Performance optimization method of phase sensitive amplifier based on double common difference measurement and dynamic gain

Abstract

The application relates to the technical field of optical processing, and discloses a phase-sensitive amplifier performance optimization method based on double common difference measurement and dynamic gain, which comprises the following steps: splitting an input light beam to obtain a transmission light beam and a reflection light beam; performing double common difference measurement on the reflection light beam to obtain in-phase measurement results and quadrature measurement results, and generating a reservation criterion and a driving criterion according to the correlation between the in-phase measurement results and the quadrature measurement results; when the sample is normal, generating a feedforward driving amount according to the measurement results corresponding to the normal sample and current dynamic gain parameters, and performing phase-sensitive amplification processing on the transmission light beam; when the sample is split, performing amplitude limiting processing or gain reduction processing on the transmission light beam according to a restricted feedforward rule; when the sample is rejected, the rejected sample is prohibited from participating in feedforward; performing state monitoring to obtain an output deviation, and updating dynamic gain parameters and driving conditions according to the output deviation. The application improves the control accuracy, continuous stability and signal transmission consistency of the phase-sensitive amplification process.

Description

Technical Field

[0001] This invention relates to the field of optical processing technology, and more specifically, to a method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain. Background Technology

[0002] The development of continuous-variable quantum optics and coherent optical information processing technologies has made phase-sensitive amplifiers increasingly valuable in weak signal detection, quantum communication, and precision measurement. Phase-sensitive amplifiers can selectively amplify specific phase components, offering significant advantages in suppressing noise, improving the readability of effective signals, and enhancing long-distance transmission performance. Therefore, they are widely used in scenarios such as continuous-variable quantum key distribution, weak signal recovery, and high-sensitivity measurement. Especially in practical systems where transmission loss, environmental disturbances, and receiver noise are unavoidable, ensuring amplification performance while maintaining output stability and control accuracy has become a key focus of ongoing research in this field.

[0003] In existing technologies, such as the invention patent CN111541536A which discloses a continuous-variable quantum key distribution scheme enhanced by phase-sensitive amplification technology, a technical approach is to amplify the transmitted optical signal in phase and attenuate it out of phase using a phase-sensitive amplifier, and then demodulate and recover the amplified optical signal at the receiving end. This scheme can improve the receiving performance under lossy transmission conditions. However, the focus of this technology is mainly on improving the amplification link itself and the receiving recovery capability. It lacks a hierarchical processing mechanism for problems such as fluctuations in the reliability of measurement results during continuous processing, inconsistencies between control strength and sample state, and difficulty in timely adjustment of the amplification state when abnormal samples continue to appear. In practical systems, when the input state, link loss, or noise level changes dynamically, if a relatively uniform amplification and control method is still used, problems such as accumulated output deviation, increased fluctuations in the amplification state, and decreased consistency in the processing of adjacent samples are likely to occur, thus affecting the overall stability and continuous operation performance of the system.

[0004] Therefore, it is necessary to design a phase-sensitive amplifier performance optimization method based on dual-difference measurement and dynamic gain to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a phase-sensitive amplifier performance optimization method based on dual-difference measurement and dynamic gain, which aims to solve the problems of output deviation accumulation, amplification state fluctuation and continuous operation stability decline caused by the reliability fluctuation of measurement results, mismatch between control strength and sample state and the continuous occurrence of abnormal samples during continuous processing of phase-sensitive amplifiers.

[0006] This invention proposes a method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain, comprising:

[0007] The input beam is split to obtain a transmitted beam and a reflected beam;

[0008] The reflected beam is subjected to dual-difference measurement to obtain in-phase measurement results and orthogonal measurement results, and retention criteria and driving criteria are generated based on the correlation between the in-phase measurement results and the orthogonal measurement results;

[0009] When the retention criterion meets the retention condition and the driving criterion meets the driving condition, the current sample is determined as a normal sample. A feedforward driving quantity is generated based on the measurement result corresponding to the normal sample and the current dynamic gain parameter, and the transmitted beam is subjected to phase-sensitive amplification. When the retention criterion meets the retention condition but the driving criterion does not meet the driving condition, the current sample is determined as a split sample, and the transmitted beam is subjected to amplitude limiting or gain reduction processing according to the restricted feedforward rule. When the retention criterion does not meet the retention condition, the current sample is determined as a rejection sample, and the rejection sample is prohibited from participating in the feedforward.

[0010] The processed output beam is monitored to obtain the output deviation, and the dynamic gain parameters and driving conditions of the next sample are updated according to the output deviation.

[0011] Furthermore, the correlation relationship refers to the direction of change, relative amplitude, and consistency of continuous sampling between the in-phase measurement results and the orthogonal measurement results.

[0012] Furthermore, when generating the retention criteria and driving criteria, the following are included:

[0013] The retention criterion is generated based on the relative amplitude and the continuous sampling consistency, and the driving criterion is generated based on the change direction and the adaptation result of the relative amplitude relative to the current dynamic gain parameter.

[0014] Furthermore, when generating the feedforward driving quantity, the measurement results corresponding to the normal sample are converted into phase driving quantity and gain driving quantity, and the transmitted beam is subjected to phase-sensitive amplification processing in the order of performing phase correction first and then gain amplification.

[0015] Furthermore, when the current dynamic gain parameter is in the enhanced gain state, the gain driving amount is increased according to the incremental release rule only when the normal samples continuously meet the driving condition; otherwise, the current dynamic gain parameter is maintained or the state is switched to normal gain.

[0016] Furthermore, the split sample is the current sample in the gain switching interval, which is the processing interval where the current dynamic gain parameter is in the enhanced gain state and the gain driving amount corresponding to the current sample is greater than the reference driving amount corresponding to the current dynamic gain parameter.

[0017] Furthermore, when following the restricted feedforward rule, it includes:

[0018] Based on the degree of deviation between the driving criterion corresponding to the split sample and the driving condition, amplitude limiting or gain reduction processing is performed;

[0019] When the deviation is within a preset limiting range, the feedforward driving amount corresponding to the split sample is limited while maintaining the current dynamic gain parameter; when the deviation is within a preset gain reduction range, the current dynamic gain parameter is reduced and the transmitted beam is reduced according to the reduced dynamic gain parameter.

[0020] Furthermore, when the current sample is determined to be a rejected sample, the feedforward driving quantity is set to zero, and the current dynamic gain parameter remains unchanged; when the number of rejected samples reaches a preset number consecutively, the dynamic gain parameter corresponding to the next sample is switched to the normal gain state.

[0021] Furthermore, when updating the dynamic gain parameter and driving condition corresponding to the next sample based on the output deviation, the process includes:

[0022] The update method is selected based on the deviation direction and continuous deviation state of the output deviation; when the output deviation corresponds to a gain deviation, the dynamic gain parameter corresponding to the next sample is updated; when the output deviation corresponds to a drive deviation, the drive condition corresponding to the next sample is updated; when the output deviation continuously exceeds the preset deviation range, the upper limit of the dynamic gain parameter corresponding to the next sample is reduced and the drive condition is tightened.

[0023] Furthermore, when the output deviation continuously exceeds the preset deviation range and the current dynamic gain parameter is in the enhanced gain state, the dynamic gain parameter corresponding to the next sample is switched to the normal gain state, and the dynamic gain parameter corresponding to the next sample is prohibited from entering the enhanced gain state; when the output deviation corresponding to a preset number of consecutive samples is within the preset deviation range, the prohibition is lifted.

[0024] Compared with the prior art, the beneficial effects of the present invention are as follows: By performing dual-difference measurements on the reflected beam, the retention criterion and driving criterion are generated using the correlation between the in-phase measurement results and the orthogonal measurement results. Based on this, the current sample is classified into normal samples, split samples, and rejection samples. This allows the transmitted beam to complete the hierarchical judgment of sample validity and driving applicability before entering the phase-sensitive amplification process. Therefore, it can avoid the problems of misadjustment and misamplification caused by directly sending samples unsuitable for normal driving into the unified amplification link. For normal samples, the feedforward driving quantity is generated through the current dynamic gain parameters and phase-sensitive amplification processing is performed, which is beneficial to maintain high amplification efficiency and control accuracy when the sample state is stable. For split samples, the conventional full-quantity processing is no longer used. Instead of using a fixed method, it performs amplitude limiting or gain reduction processing, which helps suppress output overshoot, amplification fluctuations, and processing mismatch between consecutive samples caused by inconsistencies between sample states and driving conditions. For rejected samples, they are directly prohibited from participating in feedforward, preventing abnormal samples from continuously entering subsequent control chains and amplifying errors. At the same time, the output deviation is obtained by monitoring the output beam state, and the dynamic gain parameters and driving conditions corresponding to the next sample are updated using this output deviation, thereby forming a closed-loop adjustment. This can improve the continuous stability of the phase-sensitive amplification process, the targeted nature of amplification control, and the consistency of adjacent sample processing even when the input state changes, link loss fluctuates, or noise disturbances exist, thereby improving the overall signal transmission quality and system reliability. Attached Figure Description

[0025] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0026] Figure 1 The flowchart illustrates a method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain, as provided in an embodiment of the present invention. Detailed Implementation

[0027] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0028] In some embodiments of this application, see Figure 1 As shown, this application proposes a method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain, including:

[0029] S100: Splits the input beam to obtain a transmitted beam and a reflected beam.

[0030] S200: Performs dual-difference measurements on the reflected beam to obtain in-phase and orthogonal measurement results, and generates retention and driving criteria based on the correlation between the in-phase and orthogonal measurement results.

[0031] S300: When both the retention criterion and the driving criterion meet the retention and driving conditions, the current sample is designated as a normal sample. A feedforward driving quantity is generated based on the measurement results corresponding to the normal sample and the current dynamic gain parameters, and phase-sensitive amplification is applied to the transmitted beam. When the retention criterion meets the retention condition but the driving criterion does not meet the driving condition, the current sample is designated as a split sample, and the transmitted beam is subjected to amplitude limiting or gain reduction processing according to the restricted feedforward rule. When the retention criterion does not meet the retention condition, the current sample is designated as a rejection sample, and rejection samples are prohibited from participating in the feedforward.

[0032] S400: Monitors the status of the processed output beam, obtains the output deviation, and updates the dynamic gain parameters and driving conditions for the next sample based on the output deviation.

[0033] Specifically, this embodiment is applicable to phase-sensitive amplification processing scenarios where continuous samples are input sequentially. The input beam is the beam to be processed, the transmitted beam is the beam that enters the amplification processing path after beam splitting, the reflected beam is the beam that enters the measurement path after beam splitting, and the output beam is the result beam of the transmitted beam after processing. The current sample refers to the input beam sample corresponding to the current sampling period, and the next sample refers to the input beam sample corresponding to the adjacent sampling period after the current sample. The current dynamic gain parameter is used for processing determination and amplification control of the current sample, and the dynamic gain parameter corresponding to the next sample is used for feedback update results after the current sample is processed; the two are not used interchangeably.

[0034] It should be noted that, in this application, a sample refers to a processing unit obtained from a single sampling period, a single input pulse, or a continuous input beam at a preset sampling interval. Dual-difference measurement refers to the synchronous detection of two orthogonal measurement components of the reflected beam to obtain in-phase and orthogonal measurement results respectively. This application does not limit the specific optical path implementation, as long as it can achieve the synchronous acquisition of the two orthogonal measurement components. Dynamic gain parameters include at least the current gain state, reference drive quantity, allowed drive range, enhanced gain upper limit, and corresponding gain release rules. Split samples refer to samples where the retention criterion meets the retention condition but the drive criterion does not meet the drive condition, and therefore it is not appropriate to directly perform normal feedforward according to the current dynamic gain parameters.

[0035] In step S100, the input beam is split to obtain a transmitted beam and a reflected beam. A beam splitter can be used to split the input beam into two paths according to a preset splitting ratio. The transmitted beam enters the phase-sensitive amplification processing path, and the reflected beam enters the dual-difference measurement path. The splitting ratio can be set during the debugging stage according to the measurement accuracy requirements and amplification processing requirements, preserving as much effective energy of the transmitted beam as possible while ensuring that the reflected beam has a sufficient measurement signal-to-noise ratio. The purpose of beam splitting is to separate the measurement function from the processing function, allowing the reflected beam to undertake the state recognition task and the transmitted beam to undertake the actual amplification task, providing a basis for subsequent classification processing.

[0036] In step S200, a dual-difference measurement is performed on the reflected beam to obtain in-phase and orthogonal measurement results. Based on their correlation, retention and driving criteria are generated. Dual-difference measurement refers to the simultaneous measurement of the two orthogonal channel components of the reflected beam, thereby obtaining two sets of measurement results that can jointly characterize the current sample state. The correlation includes at least the direction of change, relative amplitude, and continuous sampling consistency. The direction of change refers to the increasing, decreasing, or essentially unchanged trend of the in-phase and orthogonal measurement results between adjacent sampling periods. The relative amplitude refers to the amplitude correspondence between the in-phase and orthogonal measurement results within the same sampling period. Continuous sampling consistency refers to the degree to which the above-mentioned trends and amplitude relationships remain stable within a preset continuous sampling window. To avoid misjudgment due to instantaneous noise or single-point anomalies, a rolling sampling window can be used for continuity determination. For example, the fluctuation of the in-phase and orthogonal measurement results can be statistically analyzed within an observation window consisting of 3 to 10 consecutive samples. When the fluctuation range is within the allowable range, the continuous sampling consistency is determined to meet the requirements; otherwise, it is determined not to meet the requirements. The allowable range can be determined based on the measurement noise distribution, historical sample statistics during the stable operating phase, and control accuracy requirements.

[0037] In this embodiment, the direction of change is obtained by comparing the differences between in-phase and quadrature measurement results of adjacent sampling periods. When the difference is greater than the positive change threshold, it is determined to be an increase. When the difference is less than the negative change threshold, it is determined to be a decrease. When the absolute value of the difference is not greater than the static stability fluctuation threshold, it is determined to be basically unchanged. To avoid ambiguity caused by threshold overlap, the positive change threshold, negative change threshold, and static stability fluctuation threshold are determined during the system calibration stage based on the no-load measurement noise, detector resolution, and fluctuation boundary of stable operating samples. The static stability fluctuation threshold is not greater than the upper limit of fluctuation of stable operating samples, and the positive and negative change thresholds are respectively greater than the static stability fluctuation threshold. The relative amplitude is obtained by comparing the absolute value or normalized amplitude of in-phase and quadrature measurement results within the same sampling period. Continuous sampling consistency is obtained by counting the number of times the change direction is consistent and the relative amplitude fluctuation range within a rolling window of 3 to 10 consecutive samples. When the number of times the direction is consistent reaches 70% to 100% of the number of samples in the window, and the relative amplitude fluctuation does not exceed the amplitude fluctuation threshold, the continuous sampling consistency is determined to meet the requirements.

[0038] Furthermore, the retention criterion is established when the following conditions are met simultaneously: the relative amplitude is within the preset retention amplitude range, and the consistency of continuous sampling meets the preset consistency condition. The driving criterion is established when the following conditions are met simultaneously: the direction of change meets the preset direction consistency requirement, and the relative amplitude falls within the allowable driving range corresponding to the current dynamic gain parameter. The lower limit of the preset retention amplitude range is used to eliminate samples with insufficient effective measurement energy, and the upper limit is used to eliminate samples that significantly exceed the linear measurement range. The allowable driving range is determined based on the linear amplification range corresponding to the current dynamic gain parameter, the target output deviation tolerance, and the statistical results of historical stable samples. Thus, the retention criterion is used to determine whether the current sample has retention value for entering the subsequent processing chain, and the driving criterion is used to determine whether the current sample is suitable for directly participating in feedforward driving generation.

[0039] Based on the above correlation, the retention criterion is used to determine whether the current sample has retention value for entering the subsequent processing chain, and the driving criterion is used to determine whether the current sample is suitable for direct participation in feedforward driving generation. Preferably, the retention criterion can be generated based on the relative amplitude and continuous sampling consistency, as these two indicators are more suitable for reflecting the overall stability of the sample. The driving criterion can be generated based on the direction of change and the adaptation result of the relative amplitude relative to the current dynamic gain parameter, as these two indicators are more suitable for reflecting whether the current sample is suitable for being directly driven by the current dynamic gain parameter. The adaptation result refers to the degree of matching between the measurement state of the current sample and the target driving range corresponding to the current dynamic gain parameter, which can be determined by comparing whether the relative amplitude falls within the allowable driving range corresponding to the current dynamic gain parameter and whether the direction of change is consistent with the expected enhancement processing.

[0040] In step S300, the current sample is classified according to the retention criterion and the driving criterion. When the retention criterion meets the retention condition and the driving criterion meets the driving condition, the current sample is determined as a normal sample. For normal samples, a feedforward driving quantity is generated based on the measurement results corresponding to the normal sample and the current dynamic gain parameters, and the transmitted beam is subjected to phase-sensitive amplification. Preferably, the feedforward driving quantity includes a phase driving quantity and a gain driving quantity, wherein the phase driving quantity is used to correct the phase offset of the transmitted beam relative to the target, and the gain driving quantity is used to control the amplification of the transmitted beam. During execution, the phase driving quantity can be generated first based on the in-phase measurement results and the orthogonal measurement results to perform phase correction on the transmitted beam, and then the gain driving quantity can be generated based on the current dynamic gain parameters and the measurement results corresponding to the normal sample to perform gain amplification on the transmitted beam. This order is adopted because if the phase deviation is not suppressed first, direct amplification may amplify invalid components, thereby increasing the output error.

[0041] In this embodiment, the measurement results corresponding to normal samples are preferably converted to phase drive and gain drive quantities through a pre-calibrated correspondence. Specifically, during the system calibration phase, a mapping table is formed for the optimal phase correction and optimal gain control quantities corresponding to different in-phase and quadrature measurement results. During the operation phase, the control unit retrieves or interpolates the phase drive and gain drive quantities from the mapping table based on the in-phase and quadrature measurement results corresponding to the current normal sample. The determination of phase correction completion is preferably: the phase error falls back to the preset phase tolerance range, or the phase modulation unit reaches the preset response time. The preset phase tolerance range is preferably determined based on the phase error statistics under stable operating conditions and the upper limit of allowable output deviation.

[0042] The current dynamic gain parameters include gain state, baseline drive amount, upper limit of allowed drive range, upper limit of enhanced gain, gain release threshold, exit threshold, and corresponding adjustment step size. The normal gain state corresponds to the basic amplification state during stable operation, while the enhanced gain state corresponds to the increased amplification state allowed when samples are continuously stable and output deviation is controlled. When the current dynamic gain parameter is in the enhanced gain state, the gain drive amount is increased according to an incremental release rule only when normal samples continuously meet the drive conditions and the corresponding output deviation does not exceed the preset deviation range. The preferred incremental release rule is: after 2 to 6 consecutive samples meet the drive conditions, the gain drive amount is increased by one level by a preset step size or preset ratio until the enhanced gain upper limit is reached. Before reaching the gain release threshold, the current gain drive amount remains unchanged. When the number of samples that continuously fail to meet the drive conditions reaches the exit threshold, or when the continuous output deviation exceeds the preset deviation range, the current dynamic gain parameter is switched back to the normal gain state, thereby reducing repeated oscillations in the enhanced gain state.

[0043] In this embodiment, "normal samples continuously satisfying the driving conditions" means that within a preset continuous sample window, all consecutive samples are determined to be normal samples and the corresponding driving criteria continuously satisfy the driving conditions. The preset continuous sample window preferably contains 2 to 6 consecutive samples, and its length can be determined based on the system response speed, sample fluctuation period, and misadjustment tolerance. The incremental release rule preferably includes at least one gain release threshold. When the number of samples continuously satisfying the driving conditions reaches the gain release threshold, the gain driving amount is increased from the current level to the next level. When the gain release threshold is not reached, the existing gain driving amount remains unchanged. In this embodiment, the count value of continuously satisfying the driving conditions is incremented by 1 when the current sample is determined to be a normal sample and the driving criteria satisfy the driving conditions. When subsequent samples no longer simultaneously satisfy the above conditions, the count value is cleared and counting restarts. To avoid frequent entry and exit from the enhanced gain state, an exit threshold can also be set. When the number of samples continuously failing to meet the driving conditions in the enhanced gain state reaches the exit threshold, the current dynamic gain parameter is switched to the normal gain state. The optimal gain release threshold and exit threshold are determined based on the output overshoot risk under enhanced gain state, the convergence speed of historical sample deviation, and the allowable recovery time.

[0044] When the retention criterion meets the retention condition but the driving criterion does not meet the driving condition, the current sample is identified as a split sample. A split sample refers to a sample that still has retention value but is not suitable for direct normal feedforward driving according to the current dynamic gain parameter. Further, split samples are preferably limited to current samples in the gain switching interval. The gain switching interval refers to the processing interval where the current dynamic gain parameter is in an enhanced gain state and the gain driving amount corresponding to the current sample is greater than the reference driving amount corresponding to the current dynamic gain parameter. The reference driving amount can be understood as the basic driving level that can be maintained for a long time when the current dynamic gain parameter is at the stable operating boundary, and can be determined based on the upper limit of the linear amplification interval, the average driving level of the continuous stable operating interval, and the error tolerance.

[0045] For split samples, the full feedforward processing of normal samples is no longer directly applied. Instead, the transmitted beam is subjected to amplitude limiting or gain reduction processing according to the limited feedforward rule. Specifically, the degree of deviation is determined by comprehensively considering at least two of the following indicators: the degree of deviation between the changing direction and the preset direction consistency requirement, the magnitude of the relative amplitude exceeding the boundary of the allowable drive interval, and the degree of continuous sampling consistency falling below the preset consistency condition. For ease of engineering implementation, the degree of deviation can be divided into three levels during the system calibration phase: slight deviation, moderate deviation, and significant deviation. Slight deviation corresponds to the preset amplitude limiting interval, while moderate and significant deviations correspond to the preset gain reduction interval. For split samples within the preset amplitude limiting interval, the feedforward drive amount of the current sample is limited to a range not exceeding the reference drive amount or not exceeding the preset upper limit of the current feedforward drive amount. For split samples within the preset gain reduction interval, the current dynamic gain parameter is reduced by at least one gain level and maintained for one to several sample processing cycles after the reduction. Then, the decision to restore a higher gain state is made based on whether subsequent samples meet the drive conditions again. The boundaries between the preset amplitude limiting range and the preset gain reduction range can be determined based on the statistical results of historical mismatch samples, output overshoot risk, and allowable recovery time. To avoid continuous instability of the enhanced gain state caused by continuous split samples, a continuous split sample counting rule can also be set. When the number of consecutive split samples reaches a preset number, the reduced dynamic gain parameter is maintained until the continuous samples meet the driving conditions again, and then the determination of the enhanced gain state is restored.

[0046] In this embodiment, the preset limiting interval and preset gain reduction interval are preferably set in continuous segments, covering the deviation range corresponding to the split samples. The preset limiting interval corresponds to a slight mismatch state, and the preset gain reduction interval corresponds to a significant mismatch state, thereby avoiding unclassifiable ambiguous samples during operation. For split samples in the preset limiting interval, the limiting process is preferably implemented by setting an upper limit on the feedforward drive amount of the current sample, and the upper limit is preferably not greater than the reference drive amount corresponding to the current dynamic gain parameter. For split samples in the preset gain reduction interval, after the current dynamic gain parameter is reduced, it is preferably maintained for at least one to several sample processing cycles, and then a decision is made on whether to restore a higher gain state based on the subsequent sample state. The minimum hold period is preferably determined based on the output deviation convergence speed, the device setup time for dynamic gain parameter switching, and the system's allowed recovery time.

[0047] When the retention criteria do not meet the retention conditions, the current sample is identified as a rejected sample, and the rejected sample is prohibited from participating in the feedforward. A rejected sample refers to a sample whose overall measurement stability is insufficient, and whose continued participation in control could easily introduce significant errors. For rejected samples, the feedforward drive is set to zero, and the current dynamic gain parameter remains unchanged, thereby avoiding instantaneous jumps in the processing chain due to abnormal samples. In this embodiment, setting the feedforward drive to zero means that neither the phase drive component nor the gain drive component corresponding to the current sample is output to the modulation execution unit, thus preventing the current sample from triggering active feedforward adjustment of the transmitted beam. If the number of rejected samples reaches a preset number consecutively, the dynamic gain parameter corresponding to the next sample is switched to the normal gain state to avoid maintaining the enhanced gain state when abnormal samples occur consecutively. This preset number can be determined based on the cumulative risk of output deviation when abnormal samples occur consecutively, and is preferably determined by the debugging phase and historical operating statistics. By processing rejected samples and split samples separately, it is possible to avoid simply merging all abnormal situations into the same degradation path, thereby improving the targeting of the processing strategy.

[0048] In this embodiment, the consecutive rejection sample count is preferably incremented by 1 when the current sample is determined to be a rejection sample, and reset to zero or restored to the initial state when subsequent samples are no longer determined to be rejection samples. This ensures that the preset count represents the length of consecutive rejection samples, rather than the cumulative number of occurrences. In this way, isolated anomalies and persistent anomalies can be distinguished and handled, preventing the system from prematurely exiting the enhanced gain state due to occasional rejection samples.

[0049] In step S400, the processed output beam is monitored to obtain the output deviation, and the dynamic gain parameters and driving conditions corresponding to the next sample are updated based on the output deviation. Output deviation refers to the degree of deviation of the current output beam from the target output state. In this embodiment, the output deviation is determined by the amplitude deviation, phase deviation, and response timing deviation between the output beam and the target output state. If the phase deviation is within the allowable range but the amplitude deviation is continuously large or small, it is determined to be a gain deviation. If the amplitude deviation is within the allowable range but the phase deviation, response timing deviation, or consistency deviation after feedforward execution continuously exceeds the limit, it is determined to be a driving deviation. If both the amplitude deviation and phase deviation continuously exceed the limit, it is preferentially processed by simultaneously tightening the upper limit of the dynamic gain parameters and the driving conditions. The continuous deviation state is determined by the deviation direction and amplitude of 2 to 8 consecutive samples. When the deviation direction remains consistent and the deviation amplitude continuously exceeds the preset deviation range, the continuous deviation state is determined to be established. When the output deviation corresponds to a gain deviation, the dynamic gain parameters corresponding to the next sample are updated. When the output deviation corresponds to a driving deviation, the driving conditions corresponding to the next sample are updated. When the output deviation continuously exceeds the preset deviation range, the upper limit of the dynamic gain parameter corresponding to the next sample is simultaneously reduced and the driving conditions are tightened. The dynamic gain parameter is preferably updated stepwise according to a preset adjustment step size. The tightening of the driving conditions preferably includes narrowing the allowable driving range, increasing the consistency requirements of continuous sampling, and reducing the tolerance of the change direction.

[0050] In this embodiment, the continuous deviation state is preferably determined through a continuous deviation counting window. A continuous deviation state is determined when the output deviation direction remains consistent for 2 to 8 consecutive samples, and the deviation amplitude continuously exceeds a preset deviation range. When the output deviation corresponds to a gain deviation, the dynamic gain parameter corresponding to the next sample can be adjusted upwards, downwards, or remain unchanged according to a preset adjustment step size. When the output deviation corresponds to a drive deviation, the drive condition corresponding to the next sample can be updated by narrowing the allowable drive range, tightening the tolerance of the change direction, or increasing the continuous sampling consistency requirements. The preset adjustment step size is preferably determined based on the relationship between the output deviation and parameter correction, the convergence speed of historical sample deviations, and the target recovery time.

[0051] In this embodiment, the target output state is preferably determined based on the current dynamic gain parameter, the measurement results corresponding to normal samples, and the target response relationship established during the system calibration phase. Gain deviation preferably refers to a deviation state where the actual amplification intensity of the output beam deviates from the target amplification intensity while the phase response remains within the allowable range. Drive deviation preferably refers to a deviation state where the phase response, drive response timing, or feedforward execution result of the output beam deviates from the target state while the overall amplification intensity remains within a relatively acceptable range. When the output deviation continuously exceeds the preset deviation range and the current dynamic gain parameter is in the enhanced gain state, after the dynamic gain parameter corresponding to the next sample switches to the normal gain state, it is preferably allowed to re-enter the enhanced gain state only when the output deviations corresponding to a consecutive preset number of samples all fall back to the preset deviation range. The consecutive preset number is preferably determined based on the stable observation length required to recover from the normal gain state to the enhanced gain state, the convergence characteristics of the output deviation, and the risk of false recovery. In this embodiment, the continuous recovery count value used to de-prohibit is incremented by 1 when the current sample output deviation is within the preset deviation range. When the output deviation corresponding to any sample exceeds the preset deviation range again, the continuous recovery count value is reset to zero and counting restarts.

[0052] In some embodiments of this application, preset positive change threshold, preset negative change threshold, static stability fluctuation threshold, amplitude fluctuation threshold, preset retained amplitude range, allowed drive range, gain release threshold, exit threshold, preset amplitude limiting range, preset gain reduction range, preset deviation range, and continuous sample window length can all be determined jointly through system calibration and historical operating data statistics. Specifically, noise samples can be collected during the no-load phase, normal samples during the stable operating phase, and mismatch samples during the disturbed phase. The fluctuation boundaries, deviation distribution, and recovery time of each type of sample are statistically analyzed, and then combined with the target amplification accuracy, the linear operating range of the device, and the upper limit of the allowable output overshoot to determine each preset parameter. When updating parameters, fixed parameter table calls, graded parameter calls, or adaptive fine-tuning based on the statistical results of the most recent continuous samples can be used, but all should meet the linear operating range of the device and the target output deviation tolerance.

[0053] Specifically, in a continuous input sample sequence, retention and driving criteria are generated based on the dual-difference measurement results of the reflected beam. The current sample is then divided into three categories according to these two criteria. If the current sample is stable and suitable for driving, it enters the normal sample path, first correcting the phase and then performing gain amplification. If the current sample can be retained but is not yet suitable for normal driving, it enters the split sample path, selecting either amplitude limiting or gain reduction processing based on the degree of deviation. If the overall stability of the current sample is insufficient, it enters the rejection sample path and directly exits the feedforward. After the current sample is processed, the output beam's state is immediately monitored, and the obtained output deviation is used to correct the dynamic gain parameters and driving conditions corresponding to the next sample. Therefore, the sample state at any given moment is not processed in isolation; instead, the measurement information, processing results, and output feedback of the current sample are continuously linked, enabling sample classification, amplification control, and parameter updates to work synergistically in the continuous sample sequence.

[0054] Understandably, by performing dual-difference measurements on the reflected beam, and utilizing the correlation between in-phase and orthogonal measurement results to generate retention and driving criteria, the current sample is then classified into normal, split, and rejected samples. This allows the transmitted beam to complete the stratified judgment of sample validity and driving suitability before entering the phase-sensitive amplification process. Therefore, it avoids the problems of misadjustment and misamplification caused by directly sending samples unsuitable for normal driving into the unified amplification link. For normal samples, the feedforward driving quantity is generated using the current dynamic gain parameters and phase-sensitive amplification is performed, which helps maintain high amplification efficiency and control accuracy when the sample state is stable. For split samples, the conventional full-quantity processing method is no longer used; instead, a separate process is performed. Limiting or reducing gain helps suppress output overshoot, amplification fluctuations, and processing mismatch between consecutive samples caused by inconsistencies between sample states and driving conditions. For rejected samples, participation in feedforward is directly prohibited, preventing abnormal samples from continuously entering subsequent control chains and amplifying errors. Simultaneously, output deviation is obtained by monitoring the output beam state, and this output deviation is used to update the dynamic gain parameters and driving conditions corresponding to the next sample, thus forming a closed-loop adjustment. This can improve the continuous stability of the phase-sensitive amplification process, the targeted nature of amplification control, and the consistency of adjacent sample processing even with changes in input state, fluctuations in link loss, or noise disturbances, thereby improving the overall signal transmission quality and system reliability.

[0055] 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 it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain, characterized in that, include: The input beam is split to obtain a transmitted beam and a reflected beam; The reflected beam is subjected to dual-difference measurement to obtain in-phase measurement results and orthogonal measurement results, and retention criteria and driving criteria are generated based on the correlation between the in-phase measurement results and the orthogonal measurement results; The correlation includes at least the direction of change, relative amplitude, and consistency of continuous sampling; The direction of change is obtained by comparing the differences between in-phase and orthogonal measurement results of adjacent sampling periods. The relative amplitude is obtained by comparing the absolute value or normalized amplitude of in-phase measurement results and orthogonal measurement results within the same sampling period; The consistency of continuous sampling is obtained by statistically analyzing the number of times the change direction is consistent and the relative amplitude fluctuation range within a rolling window of 3 to 10 consecutive samples. The retention criterion is established when the following conditions are met simultaneously: the relative amplitude is within the preset retention amplitude range, and the consistency of continuous sampling meets the preset consistency condition; The driving criterion is established when the following conditions are met simultaneously: the direction of change meets the preset direction consistency requirement, and the relative amplitude falls within the allowable driving range corresponding to the current dynamic gain parameter; The lower limit of the preset retained amplitude range is used to remove samples with insufficient effective measurement energy, and the upper limit is used to remove samples that significantly exceed the linear measurement range; the allowed driving range is determined based on the linear amplification range corresponding to the current dynamic gain parameter, the target output deviation tolerance, and the statistical results of historical stable samples. When the retention criterion satisfies the retention condition and the driving criterion satisfies the driving condition, the current sample is determined as a normal sample. Based on the measurement results corresponding to the normal sample and the current dynamic gain parameters, a feedforward driving quantity is generated to perform phase-sensitive amplification processing on the transmitted beam. When the retention criterion meets the retention condition but the driving criterion does not meet the driving condition, the current sample is determined as a split sample, and the transmitted beam is subjected to amplitude limiting or gain reduction processing according to the restricted feedforward rule; when the retention criterion does not meet the retention condition, the current sample is determined as a rejection sample, and the rejection sample is prohibited from participating in the feedforward. The processed output beam is monitored to obtain the output deviation, and the dynamic gain parameters and driving conditions of the next sample are updated according to the output deviation.

2. The method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain according to claim 1, characterized in that, The correlation is the direction of change, relative amplitude, and consistency of continuous sampling between the in-phase measurement results and the orthogonal measurement results.

3. The method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain according to claim 2, characterized in that, When generating the retention criteria and driving criteria, the following are included: The retention criterion is generated based on the relative amplitude and the continuous sampling consistency, and the driving criterion is generated based on the change direction and the adaptation result of the relative amplitude relative to the current dynamic gain parameter.

4. The method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain according to claim 1, characterized in that, When generating the feedforward driving quantity, the measurement results corresponding to the normal sample are converted into phase driving quantity and gain driving quantity, and the transmitted beam is subjected to phase-sensitive amplification processing in the order of performing phase correction first and then gain amplification.

5. The method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain according to claim 4, characterized in that, When the current dynamic gain parameter is in the enhanced gain state, the gain driving amount is increased according to the incremental release rule only when the normal samples continuously meet the driving condition; otherwise, the current dynamic gain parameter is maintained or the state is switched to normal gain.

6. The method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain according to claim 1, characterized in that, The split sample is the current sample that is in the gain switching interval. The gain switching interval is the processing interval in which the current dynamic gain parameter is in the enhanced gain state and the gain driving amount corresponding to the current sample is greater than the reference driving amount corresponding to the current dynamic gain parameter.

7. The method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain according to claim 6, characterized in that, When following the restricted feedforward rule, including: Based on the degree of deviation between the driving criterion corresponding to the split sample and the driving condition, amplitude limiting or gain reduction processing is performed; When the deviation is within a preset limiting range, the feedforward driving amount corresponding to the split sample is limited while maintaining the current dynamic gain parameter; when the deviation is within a preset gain reduction range, the current dynamic gain parameter is reduced and the transmitted beam is reduced according to the reduced dynamic gain parameter.

8. The method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain according to claim 1, characterized in that, When the current sample is determined to be a rejected sample, the feedforward driving quantity is set to zero, and the current dynamic gain parameter remains unchanged; when the number of rejected samples reaches a preset number consecutively, the dynamic gain parameter corresponding to the next sample is switched to the normal gain state.

9. The method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain according to claim 1, characterized in that, When updating the dynamic gain parameter and driving condition corresponding to the next sample based on the output deviation, the following are included: The update method is selected based on the deviation direction and continuous deviation state of the output deviation; when the output deviation corresponds to a gain deviation, the dynamic gain parameter corresponding to the next sample is updated; when the output deviation corresponds to a drive deviation, the drive condition corresponding to the next sample is updated; when the output deviation continuously exceeds the preset deviation range, the upper limit of the dynamic gain parameter corresponding to the next sample is reduced and the drive condition is tightened.

10. The method for optimizing the performance of a phase-sensitive amplifier based on dual-difference measurement and dynamic gain according to claim 9, characterized in that, When the output deviation continuously exceeds the preset deviation range and the current dynamic gain parameter is in the enhanced gain state, the dynamic gain parameter corresponding to the next sample is switched to the normal gain state, and the dynamic gain parameter corresponding to the next sample is prohibited from entering the enhanced gain state; when the output deviation corresponding to a preset number of consecutive samples is within the preset deviation range, the prohibition is lifted.

Images