Loop gain adaptive adjustment system for digital pre-distortion

Abstract

The application discloses a loop gain self-adaptive adjusting system of digital pre-distortion, and relates to the technical field of loop adjusting.The technical scheme points comprise: obtaining loop real-time characteristic data of a digital pre-distortion system; judging a loop mismatch state of the digital pre-distortion system based on the loop real-time characteristic data; if the loop mismatch state is a loop non-mismatch state, maintaining the loop gain of the digital pre-distortion system as a reference adaptive value; performing gradient self-adaptive adjustment on the loop gain according to a target gain gradient adjusting step; judging a loop mismatch recovery state of the digital pre-distortion system based on gain adjustment monitoring data; if the loop mismatch recovery state is a mismatch complete recovery state, locking the loop gain as an optimal adaptive value; if the loop mismatch recovery state is a mismatch non-recovery state, continuing to perform iterative adjustment on the loop gain according to the target gain gradient adjusting step, and the effect is to guarantee efficient and stable operation in a steady state.

Description

Technical Field

[0001] This invention relates to the field of loop adjustment technology, and more specifically, to a loop gain adaptive adjustment system for digital predistortion. Background Technology

[0002] Digital predistortion technology, as a means to suppress nonlinear distortion in RF power amplifiers, is widely used in scenarios that improve power amplifier efficiency and ensure the quality of linear signal transmission. Loop gain, as a key control parameter of the digital predistortion system, directly determines the convergence speed, adaptive tracking capability, and final linearization effect of the predistortion model. Traditional digital predistortion systems often use fixed reference values ​​or manual coarse-tuning for the loop gain, which has gradually revealed significant limitations under complex real-world conditions. When the characteristics of the system input signal change dynamically, a fixed gain cannot adapt to the changed loop state, easily leading to loop mismatch problems: if the loop gain is too high, the predistortion model will overfit the instantaneous distortion characteristics, introducing additional nonlinear distortion and even causing system oscillations; if the loop gain is too low, the predistortion model will converge slowly, failing to track the distortion changes of the power amplifier in time, resulting in a significant decrease in linearization effect and making it difficult to meet the linearity requirements of signal transmission. Summary of the Invention

[0003] To address the shortcomings of existing technologies, the present invention aims to provide a digital predistortion loop gain adaptive adjustment system.

[0004] To achieve the above objectives, the present invention provides the following technical solution: A digital predistortion loop gain adaptive adjustment system includes: Acquisition module: Acquires real-time loop characteristic data of the digital predistortion system; First judgment module: judges the loop mismatch status of the digital predistortion system based on real-time loop feature data; Processing module: If the loop mismatch state is the loop no mismatch state, then the loop gain of the digital predistortion system is kept at the baseline adaptation value; if the loop mismatch state is the loop mismatch state, then the loop gain is adjusted to the temporary adjustment value and the gain adjustment monitoring data and loop linearity monitoring data are obtained. The first adjustment module evaluates the loop linearity status of the digital predistortion system based on loop linearity monitoring data, determines the target gain gradient adjustment step size based on the loop linearity status, and performs gradient-based adaptive adjustment of the loop gain based on the target gain gradient adjustment step size. The second judgment module: judges the loop mismatch recovery status of the digital predistortion system based on gain adjustment monitoring data; The second adjustment module: If the loop mismatch recovery state is a fully recovered mismatch state, the loop gain is locked to the optimal fit value; if the loop mismatch recovery state is a partially recovered mismatch state, the loop gain is iteratively adjusted according to the target gain gradient adjustment step size.

[0005] The real-time characteristic data of the loop includes the characteristic data of the loop input signal and the characteristic data of the loop feedback signal.

[0006] The loop input signal characteristic data includes the peak-to-average power ratio of the input signal, the bandwidth value of the input signal, and the acquisition timestamps corresponding to each characteristic parameter; The loop feedback signal characteristic data includes the feedback signal distortion, the feedback signal power value, and the acquisition timestamps corresponding to each characteristic parameter.

[0007] Determining the loop mismatch status of a digital predistortion system based on real-time loop feature data includes the following steps: Set the input feature threshold range and the input anomaly duration threshold; If the peak-to-average power ratio (PAPR) or bandwidth of the input signal is not within the corresponding input feature threshold range, then obtain the duration for which the PAPR or bandwidth exceeds the corresponding input feature threshold range. If the duration is greater than the input abnormal duration threshold, then determine that the PAPR or bandwidth is abnormal. If the peak-to-average power ratio (PAPR) and bandwidth of the input signal are both within the corresponding input feature threshold range, then the slope and fluctuation characteristic parameters of the PAPR and bandwidth are obtained based on the acquisition timestamps corresponding to the PAPR and bandwidth. Set the threshold for the input feature curve parameters; If the slope of change and the fluctuation characteristic parameters meet the threshold of the input characteristic curve parameters, then the input characteristic state is judged to be normal. If the slope of change and the fluctuation characteristic parameters do not meet the threshold of the input characteristic curve parameters, then the input characteristic state is judged to be abnormal.

[0008] The evaluation of the loop linearity status of a digital predistortion system based on loop linearity monitoring data includes the following steps: The loop linear response characteristics are extracted from the loop linearity monitoring data. These characteristics include the amplitude variation and phase shift trend of the loop signal transmission process. Determine whether there is nonlinear distortion in the loop signal transmission based on the loop's linear response characteristics; If there is no nonlinear distortion, then the loop linearity state is determined to be a good state; If nonlinear distortion exists, the loop linearity level is determined by combining the influence range and severity of the nonlinear distortion to obtain the loop linearity state of the digital predistortion system.

[0009] Determining the target gain gradient adjustment step size based on the loop linearity state includes the following steps: Determine the degree of anomaly in the linearity of the loop based on its linearity status; Set the corresponding gain adjustment sensitivity according to the degree of loop linearity anomaly; Based on the initial gradient adjustment step size corresponding to gain adjustment sensitivity matching; The amplitude variation pattern and phase shift trend of the loop linear response characteristics are determined, and the target gain gradient adjustment step size is obtained by dynamically calibrating the initial gradient adjustment step size based on the amplitude variation magnitude and phase shift rate.

[0010] The loop gain is adaptively adjusted in a gradient manner based on the target gain gradient adjustment step size, specifically including the following steps: The initial adjustment loop gain is obtained by performing the first gradient adjustment on the current loop gain according to the target gain gradient adjustment step size, and the initial adjustment feedback information is obtained according to the change trend of the loop gain. Based on the changing trend of the initial adjustment feedback information, the deviation tendency of the loop gain from the optimal fit state after the initial adjustment is judged, and the gradient adjustment direction is obtained. The target gain gradient adjustment step size is dynamically calibrated based on the initial adjustment feedback information to obtain the calibration adjustment step size. The initial adjustment loop gain is adaptively adjusted using a gradient method according to the gradient adjustment direction and calibration adjustment step size.

[0011] Determining the loop mismatch recovery status of a digital predistortion system based on gain adjustment monitoring data includes the following steps: Extract gain fluctuation information and loop transmission characteristic information from gain adjustment monitoring data; The gain fluctuation information is smoothed to obtain a stable gain fluctuation curve, and the loop transmission characteristic information is sorted out to obtain the loop transmission response result; The stable gain fluctuation curve is compared with the preset gain stability threshold range to obtain the gain fluctuation comparison result. The loop transmission response result is compared with the preset normal transmission response standard to obtain the transmission response comparison result. If the gain fluctuation curve is within the preset gain stability threshold range, and the loop transmission response result is consistent with the preset normal transmission response standard, then the loop mismatch recovery state is determined to be the mismatch complete recovery state. If the gain fluctuation curve exceeds the preset gain stability threshold range, or if the loop transmission response result is inconsistent with the preset normal transmission response standard, the loop mismatch recovery state is determined to be a mismatch unrecovered state.

[0012] If the loop mismatch recovery state is a complete mismatch recovery state, then the loop gain is locked to the optimal fit value, specifically including the following steps: If the loop mismatch recovery state is the complete mismatch recovery state, monitor the fluctuation of the loop gain within a preset time period and determine whether the fluctuation is within a preset stable range; If the fluctuation is within the preset stable range, then the current loop gain is confirmed to be the optimal fit value, and the loop gain locking mechanism is activated. If the fluctuation exceeds the preset stable range, the current loop gain is calibrated until the loop gain fluctuation is within the preset stable range. When the loop gain fluctuation is within the preset stable range, the current loop gain is determined as the optimal fit value, and the loop gain locking mechanism is activated.

[0013] If the loop mismatch recovery state is not the mismatch recovery state, then the loop gain is iteratively adjusted according to the target gain gradient adjustment step size, specifically including the following steps: If the loop mismatch recovery state is the mismatch unrecovered state, the loop gain is iteratively adjusted based on the target gain gradient adjustment step size. During the iterative adjustment process, the changing trend of the current loop mismatch degree is judged based on the gain adjustment monitoring data and the loop linearity monitoring data. The adjustment direction of the target gain gradient adjustment step size is calibrated according to the changing trend of the mismatch degree, and the loop gain is iteratively adjusted according to the adjustment direction.

[0014] Compared with the prior art, the present invention has the following beneficial effects: This invention distinguishes loop mismatch states based on real-time loop characteristic data. The differentiated control strategy of the processing module achieves a smooth transition between steady-state maintenance and mismatch triggering. In the loop mismatch-free state, it maintains the baseline fit value, balancing system linearity and transmission efficiency, and avoiding additional disturbances introduced by frequent adjustments. In the loop mismatch state, the gain is adjusted to a temporary adjustment value, and gain adjustment monitoring data and loop linearity monitoring data are collected. This quickly breaks the mismatch balance, providing the initiation conditions for adaptive adjustment, and provides complete performance feedback data for the adjustment process. Gradient adaptive adjustment based on loop linearity states achieves precise and adaptive gain adjustment. By evaluating the degree of linearity anomaly to match the adjustment sensitivity and initial gradient step size, and combining amplitude changes and phase shifts to dynamically calibrate the step size, the adjustment step size accurately adapts to the current loop distortion degree and dynamic characteristics, avoiding over-adjustment or under-adjustment problems caused by fixed step size adjustment. Gradient adjustment gradually approaches the optimal fit state through initial adjustment feedback calibration direction and step size, ensuring signal transmission stability. Intelligent control of the loop gain of the digital predistortion system is achieved, which not only ensures efficient and stable operation in steady state, but also improves the ability to quickly repair mismatch scenarios and effectively reduces signal nonlinear distortion. Attached Figure Description

[0015] Figure 1 A schematic diagram of a loop gain adaptive adjustment system for digital predistortion is provided for an embodiment of the present invention; Figure 2 This invention provides a schematic diagram illustrating the steps for obtaining the target gain gradient adjustment step size in a loop gain adaptive adjustment system for digital predistortion. Detailed Implementation

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0017] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0018] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0019] Reference Figures 1-2 As shown.

[0020] The embodiments further illustrate the loop gain adaptive adjustment system for digital predistortion proposed in this invention.

[0021] A digital predistortion loop gain adaptive adjustment system includes: Acquisition module: Acquires real-time loop characteristic data of the digital predistortion system; First judgment module: judges the loop mismatch status of the digital predistortion system based on real-time loop feature data; Processing module: If the loop mismatch state is the loop no mismatch state, then the loop gain of the digital predistortion system is kept at the baseline adaptation value; if the loop mismatch state is the loop mismatch state, then the loop gain is adjusted to the temporary adjustment value and the gain adjustment monitoring data and loop linearity monitoring data are obtained. The loop mismatch state is used as the criterion for differentiated control of the loop gain. When the loop mismatch state is zero, the loop gain of the digital predistortion system will be maintained at a preset baseline adaptation value. This baseline adaptation value is the optimal initial gain parameter obtained through multiple calibrations under stable operating conditions. It can ensure that the loop maintains stable signal transmission performance in zero-mismatch scenarios and avoids introducing additional disturbances due to unnecessary gain adjustments. For example, in a zero-mismatch scenario where the peak-to-average power ratio of the system input signal is stable in the range of 6 to 8, the input signal bandwidth is maintained in the range of 20MHz to 40MHz, and the feedback signal distortion is less than 0.5%, the loop gain will be maintained at the baseline adaptation value of 0.8 to ensure a balance between system linearity and transmission efficiency.

[0022] When a loop mismatch occurs, a temporary loop gain adjustment will be performed first. This adjustment will set the current loop gain to a temporary value, typically 1.1 to 1.3 times the baseline fit value. This quickly breaks the gain imbalance without causing drastic performance fluctuations due to excessive adjustment. After the temporary adjustment, the acquisition of gain adjustment monitoring data and loop linearity monitoring data will be triggered. The gain adjustment monitoring data includes the real-time value of the loop gain after adjustment, the rate of gain change, and the amplitude of gain fluctuations. The loop linearity monitoring data includes the amplitude deviation, phase offset, and proportion of nonlinear distortion components during loop signal transmission. When the system triggers a loop mismatch state due to a sudden increase in the peak-to-average power ratio of the input signal to 12 and lasting for more than 5 seconds, the processing module adjusts the loop gain from the baseline adaptation value of 0.8 to the temporary adjustment value of 1.0. Subsequently, it collects the gain fluctuation data and loop linearity data within 10 seconds after the gain adjustment. The gain change rate = (current gain - temporary adjustment value) / adjustment time, and the amplitude deviation = (output amplitude - ideal amplitude) / ideal amplitude, providing data basis for the adaptive adjustment process.

[0023] The first adjustment module evaluates the loop linearity status of the digital predistortion system based on loop linearity monitoring data, determines the target gain gradient adjustment step size based on the loop linearity status, and performs gradient-based adaptive adjustment of the loop gain based on the target gain gradient adjustment step size. The second judgment module: judges the loop mismatch recovery status of the digital predistortion system based on gain adjustment monitoring data; The second adjustment module: If the loop mismatch recovery state is a fully recovered mismatch state, the loop gain is locked to the optimal fit value; if the loop mismatch recovery state is a partially recovered mismatch state, the loop gain is iteratively adjusted according to the target gain gradient adjustment step size.

[0024] The real-time characteristic data of the loop includes the characteristic data of the loop input signal and the characteristic data of the loop feedback signal.

[0025] The loop input signal characteristic data includes the peak-to-average power ratio of the input signal, the bandwidth value of the input signal, and the acquisition timestamps corresponding to each characteristic parameter; The loop feedback signal characteristic data includes the feedback signal distortion, the feedback signal power value, and the acquisition timestamps corresponding to each characteristic parameter.

[0026] Determining the loop mismatch status of a digital predistortion system based on real-time loop feature data includes the following steps: Set the input feature threshold range and the input anomaly duration threshold; If the peak-to-average power ratio (PAPR) or bandwidth of the input signal is not within the corresponding input feature threshold range, then obtain the duration for which the PAPR or bandwidth exceeds the corresponding input feature threshold range. If the duration is greater than the input abnormal duration threshold, then determine that the PAPR or bandwidth is abnormal. If the peak-to-average power ratio (PAPR) and bandwidth of the input signal are both within the corresponding input feature threshold range, then the slope and fluctuation characteristic parameters of the PAPR and bandwidth are obtained based on the acquisition timestamps corresponding to the PAPR and bandwidth. Set the threshold for the input feature curve parameters; If the slope of change and the fluctuation characteristic parameters meet the threshold of the input characteristic curve parameters, then the input characteristic state is judged to be normal. If the slope of change and the fluctuation characteristic parameters do not meet the threshold of the input characteristic curve parameters, then the input characteristic state is judged to be abnormal.

[0027] The input feature threshold range and the input anomaly duration threshold are the core benchmarks for determining whether the input signal state is abnormal. The input feature threshold range corresponds to the normal operating range of the input signal peak-to-average power ratio (PAPR) and input signal bandwidth, respectively. The input anomaly duration threshold is used to filter out false judgments caused by transient disturbances, ensuring the stability of the state judgment. For example, the preset threshold range for the input signal PAPR is 6 to 8, the threshold range for the input signal bandwidth is 20MHz to 40MHz, and the input anomaly duration threshold is 5 seconds.

[0028] When the peak-to-average power ratio (PAPR) or bandwidth of the input signal is outside the corresponding input characteristic threshold range, a duration statistics process will be initiated to record the continuous duration for which the parameter exceeds the corresponding threshold range. If the duration exceeds the preset input abnormality duration threshold, the corresponding PAPR or bandwidth value is determined to be abnormal. If the PAPR rises to 10 at a certain moment, exceeding the preset threshold range of 6 to 8, and this over-threshold state lasts for 6 seconds, it will be determined to be abnormal because it exceeds the input abnormality duration threshold. If the PAPR only momentarily exceeds the threshold range and then quickly recovers, with a duration of 2 seconds, it will not be determined to be an abnormal state, thus avoiding misjudgment caused by momentary interference.

[0029] When both the peak-to-average ratio (PAR) and bandwidth of the input signal are within their corresponding input characteristic threshold ranges, the slope and fluctuation characteristic parameters of the PAR and bandwidth are calculated based on their corresponding acquisition timestamps to determine the dynamic trend of the parameters. The slope is calculated as (current parameter value - previous acquisition time value) / (current acquisition timestamp - previous acquisition timestamp). Within a 10-second acquisition period, the average PAR of the input signal is 7. The fluctuation characteristic parameters are obtained by calculating the sum of the squares of the differences between the PAR and the average value at each time point, dividing by the number of acquisitions, and then taking the square root. The slope is then calculated by combining the ratio of the PAR difference between adjacent times with the time difference.

[0030] The input characteristic curve parameter thresholds correspond to the allowable ranges of the slope and fluctuation characteristic parameters, respectively, and are used to determine whether the dynamic changes of the input signal are in a stable state. If both the slope and fluctuation characteristic parameters meet the input characteristic curve parameter thresholds, the input characteristic state is determined to be normal; if the slope or fluctuation characteristic parameter does not meet the input characteristic curve parameter thresholds, i.e., the slope exceeds the preset allowable slope range, or the fluctuation characteristic parameter exceeds the preset allowable fluctuation upper limit, the input characteristic state is determined to be abnormal. For example, if the preset slope threshold is -0.5 to 0.5 and the fluctuation characteristic parameter threshold is 0.3, and the slope of the input signal bandwidth value change is 0.6, it exceeds the threshold range, or the fluctuation characteristic parameter is 0.4, it exceeds the upper limit of the threshold, and the input characteristic state will be determined to be abnormal. This achieves comprehensive monitoring of the steady-state and dynamic changes of the input signal, providing a basis for subsequent judgment of loop mismatch status.

[0031] The evaluation of the loop linearity status of a digital predistortion system based on loop linearity monitoring data includes the following steps: The loop linear response characteristics are extracted from the loop linearity monitoring data. These characteristics include the amplitude variation and phase shift trend of the loop signal transmission process. Determine whether there is nonlinear distortion in the loop signal transmission based on the loop's linear response characteristics; If there is no nonlinear distortion, then the loop linearity state is determined to be a good state; If nonlinear distortion exists, the loop linearity level is determined by combining the influence range and severity of the nonlinear distortion to obtain the loop linearity state of the digital predistortion system.

[0032] First, loop linear response features are extracted based on loop linearity monitoring data. These features are the core basis for judging linearity status and mainly include two parts: amplitude variation law and phase shift trend during loop signal transmission. The amplitude variation law is obtained by the ratio of the output signal amplitude to the input signal amplitude at different frequency points. Amplitude gain = output signal amplitude / input signal amplitude. By collecting amplitude gain data at multiple frequency points within the full bandwidth, an amplitude gain versus frequency curve is formed, reflecting the linearity of amplitude transmission. The phase shift trend is obtained by calculating the difference between the output signal phase and the input signal phase, combined with the signal frequency, to obtain the phase shift rate. Phase shift rate = (output signal phase - input signal phase) / signal frequency. By plotting the phase shift rate versus frequency curve, the linearity of phase transmission is demonstrated. For example, within a signal bandwidth of 20MHz to 40MHz, input and output signals are collected at 1MHz intervals, and the amplitude gain and phase shift rate at each frequency point are calculated. Finally, complete amplitude variation curves and phase shift curves are obtained as loop linear response features.

[0033] The presence of nonlinear distortion in loop signal transmission is determined based on the linearity of the amplitude variation and phase shift trends. In an ideal linear transmission scenario, the amplitude gain should remain approximately constant across the entire bandwidth, and the phase shift rate should change linearly with frequency. If the amplitude gain curve shows significant fluctuations or jumps, or the phase shift rate curve deviates from the linear trend, nonlinear distortion is determined to exist in the loop signal transmission. When the amplitude gain at a certain frequency point suddenly increases from a reference value of 1 to 1.6, or the phase shift rate abruptly changes from 2 degrees per megahertz to 6 degrees per megahertz, and this deviation exceeds the preset allowable range, nonlinear distortion can be confirmed.

[0034] If the judgment result indicates the absence of nonlinear distortion, meaning the amplitude gain curve fluctuates minimally and the phase offset rate remains stable and linear, the loop linearity is directly determined to be in a good state. In this case, the loop signal transmission can precisely follow the linear transmission law without the need for additional gain adjustment. When the amplitude gain fluctuation range is controlled between 0.95 and 1.05, and the phase offset rate deviation is less than 0.5 degrees per megahertz, the loop linearity can be considered to be in a good state.

[0035] If the judgment result indicates the presence of nonlinear distortion, the loop linearity level will be classified based on the influence range and severity of the nonlinear distortion, thereby obtaining the loop linearity state of the digital predistortion system. The influence range is defined by the proportion of distorted frequency points to the total acquired frequency points, and can be divided into local distortion, partial distortion, and global distortion. The severity is measured by the maximum amplitude gain deviation and the maximum phase offset rate deviation, and can be divided into slight distortion, moderate distortion, and severe distortion. When the proportion of distorted frequency points is 25%, the maximum amplitude gain deviation is 0.15, and the maximum phase offset rate deviation is 0.8 degrees per megahertz, it is judged as local slight distortion, corresponding to a general loop linearity. When the proportion of distorted frequency points is 60%, the maximum amplitude gain deviation is 0.4, and the maximum phase offset rate deviation is 2.5 degrees per megahertz, it is judged as partial moderate distortion, corresponding to a poor loop linearity. When the proportion of distorted frequency points is 80%, the maximum amplitude gain deviation is 0.6, and the maximum phase offset rate deviation is 4 degrees per megahertz, it is judged as global severe distortion, corresponding to an extremely poor loop linearity. This classification will provide a decision-making basis for the gradient adaptive adjustment of loop gain.

[0036] Determining the target gain gradient adjustment step size based on the loop linearity state includes the following steps: Determine the degree of anomaly in the linearity of the loop based on its linearity status; Set the corresponding gain adjustment sensitivity according to the degree of loop linearity anomaly; Based on the initial gradient adjustment step size corresponding to gain adjustment sensitivity matching; The amplitude variation pattern and phase shift trend of the loop linear response characteristics are determined, and the target gain gradient adjustment step size is obtained by dynamically calibrating the initial gradient adjustment step size based on the amplitude variation magnitude and phase shift rate.

[0037] Firstly, the degree of linearity anomaly in the current loop is determined based on its linearity status. This transforms the qualitative linearity status into a quantifiable anomaly level, providing a basis for subsequent adjustment strategies. For example, a good linearity status corresponds to no anomaly, a moderate linearity status corresponds to a slight anomaly, a poor linearity status corresponds to a moderate anomaly, and a very poor linearity status corresponds to a severe anomaly. This correspondence allows for the definition of the severity of linearity problems.

[0038] The gain adjustment sensitivity is set according to the degree of loop linearity anomaly. Gain adjustment sensitivity represents the speed and intensity of response to linearity anomalies. Higher anomaly levels require higher gain adjustment sensitivity to ensure rapid intervention in severe anomaly scenarios; lower anomaly levels require lower gain adjustment sensitivity to prevent over-adjustment and additional system fluctuations. When the loop linearity is slightly abnormal, the gain adjustment sensitivity is set to a low level, corresponding to a sensitivity coefficient of 0.3; when the loop linearity is severely abnormal, the gain adjustment sensitivity is set to a high level, corresponding to a sensitivity coefficient of 0.9.

[0039] The initial gradient adjustment step size is matched to the set gain adjustment sensitivity. This initial gradient adjustment step size is the base step size for gain adjustment, and it is positively correlated with the gain adjustment sensitivity. That is, the higher the sensitivity, the larger the initial gradient adjustment step size, and vice versa. The initial gradient adjustment step size is calculated as: Initial Gradient Adjustment Step Size = Base Step Size × Sensitivity Coefficient. For example, if the base step size is set to 0.1, and the sensitivity coefficient is 0.3, the initial gradient adjustment step size is 0.1 × 0.3 = 0.03; if the sensitivity coefficient is 0.9, the initial gradient adjustment step size is 0.1 × 0.9 = 0.09. This formula is used to adapt the sensitivity to the initial step size.

[0040] The amplitude variation pattern and phase shift trend of the loop linear response characteristics are determined, and the amplitude variation magnitude and phase shift rate are extracted. The amplitude variation magnitude is obtained by calculating the difference between the maximum and minimum amplitude gain within the entire signal bandwidth: Amplitude Variation = Maximum Amplitude Gain - Minimum Amplitude Gain; the phase shift rate is obtained by calculating the standard deviation of the phase shift rate at each frequency point within the entire bandwidth.

[0041] The initial gradient adjustment step size is dynamically calibrated based on the amplitude change and phase shift rate to obtain the final target gain gradient adjustment step size. The calibration logic is that the larger the amplitude change and the faster the phase shift rate, the more significant the dynamic impact of linearity distortion, and the step size should be appropriately increased based on the initial step size to improve adjustment efficiency; conversely, the step size should be appropriately decreased to avoid over-adjustment. The target gain gradient adjustment step size = initial gradient adjustment step size × (1 + amplitude change weight × amplitude change + phase offset rate weight × phase offset rate), where the amplitude change weight and phase offset rate weight are system-preset calibration coefficients. For example, if the amplitude change weight is set to 0.4 and the phase offset rate weight is set to 0.6, when the amplitude change is 0.2, the phase offset rate is 1.2 degrees per megahertz, and the initial gradient adjustment step size is 0.05, the target gain gradient adjustment step size = 0.05 × (1 + 0.4 × 0.2 + 0.6 × 1.2) = 0.09. This dynamic calibration enables the target gain gradient adjustment step size to accurately match the linear distortion dynamic characteristics of the current loop, providing adjustment parameters that are more in line with the actual scenario for subsequent gradient-based adaptive adjustment.

[0042] The loop gain is adaptively adjusted in a gradient manner based on the target gain gradient adjustment step size, specifically including the following steps: The initial adjustment loop gain is obtained by performing the first gradient adjustment on the current loop gain according to the target gain gradient adjustment step size, and the initial adjustment feedback information is obtained according to the change trend of the loop gain. Based on the changing trend of the initial adjustment feedback information, the deviation tendency of the loop gain from the optimal fit state after the initial adjustment is judged, and the gradient adjustment direction is obtained. The target gain gradient adjustment step size is dynamically calibrated based on the initial adjustment feedback information to obtain the calibration adjustment step size. The initial adjustment loop gain is adaptively adjusted using a gradient method according to the gradient adjustment direction and calibration adjustment step size.

[0043] The initial loop gain is obtained by performing an initial gradient adjustment on the current loop gain based on the target gain gradient adjustment step size. The initial adjustment loop gain is then collected and generated based on the trend of the adjusted loop gain. The initial adjustment loop gain = current loop gain ± target gain gradient adjustment step size. The initial adjustment direction can be preset to positive or negative, and then determined after verification by the feedback information. If the current loop gain is 1.0 and the target gain gradient adjustment step size is 0.05, and the preset initial adjustment direction is positive, then the initial adjustment loop gain = 1.0 + 0.05 = 1.05. The adjusted loop linearity data and feedback signal distortion data are collected to form the initial adjustment feedback information, which reflects the trend of the initial adjustment's impact on loop performance.

[0044] Based on the changing trend of the feedback information after the initial adjustment, the deviation tendency of the loop gain from the optimal fit state after the initial adjustment is judged, thereby determining the gradient adjustment direction. If the loop linearity improves and the feedback signal distortion decreases after the initial adjustment, it indicates that the current adjustment direction brings the loop gain closer to the optimal fit state, and the gradient adjustment direction remains unchanged. If the loop linearity decreases and the feedback signal distortion increases after the initial adjustment, it indicates that the current adjustment direction causes the loop gain to deviate from the optimal fit state, and the gradient adjustment direction needs to be adjusted in the opposite direction. For example, if the feedback signal distortion increases from 0.8 to 0.9 and the linearity decreases from a normal state to a poor state after the initial adjustment using a positive step size, then the gradient adjustment direction should be adjusted from positive to negative.

[0045] The target gain gradient adjustment step size is dynamically calibrated based on the initial adjustment feedback information to obtain the calibration adjustment step size, which adapts to the actual adjustment requirements of the current loop. The calibration process is achieved through a feedback coefficient, which is determined by the improvement in loop linearity or the rate of change in distortion after the initial adjustment. The calibration adjustment step size = target gain gradient adjustment step size × feedback coefficient. The rules for selecting the feedback coefficient are as follows: if the linearity improvement is greater than 30% or the distortion decrease is greater than 20%, the feedback coefficient is between 1.1 and 1.3, and the step size is appropriately increased to accelerate the adjustment efficiency; if the linearity improvement is between 10% and 30% or the distortion decrease is between 5% and 20%, the feedback coefficient is between 0.9 and 1.1, keeping the step size stable; if there is no improvement in linearity or an increase in distortion, the feedback coefficient is between 0.7 and 0.9, and the step size is decreased to avoid over-adjustment. For example, if the target gain gradient adjustment step size is 0.05, and the distortion decreases by 25% after the initial adjustment, and the feedback coefficient is selected as 1.2, then the calibration adjustment step size = 0.05 × 1.2 = 0.06; if the distortion increases by 15% after the initial adjustment, and the feedback coefficient is selected as 0.8, then the calibration adjustment step size = 0.05 × 0.8 = 0.04.

[0046] The loop gain will be initially adjusted using a gradient-based adaptive adjustment according to the direction and step size, thus entering the iterative optimization stage. After each iteration, loop performance data is collected to form new feedback information. The deviation tendency is then reassessed to calibrate the adjustment direction, and the calibration step size is dynamically adjusted based on the feedback effect, gradually bringing the loop gain closer to the optimal fit. For example, if the initial loop gain is 1.05, the gradient adjustment direction is determined to be negative, and the calibration step size is 0.06, then the loop gain after the first iteration = 1.05 - 0.06 = 0.99. After the adjustment is completed, new feedback information is collected. If the distortion drops to 0.5 and the linearity improves to a good state, the calibration step size can be appropriately reduced to 0.03, and fine-tuning continues until the loop gain stabilizes at the optimal value. If the distortion is still too high, the current step size is maintained and iteration continues until the loop linearity reaches the preset standard, achieving adaptive adjustment of the loop gain.

[0047] Determining the loop mismatch recovery status of a digital predistortion system based on gain adjustment monitoring data includes the following steps: Extract gain fluctuation information and loop transmission characteristic information from gain adjustment monitoring data; The gain fluctuation information is smoothed to obtain a stable gain fluctuation curve, and the loop transmission characteristic information is sorted out to obtain the loop transmission response result; The stable gain fluctuation curve is compared with the preset gain stability threshold range to obtain the gain fluctuation comparison result. The loop transmission response result is compared with the preset normal transmission response standard to obtain the transmission response comparison result. If the gain fluctuation curve is within the preset gain stability threshold range, and the loop transmission response result is consistent with the preset normal transmission response standard, then the loop mismatch recovery state is determined to be the mismatch complete recovery state. If the gain fluctuation curve exceeds the preset gain stability threshold range, or if the loop transmission response result is inconsistent with the preset normal transmission response standard, the loop mismatch recovery state is determined to be a mismatch unrecovered state.

[0048] Two key data points are extracted from the gain adjustment monitoring data: gain fluctuation information and loop transmission characteristic information. Gain fluctuation information refers to the continuous numerical sequence of loop gain changes over time or the number of adjustment iterations during gradient adaptive adjustment, including the instantaneous loop gain value after each adjustment and the gain difference between two adjacent adjustments. Loop transmission characteristic information covers the amplitude gain, phase offset rate, feedback signal distortion, and effective transmission bandwidth utilization of the adjusted loop signal. These information collectively constitute the basic data source for determining whether loop mismatch has been recovered.

[0049] The extracted gain fluctuation information is smoothed to eliminate interference from instantaneous noise or single adjustment disturbances, thus obtaining a stable gain fluctuation curve. Smoothing typically employs a moving average filtering algorithm. The smoothed gain value is calculated as (the sum of the n instantaneous gain values ​​before the current moment) / n, where n is the length of the moving window, set according to the system sampling frequency. For example, setting n to 5 means averaging the gain values ​​at the current moment and the previous four moments to obtain the smoothed gain sequence. This is then plotted as a gain fluctuation curve that changes over time or iterations, visually reflecting the overall trend of gain variation. Simultaneously, loop transmission characteristic information is organized and integrated. The scattered amplitude gain and phase offset data at various frequency points are summarized into a full-bandwidth transmission response curve. Distortion and bandwidth utilization are calculated as averages, resulting in a unified loop transmission response result.

[0050] The stable gain fluctuation curve is compared with a preset gain stability threshold range to obtain the gain fluctuation comparison result. The preset gain stability threshold range is the allowable gain fluctuation range of the system under mismatch-free conditions, for example, set to 0.95 to 1.05. If all values ​​of the smoothed gain fluctuation curve fall within this range, the gain fluctuation comparison result is considered qualified; if any value exceeds this range, it is considered unqualified. The smoothed loop transmission response result is compared with a preset normal transmission response standard. The normal transmission response standard includes quantitative indicators such as the upper limit of amplitude gain fluctuation, the upper limit of phase offset rate deviation, and the upper limit of feedback signal distortion. For example, the amplitude gain fluctuation is set to ≤0.05, the phase offset rate deviation to ≤0.5 degrees per megahertz, and the feedback signal distortion to ≤0.3%. If all indicators of the loop transmission response result meet the standard, the transmission response comparison result is considered qualified; if any indicator does not meet the standard, it is considered unqualified.

[0051] The loop mismatch recovery status is determined based on a comprehensive assessment of the two comparison results. If the gain fluctuation curve is within the preset gain stability threshold range, and the loop transmission response result is completely consistent with the preset normal transmission response standard (i.e., both comparison results are qualified), then the loop mismatch recovery status is determined to be a complete mismatch recovery status, indicating that the current loop gain has been adjusted to a suitable state, and the loop performance has returned to a normal level. For example, if the smoothed gain fluctuation curve is consistently between 0.97 and 1.02, the amplitude gain fluctuation is 0.03, the phase offset rate deviation is 0.2 degrees per megahertz, and the feedback signal distortion is 0.2%, all meeting the preset standards, then the mismatch is determined to be completely recovered. If the gain fluctuation curve exceeds the preset gain stability threshold range, or the loop transmission response result is inconsistent with the preset normal transmission response standard (i.e., any one of the two comparison results is unqualified), then the loop mismatch recovery status is determined to be a mismatch not recovered status, indicating that the current loop gain adjustment has not yet achieved the desired effect and iterative adjustment is required. For example, if the smoothed gain fluctuation curve shows a value of 1.08, or the feedback signal distortion is 0.4%, it is determined that the mismatch has not been recovered, and the gradient adaptive adjustment process needs to be continued.

[0052] If the loop mismatch recovery state is a complete mismatch recovery state, then the loop gain is locked to the optimal fit value, specifically including the following steps: If the loop mismatch recovery state is the complete mismatch recovery state, monitor the fluctuation of the loop gain within a preset time period and determine whether the fluctuation is within a preset stable range; If the fluctuation is within the preset stable range, then the current loop gain is confirmed to be the optimal fit value, and the loop gain locking mechanism is activated. If the fluctuation exceeds the preset stable range, the current loop gain is calibrated until the loop gain fluctuation is within the preset stable range. When the loop gain fluctuation is within the preset stable range, the current loop gain is determined as the optimal fit value, and the loop gain locking mechanism is activated.

[0053] First, the loop gain fluctuation monitoring process is initiated within a preset duration. The preset duration is set according to the actual operating requirements of the system, such as 30 seconds or 10 consecutive adjustment iterations. Within this duration, the real-time values ​​of the loop gain are continuously collected to form a gain change sequence. To quantify the degree of fluctuation, the gain fluctuation amplitude and the fluctuation standard deviation are calculated. The gain fluctuation amplitude = the maximum loop gain value within the monitoring duration - the minimum loop gain value within the monitoring duration. These two indicators are used to comprehensively determine whether the fluctuation is within the preset stable range. The preset stable range is set according to the system accuracy requirements, for example, the gain fluctuation amplitude is set to ≤0.05 and the fluctuation standard deviation is set to ≤0.02.

[0054] If the loop gain fluctuation is within the preset stable range, meaning both the fluctuation amplitude and standard deviation meet the preset threshold requirements, the current loop gain will be directly confirmed as the optimal fit value. This optimal fit value is the best gain parameter that meets the loop linearity and stability requirements after gradient adjustment and stability verification. Upon confirmation, the system will activate the loop gain locking mechanism, pausing subsequent gradient-adaptive adjustment processes, keeping the loop gain constant, and retaining only basic state monitoring functions until the next loop mismatch is detected, at which point the lock will be released, and a new round of adjustment will begin. For example, if the loop gain remains between 0.99 and 1.01 within a 30-second monitoring period, with a fluctuation amplitude of 0.02 and a standard deviation of 0.01, both within the preset stable range, then the current gain of 1.00 is confirmed as the optimal fit value, and the locking mechanism is activated.

[0055] If the loop gain fluctuation exceeds the preset stable range, i.e., the fluctuation amplitude or standard deviation exceeds the preset threshold, it indicates that although the current loop gain has temporarily restored the mismatch, there is still a risk of instability. An amplitude calibration operation will then be performed on the current loop gain. Amplitude calibration reduces the gain fluctuation range by fine-tuning the gain value. Specific calibration methods include: averaging calibration, where the average loop gain over the monitoring period is used as the new initial gain value, and the calibrated loop gain equals the average loop gain over the monitoring period; or step calibration, where the gain is fine-tuned in increments of 0.01 towards the center of the fluctuation until the gain fluctuation falls within the preset stable range. After one calibration, the fluctuation monitoring process will be restarted for the preset duration, and the fluctuation will be repeatedly assessed. If the fluctuation still exceeds the stable range, calibration will continue until the loop gain fluctuation is completely within the preset stable range. Once the fluctuation meets the requirements, the loop gain at this point will be determined as the optimal fit value, and a loop gain locking mechanism will be activated to ensure stable linear transmission performance during long-term operation. For example, if the loop gain fluctuates from 0.97 to 1.06 within the monitoring period, with a fluctuation range of 0.09, exceeding the preset threshold of 0.05, and the average gain is 1.015, this is used as the calibrated gain. After monitoring for 30 seconds, the gain fluctuation range shrinks to 1.01 to 1.02, with a fluctuation range of 0.01, which meets the stability requirements. At this point, 1.015 is confirmed as the optimal adaptation value and locking is initiated.

[0056] If the loop mismatch recovery state is not the mismatch recovery state, then the loop gain is iteratively adjusted according to the target gain gradient adjustment step size, specifically including the following steps: If the loop mismatch recovery state is the mismatch unrecovered state, the loop gain is iteratively adjusted based on the target gain gradient adjustment step size. During the iterative adjustment process, the changing trend of the current loop mismatch degree is judged based on the gain adjustment monitoring data and the loop linearity monitoring data. The adjustment direction of the target gain gradient adjustment step size is calibrated according to the changing trend of the mismatch degree, and the loop gain is iteratively adjusted according to the adjustment direction.

[0057] When the loop mismatch recovery state is determined to be in an unrecovered mismatch state, it means that the current gain adjustment has not yet restored the loop performance to the normal standard. Therefore, continuous iterative adjustment of the loop gain is initiated based on the target gain gradient adjustment step size. The initial operation of the iterative adjustment follows gradient adjustment logic: the current iteration loop gain = the previous iteration loop gain ± the target gain gradient adjustment step size. The initial adjustment direction follows the direction determined in the previous gradient adjustment stage, and gain optimization is gradually advanced based on this.

[0058] During the iterative adjustment process, gain adjustment monitoring data and loop linearity monitoring data are collected. Gain adjustment monitoring data includes the instantaneous loop gain value after each iteration and the gain change between adjacent iterations. Loop linearity monitoring data includes feedback signal distortion, amplitude change magnitude, and phase shift rate. A mismatch quantification model is constructed based on these two types of data: Mismatch Degree = Feedback Signal Distortion × Weight 1 + Amplitude Change Magnitude × Weight 2 + Phase Shift Rate × Weight 3. For example, weight 1 is set to 0.4, weight 2 to 0.3, and weight 3 to 0.3. The trend of the current loop mismatch degree is determined by comparing the mismatch degree values ​​between two adjacent iterations, specifically through the mismatch degree change rate: Mismatch Degree Change Rate = (Current Iteration Mismatch Degree - Previous Iteration Mismatch Degree) / Previous Iteration Mismatch Degree × 100%. A negative change rate indicates that the mismatch degree is decreasing; a positive change rate indicates that the mismatch degree is increasing; and a change rate close to 0 indicates that the mismatch degree has not significantly improved.

[0059] The adjustment direction of the target gain gradient adjustment step size will be dynamically calibrated based on the changing trend of the mismatch degree. If the rate of change of the mismatch degree is negative, meaning the mismatch degree is continuously decreasing, it indicates that the current adjustment direction can push the loop closer to the optimal fit state, so the original adjustment direction remains unchanged. If the rate of change of the mismatch degree is positive, meaning the mismatch degree is continuously increasing, it indicates that the current adjustment direction is causing the loop to deviate from the optimal fit state, so the adjustment direction is reversed, i.e., the original positive adjustment is changed to negative adjustment, and the original negative adjustment is changed to positive adjustment. If the rate of change of the mismatch degree is close to 0, meaning the mismatch degree has no significant change, it indicates that the current adjustment direction or step size cannot effectively improve performance. First, the adjustment direction is switched, and then the target gain gradient adjustment step size is finely adjusted to explore a better adjustment path. When the mismatch degree decreases from 0.7 to 0.5 after two consecutive iterations, with a mismatch degree change rate of -28.57%, the original adjustment direction is maintained. If the mismatch degree increases from 0.7 to 0.8, with a mismatch degree change rate of 14.29%, the adjustment direction is reversed, and if the original adjustment was to increase the gain, it is changed to decrease the gain.

[0060] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0061] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0062] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A digital predistortion loop gain adaptive adjustment system, characterized in that, include: Acquisition module: Acquires real-time loop characteristic data of the digital predistortion system; First judgment module: judges the loop mismatch status of the digital predistortion system based on real-time loop feature data; Processing module: If the loop mismatch state is the loop no mismatch state, then the loop gain of the digital predistortion system is kept at the baseline adaptation value; if the loop mismatch state is the loop mismatch state, then the loop gain is adjusted to the temporary adjustment value and the gain adjustment monitoring data and loop linearity monitoring data are obtained. The first adjustment module evaluates the loop linearity status of the digital predistortion system based on loop linearity monitoring data, determines the target gain gradient adjustment step size based on the loop linearity status, and performs gradient-based adaptive adjustment of the loop gain based on the target gain gradient adjustment step size. The second judgment module: judges the loop mismatch recovery status of the digital predistortion system based on gain adjustment monitoring data; The second adjustment module: If the loop mismatch recovery state is a fully recovered mismatch state, the loop gain is locked to the optimal fit value; if the loop mismatch recovery state is a partially recovered mismatch state, the loop gain is iteratively adjusted according to the target gain gradient adjustment step size.

2. The digital predistortion loop gain adaptive adjustment system according to claim 1, characterized in that, The real-time characteristic data of the loop includes the characteristic data of the loop input signal and the characteristic data of the loop feedback signal.

3. The digital predistortion loop gain adaptive adjustment system according to claim 1, characterized in that, The loop input signal characteristic data includes the peak-to-average power ratio of the input signal, the bandwidth value of the input signal, and the acquisition timestamps corresponding to each characteristic parameter; The loop feedback signal characteristic data includes the feedback signal distortion, the feedback signal power value, and the acquisition timestamps corresponding to each characteristic parameter.

4. The digital predistortion loop gain adaptive adjustment system according to claim 3, characterized in that, Determining the loop mismatch status of a digital predistortion system based on real-time loop feature data includes the following steps: Set the input feature threshold range and the input anomaly duration threshold; If the peak-to-average power ratio (PAPR) or bandwidth of the input signal is not within the corresponding input feature threshold range, then obtain the duration for which the PAPR or bandwidth exceeds the corresponding input feature threshold range. If the duration is greater than the input abnormal duration threshold, then determine that the PAPR or bandwidth is abnormal. If the peak-to-average power ratio (PAPR) and bandwidth of the input signal are both within the corresponding input feature threshold range, then the slope and fluctuation characteristic parameters of the PAPR and bandwidth are obtained based on the acquisition timestamps corresponding to the PAPR and bandwidth. Set the threshold for the input feature curve parameters; If the slope of change and the fluctuation characteristic parameters meet the threshold of the input characteristic curve parameters, then the input characteristic state is judged to be normal. If the slope of change and the fluctuation characteristic parameters do not meet the threshold of the input characteristic curve parameters, then the input characteristic state is judged to be abnormal.

5. The digital predistortion loop gain adaptive adjustment system according to claim 1, characterized in that, The evaluation of the loop linearity status of a digital predistortion system based on loop linearity monitoring data includes the following steps: The loop linear response characteristics are extracted from the loop linearity monitoring data. These characteristics include the amplitude variation and phase shift trend of the loop signal transmission process. Determine whether there is nonlinear distortion in the loop signal transmission based on the loop's linear response characteristics; If there is no nonlinear distortion, then the loop linearity state is determined to be a good state; If nonlinear distortion exists, the loop linearity level is determined by combining the influence range and severity of the nonlinear distortion to obtain the loop linearity state of the digital predistortion system.

6. The digital predistortion loop gain adaptive adjustment system according to claim 5, characterized in that, Determining the target gain gradient adjustment step size based on the loop linearity state includes the following steps: Determine the degree of anomaly in the linearity of the loop based on its linearity status; Set the corresponding gain adjustment sensitivity according to the degree of loop linearity anomaly; Based on the initial gradient adjustment step size corresponding to gain adjustment sensitivity matching; The amplitude variation pattern and phase shift trend of the loop linear response characteristics are determined, and the target gain gradient adjustment step size is obtained by dynamically calibrating the initial gradient adjustment step size based on the amplitude variation magnitude and phase shift rate.

7. The digital predistortion loop gain adaptive adjustment system according to claim 6, characterized in that, The loop gain is adaptively adjusted in a gradient manner based on the target gain gradient adjustment step size, specifically including the following steps: The initial adjustment loop gain is obtained by performing the first gradient adjustment on the current loop gain according to the target gain gradient adjustment step size, and the initial adjustment feedback information is obtained according to the change trend of the loop gain. Based on the changing trend of the initial adjustment feedback information, the deviation tendency of the loop gain from the optimal fit state after the initial adjustment is judged, and the gradient adjustment direction is obtained. The target gain gradient adjustment step size is dynamically calibrated based on the initial adjustment feedback information to obtain the calibration adjustment step size. The initial adjustment loop gain is adaptively adjusted using a gradient method according to the gradient adjustment direction and calibration adjustment step size.

8. The digital predistortion loop gain adaptive adjustment system according to claim 1, characterized in that, Determining the loop mismatch recovery status of a digital predistortion system based on gain adjustment monitoring data includes the following steps: Extract gain fluctuation information and loop transmission characteristic information from gain adjustment monitoring data; The gain fluctuation information is smoothed to obtain a stable gain fluctuation curve, and the loop transmission characteristic information is sorted out to obtain the loop transmission response result; The stable gain fluctuation curve is compared with the preset gain stability threshold range to obtain the gain fluctuation comparison result. The loop transmission response result is compared with the preset normal transmission response standard to obtain the transmission response comparison result. If the gain fluctuation curve is within the preset gain stability threshold range, and the loop transmission response result is consistent with the preset normal transmission response standard, then the loop mismatch recovery state is determined to be the mismatch complete recovery state. If the gain fluctuation curve exceeds the preset gain stability threshold range, or if the loop transmission response result is inconsistent with the preset normal transmission response standard, the loop mismatch recovery state is determined to be a mismatch unrecovered state.

9. The digital predistortion loop gain adaptive adjustment system according to claim 8, characterized in that, If the loop mismatch recovery state is a complete mismatch recovery state, then the loop gain is locked to the optimal fit value, specifically including the following steps: If the loop mismatch recovery state is the complete mismatch recovery state, monitor the fluctuation of the loop gain within a preset time period and determine whether the fluctuation is within a preset stable range; If the fluctuation is within the preset stable range, then the current loop gain is confirmed to be the optimal fit value, and the loop gain locking mechanism is activated. If the fluctuation exceeds the preset stability range, the current loop gain is calibrated until the loop gain fluctuation is within the preset stability range. When the loop gain fluctuation is within the preset stability range, the current loop gain is determined as the optimal fit value, and the loop gain locking mechanism is activated.

10. The digital predistortion loop gain adaptive adjustment system according to claim 9, characterized in that, If the loop mismatch recovery state is not the mismatch recovery state, then the loop gain is iteratively adjusted according to the target gain gradient adjustment step size, specifically including the following steps: If the loop mismatch recovery state is the mismatch unrecovered state, the loop gain is iteratively adjusted based on the target gain gradient adjustment step size. During the iterative adjustment process, the changing trend of the current loop mismatch degree is judged based on the gain adjustment monitoring data and the loop linearity monitoring data. The adjustment direction of the target gain gradient adjustment step size is calibrated according to the changing trend of the mismatch degree, and the loop gain is iteratively adjusted according to the adjustment direction.

Images