A method for capturing a wide range of carrier frequency deviation and tracking low steady-state errors

Abstract

The present invention discloses a method for large-scale carrier frequency deviation capture and low steady-state error tracking, and relates to the technical field of wireless signal tracking. The method for large-scale carrier frequency deviation capture and low steady-state error tracking mainly includes two steps: large-scale carrier frequency deviation capture and low steady-state error tracking. Through deep learning frequency deviation prediction and frequency domain parallel capture technology, the capture range and speed of the carrier frequency deviation are significantly improved, and the wide range of frequency deviation changes can be quickly adapted, thereby enhancing the robustness and flexibility of the system. At the same time, the adaptive adjustment mechanism enables the search step size and decision threshold to be dynamically optimized according to the actual environment, further improving the capture efficiency. In the low steady-state error tracking stage, the steady-state error is effectively reduced and the tracking accuracy is improved through fine frequency difference estimation and compensation, phase noise suppression, and dual-loop carrier tracking technology.

Description

Technical Field

[0001] The present invention relates to the technical field of wireless signal tracking, and in particular to a method for large-range capture of carrier frequency deviation and low steady-state error tracking. Background Art

[0002] In the field of wireless communication and microwave communication, in high-dynamic scenarios such as satellite communication and high-speed mobile devices, the carrier frequency will shift significantly due to the Doppler effect caused by relative motion, requiring the receiver to have the ability to quickly and accurately capture a wide range of frequency deviations. At the same time, with the continuous increase in communication rates, higher requirements are also placed on the accuracy and stability of carrier synchronization. In order to ensure the continuity and reliability of communication, and to meet the application requirements of high-precision positioning, data transmission, etc., the system needs to achieve low steady-state error tracking to reduce error accumulation during the synchronization process and improve synchronization accuracy and system stability. In addition, in wireless communication and microwave communication systems, the pressure of limited resources and energy conservation and emission reduction cannot be ignored. Low steady-state error tracking can also help optimize system resource utilization and reduce unnecessary adjustments and energy consumption. Therefore, large-scale carrier frequency deviation capture and low steady-state error tracking have become key challenges in the design and optimization of modern communication systems.

[0003] Publication No. CN106603451B discloses a method for estimating high dynamic Doppler frequency deviation and frequency deviation change rate based on delayed autocorrelation, which is an effective method for frequency locking and capturing high dynamic low information rate spread spectrum signals. The Doppler frequency deviation and Doppler frequency deviation change rate generated by low information rate high dynamic spread spectrum signals are very large, which increases the difficulty of capturing and tracking spread spectrum signals. This method demodulates the high dynamic spread spectrum received signal through a flat method to remove the influence of the spread spectrum code, and then uses the delayed autocorrelation method to obtain the frequency deviation and frequency deviation change rate to assist in the capture of high dynamic spread spectrum signals. The receivable frequency deviation and frequency deviation change rate range of low information rate high dynamic spread spectrum signals are greatly improved, enabling high-speed and ultra-high-speed aircraft and other platforms to establish spread spectrum communication links with strong anti-interference capabilities.

[0004] Although the existing technology can effectively estimate the Doppler frequency deviation and frequency deviation change rate of high-dynamic low-information-rate spread spectrum signals, there is still room for improvement in capture efficiency. Especially in complex and changeable communication environments, the lack of intelligent prediction and adjustment mechanisms makes it impossible to quickly and accurately lock the carrier frequency deviation. In addition, in the low steady-state error tracking stage, the existing technology also faces challenges in tracking accuracy and robustness. Especially in high-dynamic environments, due to factors such as the Doppler effect and channel interference, the carrier frequency of the received signal may change rapidly, which places higher requirements on the accuracy and stability of the tracking system. Summary of the invention

[0005] In view of the deficiencies of the prior art, the present invention provides a method for large-range carrier frequency deviation capture and low steady-state error tracking, which solves the deficiencies of the prior art.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a method for large-scale carrier frequency deviation capture and low steady-state error tracking, the steps mainly include two steps of large-scale carrier frequency deviation capture and low steady-state error tracking, wherein the step of large-scale carrier frequency deviation capture includes:

[0007] S1 initialization and preprocessing: configure receiver parameters, filter, amplify and down-convert the signal;

[0008] S2 frequency deviation prediction based on deep learning: Use historical frequency deviation data to predict the current frequency deviation range;

[0009] S3 frequency domain parallel capture: frequency deviation is quickly detected through FFT fast Fourier transform and compressed sensing theory;

[0010] S4 adaptive adjustment: adaptively adjust the search step size and decision threshold according to external environmental factors;

[0011] The steps for low steady-state error tracking include:

[0012] S5 initial synchronization: Use the captured frequency to perform initial synchronization of the local carrier and configure the phase-locked loop PLL and frequency-locked loop FLL;

[0013] S6 Frequency difference estimation and compensation: Use algorithms to estimate frequency difference and compensate, and take external environmental factors into account to correct the error model;

[0014] S7 Phase Noise Suppression: Use adaptive filters to suppress phase noise;

[0015] S8 dual-loop carrier tracking: Combine FLL and PLL for carrier tracking and adjust parameters in real time;

[0016] S9 adaptively adjusts tracking parameters: adjusts PLL parameters according to environmental changes, monitors tracking errors and makes continuous adjustments.

[0017] Preferably, step S1 initialization and preprocessing specifically includes:

[0018] S1.1 Initialization settings:

[0019] Configure the receiver's hardware parameters: set the sampling rate to S, the number of FFT points to N, and the search step to Δf s ;

[0020] Set the initial parameters of the adaptive control algorithm: the learning rate is η, the adjustment step size is Δf a , the decision threshold is θ d ;

[0021] S1.2 Signal preprocessing:

[0022] Bandpass filter the received signal to remove out-of-band noise and interference;

[0023] Amplify the filtered signal with a magnification of A;

[0024] Down-convert the amplified signal to the intermediate frequency f IF to be processed.

[0025] Preferably, the frequency deviation prediction based on deep learning in step S2 specifically includes:

[0026] S2.1 uses a deep learning model to learn historical frequency deviation data and predict the current frequency deviation range:

[0027] Assume that the historical frequency deviation data is {fh 1 , fh 2 ,…,fh m}, where fh represents the historical frequency deviation data, m represents the number of historical frequency deviation data; the predicted frequency deviation range output by the deep learning model is [f min , f max ], then the prediction process is expressed as: [f min , f max ]=DeepLearningModel({fh 1 , fh 2 ,…,fh m});

[0028] S2.2 uses the predicted frequency offset range as the initial search range for frequency domain parallel capture to reduce the search time.

[0029] Preferably, step S3 of frequency domain parallel capture specifically includes:

[0030] S3.1 performs FFT fast Fourier transform on the preprocessed signal to obtain a frequency domain signal;

[0031] S3.2 uses compressed sensing theory to sparsely represent the signal in the frequency domain and quickly detect the frequency deviation: Let the received signal be r(t), where t represents time. After FFT transformation, the frequency domain signal R(f) is obtained. By solving the sparse optimization problem, the sparse frequency deviation vector S is obtained. The sparse representation formula of compressed sensing is expressed as:

[0032] R(f)=Φ·S+N;

[0033] Where Φ is the measurement matrix, S is the sparse frequency offset vector, and N is the noise;

[0034] The frequencies corresponding to the non-zero elements in the sparse frequency offset vector S are taken as the capture results.

[0035] Preferably, step S4 of adaptive adjustment specifically includes:

[0036] S4.1 Monitor external environmental factors, denoted as E;

[0037] S4.2 Adaptively adjust the search step size to Δf according to environmental changes s ′(E):

[0038] Δf s ′(E) = f(E 1 , E 2 , ..., E n );

[0039] Among them, f() is the established mathematical relationship model, E 1 , E 2 , ..., E n are the different external environmental factors monitored;

[0040] S4.3 Real-time monitoring of capture results, adjusting the search range to F according to the result feedback search ′(f captured ,θ d ) and the decision threshold to θ d ′(f captured ), where f captured Represents the captured frequency value, F search ′() indicates the adjusted search range;

[0041] The search range adjustment method is: set a signal strength threshold, when the captured signal strength exceeds the signal strength threshold, the current search range is reduced by the set ratio; when the signal strength is lower than the signal strength threshold, the search range is expanded by the set ratio;

[0042] The decision threshold adjustment method is as follows: define a signal quality indicator, which is the signal-to-noise ratio SNR or the bit error rate BER; set a signal quality threshold according to the actual application scenario and the performance requirements of the capture system; when the captured signal quality exceeds the signal quality threshold, increase the current decision threshold according to the set ratio; when the signal quality is lower than the signal quality threshold, reduce the decision threshold according to the set ratio.

[0043] Preferably, step S5 initial synchronization specifically includes:

[0044] S5.1 Initial carrier synchronization:

[0045] S5.1.1 Using the captured carrier frequency f captured, perform initial synchronization on the local carrier and set the initial frequency of the local carrier to , the phase is ;

[0046] S5.1.2 Configure the initial parameters of the phase-locked loop (PLL) and the frequency-locked loop (FLL): Set the transfer function of the PLL loop filter to HPLL(s) and the gain to KPLL; Set the transfer function of the FLL loop filter to HFLL(s) and the gain to KFLL;

[0047] S5.2 Initial carrier synchronization optimization: Use auxiliary information for preliminary prediction and calibration; then use the MLE algorithm to perform fine synchronization parameter estimation; including:

[0048] S5.2.1 Use of auxiliary information: The auxiliary information includes the nominal carrier frequency fc specified in the system configuration, the frequency offset statistics in the historical synchronization data, or any other reliable source that can provide information about the received signal frequency; based on the auxiliary information, if the historical data is available, the frequency offset Δf is estimated by using the statistical average method; if the system configuration and signal characteristics are known, a system prediction model is established to estimate the frequency offset Δf;

[0049] S5.2.2 MLE algorithm: In the initial carrier synchronization stage, the synchronization algorithm based on maximum likelihood estimation (MLE) is used to estimate the frequency and phase difference between the received signal and the local signal;

[0050] Assume the received signal is:

[0051] ;

[0052] Where: A is the amplitude of the signal; s(t) is the transmitted baseband signal; τ is the time delay; f c is the center frequency of the carrier; Δf is the frequency offset; φ is the phase offset; n(t) is the noise;

[0053] The local carrier signal l(t) is expressed as:

[0054] ;

[0055] Where: f LO is the frequency of the local carrier; φ LO is the phase of the local carrier;

[0056] The goal of the MLE algorithm is to find the frequency offset Δf and phase offset φ that maximize the likelihood function.

[0057] Preferably, step S6 of frequency difference estimation and compensation specifically includes:

[0058] S6.1 estimates the frequency difference of the received signal using a frequency difference estimation algorithm, denoted as Δf est =f est_algo (r(t), l(t)), where Δf est Indicates the frequency difference value, f est_algo () represents the frequency difference estimation algorithm;

[0059] S6.2 Set the frequency difference Δf est Input to the adaptive control algorithm to adjust the local carrier frequency;

[0060] S6.3 Establish the error model of carrier frequency deviation, considering the influence of external environmental factor E on carrier frequency deviation, denoted as Δf error =f error_model (E), where Δf error Indicates the actual carrier frequency error (Actual Frequency Error), that is, the actual difference between the carrier frequency of the received signal and the local carrier frequency, f error_model () represents the error model of carrier frequency deviation;

[0061] S6.4 Update and correct the error model based on the real-time tracking results to obtain the compensated frequency difference value Δf comp =Δf est −Δf error .

[0062] Preferably, the phase noise suppression in step S7 specifically includes:

[0063] S7.1 uses an adaptive filter to suppress phase noise and reduce its impact on carrier tracking;

[0064] S7.2 Set the received signal to , where s(t) is the baseband signal, ω 0 is the carrier frequency, ϕ(t) is the phase noise, and n(t) is the noise;

[0065] The input of S7.3 adaptive filter is the received signal r(t), and the output is the estimated phase noise ;

[0066] S7.4 subtracts the estimated phase noise from the received signal to obtain a signal with suppressed phase noise for subsequent carrier tracking.

[0067] Preferably, step S8 of dual-loop carrier tracking specifically includes:

[0068] S8.1 uses a second-order frequency-locked loop (FLL) to assist a third-order phase-locked loop (PLL) for carrier tracking;

[0069] S8.2 adjusts the parameters of FLL and PLL in real time according to the tracking results.

[0070] Preferably, step S9 of adaptively adjusting tracking parameters specifically includes:

[0071] S9.1 adaptively adjusts the parameters of the PLL loop filter according to the monitored external environmental factor E, denoted as HPLL′(s, E);

[0072] S9.2 monitors the tracking error in real time and further adjusts the search range to F according to the error size track ′(e,θ d ) and the decision threshold to θ d ′(e), where e is the tracking error;

[0073] S9.3 verifies the tracked carrier frequency and phase to confirm that they meet the accuracy and stability requirements of the communication system; after the verification, the tracking results are output to other parts of the subsequent communication system for processing.

[0074] The present invention provides a method for capturing a wide range of carrier frequency deviation and tracking a low steady-state error. Compared with the prior art, it has the following beneficial effects:

[0075] 1. This method of large-range carrier frequency deviation capture and low steady-state error tracking significantly improves the capture range and speed of carrier frequency deviation through deep learning frequency deviation prediction and frequency domain parallel capture technology, can quickly adapt to a wide range of frequency deviation changes, and enhances the robustness and flexibility of the system. At the same time, the adaptive adjustment mechanism enables the search step size and decision threshold to be dynamically optimized according to the actual environment, further improving the capture efficiency. In the low steady-state error tracking stage, the steady-state error is effectively reduced and the tracking accuracy is improved through precise frequency difference estimation and compensation, phase noise suppression, and dual-loop carrier tracking technology. In addition, real-time adaptive adjustment of tracking parameters ensures that the system can continuously and stably track the carrier frequency, improving overall performance and stability. These improvements make this method have a wider application prospect in complex environments.

[0076] 2. This method of large-range carrier frequency deviation capture and low steady-state error tracking ensures the accuracy and stability of the received signal through sophisticated initialization and preprocessing steps, laying a solid foundation for subsequent steps. The frequency deviation prediction technology based on deep learning can make full use of historical data to predict the current frequency deviation range, thereby greatly reducing the search time and improving the capture efficiency. Frequency domain parallel capture combined with compressed sensing theory realizes fast and accurate detection of frequency deviation, further improving the capture performance. In addition, the method also introduces an adaptive adjustment mechanism to monitor and adjust the search step, search range and decision threshold in real time according to external environmental factors, so that the capture system can dynamically adapt to environmental changes and maintain high performance and stability. This adaptive adjustment not only improves the capture accuracy, but also effectively reduces the misjudgment rate and enhances the robustness of the system.

[0077] 3. The method of large-scale carrier frequency deviation capture and low steady-state error tracking, in the low steady-state error tracking stage, through a fine initial synchronization step, including the use of auxiliary information and MLE algorithm for fine synchronization parameter estimation, significantly improves the accuracy and stability of carrier synchronization. At the same time, in the frequency difference estimation and compensation step, an advanced frequency difference estimation algorithm is adopted in combination with an adaptive control algorithm and an error model to achieve accurate estimation and compensation of the frequency difference, effectively reducing the carrier frequency deviation error. The phase noise suppression step uses an adaptive filter to effectively reduce the impact of phase noise on carrier tracking and improve signal quality. In addition, the dual-loop carrier tracking step combines the advantages of FLL and PLL to achieve fast and accurate tracking of carrier frequency and phase, and at the same time adjusts the loop parameters in real time according to the tracking results, enhancing the adaptability and stability of the system. The adaptive adjustment of tracking parameters step monitors and adjusts the tracking parameters in real time according to external environmental factors, further improving the robustness and tracking accuracy of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 It is a schematic diagram of the overall steps of the present invention;

[0079] Figure 2 A schematic diagram of the steps of capturing a large range of carrier frequency deviations according to the present invention;

[0080] Figure 3 It is a schematic flow chart of the steps of low steady-state error tracking of the present invention. DETAILED DESCRIPTION

[0081] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0082] See also Figure 1-Figure 3 , the present invention provides the following three technical solutions:

[0083] Figure 1 The first embodiment is shown: a method for large-scale carrier frequency offset capture and low steady-state error tracking, the steps mainly include two steps of large-scale carrier frequency offset capture and low steady-state error tracking, wherein the step of large-scale carrier frequency offset capture includes:

[0084] S1 initialization and preprocessing: configure receiver parameters, filter, amplify and down-convert the signal;

[0085] S2 frequency deviation prediction based on deep learning: Use historical frequency deviation data to predict the current frequency deviation range;

[0086] S3 frequency domain parallel capture: frequency deviation is quickly detected through FFT fast Fourier transform and compressed sensing theory;

[0087] S4 adaptive adjustment: adaptively adjust the search step size and decision threshold according to external environmental factors;

[0088] The steps for low steady-state error tracking include:

[0089] S5 initial synchronization: Use the captured frequency to perform initial synchronization of the local carrier and configure the phase-locked loop PLL and frequency-locked loop FLL;

[0090] S6 Frequency difference estimation and compensation: Use algorithms to estimate frequency difference and compensate, and take external environmental factors into account to correct the error model;

[0091] S7 Phase Noise Suppression: Use adaptive filters to suppress phase noise;

[0092] S8 dual-loop carrier tracking: Combine FLL and PLL for carrier tracking and adjust parameters in real time;

[0093] S9 adaptively adjusts tracking parameters: adjusts PLL parameters according to environmental changes, monitors tracking errors and makes continuous adjustments.

[0094] This method significantly improves the capture range and speed of carrier frequency deviation through deep learning frequency deviation prediction and frequency domain parallel capture technology, can quickly adapt to a wide range of frequency deviation changes, and enhances the robustness and flexibility of the system. At the same time, the adaptive adjustment mechanism enables the search step size and decision threshold to be dynamically optimized according to the actual environment, further improving the capture efficiency. In the low steady-state error tracking stage, the steady-state error is effectively reduced and the tracking accuracy is improved through precise frequency difference estimation and compensation, phase noise suppression, and dual-loop carrier tracking technology. In addition, real-time adaptive adjustment of tracking parameters ensures that the system can continuously and stably track the carrier frequency, improving the overall performance and stability. These improvements make this method have a wider application prospect in complex environments.

[0095] Figure 2 A second implementation is shown, which is mainly different from the first implementation in that the steps of capturing a large range of carrier frequency deviations include:

[0096] Step S1 initialization and preprocessing specifically includes:

[0097] S1.1 Initialization settings:

[0098] Configure the receiver's hardware parameters: set the sampling rate to S, the number of FFT points to N, and the search step to Δf s ;

[0099] Set the initial parameters of the adaptive control algorithm: the learning rate is η, the adjustment step size is Δf a , the decision threshold is θ d ;

[0100] S1.2 Signal preprocessing:

[0101] Bandpass filter the received signal to remove out-of-band noise and interference;

[0102] Amplify the filtered signal with a magnification of A;

[0103] Down-convert the amplified signal to the intermediate frequency f IF to be processed.

[0104] Step S2 frequency deviation prediction based on deep learning specifically includes:

[0105] S2.1 uses deep learning models (such as LSTM, GRU, etc.) to learn historical frequency deviation data and predict the current frequency deviation range:

[0106] Assume that the historical frequency deviation data is {fh 1 , fh 2 ,…,fh m}, where fh represents the historical frequency deviation data, mrepresents the number of historical frequency deviation data; the predicted frequency deviation range output by the deep learning model is [f min , f max ], then the prediction process is expressed as: [f min , f max ]=DeepLearningModel({fh 1 , fh 2 ,…,fh m});

[0107] S2.2 uses the predicted frequency offset range as the initial search range for frequency domain parallel capture to reduce the search time.

[0108] Step S3 of frequency domain parallel capture specifically includes:

[0109] S3.1 performs FFT fast Fourier transform on the preprocessed signal to obtain a frequency domain signal;

[0110] S3.2 uses compressed sensing theory to sparsely represent the signal in the frequency domain and quickly detect the frequency deviation: Let the received signal be r(t), where t represents time. After FFT transformation, the frequency domain signal R(f) is obtained. By solving the sparse optimization problem, the sparse frequency deviation vector S is obtained. The sparse representation formula of compressed sensing is expressed as:

[0111] R(f)=Φ·S+N;

[0112] Where Φ is the measurement matrix, S is the sparse frequency offset vector, and N is the noise;

[0113] The frequencies corresponding to the non-zero elements in the sparse frequency offset vector S are taken as the capture results to improve the capture accuracy.

[0114] Step S4 of adaptive adjustment specifically includes:

[0115] S4.1 Monitor external environmental factors, such as temperature, humidity, etc., denoted as E;

[0116] S4.2 Adaptively adjust the search step size to Δf according to environmental changes s ′(E):

[0117] Δf s ′(E) = f(E 1 , E 2 , ..., E n );

[0118] Among them, f() is the established mathematical relationship model, E 1 , E 2 , ..., E n are the different external environmental factors monitored (such as temperature, humidity, etc.);

[0119] S4.3 Real-time monitoring of capture results, adjusting the search range to F according to the result feedback search ′(f captured ,θ d ) and the decision threshold to θ d ′(f captured ), where f captured Represents the captured frequency value, F search ′() indicates the adjusted search range;

[0120] The search range is dynamically adjusted according to the captured signal strength. When the signal strength is strong, it indicates that the capture system has approached or has captured the target signal. At this time, the search range can be appropriately narrowed to improve the capture accuracy and reduce unnecessary search overhead. On the contrary, when the signal strength is weak, the search range needs to be expanded to increase the possibility of capturing the target signal. The search range adjustment method is: set a signal strength threshold. When the captured signal strength exceeds the signal strength threshold, the current search range is narrowed by a set ratio (such as 50%). When the signal strength is lower than the signal strength threshold, the search range is expanded by a set ratio (such as 150%). The specific ratio value is adjusted according to the actual application scenario and the performance of the capture system.

[0121] The decision threshold is adaptively adjusted according to the quality of the captured signal; when the signal quality is high, the decision threshold can be appropriately increased to reduce the possibility of false capture; when the signal quality is low, the decision threshold needs to be lowered to increase the chance of capturing the target signal; the decision threshold adjustment method is: define a signal quality indicator, the signal quality indicator is the signal-to-noise ratio SNR or the bit error rate BER; set a signal quality threshold according to the actual application scenario and the performance requirements of the capture system; when the captured signal quality exceeds the signal quality threshold, increase the current decision threshold by a set ratio (such as 10%); when the signal quality is lower than the signal quality threshold, reduce the decision threshold by a set ratio (such as 20%); the specific value of the ratio is also adjusted according to the actual situation.

[0122] Through sophisticated initialization and preprocessing steps, the accuracy and stability of the received signal are ensured, laying a solid foundation for subsequent steps. The frequency deviation prediction technology based on deep learning can make full use of historical data to predict the current frequency deviation range, thereby greatly reducing the search time and improving the capture efficiency. Frequency domain parallel capture combined with compressed sensing theory realizes fast and accurate detection of frequency deviation, further improving the capture performance. In addition, the method also introduces an adaptive adjustment mechanism to monitor and adjust the search step, search range and decision threshold in real time according to external environmental factors, so that the capture system can dynamically adapt to environmental changes and maintain high performance and stability. This adaptive adjustment not only improves the capture accuracy, but also effectively reduces the misjudgment rate and enhances the robustness of the system.

[0123] Figure 3 A third embodiment is shown, which is mainly different from the first embodiment in that the steps of low steady-state error tracking include:

[0124] Step S5 initial synchronization specifically includes:

[0125] S5.1 Initial carrier synchronization:

[0126] S5.1.1 Using the captured carrier frequency f captured , perform initial synchronization on the local carrier and set the initial frequency of the local carrier to , the phase is ;

[0127] S5.1.2 Configure the initial parameters of the phase-locked loop (PLL) and the frequency-locked loop (FLL): Set the transfer function of the PLL loop filter to HPLL(s) and the gain to KPLL; Set the transfer function of the FLL loop filter to HFLL(s) and the gain to KFLL;

[0128] S5.2 Initial carrier synchronization optimization: Use auxiliary information for preliminary prediction and calibration; then use the MLE algorithm to perform fine synchronization parameter estimation; including:

[0129] S5.2.1 Use of auxiliary information: The auxiliary information includes the nominal carrier frequency fc specified in the system configuration, the frequency offset statistics in the historical synchronization data, or any other reliable source that can provide information about the received signal frequency; based on the auxiliary information, if the historical data is available, the frequency offset Δf is estimated by using the statistical average value; if the system configuration and signal characteristics are known, a system prediction model is established to estimate the frequency offset Δf; for example, if the nominal carrier frequency fc is specified in the system configuration, and it is known that the receiving device may have a certain frequency offset due to certain reasons (such as temperature changes, aging, etc.), an initial frequency offset Δf can be estimated based on these factors;

[0130] S5.2.2 MLE algorithm: During the initial carrier synchronization phase, a synchronization algorithm based on maximum likelihood estimation (MLE) is used to estimate the frequency and phase difference between the received signal and the local signal;

[0131] Assume the received signal is:

[0132] ;

[0133] Where: A is the amplitude of the signal; s(t) is the transmitted baseband signal; τ is the time delay; f c is the center frequency of the carrier; Δf is the frequency offset; φ is the phase offset; n(t) is the noise;

[0134] The local carrier signal l(t) is expressed as:

[0135] ;

[0136] Where: f LO is the frequency of the local carrier; φ LO is the phase of the local carrier;

[0137] The goal of the MLE algorithm is to find the frequency offset Δf and phase offset φ that maximize the likelihood function; the likelihood function is expressed as a function of the correlation between the received signal and the local signal; by maximizing the likelihood function, the estimated values ​​of the frequency and phase can be obtained.

[0138] Step S6 of frequency difference estimation and compensation specifically includes:

[0139] S6.1 estimates the frequency difference of the received signal using a frequency difference estimation algorithm, denoted as Δf est =f est_algo (r(t), l(t)), where Δf est Indicates the frequency difference value, f est_algo () represents the frequency difference estimation algorithm;

[0140] S6.2 Set the frequency difference Δf est Input to the adaptive control algorithm to adjust the local carrier frequency;

[0141] S6.3 Establish the error model of carrier frequency deviation, considering the influence of external environmental factor E on carrier frequency deviation, denoted as Δf error =f error_model (E), where Δf error Indicates the actual carrier frequency error (Actual Frequency Error), that is, the actual difference between the carrier frequency of the received signal and the local carrier frequency, f error_model () represents the error model of carrier frequency deviation;

[0142] S6.4 Update and correct the error model based on the real-time tracking results to obtain the compensated frequency difference value Δf comp =Δf est −Δf error .

[0143] Step S7, phase noise suppression specifically includes:

[0144] S7.1 uses an adaptive filter to suppress phase noise and reduce its impact on carrier tracking;

[0145] S7.2 Set the received signal to , where s(t) is the baseband signal, ω 0is the carrier frequency, ϕ(t) is the phase noise, and n(t) is the noise;

[0146] The input of S7.3 adaptive filter is the received signal r(t), and the output is the estimated phase noise ;

[0147] S7.4 subtracts the estimated phase noise from the received signal to obtain a signal with suppressed phase noise for subsequent carrier tracking.

[0148] Step S8 dual-loop carrier tracking specifically includes:

[0149] S8.1 uses a second-order frequency-locked loop (FLL) to assist a third-order phase-locked loop (PLL) for carrier tracking;

[0150] S8.2 According to the tracking result, the parameters of the FLL and PLL are adjusted in real time, such as increasing the gain of the FLL to KFLL′(E), keeping the gain of the PLL KPLL unchanged or adjusting it as needed;

[0151] S8.2.1FLL parameter adjustment:

[0152] Loop bandwidth adjustment: The loop bandwidth of the FLL can be adjusted dynamically according to the size and change rate of the frequency error. Large frequency errors or rapidly changing frequencies may require a wider loop bandwidth to speed up tracking, but too wide a bandwidth may introduce more noise.

[0153] Gain adjustment: The gain of the FLL (KFLL) can also be adjusted based on the tracking results; increasing the gain can speed up the frequency locking speed, but it may also cause loop instability; therefore, it is necessary to find a balance between stability and tracking speed;

[0154] S8.2.2 PLL parameter adjustment:

[0155] Loop filter parameter adjustment: The loop filter parameters of the PLL (HPLL(s)) are crucial to the accuracy and stability of phase tracking; according to the magnitude and rate of change of the phase error, the order, bandwidth, and gain of the loop filter can be adjusted;

[0156] Numerically controlled oscillator parameter adjustment: The parameters of the numerically controlled oscillator (NCO), such as frequency and phase, also need to be adjusted in real time based on the tracking results of the PLL to ensure synchronization between the local carrier and the received signal;

[0157] S8.2.3 Collaborative Adjustment:

[0158] In the dual-loop structure, FLL and PLL work together; therefore, their mutual influence needs to be considered when adjusting parameters;

[0159] For example, after the FLL quickly locks the frequency, its loop bandwidth can be gradually reduced and the loop bandwidth of the PLL can be increased to improve the accuracy of phase tracking;

[0160] At the same time, the gains and filter parameters of the two loops need to be dynamically adjusted according to the stability of the tracking results to ensure the stability and accuracy of the entire tracking loop;

[0161] Step S9 of adaptively adjusting tracking parameters specifically includes:

[0162] S9.1 adaptively adjusts the parameters of the PLL loop filter according to the monitored external environmental factor E, denoted as HPLL′(s, E);

[0163] S9.2 monitors the tracking error in real time and further adjusts the search range to F according to the error size track ′(e,θ d ) and the decision threshold to θ d ′(e), where e is the tracking error;

[0164] S9.3 verifies the tracked carrier frequency and phase to confirm that they meet the accuracy and stability requirements of the communication system; after the verification, the tracking results are output to other parts of the subsequent communication system for processing.

[0165] In the low steady-state error tracking stage, the accuracy and stability of carrier synchronization are significantly improved through a fine initial synchronization step, including the use of auxiliary information and MLE algorithm for fine synchronization parameter estimation. At the same time, in the frequency difference estimation and compensation step, an advanced frequency difference estimation algorithm is used in combination with an adaptive control algorithm and an error model to achieve accurate estimation and compensation of the frequency difference, effectively reducing the carrier frequency offset error. The phase noise suppression step uses an adaptive filter to effectively reduce the impact of phase noise on carrier tracking and improve signal quality. In addition, the dual-loop carrier tracking step combines the advantages of FLL and PLL to achieve fast and accurate tracking of carrier frequency and phase, while adjusting the loop parameters in real time according to the tracking results, enhancing the adaptability and stability of the system. The adaptive adjustment of tracking parameters step monitors and adjusts the tracking parameters in real time according to external environmental factors, further improving the robustness and tracking accuracy of the system.

[0166] Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0167] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0168] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for large-range carrier frequency deviation capture and low steady-state error tracking, characterized in that: The steps include two steps: large-range carrier frequency deviation capture and low steady-state error tracking. The step of large-range carrier frequency deviation capture includes: S1 initialization and preprocessing: configure receiver parameters, filter, amplify and down-convert the signal; S2 frequency deviation prediction based on deep learning: Use historical frequency deviation data to predict the current frequency deviation range, including: S2.1 uses a deep learning model to learn historical frequency deviation data and predict the current frequency deviation range; S2.2 uses the predicted frequency offset range as the initial search range for frequency domain parallel capture to reduce the search time; S3 frequency domain parallel capture: frequency deviation is quickly detected through FFT fast Fourier transform and compressed sensing theory, including: S3.1 performs FFT fast Fourier transform on the preprocessed signal to obtain a frequency domain signal; S3.2 uses compressed sensing theory to sparsely represent the signal in the frequency domain and quickly detect the frequency deviation; S4 adaptive adjustment: adaptively adjust the search step size and decision threshold according to external environmental factors; The steps for low steady-state error tracking include: S5 initial synchronization: Use the captured frequency to perform initial synchronization of the local carrier and configure the phase-locked loop (PLL) and frequency-locked loop (FLL) parameters; S6 Frequency difference estimation and compensation: Use algorithms to estimate frequency difference and compensate, and take external environmental factors into account to correct the error model; S7 Phase Noise Suppression: Use adaptive filters to suppress phase noise; S8 dual-loop carrier tracking: Combine FLL and PLL for carrier tracking and adjust parameters in real time; S9 adaptively adjusts tracking parameters: adjusts PLL parameters according to environmental changes, monitors tracking errors and makes continuous adjustments.

2. The method for large-range carrier frequency offset capture and low steady-state error tracking according to claim 1, characterized in that: Step S1 initialization and preprocessing specifically includes: S1.1 Initialization settings: Configure the receiver's hardware parameters: set the sampling rate to S, the number of FFT points to N, and the search step to Δf s ; Set the initial parameters of the adaptive control algorithm: the learning rate is η, the adjustment step size is Δf a , the decision threshold is θ d ; S1.2 Signal preprocessing: Bandpass filter the received signal to remove out-of-band noise and interference; Amplify the filtered signal with a magnification of A; Down-convert the amplified signal to the intermediate frequency f IF to be processed.

3. The method for large-range carrier frequency offset capture and low steady-state error tracking according to claim 1, characterized in that: Step S2: Frequency deviation prediction based on deep learning: Assume that the historical frequency deviation data is {fh1, fh2, ..., fh m }, where fh represents the historical frequency deviation data, and m represents the number of historical frequency deviation data; the predicted frequency deviation range output by the deep learning model is [f min , f max ], then the prediction process is expressed as: [f min , f max ]=DeepLearningModel({fh1, fh2,…, fh m }).

4. A method for large-scale carrier frequency deviation capture and low steady-state error tracking according to claim 1, characterized in that: Step S3: frequency domain parallel capture: Assume that the received signal is r(t), where t represents time. After FFT transformation, the frequency domain signal R(f) is obtained. By solving the sparse optimization problem, the sparse frequency offset vector S is obtained. The sparse representation formula of compressed sensing is expressed as: R(f)=Φ·S+N; Where Φ is the measurement matrix, S is the sparse frequency offset vector, and N is the noise; The frequencies corresponding to the non-zero elements in the sparse frequency offset vector S are taken as the capture results.

5. A method for large-scale carrier frequency deviation capture and low steady-state error tracking according to claim 1, characterized in that: Step S5 initial synchronization specifically includes: S5.1 Initial carrier synchronization: S5.1.1 Using the captured carrier frequency f captured , perform initial synchronization on the local carrier and set the initial frequency of the local carrier to , the phase is ; S5.1.2 Configure the initial parameters of the phase-locked loop (PLL) and the frequency-locked loop (FLL): Set the transfer function of the PLL loop filter to HPLL(s) and the gain to KPLL; Set the transfer function of the FLL loop filter to HFLL(s) and the gain to KFLL; S5.2 Initial carrier synchronization optimization: Use auxiliary information for preliminary prediction and calibration; then use the MLE algorithm to perform fine synchronization parameter estimation; including: S5.2.1 Use of auxiliary information: The auxiliary information includes the nominal carrier frequency fc specified in the system configuration, the frequency offset statistics in the historical synchronization data, or any other reliable source that can provide information about the received signal frequency; based on the auxiliary information, if the historical data is available, the frequency offset Δf is estimated by using the statistical average method; if the system configuration and signal characteristics are known, a system prediction model is established to estimate the frequency offset Δf; S5.2.2 MLE algorithm: In the initial carrier synchronization stage, the synchronization algorithm based on maximum likelihood estimation (MLE) is used to estimate the frequency and phase difference between the received signal and the local signal; Assume the received signal is: ； Where: A is the amplitude of the signal; s(t) is the transmitted baseband signal; τ is the time delay; Δf is the frequency offset; φ is the phase offset; n(t) is the noise; The local carrier signal l(t) is expressed as: ； Where: f LO is the frequency of the local carrier; φ LO is the phase of the local carrier; The goal of the MLE algorithm is to find the frequency offset Δf and phase offset φ that maximize the likelihood function.

6. A method for large-range carrier frequency deviation capture and low steady-state error tracking according to claim 5, characterized in that: Step S6 of frequency difference estimation and compensation specifically includes: S6.1 estimates the frequency difference of the received signal using a frequency difference estimation algorithm, denoted as Δf est =f est_algo (r(t), l(t)), where Δf est Indicates the frequency difference value, f est_algo () represents the frequency difference estimation algorithm; S6.2 Set the frequency difference Δf est Input to the adaptive control algorithm to adjust the local carrier frequency; S6.3 Establish the error model of carrier frequency deviation, considering the influence of external environmental factor E on carrier frequency deviation, denoted as Δf error =f error_model (E), where Δf error represents the actual carrier frequency deviation error, that is, the actual difference between the carrier frequency of the received signal and the local carrier frequency, f error_model () represents the error model of carrier frequency deviation; S6.4 Update and correct the error model based on the real-time tracking results to obtain the compensated frequency difference value Δf comp =Δf est −Δf error .

7. A method for large-scale carrier frequency deviation capture and low steady-state error tracking according to claim 1, characterized in that: Step S7 phase noise suppression specifically includes: S7.1 uses an adaptive filter to suppress phase noise and reduce its impact on carrier tracking; S7.2 Set the received signal to , where s(t) is the baseband signal, ω0 is the carrier frequency, ϕ(t) is the phase noise, and n(t) is the noise; The input of S7.3 adaptive filter is the received signal r(t), and the output is the estimated phase noise ; S7.4 subtracts the estimated phase noise from the received signal to obtain a signal with suppressed phase noise for subsequent carrier tracking.

8. A method for large-range carrier frequency offset capture and low steady-state error tracking according to claim 1, characterized in that: Step S8 dual-loop carrier tracking specifically includes: S8.1 uses a second-order frequency-locked loop (FLL) to assist a third-order phase-locked loop (PLL) for carrier tracking. S8.2 adjusts the parameters of FLL and PLL in real time according to the tracking results.

Images