power amplifier makes it difficult for existing blind receiver systems to meet the practical application requirements in terms of acquisition accuracy, convergence speed, and versatility.

[0024] To address the aforementioned issues, this embodiment provides an adaptive blind demodulation method based on a satellite communication system. Specifically, this embodiment first performs parameter estimation, frequency offset compensation, and timing synchronization processing on the original baseband sampled data sequence to obtain a single-symbol-rate data symbol sequence. This effectively eliminates the impact of sampling deviation, carrier frequency offset, and timing errors on subsequent demodulation, improving signal synchronization accuracy and data reliability. Next, based on the single-symbol-rate data symbol sequence, iterative processing of hierarchical phase-locked loop carrier coarse synchronization is performed to obtain the target convergent constellation diagram symbol sequence. This achieves rapid coarse adjustment of the carrier phase and constellation diagram convergence, significantly reducing demodulation distortion caused by phase ambiguity and noise interference. Then, based on the target convergent constellation diagram symbol sequence, high-order cumulant features are extracted, and modulation scheme identification and confidence level determination are performed. This process iterates until preset decision conditions are met to obtain the target modulation identification result. Even in complex channel environments, this method ensures the accuracy and robustness of modulation identification, avoiding the deterioration of demodulation performance due to misidentification. Finally, based on the target modulation identification results, a combined process of carrier fine phase-locking and blind equalization with modulation mode adaptation is performed to obtain the demodulated symbol sequence. This achieves fine carrier phase correction and channel distortion equalization compensation, further improving the signal-to-noise ratio and decision correctness of the demodulated symbols. This invention, through its overall architecture of step-by-step synchronization, hierarchical carrier locking, adaptive modulation identification, and fine demodulation combined processing, achieves efficient blind demodulation without prior information in complex time-varying channels, low signal-to-noise ratio, and multi-modulation system coexistence scenarios in satellite communication. This effectively improves demodulation accuracy, convergence speed, and system adaptability, enhancing the stability and transmission reliability of satellite communication links.

[0025] The adaptive blind demodulation method based on a satellite communication system provided in this embodiment can be applied to smart terminals, such as... Figure 1 As shown, the specific steps include the following: Step S100: Based on the original baseband sampled data sequence, perform parameter estimation, frequency offset compensation and timing synchronization processing to obtain a single symbol rate data symbol sequence.

[0026] In this embodiment, this step serves as a fundamental pre-processing step in the entire data processing flow. Its core is to sequentially complete three key processing operations—parameter estimation, frequency offset compensation, and timing synchronization—based on the acquired original baseband sampled data sequence, ultimately outputting a single-symbol-rate data symbol sequence. The core purpose of this step is to resolve potential issues such as parameter deviations, frequency offsets, and timing misalignments in the original baseband sampled data, providing accurate and standardized input data for subsequent data processing stages. Specifically, parameter estimation accurately captures the core characteristic parameters of the original data, providing a reliable basis for subsequent compensation and synchronization operations; frequency offset compensation effectively eliminates frequency offsets caused by equipment differences and channel interference during signal transmission, avoiding signal distortion; and timing synchronization calibrates the time reference for data sampling, ensuring precise alignment between the sampling point and the symbol period. The synergistic effect of these three steps significantly improves the integrity and accuracy of the data symbols, reduces errors in subsequent processing stages, and simplifies the complexity of subsequent data parsing and processing. This provides strong support for the stable operation and processing efficiency of the entire system, ensuring that subsequent steps can be carried out in an orderly manner based on the standardized single-symbol-rate data symbol sequence, guaranteeing the stability and reliability of the overall processing effect.

[0027] Specifically, step S100 includes the following steps: Step S101: Perform spectral analysis on the original baseband sampling data sequence to obtain the signal bandwidth and signal center frequency, and obtain the coarse frequency offset based on the signal center frequency; Step S102: Based on the coarse frequency offset, frequency offset compensation is performed on the original baseband sampled data sequence by digital spectrum shifting to obtain the coarse compensation signal sequence; Step S103: Calculate the symbol rate estimate based on the signal bandwidth, and perform downsampling on the coarse-compensated signal sequence according to the symbol rate estimate to obtain an integer multiple symbol rate downsampled signal sequence and perform matched filtering. Step S104: Perform timing error detection and interpolation adjustment on the filtered integer multiple symbol rate sampled signal sequence to obtain a single symbol rate data symbol sequence.

[0028] In one implementation, such as Figure 2 and 3 As shown, the baseband data (i.e., the raw baseband sampled data sequence) is first received. Spectral analysis is then performed on the raw baseband sampled data sequence output from the receiver front-end to obtain key signal parameters. The baseband model of the received signal can be represented as: ,in The baseband signal output from the receiver front end (raw baseband sampled data sequence), where n is the sampling index point. , The original data sent by the sending end; The carrier frequency offset contained in the received signal, in Hz; The initial phase of the received signal, in rad. For the ADC sampling period, satisfying , This refers to the ADC sampling rate; Additive white Gaussian noise with a mean of 0 and a variance of 0. To accurately extract the bandwidth and center frequency from the signal, the Fast Fourier Transform (FFT) algorithm is used to calculate the signal's power spectrum. Specifically, for The formula for performing an FFT operation is as follows: ,in, For FFT points, For the first Power spectrum values ​​at each frequency point To improve estimation performance, especially in low signal-to-noise ratio environments, energy superposition is performed on the power spectra of multi-packet data to smooth noise fluctuations and make the signal spectral characteristics more prominent. After obtaining the smoothed power spectrum, a 3dB bandwidth is extracted, and by searching the power spectrum, two frequency points where the power spectrum value drops to half of its maximum value are found. and ,satisfy Based on these two frequency indexes, bandwidth estimation can be performed to calculate the signal bandwidth. The signal bandwidth It is determined by the difference in physical frequencies corresponding to the two frequency points, that is: Simultaneously, the center frequency of the received signal is calculated based on the spectral distribution. Since this embodiment uses baseband signal reception, the coarse frequency offset of the received signal can be obtained by calculating the signal bandwidth. Coarse frequency offset compensation is then performed on the signal. Based on the coarse frequency offset value obtained in the above steps, the original baseband sampled data sequence is compensated for frequency offset using digital spectrum shifting technology. By shifting the spectrum of the received signal, the carrier frequency offset caused by the Doppler effect or the difference between the transmitting and receiving crystal oscillators is eliminated, thereby obtaining the coarsely compensated signal sequence. This operation shifts the signal spectrum to the center of the baseband, reducing the phase rotation effect in subsequent processing and ensuring the stability of signal demodulation.

[0029] In one implementation, based on the calculated signal bandwidth Furthermore, the symbol rate estimate is derived and calculated. Based on the mapping relationship between Nyquist bandwidth and symbol rate, and combined with a preset roll-off factor... , ( The range of values ​​is In this embodiment ), signal bandwidth With symbol rate Satisfying the relation Therefore, the symbol rate estimate is... Based on this symbol rate estimate The coarse compensation signal sequence is downsampled to obtain To accommodate subsequent timing synchronization algorithms, the sampling rate after downsampling is set to... This means that each symbol period contains two sampling points. Downsampling at this integer multiple of the symbol rate not only reduces the data rate and alleviates the processing burden, but also prepares the data format for the Gardner timing error detection algorithm. The downsampled signal is then subjected to matched filtering to maximize the signal-to-noise ratio and suppress inter-symbol interference. Specifically, root-raised cosine matched filtering is applied to the downsampled signal, with the roll-off coefficient of the matched filter consistent with that of the symbol rate estimation stage, suppressing out-of-band noise and optimizing the signal waveform.

[0030] In one implementation, timing error detection and interpolation adjustment are performed on the filtered signal sequence downsampled at integer multiples of the symbol rate. This embodiment uses the Gardner algorithm for timing synchronization, which is suitable for signal processing with two sampling points per symbol. The formula for calculating the timing error is: In the formula, Indicates the first The interpolated sample value at each symbol interpolation time, between two consecutive symbols. The difference can represent the direction of the timing error; Indicates the first One and The interpolated sample value at the midpoint of each symbol, which represents the magnitude of the error. , These represent the real and imaginary parts of the data, respectively. Based on the calculated timing error, the interpolation offset of the interpolation filter is adjusted. The signal is resampled. The output of the interpolation filter... This is the final recovered single-symbol-rate data symbol sequence, and its calculation process is expressed as follows: ,in, For single-symbol-rate data symbol sequences, These are the interpolation filter coefficients. Let the filter order be . This is the interpolation offset (adjusted by timing error). For symbol period ( , (This refers to the number of sampling points per symbol, which is 2 in this embodiment). After the timing synchronization process described above, the timing deviation τ is eliminated, and the final single-symbol-rate data symbol sequence is obtained. It can be represented as: , For single-symbol-rate data symbols after timed synchronization, To send symbols, For residual frequency offset, For the initial phase, For symbol period, It is additive white Gaussian noise. At this time, The method eliminates timing jitter between symbols, providing a high-quality input signal for subsequent carrier synchronization and demodulation. Overall, through cascaded processing from spectrum analysis to timing synchronization, this method achieves blind rate matching and high-precision synchronization under unknown signal parameters, significantly improving the receiving performance of high and low orbit satellite communication systems.

[0031] Step S200: Based on the single symbol rate data symbol sequence, perform hierarchical phase-locked loop carrier coarse synchronization iterative processing to obtain the target convergent constellation diagram symbol sequence.

[0032] In this embodiment, this step, as the core step of carrier phase recovery, takes the single-symbol-rate data symbol sequence output from the previous steps and performs iterative coarse synchronization processing by constructing a hierarchical phase-locked loop (PLL) architecture. This process gradually approximates and corrects residual phase deviations in the signal, ultimately outputting the target converged constellation symbol sequence. The core function of this step is to accurately recognize and calibrate the phase of the pre-processed signal, effectively overcoming disturbances to the signal phase caused by factors such as channel fading, noise interference, and residual frequency offset, and avoiding symbol decision errors due to phase ambiguity. The hierarchical PLL achieves layered optimization of the coarse synchronization process, gradually narrowing the phase error range through multiple iterations. Compared to traditional single-loop PLL mechanisms, it has a faster convergence speed and stronger noise robustness. The iterative processing design can dynamically adapt to phase change characteristics under complex channel environments, ensuring the stability and adaptability of the synchronization process. After this step, the constellation clustering of the target convergence constellation diagram symbol sequence is significantly improved, and the phase error is greatly reduced. This provides a high-quality, phase-regular signal foundation for subsequent high-frequency and efficient symbol demodulation and decoding steps, effectively reducing the overall system bit error rate, ensuring the reliability and efficiency of data transmission, and driving the entire processing flow toward accurate and efficient target convergence.

[0033] Specifically, step S200 includes the following steps: Step S201: Initialize the first stage of the hierarchical phase-locked loop, clear the integral term of the loop filter and the cumulative phase of the numerically controlled oscillator, and configure the second-order loop filter coefficients specific to the coarse phase-locked stage. Step S202: Perform continuous iterative processing on the single symbol rate data symbol sequence. For each current single symbol rate data symbol in the sequence, perform phase rotation compensation on the current single symbol rate data symbol using the current accumulated phase of the numerically controlled oscillator. Perform phase detection operation on the compensated symbol to obtain the phase detection error. Input the phase detection error into the second-order loop filter, update the integral term of the loop filter and obtain the phase adjustment amount. Update the accumulated phase of the numerically controlled oscillator using the phase adjustment amount. Use a sliding window to calculate the moving average variance of the phase detection error and compare it with the preset steady-state judgment threshold. When the preset convergence condition is met, use the symbols output at this time and subsequently after phase rotation compensation as the target convergent constellation diagram symbol sequence.

[0034] In one implementation, such as Figure 2-4 As shown, firstly, the first stage (i.e., the coarse phase-locked loop) of the hierarchical phase-locked loop is initialized by clearing the integral term of the loop filter. and the accumulated phase of the numerically controlled oscillator (NCO) The coefficients of the second-order loop filter in the coarse phase-locked loop stage are configured according to the system design requirements. and ,in For integral gain, The proportional gain is used to establish the basic dynamic response characteristics of the loop. Next, the input single-symbol-rate data symbol sequence is continuously iterated. For each current single-symbol-rate data symbol in the sequence, the accumulated phase of the NCO at the current time is used. Phase rotation compensation is applied to eliminate the current frequency and phase deviations, thus obtaining a constellation symbol that has undergone preliminary correction. Its mathematical expression is: Subsequently, the initial constellation symbol was... Perform QPSK hard decision to obtain the closest ideal constellation point. And calculate the conjugate of the ideal constellation point. With preliminary constellation symbols The product of the two terms is used to obtain the real-time coarse phase-locked loop (PLL) error by extracting the imaginary part of the product result. The process follows the formula: This accurately reflects the phase deviation of the current symbol. Then, the calculated phase detection error is... The input is fed into a second-order loop filter, and then processed according to the recursive formula of the integral term. Update the integral term of the loop filter The total phase adjustment is calculated by combining the proportional term, and then the NCO cumulative phase recursive formula is applied. Update the cumulative phase of NCO The updated phase will be used for phase rotation of the next single-symbol-rate data symbol in the single-symbol-rate data symbol sequence, forming a closed-loop feedback control. Simultaneously, to accurately monitor the loop's locking state, a sliding window statistical method is used to calculate the moving average variance of the phase detection error signal based on the real-time generated phase detection error. The specific calculation involves using a length of ( The sliding window, according to the formula Statistical analysis shows that this variance value can effectively characterize the degree of fluctuation in loop error. Finally, the calculated moving average variance... Compared with the preset steady-state determination threshold ( Compare the results; if the moving average variance is greater than or equal to the preset steady-state judgment threshold (…), then… If the loop has not yet converged, then the next single-symbol-rate data symbol is read, and the above-mentioned phase rotation, decision phase detection, filter update, and variance statistical iteration process is repeated; otherwise, if the moving average variance is less than the preset steady-state judgment threshold ( If the carrier coarse phase-locked loop (CLL) is considered to have converged, the loop has entered a steady state, and the symbols output at this point and subsequently after phase rotation compensation are used as the target converged constellation diagram symbol sequence. This ensures that frequency and phase offset corrections can be completed quickly and accurately during the CLL phase, providing a high-quality signal foundation for subsequent processing. Once the loop has entered a steady state, a trigger is immediately output to initiate the subsequent modulation scheme identification process. Modulation identification is only initiated after the carrier coarse phase-locked loop is completed and the constellation diagram has stably converged. This fundamentally avoids the problem of misjudgment in modulation identification when the constellation diagram diverges, significantly improving the reliability of modulation identification. The statistical method of moving average variance effectively filters out error fluctuations caused by noise, avoiding steady-state misjudgments caused by single error jumps and ensuring the accuracy of the trigger timing.

[0035] Step S300: Based on the target convergent constellation diagram symbol sequence, extract high-order cumulant features and perform modulation mode identification and confidence judgment, iterating until the preset judgment conditions are met to obtain the target modulation identification result.

[0036] In this embodiment, as Figure 2-4 As shown, firstly, when the receiver loop enters a steady state, the modulation identification unit is triggered to start collecting a preset number of data points (the preset number of data points is not specified in the original text). The compensated target convergence constellation diagram symbol sequence As the current identification samples, and based on these samples, higher-order cumulative features are calculated. The specific calculation process is based on statistical principles and utilizes the expectation operator. Solve for the second-, fourth-, and sixth-order cumulants respectively, their mathematical definitions are as follows: , , , , The cumulative calculation results obtained through the above formula can reflect the geometric distribution characteristics of the signal constellation diagram; then, the normalized fourth-order cumulative is extracted from the calculation results as the primary classification criterion to distinguish the signals corresponding to the identification samples into major categories, thus obtaining the major category distinction results. Specifically: using... The signal is categorized into major classes: values ​​close to 2.0 are BPSK, values ​​close to 1.0 are PSK (QPSK / 8PSK), values ​​between 0.6 and 0.8 are QAM, and values ​​between 0.3 and 0.5 are APSK. After obtaining the major class classification results, the modulation scheme is further subdivided to identify the current modulation scheme. Table 1 shows the theoretical values ​​of the integrals for each modulation scheme and the notes used for modulation scheme subdivision identification. The modulation schemes can be classified into different categories based on the information in Table 1. Specifically: If the major category classification result is PSK, calculate the fourth-order phase moment of the identified sample, and based on the fourth-order phase moment, distinguish the signal corresponding to the identified sample into QPSK and 8PSK types to obtain the subdivision modulation result, as shown in Table 1; if the major category classification result is QAM, based on the normalized fourth-order and sixth-order cumulants in the cumulant calculation result, distinguish the signal corresponding to the identified sample into 16QAM and 32QAM types to obtain the subdivision modulation result, as shown in Table 1; if the major category classification result is APSK, based on the normalized sixth-order cumulants in the cumulant calculation result, distinguish the signal corresponding to the identified sample into 16APSK and 32APSK types to obtain the subdivision modulation result, as shown in Table 1; based on the subdivision modulation result, obtain the current modulation method. Using the current modulation scheme as input, a confidence decision mechanism is introduced to determine whether the recognition results of the recognition sample are consistent across multiple consecutive attempts (typically two times), and whether the deviation between the recognition feature quantity and the theoretical value of the corresponding modulation scheme is less than 5%, thus obtaining a confidence decision result. If the confidence decision result meets the preset confidence decision conditions, the recognition result of the recognition sample is determined to be valid, and the target modulation recognition result is obtained. The target modulation identification results cover BPSK, QPSK, 8PSK, 16QAM, 32QAM, 16APSK, and 32APSK. If the confidence judgment result does not meet the preset confidence judgment conditions, the target convergent constellation diagram symbol sequence is re-acquired as an identification sample. The process returns to the previous steps of acquiring a preset number of target convergent constellation diagram symbol sequences as the current identification sample and calculating their second-order, fourth-order, and sixth-order cumulant features, repeating the entire modulation identification process. This embodiment completes the identification based on steady-state convergent constellation diagram symbols, comprehensively covering the seven mainstream modulation methods of terrestrial and satellite communications. In particular, it achieves accurate identification of the 16APSK and 32APSK multi-loop modulation signals specific to satellite communications through high-order cumulant features, solving the problem of poor adaptability of traditional identification schemes to APSK signals.

[0037] Table 1

[0038] Step S400: Based on the target modulation identification result, perform joint processing of carrier fine phase-locking and blind equalization with modulation mode adaptation to obtain the demodulation symbol sequence.

[0039] In this embodiment, this step is the core and crucial step in the entire demodulation process. Following the target modulation identification result obtained above, it performs adaptive carrier phase-locked loop (PLL) and blind equalization combined processing, ultimately outputting a demodulated symbol sequence. The core logic relies on the target modulation identification result to achieve adaptive coordination between the PLL and blind equalization processes. This allows for matching the current modulation scheme without manual intervention, ensuring a high degree of compatibility between the processing and the modulation type. Specifically, PLL is used to accurately calibrate the carrier's frequency and phase deviations, resolving demodulation distortion caused by inaccurate carrier synchronization. Blind equalization compensates for inter-symbol interference generated during channel transmission, improving the accuracy of symbol decision. The combined processing of these two steps complements each other, effectively avoiding the limitations of single processing stages. This step enables stable demodulation under different modulation schemes, significantly improving the adaptability and robustness of the demodulation system. It ensures the high integrity and accuracy of the output demodulated symbol sequence, providing a reliable foundation for subsequent signal decoding and data recovery. Simultaneously, it simplifies the system debugging process and improves overall demodulation efficiency.

[0040] Specifically, step S400 includes the following steps: Step S401: Based on the target modulation recognition result, adaptively switch to the fine phase detector that perfectly matches the target modulation recognition result, and adjust the coefficients of the second-order loop filter to the values ​​specific to the fine phase-locked stage. During the switching process, keep the cumulative phase of the numerically controlled oscillator and the integral term of the loop filter unchanged. Step S402: Based on the switched refined phase detector and the adjusted second-order loop filter, perform carrier fine phase-locking iterative operations on the single-symbol-rate data symbol sequence to obtain a stable constellation symbol sequence; Step S403: According to the target modulation recognition result, match and select the exclusive constant modulus value corresponding to the target modulation recognition result, and perform blind equalization processing on the stable constellation symbol sequence using the exclusive constant modulus algorithm corresponding to the exclusive constant modulus value to obtain an equalized symbol sequence; Step S404: Based on the decision-directed phase-locked loop, perform phase compensation processing on the equalized symbol sequence, smooth the phase compensation error through a first-order loop filter and update the phase compensation value to correct the phase ambiguity and obtain a demodulated symbol sequence with accurate phase.

[0041] In one implementation, as Figure 2-4 shown, first, according to the target modulation recognition result , quickly switch to a refined phase detector that exactly matches the target modulation recognition result through a multiplexer, and synchronously adjust the coefficients of the second-order loop filter to the exclusive values for the fine phase-locking stage (K1'<K1, K2'<K2), thereby narrowing the loop bandwidth, reducing the steady-state phase jitter, and adapting to the steady-state tracking requirements of the fine phase-locking stage. For different target modulation recognition results, this embodiment designs a group of exclusive phase detectors covering all target modulation recognition results: for the BPSK modulation method, switch to the Costas loop phase detector adapted to its constant envelope characteristic, and use the multiplication of the real part of the symbol and the imaginary part to achieve blind phase discrimination. The error calculation formula is: , where are the real part and the imaginary part of the received signal respectively, is the sign function. Since the BPSK signal has only real part information and the imaginary part reflects the phase deviation, multiplying the real part symbol by the imaginary part can obtain an error signal proportional to the phase error; for the QPSK modulation method, utilize the symmetry of its constellation points, reuse the decision-directed phase detector, and directly obtain the phase error by taking the imaginary part of the conjugate product of the hard decision result and the received symbol. The error calculation formula is: , where is the conjugate of the hard decision result to the QPSK standard constellation point (normalized ), denotes taking the imaginary part of the complex number; for the 8PSK modulation method, switch to the octave loop phase detector, and utilize the characteristic that the phase of the constellation point is . After octaving, all points are mapped to the positive real axis, and the imaginary part only reflects the 8-fold phase error, thus perfectly eliminating the modulation information and adapting to its uniform phase distribution. The error calculation formula is: , where is the constellation symbol The calculation is performed to the power of 8. For rectangular constellation structures such as 16QAM / 32QAM, the polarity decision phase detector for rectangular constellations is switched to calculate the phase detection error using only the outermost loop point of the constellation diagram with the strongest noise immunity. This avoids noise interference from the inner loop point and improves the synchronization robustness of higher-order QAM. The error calculation formula is as follows: ,in, for The conjugate of the hard decision to the standard point of the outermost ring of the QAM constellation. The QAM power threshold is set to 1.5 to 2 times the average signal power Pavg (typically 1.8 Pavg), preset based on statistical signal power distribution. For concentric ring structures like 16APSK / 32APSK, a dedicated polarity decision phase detector for multi-ring structures is used to accurately select the effective constellation points of the outermost ring to avoid mutual interference between signals from different ring layers. The threshold is set according to the ring structure differences, and the error calculation formula is: ,in, for The conjugate of the hard decision to the standard point of the outermost ring of the APSK constellation. The APSK power threshold must be between the squares of the outermost and second outermost ring radii: Next, during the phase detector switching and loop filter coefficient adjustment process, the cumulative phase of the numerically controlled oscillator (NCO) and the integral term of the loop filter remain completely unchanged to avoid phase jumps during switching, achieving smooth switching with continuous phase transitions and preventing loop lockout and demodulation interruption. Subsequently, based on the switched precise phase detector and the adjusted second-order loop filter, carrier precise phase-locked iteration is performed on the single symbol rate data symbol sequence. The phase error is accurately calculated using the error calculation formula corresponding to each dedicated phase detector. The error signal is processed using the second-order loop filter to accurately compensate for the residual frequency offset and phase noise in the single symbol rate data symbol sequence, ultimately obtaining a stable and clear constellation symbol sequence. This provides a high signal-to-noise ratio signal foundation for subsequent channel equalization, achieving adaptive high-precision carrier synchronization for multiple modulated signals. By employing smooth switching processing, phase jumps and loop lockout problems during phase detector switching are avoided, ensuring the continuity and stability of the carrier synchronization process and preventing demodulation interruptions due to parameter switching. The hierarchical architecture of coarse phase-locked loop and fine phase-locked loop is adopted, which balances the speed of carrier acquisition and the accuracy of synchronization tracking. It not only solves the carrier start-up problem under unknown modulation mode, but also achieves high-precision carrier synchronization of high-order modulation signals, providing a high signal-to-noise ratio constellation diagram symbol for subsequent equalization.

[0042] Then, based on the modulation recognition results, the constant modulus algorithm (CMA) parameters are matched, and the constant modulus value specific to the corresponding modulation recognition result is selected. (Different modulation schemes have different preset values), resulting in a dedicated constant modulus algorithm; this dedicated constant modulus algorithm is used to perform blind equalization on the constellation diagram symbol sequence to eliminate inter-symbol interference in the signal, where the cost function of CMA is: The CMA coefficient update formula is: ,in, Let cost function be The symbol after balance. This is the constant modulus value (the preset value varies depending on the modulation method). Equalizer coefficients To update the step size of the coefficients, This is the output of the equalizer. Under different modulation schemes, other equalization methods can be used, such as 16APSK, 32APSK, or a region-based equalization method. A dedicated constant modulus algorithm is used to perform blind equalization on the constellation diagram symbol sequence to eliminate inter-symbol interference in the signal, resulting in the equalized symbol sequence. After equalization, since the equalization algorithm introduces phase deviation, phase compensation of the received signal is required. A decision-guided phase-locked loop (PLL) is used for phase deviation compensation. The error formula for the PLL phase detector is: ,in To compensate for phase error, Symbol after equilibrium The hard verdict, for The conjugate of the equation is then applied. A first-order loop filter is used to smooth the phase compensation error, updating the phase compensation value. This compensation value is then used to correct the residual phase of the equalized symbol sequence, resolving the inherent phase ambiguity problem of the CMA algorithm and obtaining a phase-accurate, distortion-free demodulated symbol sequence. .

[0043] Overall, this embodiment achieves fast and accurate switching of the phase detector through a multiplexer. Combined with dynamic adjustment of the loop filter coefficients, it realizes adaptive high-precision carrier synchronization under multiple modulation modes, effectively solving the problem that traditional single phase detectors cannot adapt to multiple modulation modes. At the same time, by keeping the NCO cumulative phase and the loop filter integral term unchanged, the phase continuity of the switching process is ensured, avoiding loop lockout and demodulation interruption. Combined with the phase compensation of CMA blind equalization and decision-guided phase-locked loop, the phase deviation introduced by inter-symbol interference and equalization is further eliminated, realizing stable and high-precision signal demodulation, and improving the communication reliability and system robustness in multi-modulation signal environments.

[0044] In one implementation, after obtaining the demodulated symbol sequence, the signal-to-noise ratio (SNR) of the demodulated symbol sequence is further estimated using either the error vector magnitude method or the M2M4 blind estimation algorithm. Specifically, if the error vector magnitude method is used, based on the demodulated symbol... According to the formula for calculating the error vector magnitude (EVM): The error vector magnitude is calculated, and then converted using the SNR conversion formula: The EVM value is converted into a signal-to-noise ratio estimate (SNRest). If the M2M4 blind estimation algorithm is used, the SNR estimate is derived by calculating the second and fourth moments (average power and signal fluctuation characteristics) of the signal based on the characteristics of the second and fourth moments of the demodulated symbol. This algorithm does not rely on an ideal reference symbol, is a blind estimation algorithm, is simple to operate and easy to implement in engineering, and is especially suitable for scenarios with complex channel characteristics such as scatter communication. Then, the synchronously output demodulated symbol is converted into a signal-to-noise ratio estimate (SNRest). The signal-to-noise ratio (SNR) estimate (SNRest) is used for subsequent decoding or analysis. This embodiment obtains the SNR estimate through quantization calculation, providing data support for signal performance evaluation and channel quality analysis. The phase-corrected demodulated symbols are directly used in subsequent channel decoding, information parsing, and signal monitoring and analysis, completing the entire processing chain from baseband signal input to demodulation result output. SNR estimation achieves a closed-loop performance model for the entire blind demodulation chain, not only outputting the demodulated symbol results but also simultaneously quantifying demodulation performance and signal quality. This provides complete quantitative data for applications such as satellite signal monitoring and spectrum analysis, significantly improving the convenience of engineering applications.

[0045] In summary, this embodiment first performs parameter estimation, frequency offset compensation, and timing synchronization processing based on the original baseband sampled data sequence to obtain a single symbol rate data symbol sequence. This effectively eliminates the impact of sampling deviation, carrier frequency offset, and timing error on subsequent demodulation, improving signal synchronization accuracy and data reliability. Next, based on the single symbol rate data symbol sequence, iterative processing of hierarchical phase-locked loop carrier coarse synchronization is performed to obtain the target convergent constellation diagram symbol sequence. This achieves rapid coarse adjustment of the carrier phase and constellation diagram convergence, significantly reducing demodulation distortion caused by phase ambiguity and noise interference. Then, based on the target convergent constellation diagram symbol sequence, high-order cumulant features are extracted, and modulation scheme identification and confidence judgment are performed. This process iterates until preset judgment conditions are met to obtain the target modulation identification result. Even in complex channel environments, the accuracy and robustness of modulation identification are guaranteed, avoiding the degradation of demodulation performance due to misidentification. Finally, based on the target modulation identification result, joint processing of modulation scheme adaptive carrier fine phase-locking and blind equalization is performed to obtain the demodulated symbol sequence. This achieves fine carrier phase correction and channel distortion equalization compensation, further improving the signal-to-noise ratio and decision correctness of the demodulated symbols. This invention, through its overall architecture of step-by-step synchronization, hierarchical carrier locking, adaptive modulation identification, and fine demodulation joint processing, achieves efficient blind demodulation without prior information in complex time-varying channels, low signal-to-noise ratio, and multiple modulation schemes coexisting in satellite communication scenarios. This effectively improves demodulation accuracy, convergence speed, and system adaptability, and enhances the stability and transmission reliability of satellite communication links.

[0046] like Figure 5 As shown in the illustration, this embodiment also provides an adaptive blind demodulation system based on a satellite communication system. This system includes: a single-symbol-rate data symbol sequence acquisition module 10, a target convergence constellation diagram symbol sequence acquisition module 20, a target modulation identification result acquisition module 30, and a demodulation symbol sequence acquisition module 40. Specifically, the single-symbol-rate data symbol sequence acquisition module 10 is used to perform parameter estimation, frequency offset compensation, and timing synchronization processing based on the original baseband sampled data sequence to obtain a single-symbol-rate data symbol sequence. The target convergence constellation diagram symbol sequence acquisition module 20 is used to perform hierarchical phase-locked loop carrier coarse synchronization iterative processing based on the single-symbol-rate data symbol sequence to obtain a target convergence constellation diagram symbol sequence. The target modulation identification result acquisition module 30 is used to extract high-order cumulant features based on the target convergence constellation diagram symbol sequence and perform modulation mode identification and confidence judgment, iterating until a preset judgment condition is met to obtain a target modulation identification result. The demodulation symbol sequence acquisition module 40 is used to perform adaptive carrier fine phase-locking and blind equalization joint processing based on the target modulation identification result to obtain a demodulation symbol sequence.

[0047] In one implementation, the single symbol rate data symbol sequence acquisition module 10 includes: The coarse frequency offset acquisition unit is used to perform spectral analysis on the original baseband sampled data sequence to obtain the signal bandwidth and signal center frequency, and obtain the coarse frequency offset based on the signal center frequency; The coarse compensation signal sequence acquisition unit is used to perform frequency offset compensation on the original baseband sampled data sequence based on the coarse frequency offset value by digital spectrum shifting to obtain the coarse compensation signal sequence. An integer multiple symbol rate downsampled signal sequence acquisition and matched filtering unit is used to calculate a symbol rate estimate based on the signal bandwidth, and perform downsampling processing on the coarse compensation signal sequence according to the symbol rate estimate to obtain an integer multiple symbol rate downsampled signal sequence and perform matched filtering. The single symbol rate data symbol sequence acquisition unit is used to perform timing error detection and interpolation adjustment on the sampled signal sequence at an integer multiple symbol rate after filtering, so as to obtain the single symbol rate data symbol sequence.

[0048] In one implementation, the target convergent constellation diagram symbol sequence acquisition module 20 includes: The initialization unit is used to initialize the first stage of the hierarchical phase-locked loop, clear the integral term of the loop filter and the cumulative phase of the numerically controlled oscillator, and configure the second-order loop filter coefficients specific to the coarse phase-locked stage. The target convergent constellation diagram symbol sequence acquisition unit is used to continuously iterate the single-symbol-rate data symbol sequence. For each current single-symbol-rate data symbol in the sequence, the current accumulated phase of the numerically controlled oscillator is used to perform phase rotation compensation on the current single-symbol-rate data symbol. A phase detection operation is performed on the compensated symbol to obtain the phase detection error. The phase detection error is input into a second-order loop filter to update the integral term of the loop filter and obtain the phase adjustment amount. The accumulated phase of the numerically controlled oscillator is updated using the phase adjustment amount. The sliding window is used to calculate the moving average variance of the phase detection error and compare it with a preset steady-state judgment threshold. When the preset convergence condition is met, the symbols output at this time and subsequently after phase rotation compensation are used as the target convergent constellation diagram symbol sequence.

[0049] In one implementation, the target modulation recognition result acquisition module 30 includes: The target modulation recognition result acquisition unit is used to collect a preset number of target convergent constellation diagram symbol sequences as current recognition samples in each iteration, based on the target convergent constellation diagram symbol sequence, and calculate the cumulative quantity feature of the current recognition sample; perform signal class differentiation and modulation mode subdivision recognition based on the cumulative quantity feature to obtain the current modulation mode, and perform confidence judgment on the current modulation mode until the preset confidence judgment condition is met to obtain the target modulation recognition result.

[0050] In one implementation, the target modulation recognition result acquisition unit includes: The category distinction result acquisition subunit is used to extract normalized fourth-order cumulants based on the cumulant features, and to perform category distinction on the signals corresponding to the identification samples based on the normalized fourth-order cumulants to obtain category distinction results. The first subunit is obtained by subdividing the modulation results. If the classification result is PSK, the fourth phase moment of the identified sample is calculated, and the signal corresponding to the identified sample is classified into QPSK and 8PSK types based on the fourth phase moment to obtain the subdivided modulation results. The second subunit is obtained from the subdivision modulation result. If the classification result is QAM, the signal corresponding to the identification sample is classified into 16QAM and 32QAM based on the normalized fourth-order cumulant and sixth-order cumulant in the cumulant calculation result, and the subdivision modulation result is obtained. The third subunit is obtained from the subdivision modulation result. If the classification result is APSK, the signal corresponding to the identification sample is classified into 16APSK and 32APSK types based on the sixth-order cumulant in the cumulant calculation result, and the subdivision modulation result is obtained. The current modulation mode acquisition subunit is used to obtain the current modulation mode based on the subdivision modulation results.

[0051] In one implementation, the demodulation symbol sequence acquisition module 40 includes: The switching and coefficient adjustment unit is used to adaptively switch to the fine phase detector that perfectly matches the target modulation recognition result according to the target modulation recognition result, and adjust the coefficient of the second-order loop filter to the value specific to the fine phase-locked stage, while keeping the cumulative phase of the numerically controlled oscillator and the integral term of the loop filter unchanged during the switching process; The stable constellation diagram symbol sequence acquisition unit is used to perform carrier fine phase-locked iteration operation on the single symbol rate data symbol sequence based on the switched fine phase detector and the adjusted second-order loop filter to obtain the stable constellation diagram symbol sequence. The equalized symbol sequence acquisition unit is used to match and select the exclusive constant modulus value corresponding to the target modulation recognition result according to the target modulation recognition result, and use the exclusive constant modulus algorithm corresponding to the exclusive constant modulus value to perform blind equalization processing on the stable constellation diagram symbol sequence to obtain the equalized symbol sequence. The demodulated symbol sequence acquisition unit is used to perform phase compensation processing on the equalized symbol sequence based on the decision-guided phase-locked loop, smooth the phase compensation error and update the phase compensation value through a first-order loop filter, correct the phase ambiguity, and obtain a demodulated symbol sequence with accurate phase.

[0052] In one implementation, the system further includes: The signal-to-noise ratio (SNR) estimation module is used to estimate the SNR of the demodulated symbol sequence using the error vector magnitude method or the M2M4 blind estimation algorithm, and obtain the SNR estimate.

[0053] The working principle of each module in the adaptive blind demodulation system based on the satellite communication system in this embodiment is the same as that of each step in the above method embodiment, and will not be repeated here.

[0054] Based on the above embodiments, the present invention also provides a terminal, the principle block diagram of which can be as follows: Figure 6 As shown. The terminal may include one or more processors 100 ( Figure 6 (Only one is shown in the image), memory 101, and a computer program 102 stored in memory 101 and executable on one or more processors 100, such as an adaptive blind demodulation program based on a satellite communication system. When one or more processors 100 execute computer program 102, they can implement the various steps in the embodiments of the adaptive blind demodulation method based on a satellite communication system. Alternatively, when one or more processors 100 execute computer program 102, they can implement the functions of various modules / units in the embodiments of the adaptive blind demodulation method based on a satellite communication system, which is not limited here.

[0055] In one embodiment, the processor 100 may 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 may be a microprocessor or any conventional processor.

[0056] In one embodiment, memory 101 may be an internal storage unit of an electronic device, such as a hard drive or RAM. Memory 101 may also be an external storage device of the electronic device, such as a plug-in hard drive, smart media card (SMC), secure digital (SD) card, flash card, etc. Furthermore, memory 101 may include both internal and external storage units. Memory 101 is used to store computer programs and other programs and data required by the terminal. Memory 101 can also be used to temporarily store data that has been output or will be output.

[0057] Those skilled in the art will understand that Figure 6 The schematic diagram shown is merely a partial structural diagram related to the present invention and does not constitute a limitation on the terminal to which the present invention is applied. A specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0058] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, operational databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual operating data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0059] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An adaptive blind demodulation method based on a satellite communication system, characterized in that, The method includes: Based on the original baseband sampled data sequence, parameter estimation, frequency offset compensation and timing synchronization processing are performed to obtain a single symbol rate data symbol sequence; Based on the single symbol rate data symbol sequence, perform hierarchical phase-locked loop carrier coarse synchronization iterative processing to obtain the target convergent constellation diagram symbol sequence; Based on the target convergent constellation diagram symbol sequence, high-order cumulant features are extracted and modulation mode identification and confidence judgment are performed. The process is iterated until the preset judgment conditions are met to obtain the target modulation identification result. Based on the target modulation identification results, a carrier phase-locked loop and blind equalization process with adaptive modulation mode are performed to obtain the demodulated symbol sequence.

2. The adaptive blind demodulation method based on a satellite communication system according to claim 1, characterized in that, The process of parameter estimation, frequency offset compensation, and timing synchronization based on the original baseband sampled data sequence yields a single-symbol-rate data symbol sequence, including: Spectral analysis is performed on the original baseband sampled data sequence to obtain the signal bandwidth and signal center frequency, and coarse frequency offset is obtained based on the signal center frequency; Based on the coarse frequency offset, the original baseband sampled data sequence is compensated for the frequency offset by digital spectrum shifting to obtain the coarse compensated signal sequence; Based on the signal bandwidth, a symbol rate estimate is calculated, and the coarsely compensated signal sequence is downsampled according to the symbol rate estimate to obtain an integer multiple symbol rate downsampled signal sequence and then matched filtering is performed. Timing error detection and interpolation adjustment are performed on the filtered sampled signal sequence at integer multiple symbol rates to obtain a single symbol rate data symbol sequence.

3. The adaptive blind demodulation method based on a satellite communication system according to claim 1, characterized in that, The step of performing hierarchical phase-locked loop carrier coarse synchronization iterative processing based on the single symbol rate data symbol sequence to obtain the target convergent constellation diagram symbol sequence includes: The first stage of the hierarchical phase-locked loop is initialized, the integral term of the loop filter and the cumulative phase of the numerically controlled oscillator are cleared, and the coefficients of the second-order loop filter specific to the coarse phase-locking stage are configured. The single-symbol-rate data symbol sequence is continuously iterated. For each current single-symbol-rate data symbol in the sequence, phase rotation compensation is performed on the current single-symbol-rate data symbol using the current accumulated phase of the numerically controlled oscillator. A phase detection operation is performed on the compensated symbol to obtain the phase detection error. The phase detection error is input into a second-order loop filter to update the integral term of the loop filter and obtain the phase adjustment amount. The accumulated phase of the numerically controlled oscillator is updated using the phase adjustment amount. The sliding window is used to calculate the moving average variance of the phase detection error and compare it with a preset steady-state judgment threshold. When the preset convergence condition is met, the symbols output at this time and subsequently after phase rotation compensation are taken as the target convergent constellation diagram symbol sequence.

4. The adaptive blind demodulation method based on a satellite communication system according to claim 1, characterized in that, The process involves extracting high-order cumulant features from the target convergent constellation diagram symbol sequence and performing modulation scheme identification and confidence judgment, iterating until a preset judgment condition is met to obtain the target modulation identification result, including: In each iteration, based on the target convergent constellation diagram symbol sequence, a preset number of target convergent constellation diagram symbol sequences are collected as the current identification samples, and the cumulative quantity features of the current identification samples are calculated; based on the cumulative quantity features, signal class differentiation and modulation mode subdivision identification are performed to obtain the current modulation mode, and a confidence decision is performed on the current modulation mode until the preset confidence decision conditions are met, and the target modulation identification result is obtained.

5. The adaptive blind demodulation method based on a satellite communication system according to claim 4, characterized in that, The step of performing signal category differentiation and modulation mode subdivision identification based on the accumulated quantity characteristics to obtain the current modulation mode includes: Based on the accumulated feature, a normalized fourth-order accumulated value is extracted. Based on the normalized fourth-order accumulated value, the signal corresponding to the identification sample is classified into major categories to obtain the major category classification result. If the major category classification result is PSK, calculate the fourth phase moment of the identified sample, and distinguish the QPSK and 8PSK types of the signal corresponding to the identified sample based on the fourth phase moment to obtain the subdivision modulation result; If the major category classification result is QAM, based on the normalized fourth-order cumulant and sixth-order cumulant in the cumulant calculation result, the signal corresponding to the identification sample is classified into 16QAM and 32QAM types to obtain the subdivision modulation result. If the major category classification result is APSK, the signal corresponding to the identification sample is classified into 16APSK and 32APSK types based on the sixth-order cumulant in the cumulant calculation result to obtain the subdivision modulation result; The current modulation scheme is obtained based on the subdivision modulation results.

6. The adaptive blind demodulation method based on a satellite communication system according to claim 1, characterized in that, The step of performing carrier phase-locked loop and blind equalization joint processing based on the target modulation identification result to obtain the demodulated symbol sequence includes: Based on the target modulation recognition result, the system adaptively switches to a fine phase detector that perfectly matches the target modulation recognition result, and adjusts the coefficients of the second-order loop filter to values ​​specific to the fine phase-locked stage. During the switching process, the cumulative phase of the numerically controlled oscillator and the integral term of the loop filter remain unchanged. Based on the switched fine phase detector and the adjusted second-order loop filter, a carrier fine phase-locked iteration operation is performed on the single symbol rate data symbol sequence to obtain a stable constellation diagram symbol sequence. Based on the target modulation recognition result, a unique constant modulus value corresponding to the target modulation recognition result is matched and selected. The unique constant modulus algorithm corresponding to the unique constant modulus value is used to perform blind equalization processing on the stable constellation diagram symbol sequence to obtain the equalized symbol sequence. Based on the decision-guided phase-locked loop, phase compensation processing is performed on the equalized symbol sequence. The phase compensation error is smoothed and the phase compensation value is updated through a first-order loop filter to correct phase ambiguity and obtain a demodulated symbol sequence with accurate phase.

7. The adaptive blind demodulation method based on a satellite communication system according to claim 1, characterized in that, After obtaining the demodulated symbol sequence, the process further includes: The signal-to-noise ratio (SNR) of the demodulated symbol sequence is estimated using the error vector magnitude method or the M2M4 blind estimation algorithm to obtain the SNR estimate.

8. An adaptive blind demodulation system based on a satellite communication system, characterized in that, The system includes: The single symbol rate data symbol sequence acquisition module is used to perform parameter estimation, frequency offset compensation and timing synchronization processing based on the original baseband sampled data sequence to obtain the single symbol rate data symbol sequence. The target convergence constellation diagram symbol sequence acquisition module is used to perform hierarchical phase-locked loop carrier coarse synchronization iterative processing based on the single symbol rate data symbol sequence to obtain the target convergence constellation diagram symbol sequence; The target modulation recognition result acquisition module is used to extract high-order cumulative features based on the target convergent constellation diagram symbol sequence and perform modulation mode recognition and confidence judgment, iterating until the preset judgment conditions are met to obtain the target modulation recognition result. The demodulation symbol sequence acquisition module is used to perform joint processing of carrier fine phase-locking and blind equalization with modulation mode adaptation based on the target modulation identification result to obtain the demodulation symbol sequence.

9. A terminal, characterized in that, The terminal includes a memory, a processor, and an adaptive blind demodulation program based on a satellite communication system stored in the memory and executable on the processor. When the processor executes the adaptive blind demodulation program based on a satellite communication system, it implements the steps of the adaptive blind demodulation method based on a satellite communication system as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores an adaptive blind demodulation program based on a satellite communication system. When the adaptive blind demodulation program based on a satellite communication system is executed by a processor, it implements the steps of the adaptive blind demodulation method based on a satellite communication system as described in any one of claims 1-7.

Images