Low-PAPR self-adaptive spectrum splicing method and device for broadband microwave component characterization

Abstract

The invention relates to a low-PAPR adaptive spectrum splicing method and device for broadband microwave component characterization. According to the method, starting from a bottom layer mechanism of waveform synthesis, the local signal-to-noise ratio of a splicing boundary is remarkably improved through a three-section type power enhancement framework, and the phase continuity under the condition that thermal noise and hardware defects exist is ensured. The phase symmetry constraint eliminates the dependence of a traditional method on a complex delay estimation algorithm, and seamless splicing of zero phase difference is achieved. The signal synthesized by adopting the method is low in phase mismatch error, the amplitude discontinuity can be ignored, the PAPR performance is superior to that of the prior art, and a steady solution is provided for high-precision nonlinear measurement of millimeter wave and terahertz frequency bands.

Description

Technical Field

[0001] This application relates to the field of microwave measurement and signal processing technology, and in particular to a low PAPR adaptive spectrum stitching method and apparatus for characterizing broadband microwave components. Background Technology

[0002] As 6G wireless communication networks evolve towards millimeter-wave and terahertz bands, the linearity and bandwidth requirements for RF front-end components have increased unprecedentedly. To ensure the fidelity of communication systems, ultra-wideband signals with high spectral purity and low peak-to-average power ratio (PAPR) are needed to characterize key components such as power amplifiers. However, traditional time-domain instruments, such as oscilloscopes, have limitations in signal-to-noise ratio and impedance matching, failing to meet the demands of high-precision microwave metrology. In contrast, frequency-domain analysis methods based on vector network analyzers (VNAs) offer higher dynamic range and sensitivity, but rely on low PAPR wideband standard signals for system error calibration.

[0003] However, generating such wideband standard signals faces significant challenges. Limited by the sampling rate and storage depth of digital-to-analog converters (DACs), direct synthesis of ultra-wideband signals is often impractical; therefore, frequency domain stitching has become the mainstream solution. Existing stitching methods typically estimate phase delay by overlapping tones, falling under the category of post-processing and failing to fundamentally address the quality issues of the excitation signal itself. Specifically, existing methods suffer from the following difficulties: 1) They fail to effectively minimize the PAPR of each segment, leading to hardware saturation of the high-amplitude peak driving signal and introducing nonlinear amplitude-phase modulation distortion that is difficult to eliminate through post-processing; 2) Traditional PAPR reduction algorithms (such as peak clipping filtering and selective mapping) rely on free phase adjustment, while frequency domain stitching requires maintaining strict phase and amplitude continuity at the boundaries, creating a theoretical conflict; 3) Existing stitching methods typically require complex post-processing delay estimation, resulting in high computational costs and sensitivity to noise. Summary of the Invention

[0004] Therefore, it is necessary to provide a low PAPR adaptive spectrum splicing method and apparatus for broadband microwave component characterization that can achieve high-fidelity test signal synthesis with ultra-wideband, low PAPR and intrinsic coherence, in order to address the above-mentioned technical problems.

[0005] A low PAPR adaptive spectrum stitching method for characterizing broadband microwave components, the method comprising:

[0006] Based on the multi-tone signal model, an initial signal is constructed, and the frequency domain vector of the initial signal is used as the signal to be synthesized. The multiple spectral tones corresponding to each frequency band in the signal to be synthesized are divided into three continuous sets: an initial overlapping segment set, an intermediate segment set, and an end overlapping segment set. The initial overlapping segment set is used to overlap and connect with the end overlapping segment set of the previous frequency band, and the end overlapping segment set is used to overlap and connect with the initial overlapping segment set of the next frequency band. The initial overlapping segment set and the end overlapping segment set are assigned higher power weights than the intermediate segment set. The total power is kept constant by a normalization constant to generate the target amplitude spectrum. Based on the target amplitude spectrum and the three consecutive sets, spectral constraints are set to ensure splicing coherence, balance PAPR and spectral distortion, and maintain power conservation. Under the constraints of the spectral constraints, an adaptive stitching sensing iterative projection algorithm is executed on the frequency domain vector of the signal to be synthesized to achieve a synergistic optimization that minimizes PAPR and satisfies the spectral constraints, thereby obtaining the optimal waveform segments corresponding to different center frequencies. The optimal waveform segments are arranged and spliced ​​in frequency order. During this process, the overlapping segments at the end of the previous frequency band are overlapped and connected with the initial overlapping segments of the next frequency band to complete the splicing of the ultra-wideband signal and obtain the ultra-wideband low PAPR excitation signal.

[0007] In one embodiment, the signal to be synthesized is represented as:

[0008] In the above formula, This indicates the length of the signal to be synthesized. Indicates inclusion A frequency domain vector with non-zero pitch and zero-filled points.

[0009] In one embodiment, the initial overlapping segment set and the final overlapping segment set are assigned higher power weights than the intermediate segment set. A normalization constant ensures the conservation of total power, generating the target amplitude spectrum. This process is represented as follows:

[0010] in,

[0011] In the above formula, This represents the target amplitude spectrum. Indicates the power boost ratio. , , These represent the initial set of overlapping segments, the middle set of overlapping segments, and the final set of overlapping segments, respectively. This represents the number of spectral tones in the initial set of overlapping segments and the final set of overlapping segments. This indicates the number of spectral tones in the set of intermediate segments. Represents the normalization constant. This indicates the total number of spectral tones in each frequency band.

[0012] In one embodiment, the spectral constraints include three hard constraints, namely: Phase symmetry constraint, which forces the initial overlapping segment set and the last overlapping segment set within a single frequency band to be completely consistent in phase; Amplitude fidelity constraint, wherein the deviation between the synthesized amplitude and the target amplitude spectrum is limited to a preset Euclidean sphere neighborhood; The power conservation constraint forces the total power of signals in each frequency band to be maintained at a preset target total power.

[0013] In one embodiment, the adaptive stitching sensing iterative projection algorithm performs time-domain projection processing and frequency-domain constraint recovery sequentially during each iteration. When performing the time-domain projection processing, the inverse fast Fourier transform is performed on the frequency domain vector obtained in the previous iteration to obtain the time-domain signal. The peak-shaving threshold is calculated based on the current adaptive peak-shaving ratio, and a hard peak-shaving operation is performed on the time-domain signal. During the frequency domain constraint recovery, the peak-shaving time domain signal is converted back to the frequency domain through a fast Fourier transform and decomposed to obtain amplitude and phase. Phase consistency is enforced based on the phase symmetry constraint, and weighted projection is performed using an adaptive relaxation factor based on the amplitude fidelity constraint. Then, the frequency domain vector is scaled based on the power conservation constraint to obtain the frequency domain vector of the current iteration.

[0014] In one embodiment, during each iteration: The moving average improvement rate of the peak-to-average power ratio is calculated in real time. When the moving average improvement rate is less than a preset stagnation threshold, the peak reduction ratio and the relaxation factor are adaptively updated.

[0015] In one embodiment, when the peak reduction ratio and the relaxation factor are adaptively updated: Add the preset smoothing factor to the current peak reduction ratio to obtain the updated peak reduction ratio; The updated relaxation factor is obtained by subtracting the preset smoothing factor from the current relaxation factor.

[0016] In one embodiment, during each iteration: The local fluctuations of the mean peak-to-average power ratio are calculated using a short-time observation window, and the local fluctuations are marked as a steady state when they are below a preset convergence threshold. The percentage of steady-state iterations is statistically analyzed using a long-term observation window. When the percentage exceeds a preset confidence level, it is determined that a global steady state has been reached and the iteration is terminated.

[0017] This application also provides a low PAPR adaptive spectrum splicing device characterized by broadband microwave components, the device comprising: The signal to be synthesized module is used to construct an initial signal based on a multi-tone signal model, and use the frequency domain vector of the initial signal as the signal to be synthesized. The spectrum tone division module is used to divide the multiple spectrum tones corresponding to each frequency band in the signal to be synthesized into three consecutive sets: an initial overlapping segment set, an intermediate segment set, and an end overlapping segment set. The initial overlapping segment set is used to overlap and connect with the end overlapping segment set of the previous frequency band, and the end overlapping segment set is used to overlap and connect with the initial overlapping segment set of the next frequency band. The module assigns a higher power weight to the initial overlapping segment set and the end overlapping segment set than to the intermediate segment set, and ensures the total power is conserved through a normalization constant to generate the target amplitude spectrum. The spectrum constraint setting module is used to set spectrum constraints based on the target amplitude spectrum and the three consecutive sets to ensure splicing coherence, balance PAPR and spectrum distortion, and maintain power conservation. The optimal waveform segment solution module is used to perform an adaptive stitching sensing iterative projection algorithm on the frequency domain vector of the signal to be synthesized under the constraints of the spectral constraints, so as to achieve the cooperative optimization of minimizing PAPR and satisfying the spectral constraints, and obtain the optimal waveform segments corresponding to different center frequencies. The ultra-wideband low PAPR excitation signal generation module is used to arrange and splice the optimal waveform segments in frequency order. During this process, the set of overlapping segments at the end of the previous frequency band overlaps and connects with the set of overlapping segments at the beginning of the next frequency band to complete the splicing of the ultra-wideband signal and obtain the ultra-wideband low PAPR excitation signal.

[0018] A computer device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps in the low PAPR adaptive spectrum stitching method characterized by the above-described broadband microwave component.

[0019] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the low PAPR adaptive spectrum stitching method characterized by the above-described broadband microwave component.

[0020] The low PAPR adaptive spectrum stitching method and apparatus characterized by the aforementioned broadband microwave components construct an initial signal based on a multi-tone signal model, and uses the frequency domain vector of this signal as the signal to be synthesized. The multiple spectral tones corresponding to each frequency band in the signal to be synthesized are divided into three continuous sets: an initial overlapping segment set for overlapping with the end overlapping segment set of the previous frequency band, an intermediate segment set, and an end overlapping segment set for overlapping with the initial overlapping segment set of the next frequency band. The target amplitude spectrum is obtained based on the initial overlapping segment set and the end overlapping segment set. Based on the target amplitude spectrum and the three divided sets... A continuous set is set to ensure splicing coherence, balance PAPR and spectral distortion, and maintain power conservation. Under the constraints of the spectral constraints, an adaptive splicing sensing iterative projection algorithm is executed on the frequency domain vector of the signal to be synthesized to achieve collaborative optimization that minimizes PAPR and satisfies the spectral constraints, obtaining the optimal waveform segments corresponding to different center frequencies. The optimal waveform segments are arranged and spliced ​​in frequency order. During this process, the overlapping segment set of the previous frequency band is overlapped and connected with the initial overlapping segment set of the next frequency band to complete the splicing of the ultra-wideband signal and obtain the ultra-wideband low PAPR excitation signal.

[0021] This method can significantly reduce the peak-to-average power ratio (PAPR) of the excitation signal while effectively ensuring the continuity of the ultra-wideband signal spectrum splicing and power conservation. This reduces the nonlinear distortion of the signal during transmission in the broadband microwave component, improves the dynamic range and measurement accuracy of the system, and simultaneously balances spectrum distortion control and computational efficiency through an adaptive iterative strategy, providing a high-quality, low PAPR excitation signal for accurate characterization of the broadband microwave component. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating a low PAPR adaptive spectrum stitching method characterized by a broadband microwave component in one embodiment. Figure 2 This is a schematic diagram of a three-segment spectrum splicing architecture in one embodiment; Figure 3 This is a flowchart illustrating the adaptive stitching sensing iterative projection algorithm in one embodiment; Figure 4 This is a schematic diagram comparing the time-domain waveform envelopes optimized using this method under different initial phase conditions in one embodiment. Figure 5 This is a schematic diagram comparing the spectral amplitudes after optimization using this method under different initial phase conditions in one embodiment. Figure 6 This is a schematic diagram of the phase distribution characteristics of a signal optimized using this method in one embodiment; Figure 7 This is a schematic diagram of the phase error statistical distribution of the signal after optimization using this method in one embodiment; Figure 8 This is a schematic diagram illustrating the results of an existing method in a hardware-in-the-loop experiment in one embodiment. Figure 9 This is a schematic diagram illustrating the results of the method in a hardware-in-the-loop experiment in one embodiment; Figure 10 This is a structural block diagram of a low PAPR adaptive spectrum splicing device characterized by a broadband microwave component in one embodiment. Figure 11 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0024] This application addresses the signal distortion and splicing discontinuities in broadband microwave component characterization by proposing a low PAPR adaptive spectrum splicing method. Unlike existing post-processing alignment methods, this method introduces the concept of "splicing as design," proposing a three-segment spectrum architecture and an adaptive iterative projection algorithm. Power enhancement is introduced in the overlapping region to improve the signal-to-noise ratio, symmetrical phase constraints are used to ensure intrinsic coherence, and a dynamic parameter control mechanism is employed to jointly optimize spectral fidelity and envelope ripple in the non-convex solution space. The specific steps include: Step S100: Based on the multi-tone signal model, construct an initial signal and use the frequency domain vector of the initial signal as the signal to be synthesized.

[0025] Step S110: Divide the multiple spectral tones corresponding to each frequency band in the signal to be synthesized into three continuous sets: the initial overlapping segment set, the middle segment set, and the final overlapping segment set. The initial overlapping segment set is used to overlap and connect with the final overlapping segment set of the previous frequency band, and the final overlapping segment set is used to overlap and connect with the initial overlapping segment set of the next frequency band. Assign power weights higher than those of the middle segment set to the initial overlapping segment set and the final overlapping segment set. Ensure the total power is conserved through a normalization constant to generate the target amplitude spectrum.

[0026] Step S120: Based on the target amplitude spectrum and the three consecutive sets, set spectral constraints to ensure splicing coherence, balance PAPR and spectral distortion, and maintain power conservation.

[0027] Step S130: Under the constraint of the spectrum constraint, the adaptive stitching sensing iterative projection algorithm is executed on the frequency domain vector of the signal to be synthesized to achieve the synergistic optimization of minimizing PAPR and satisfying the spectrum constraint, so as to obtain the optimal waveform segment corresponding to each different center frequency.

[0028] Step S140: Arrange and splice the optimal waveform segments in frequency order. During this process, make the set of overlapping segments at the end of the previous frequency band overlap and connect with the set of overlapping segments at the beginning of the next frequency band to complete the splicing of the ultra-wideband signal and obtain the ultra-wideband low PAPR excitation signal.

[0029] In step S100, a multi-tone waveform is used as the basic signal model. Let the continuous-time complex envelope signal to be synthesized be... Depend on Each frequency interval is It is formed by superimposing sine waves. To accurately generate and capture the peak characteristics of analog signals in the digital domain, oversampling is required. The oversampling factor is defined as... (In this embodiment, take) ), the construction length is Discrete sequences ( Discrete-time signals The signal to be synthesized is generated by inverse fast Fourier transform, and thus the signal is represented as follows: (1) In formula (1), Indicates the length of the signal to be synthesized. Indicates inclusion A frequency domain vector with non-zero pitch and zero-filled points.

[0030] In this application, the optimization objective is to minimize the PAPR (Peak-to-Average Power Ratio) of the discrete signal shown in Equation (1). PAPR is an important indicator used in the fields of communication and signal processing to measure the fluctuation characteristics of a signal. It is defined as the ratio of the peak power to the average power of the signal.

[0031] Specifically, the PAPR of minimizing formula (1) is expressed as: (2) In step S110, to achieve seamless splicing without post-processing, this method involves a "splicing as design" spectral structure. For example... Figure 2 As shown, the total number is The spectral tones are divided into three consecutive sets: the initial overlapping segment set, the middle segment set, and the final overlapping segment set. The initial overlapping segment set... Include index The former A tone, used to represent the set of overlapping segments at the end of the previous frequency band. Overlapping. Set of intermediate segments Include index The middle Each tone represents an independent carrier portion of that frequency band. The set of overlapping segments at the end. Include index After A set of initial overlapping segments with the next frequency band. overlapping.

[0032] In this embodiment, the total number of pitches in each frequency band satisfies .

[0033] Furthermore, to improve the robustness of the splicing boundary in the presence of thermal noise and filter roll-off, this method introduces a power boost ratio in the overlapping region. (For example, take) (Approximately 7dB gain) means that the initial and final overlapping segments are assigned higher power weights than the middle segment set. A normalization constant is used to ensure total power conservation, thus generating the target amplitude spectrum. , is represented as: (3) in,

[0034] In the above formula, Indicates the target amplitude spectrum. Indicates the power boost ratio. , , These represent the initial set of overlapping segments, the middle set of overlapping segments, and the final set of overlapping segments, respectively. This represents the number of spectral tones in the initial and final sets of overlapping segments. This indicates the number of spectral tones in the middle segment set. Represents the normalization constant. This indicates the total number of spectral tones in each frequency band.

[0035] Specifically, normalization constant The calculation formula shows that as the power of the overlapping section increases ( Increase), overall amplitude benchmark The power will be reduced accordingly to maintain power conservation.

[0036] In step S120, this method models the waveform synthesis problem as finding the optimal complex vector. The process involves setting spectral constraints based on the target amplitude spectrum and three consecutive sets to ensure splicing coherence, balance PAPR and spectral distortion, and maintain power conservation during the solution process. These spectral constraints include three hard constraints: a phase symmetry constraint, which forces the initial and final overlapping segments within a single frequency band to be in phase; an amplitude fidelity constraint, which limits the deviation between the synthesized amplitude and the target amplitude spectrum to within a predetermined Euclidean sphere neighborhood; and a power conservation constraint, which forces the total power of the signals in each frequency band to be maintained at a predetermined target total power.

[0037] Specifically, constraint (C1), or phase symmetry, requires that the initial overlapping segment and the final overlapping segment have completely identical phases in order to ensure the inherent coherence of the waveform block in the frequency domain. This constraint eliminates the splicing discontinuities caused by phase randomness in traditional methods, and is expressed as: (4) Specifically, constraint (C2) refers to amplitude fidelity: allowing the synthesized amplitude to deviate from the target amplitude. However, it must be restricted to the neighborhood of the Euclidean sphere. This relaxation space provides degrees of freedom for reducing PAPR, expressed as: (5) Specifically, constraint (C3) is power conservation: strictly limiting total power. This prevents the algorithm from altering PAPR performance by simply scaling the signal amplitude.

[0038] In step S130, for the constraint optimization problem mentioned above, this method proposes an Adaptive Stitching-Aware Iterative Projection Algorithm (ASIPA). Its core lies in finding the intersection of the time-domain peak-clipping constraint set and the frequency-domain stitching constraint set through alternating projection. In each iteration, time-domain projection processing and frequency-domain constraint recovery are performed sequentially. Specifically, during time-domain projection processing, an inverse fast Fourier transform is performed on the frequency-domain vector obtained from the previous iteration to obtain the time-domain signal. A peak-clipping threshold is calculated based on the current adaptive peak-clipping ratio, and a hard peak-clipping operation is performed on the time-domain signal. During frequency-domain constraint recovery, the peak-clipping time-domain signal is converted back to the frequency domain using a fast Fourier transform and decomposed to obtain amplitude and phase. Phase consistency enforcement is performed based on phase symmetry constraints, weighted projection is performed using an adaptive relaxation factor based on amplitude fidelity constraints, and the frequency-domain vector is scaled based on power conservation constraints to obtain the frequency-domain vector for the current iteration.

[0039] Specifically, during time-domain projection processing, in the first... In the next iteration, the time-domain signal is calculated. Based on the current adaptive peak reduction ratio Calculate dynamic threshold Perform hard clipping on the signal: (6) The PAPR is reduced by time-domain projection processing, but out-of-band spectral regeneration and in-band distortion are introduced.

[0040] Specifically, during frequency domain constraint recovery, the following will be performed: Transform back to the frequency domain to obtain And decomposed into amplitude and phase During this process, phase consistency enforcement (constraint C1) is performed, and symmetry is enforced through vector averaging for overlapping regions. The composite phase is calculated using the following formula: (7) Next, the calculated composite phase Simultaneously assign a value to the initial segment and the last paragraph The phase remains unchanged in the middle segment.

[0041] Furthermore, perform amplitude relaxation projection (constraint C2) using an adaptive relaxation factor. The target spectrum and the peak-shaving spectrum are weighted and represented as follows: (8) Specifically, adaptive relaxation factor The larger the value, the more peak clipping features are retained (lower PAPR), but the greater the spectral distortion.

[0042] Furthermore, execute power normalization (i.e., constraint C3): for Scaling is performed to restore its total energy. .

[0043] To address the problem of non-convex optimization easily getting trapped in local optima, this method employs a dynamic parameter evolution strategy. Specifically, during each iteration, the moving average improvement rate of the peak-to-average power ratio is calculated in real time. When the moving average improvement rate falls below a preset stagnation threshold, the peak reduction ratio and the relaxation factor are adaptively updated.

[0044] Specifically, in each iteration, two parameters are adaptively updated. The first dynamic parameter is the peak reduction ratio. From the smaller (Radical peak shaving) gradually increases to (Reduce distortion). The second dynamic parameter is the relaxation factor. From the larger (Allowing significant deviations) Gradually decrease to (Strictly approximates the target spectrum).

[0045] Specifically, during each iteration, the moving average improvement rate of PAPR is calculated in real time. .when When the algorithm enters a stagnant period, parameter updates are triggered. During adaptive updates of the peak reduction ratio and the relaxation factor: the current peak reduction ratio is added to a preset smoothing factor to obtain the updated peak reduction ratio, and the current relaxation factor is subtracted from the preset smoothing factor to obtain the updated relaxation factor. The update process is specifically expressed as follows: (9) (10) In formulas (9) and (10), Set a smoothing factor (e.g., 0.01) to ensure a smooth parameter trajectory and avoid oscillations.

[0046] In this embodiment, to accurately identify the global convergence state of the algorithm and prevent premature termination or unnecessary redundant iterations caused by local fluctuations, a combined long- and short-time window steady-state detection mechanism is employed. During each iteration: the local fluctuations in the peak-to-average power ratio are calculated using a short-time observation window; when the local fluctuations are below a preset convergence threshold, it is marked as a steady state. The proportion of iterations in the steady state is statistically analyzed using a long-time observation window; when this proportion exceeds a preset confidence level, a global steady state is determined to have been reached, and the iteration terminates.

[0047] Specifically, firstly, utilize a shorter observation window. Calculate the local fluctuations of the PAPR mean If the fluctuation is below the strict convergence threshold If the current iteration is stable, it is marked as a stable state, which is then used as a criterion for local stability. Furthermore, a long window with a time span much larger than the short window is introduced. (For example, covering 300 iterations), used to count the percentage of iterations marked as stable within this long window. Only when this percentage exceeds the set confidence level... The algorithm is considered to have reached a global steady state and the iteration terminates only when the success rate reaches 95% (for example). This dual-window mechanism combines sensitivity and robustness, effectively filtering out transient disturbances during the optimization process and ensuring that the final output waveform solution is statistically convergent and stable.

[0048] In step S140, multiple optimal waveform segments generated in step S130 are processed, each corresponding to a different center frequency. Arranged sequentially in the frequency domain. Since all waveform segments satisfy the phase symmetry constraint (constraint C1) and the overlapping area has a high signal-to-noise ratio, the previous segment is directly... With the next paragraph Coherent splicing can be achieved by aligning and superimposing the signals to generate the final ultra-wideband excitation signal.

[0049] To verify the effectiveness of the proposed adaptive stitching sensing iterative projection algorithm in broadband waveform synthesis, numerical simulations were first performed. The simulation model adopted a baseband equivalent model, and the main parameters were designed to simulate actual physical limitations and establish rigorous theoretical benchmarks. Next, experimental measurements were conducted using modular instruments and a vector network analyzer. The system parameters and algorithm control parameters in this embodiment are shown in Tables 1 and 2, respectively.

[0050] Table 1 ASIPA Algorithm System Parameters

[0051] Table 2 Dynamic Control and Dual-Window Parameters of ASIPA Algorithm

[0052] To verify the effectiveness of this method in reducing PAPR and maintaining spectral splicing continuity in broadband waveform synthesis. Figure 4 , Figure 5 The optimized time-domain waveforms and spectral amplitudes under different initial phase conditions are presented for comparison. Experimental results show that, in this method, whether using a random phase sequence with an initial PAPR as high as 9.5 dB or a Schroeder phase sequence with an initial PAPR of 2.6 dB, after adaptive iterative projection, the PAPR of the final waveform converges to an extremely low level of approximately 2.2 dB, and the preset three-segment target spectral structure is accurately reconstructed in the spectrum; in particular, the power of the overlapping segment achieves the expected result. The algorithm achieves a power ratio increase of approximately 7 dB, while maintaining a flat mid-range. The measured power ratio deviates from the theoretical target by only 0.032%, demonstrating that the algorithm can maintain extremely high amplitude fidelity even under extremely low PAPR constraints. This effectively resolves the contradiction between low PAPR and strict spectral constraints in existing technologies.

[0053] Figure 6 and Figure 7 This indicates the phase distribution characteristics of the optimized signal and the verification of the phase difference at the splicing boundary. Figure 6 The middle segment exhibits a pseudo-random phase distribution that is conducive to PAPR suppression, while the overlapping segment exhibits a strictly controlled phase trajectory. Figure 7 The phase difference between the initial overlapping segment and the final overlapping segment was further verified. It is strictly zero across the entire bandwidth. Simulation results show that this method achieves endogenous coherence of "design as splicing" through built-in phase symmetry constraints, ensuring that there will be no phase discontinuity when subsequent frequency bands are cascaded, thus eliminating the need for complex post-processing calibration as required by existing methods.

[0054] Figures 8 to 9 This indicates the comparison results between this method and other existing methods in hardware-in-the-loop experiments. Figure 8 The results show that existing methods, due to the lack of joint optimization, result in high PAPR of the synthesized waveform, leading to hardware nonlinear distortion and significant amplitude ripple at the splicing boundaries, with mismatch errors as high as 4% and 8%, respectively. In contrast, the waveform generated by our method has a flat amplitude, and the relative amplitude mismatch error is consistently controlled within 0.8%. Figure 9 Phase error analysis shows that the phase mismatch error at the splicing point of the proposed method is strictly less than 0.3 degrees, exhibiting perfect linearity, while existing methods suffer from nonlinear phase jumps of 2 to 4 degrees. Experimental results demonstrate that the proposed method can effectively eliminate amplitude and phase distortion introduced by hardware nonlinearity, achieving high-precision broadband waveform synthesis.

[0055] In summary, experiments have demonstrated that this method, by establishing a three-segment spectrum splicing model, constructing phase symmetry constraints and amplitude projection operators, jointly iteratively optimizing the time-domain envelope and frequency-domain spectrum, and introducing a joint long-short time window steady-state detection mechanism, solves the problem of splicing boundary distortion caused by excessively high PAPR in broadband signal generation. In particular, it eliminates phase discontinuities caused by hardware nonlinear effects, achieving high-precision, low-PAPR seamless waveform synthesis in ultra-wideband microwave component characterization scenarios.

[0056] In the low PAPR adaptive spectrum stitching method characterized by the above broadband microwave components, the signal to be synthesized is first set to contain... Each tone has a frequency interval of 1 / 2. To accurately capture peak reproduction between Nyquist sampling points, an oversampling factor is defined. , build A discrete-time signal model of a point. Then, a three-segment spectrum splicing architecture is constructed. The spectral tones are divided into three consecutive sets: the initial overlapping segment. (including the previous) (tones), middle section (including the middle) (each tone) and the final overlapping section (including) (Each tone). This architecture aims to directly meet splicing requirements through design. Next, spectral amplitude and phase constraints are set. To enhance the noise immunity of the splicing boundaries, overlapping segments... and Assign a higher power weight than the middle section Simultaneously, phase symmetry constraints are applied, meaning the phase profiles of the initial and final overlapping segments remain consistent, thus enabling intrinsic coherent stitching without complex post-processing delay estimation.

[0057] Furthermore, an adaptive stitching sensing iterative projection algorithm is executed. This algorithm performs time-domain projection and frequency-domain constraint recovery once in each iteration. The time-domain projection transforms the frequency-domain signal to the time domain, and the dynamic peak reduction ratio is used as the basis for the transformation. The peak clipping threshold is calculated, and the signal amplitude is nonlinearly limited to reduce PAPR. Frequency domain constraint recovery transforms the clipped signal back to the frequency domain, extracting components within the effective bandwidth. Simultaneously, in each iteration, phase consistency updates preserve the clipped phase in intermediate segments, and for overlapping segments, the phasor mean at corresponding frequency points is calculated to perform symmetry, ensuring smooth splicing boundaries. Amplitude fidelity updates utilize relaxation factors. A weighted projection is performed between the target spectral amplitude and the peak-clipping amplitude to balance PAPR reduction and spectral distortion.

[0058] This method also implements a joint steady-state detection mechanism combining dynamic parameter control and long / short time windows. During the iteration process, the peak reduction ratio is adaptively adjusted based on the convergence state. and relaxation factor The algorithm smoothly transitions from an aggressive exploration phase to a conservative refinement phase. Simultaneously, it employs a dual-window patient stopping (DWPS) criterion, which incorporates both local stability and global convergence, to prevent the algorithm from getting trapped in local minima or terminating prematurely.

[0059] This method starts from the underlying mechanism of waveform synthesis and significantly improves the local signal-to-noise ratio at the splicing boundary through a three-segment power enhancement architecture, ensuring phase continuity even in the presence of thermal noise and hardware defects. The proposed phase symmetry constraint eliminates the dependence of traditional methods on complex delay estimation algorithms, achieving seamless splicing with "zero phase difference". Experimental verification shows that the phase mismatch error of the synthesized signal is less than 0.3 degrees, the amplitude discontinuity is negligible, and the PAPR performance is superior to existing technologies, providing a robust solution for high-precision nonlinear measurements in the millimeter-wave and terahertz bands. This method solves the problems of nonlinear distortion and phase discontinuity at the splicing boundary caused by high PAPR in existing technologies. It achieves high-fidelity test signal synthesis with ultra-wideband, low PAPR and intrinsic coherence.

[0060] It should be understood that, although Figure 1The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0061] In one embodiment, such as Figure 10 As shown, a low PAPR adaptive spectrum stitching device characterized by broadband microwave components is provided, comprising: a signal construction module 200, a spectrum tone division module 210, a spectrum constraint setting module 220, an optimal waveform segment solving module 230, and an ultra-wideband low PAPR excitation signal generation module 240, wherein: The signal construction module 200 is used to construct an initial signal based on a multi-tone signal model, and use the frequency domain vector of the initial signal as the signal to be synthesized.

[0062] The spectrum tone division module 210 is used to divide the multiple spectrum tones corresponding to each frequency band in the signal to be synthesized into three continuous sets, namely the initial overlapping segment set, the middle segment set, and the final overlapping segment set. The initial overlapping segment set is used to overlap and connect with the final overlapping segment set of the previous frequency band, and the final overlapping segment set is used to overlap and connect with the initial overlapping segment set of the next frequency band. The initial overlapping segment set and the final overlapping segment set are assigned power weights higher than those of the middle segment set. The total power is kept constant by a normalization constant to generate the target amplitude spectrum.

[0063] The spectrum constraint setting module 220 is used to set spectrum constraints based on the target amplitude spectrum and the three consecutive sets to ensure splicing coherence, balance PAPR and spectrum distortion, and maintain power conservation.

[0064] The optimal waveform segment solution module 230 is used to perform an adaptive stitching sensing iterative projection algorithm on the frequency domain vector of the signal to be synthesized under the constraints of the spectral constraints, so as to achieve the cooperative optimization of minimizing PAPR and satisfying the spectral constraints, and obtain the optimal waveform segments corresponding to different center frequencies.

[0065] The ultra-wideband low PAPR excitation signal generation module 240 is used to arrange and splice the optimal waveform segments in frequency order. During this process, the set of overlapping segments at the end of the previous frequency band overlaps and connects with the set of overlapping segments at the beginning of the next frequency band to complete the splicing of the ultra-wideband signal and obtain the ultra-wideband low PAPR excitation signal.

[0066] Specific limitations regarding the low PAPR adaptive spectrum stitching device characterized by broadband microwave components can be found in the limitations of the low PAPR adaptive spectrum stitching method characterized by broadband microwave components described above, and will not be repeated here. Each module in the aforementioned low PAPR adaptive spectrum stitching device characterized by broadband microwave components can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in hardware or independently of the processor in a computer device, or stored in software in the memory of a computer device, so that the processor can call and execute the operations corresponding to each module.

[0067] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 11 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements a low PAPR adaptive spectrum stitching method characterized by a broadband microwave component. The display screen can be a liquid crystal display (LCD) or an e-ink display. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device casing, or an external keyboard, touchpad, or mouse.

[0068] Those skilled in the art will understand that Figure 11 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0069] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps: Based on the multi-tone signal model, an initial signal is constructed, and the frequency domain vector of the initial signal is used as the signal to be synthesized. The multiple spectral tones corresponding to each frequency band in the signal to be synthesized are divided into three continuous sets: an initial overlapping segment set, an intermediate segment set, and an end overlapping segment set. The initial overlapping segment set is used to overlap and connect with the end overlapping segment set of the previous frequency band, and the end overlapping segment set is used to overlap and connect with the initial overlapping segment set of the next frequency band. The initial overlapping segment set and the end overlapping segment set are assigned higher power weights than the intermediate segment set. The total power is kept constant by a normalization constant to generate the target amplitude spectrum. Based on the target amplitude spectrum and the three consecutive sets, spectral constraints are set to ensure splicing coherence, balance PAPR and spectral distortion, and maintain power conservation. Under the constraints of the spectral constraints, an adaptive stitching sensing iterative projection algorithm is executed on the frequency domain vector of the signal to be synthesized to achieve a synergistic optimization that minimizes PAPR and satisfies the spectral constraints, thereby obtaining the optimal waveform segments corresponding to different center frequencies. The optimal waveform segments are arranged and spliced ​​in frequency order. During this process, the overlapping segments at the end of the previous frequency band are overlapped and connected with the initial overlapping segments of the next frequency band to complete the splicing of the ultra-wideband signal and obtain the ultra-wideband low PAPR excitation signal.

[0070] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Based on the multi-tone signal model, an initial signal is constructed, and the frequency domain vector of the initial signal is used as the signal to be synthesized. The multiple spectral tones corresponding to each frequency band in the signal to be synthesized are divided into three continuous sets: an initial overlapping segment set, an intermediate segment set, and an end overlapping segment set. The initial overlapping segment set is used to overlap and connect with the end overlapping segment set of the previous frequency band, and the end overlapping segment set is used to overlap and connect with the initial overlapping segment set of the next frequency band. The initial overlapping segment set and the end overlapping segment set are assigned higher power weights than the intermediate segment set. The total power is kept constant by a normalization constant to generate the target amplitude spectrum. Based on the target amplitude spectrum and the three consecutive sets, spectral constraints are set to ensure splicing coherence, balance PAPR and spectral distortion, and maintain power conservation. Under the constraints of the spectral constraints, an adaptive stitching sensing iterative projection algorithm is executed on the frequency domain vector of the signal to be synthesized to achieve a synergistic optimization that minimizes PAPR and satisfies the spectral constraints, thereby obtaining the optimal waveform segments corresponding to different center frequencies. The optimal waveform segments are arranged and spliced ​​in frequency order. During this process, the overlapping segments at the end of the previous frequency band are overlapped and connected with the initial overlapping segments of the next frequency band to complete the splicing of the ultra-wideband signal and obtain the ultra-wideband low PAPR excitation signal.

[0071] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0072] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0073] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A low PAPR adaptive spectrum stitching method for characterizing broadband microwave components, characterized in that, The method includes: Based on the multi-tone signal model, an initial signal is constructed, and the frequency domain vector of the initial signal is used as the signal to be synthesized. The multiple spectral tones corresponding to each frequency band in the signal to be synthesized are divided into three continuous sets: an initial overlapping segment set, an intermediate segment set, and an end overlapping segment set. The initial overlapping segment set is used to overlap and connect with the end overlapping segment set of the previous frequency band, and the end overlapping segment set is used to overlap and connect with the initial overlapping segment set of the next frequency band. The initial overlapping segment set and the end overlapping segment set are assigned higher power weights than the intermediate segment set. The total power is kept constant by a normalization constant to generate the target amplitude spectrum. Based on the target amplitude spectrum and the three consecutive sets, spectral constraints are set to ensure splicing coherence, balance PAPR and spectral distortion, and maintain power conservation. Under the constraints of the spectral constraints, an adaptive stitching sensing iterative projection algorithm is executed on the frequency domain vector of the signal to be synthesized to achieve a synergistic optimization that minimizes PAPR and satisfies the spectral constraints, thereby obtaining the optimal waveform segments corresponding to different center frequencies. The optimal waveform segments are arranged and spliced ​​in frequency order. During this process, the overlapping segments at the end of the previous frequency band are overlapped and connected with the initial overlapping segments of the next frequency band to complete the splicing of the ultra-wideband signal and obtain the ultra-wideband low PAPR excitation signal.

2. The low PAPR adaptive spectrum stitching method for characterizing broadband microwave components according to claim 1, characterized in that, The signal to be synthesized is represented as follows: In the above formula, This indicates the length of the signal to be synthesized. Indicates inclusion A frequency domain vector with non-zero pitch and zero-filled points.

3. The low PAPR adaptive spectrum stitching method for characterizing broadband microwave components according to claim 2, characterized in that, Assign higher power weights than the intermediate segment set to the initial overlapping segment set and the final overlapping segment set, and ensure the total power is conserved by a normalization constant to generate the target amplitude spectrum. The process is as follows: in, In the above formula, This represents the target amplitude spectrum. Indicates the power boost ratio. , , These represent the initial set of overlapping segments, the middle set of overlapping segments, and the final set of overlapping segments, respectively. This represents the number of spectral tones in the initial set of overlapping segments and the final set of overlapping segments. This indicates the number of spectral tones in the set of intermediate segments. Represents the normalization constant. This indicates the total number of spectral tones in each frequency band.

4. The low PAPR adaptive spectrum stitching method for characterizing broadband microwave components according to claim 3, characterized in that, The spectral constraints include three hard constraints, namely: Phase symmetry constraint, which forces the initial overlapping segment set and the last overlapping segment set within a single frequency band to be completely consistent in phase; Amplitude fidelity constraint, wherein the deviation between the synthesized amplitude and the target amplitude spectrum is limited to within a preset Euclidean sphere neighborhood; The power conservation constraint forces the total power of signals in each frequency band to be maintained at a preset target total power.

5. The low PAPR adaptive spectrum stitching method for characterizing broadband microwave components according to claim 4, characterized in that, The adaptive stitching sensing iterative projection algorithm performs time-domain projection processing and frequency-domain constraint recovery sequentially during each iteration. When performing the time-domain projection processing, the inverse fast Fourier transform is performed on the frequency domain vector obtained in the previous iteration to obtain the time-domain signal. The peak-shaving threshold is calculated based on the current adaptive peak-shaving ratio, and a hard peak-shaving operation is performed on the time-domain signal. During the frequency domain constraint recovery, the peak-shaving time domain signal is converted back to the frequency domain through a fast Fourier transform and decomposed to obtain amplitude and phase. Phase consistency is enforced based on the phase symmetry constraint, and weighted projection is performed using an adaptive relaxation factor based on the amplitude fidelity constraint. Then, the frequency domain vector is scaled based on the power conservation constraint to obtain the frequency domain vector of the current iteration.

6. The low PAPR adaptive spectrum stitching method for characterizing broadband microwave components according to claim 5, characterized in that, In each iteration: The moving average improvement rate of the peak-to-average power ratio is calculated in real time. When the moving average improvement rate is less than a preset stagnation threshold, the peak reduction ratio and the relaxation factor are adaptively updated.

7. The low PAPR adaptive spectrum stitching method for characterizing broadband microwave components according to claim 6, characterized in that, When the peak reduction ratio and the relaxation factor are adaptively updated: Add the preset smoothing factor to the current peak reduction ratio to obtain the updated peak reduction ratio; The updated relaxation factor is obtained by subtracting the preset smoothing factor from the current relaxation factor.

8. The low PAPR adaptive spectrum stitching method for characterizing broadband microwave components according to claim 5, characterized in that, In each iteration: The local fluctuations of the mean peak-to-average power ratio are calculated using a short-time observation window, and the local fluctuations are marked as a steady state when they are below a preset convergence threshold. The percentage of steady-state iterations is statistically analyzed using a long-term observation window. When the percentage exceeds a preset confidence level, it is determined that a global steady state has been reached and the iteration is terminated.

9. A low PAPR adaptive spectrum splicing device characterized by broadband microwave components, characterized in that, The device includes: The signal to be synthesized module is used to construct an initial signal based on a multi-tone signal model, and use the frequency domain vector of the initial signal as the signal to be synthesized. The spectrum tone division module is used to divide the multiple spectrum tones corresponding to each frequency band in the signal to be synthesized into three consecutive sets: an initial overlapping segment set, an intermediate segment set, and an end overlapping segment set. The initial overlapping segment set is used to overlap and connect with the end overlapping segment set of the previous frequency band, and the end overlapping segment set is used to overlap and connect with the initial overlapping segment set of the next frequency band. The module assigns a higher power weight to the initial overlapping segment set and the end overlapping segment set than to the intermediate segment set, and ensures the total power is conserved through a normalization constant to generate the target amplitude spectrum. The spectrum constraint setting module is used to set spectrum constraints based on the target amplitude spectrum and the three consecutive sets to ensure splicing coherence, balance PAPR and spectrum distortion, and maintain power conservation. The optimal waveform segment solution module is used to perform an adaptive stitching sensing iterative projection algorithm on the frequency domain vector of the signal to be synthesized under the constraints of the spectral constraints, so as to achieve the cooperative optimization of minimizing PAPR and satisfying the spectral constraints, and obtain the optimal waveform segments corresponding to different center frequencies. The ultra-wideband low PAPR excitation signal generation module is used to arrange and splice the optimal waveform segments in frequency order. During this process, the set of overlapping segments at the end of the previous frequency band overlaps and connects with the set of overlapping segments at the beginning of the next frequency band to complete the splicing of the ultra-wideband signal and obtain the ultra-wideband low PAPR excitation signal.

Images