Method, system, device, processor and storage medium for realizing wideband signal transmitting equipment amplitude and phase flatness analysis based on comb spectrum

Abstract

The present application relates to a kind of based on comb spectrum implementation wideband signal transmitting device amplitude phase flatness analytical method, comprising the following steps acquisition measured link's time domain complex IQ sampling data;Perform fast Fourier transform;Zero frequency component is moved to spectrum center;Calculate the index of each target frequency in spectrum array;The amplitude and phase at each index are calculated;Calculate preliminary amplitude error and phase error;Get amplitude flatness distortion;Interlaced phase correction is carried out;Phase unwrapping is carried out;Get phase flatness distortion;Filter out abnormal value;Output amplitude flatness distortion and phase flatness distortion result.The wideband signal transmitting device amplitude phase flatness analytical method based on comb spectrum implementation of the present application, system, device, processor and its computer readable storage medium, around the flatness calibration demand of wideband wireless signal transmitting device, obtains the flatness information of transmitting device from comb spectrum, realizes the asynchronous acquisition and efficient analysis without synchronization.

Description

Technical Field

[0001] This invention relates to the field of wireless communication, and more particularly to the field of signal processing and analysis. Specifically, it relates to a method, system, apparatus, processor, and computer-readable storage medium for analyzing the amplitude and phase flatness of broadband signal transmitting devices based on comb spectrum. Background Technology

[0002] With the rapid evolution of modern communication technologies, especially the development of systems for 5G-Advanced and 6G, the bandwidth of signal transmission is growing at an unprecedented rate. In ultra-wideband communication systems, the uniformity of amplitude and phase response of signal transmitting equipment across the entire operating frequency band, i.e., flatness, becomes a key indicator determining system performance. Any amplitude and phase flatness will directly lead to signal waveform distortion. For signals using complex modulation methods such as high-order quadrature amplitude modulation (QAM), this distortion will worsen the error vector amplitude (EVM), increase the bit error rate, and in severe cases, even lead to communication link interruption. Therefore, accurate amplitude and phase flatness calibration of broadband signal transmitting equipment is a necessary step to ensure its performance meets standards.

[0003] like Figure 1 As shown, the basic principle of flatness calibration is to construct a closed-loop measurement and compensation system: First, the transmitting equipment transmits a reference signal with known amplitude and phase characteristics under ideal conditions. Then, a precisely calibrated signal acquisition device with sufficiently ideal flatness receives this signal. By comparing and analyzing the difference between the acquired actual signal and the original ideal signal, the amplitude and phase distortion introduced by the transmission link is accurately measured. Finally, this distortion information is fed back to the digital baseband processing unit of the transmitting equipment, and reverse compensation is performed through pre-distortion technology, thereby correcting the flatness of the transmitted signal and ensuring that its output signal at the air interface has flat amplitude and phase characteristics.

[0004] In existing technical practices, there are various schemes for flatness calibration, but they all have their own limitations.

[0005] A traditional and widely used method is to use a frequency sweep signal. This method involves sequentially transmitting a series of single-tone sinusoidal signals across the entire target frequency band and measuring their amplitude response point by point, ultimately stitching them together to form a complete amplitude flatness curve. The main drawback of this method is its inability to acquire phase information. Because each frequency point is measured independently during the sweep process, the signal's phase reference is lost during frequency switching, making it impossible to establish a continuous phase relationship across the entire frequency band. For modern communication systems, nonlinear phase distortion (i.e., group delay variation) is also a significant factor contributing to signal quality degradation; simply calibrating amplitude flatness is insufficient and cannot meet the requirements of high-performance systems. Furthermore, the frequency sweep measurement process is time-consuming, especially in broadband testing scenarios requiring high frequency resolution, making it inefficient and unsuitable for fast-paced production testing environments.

[0006] To simultaneously acquire amplitude and phase information, a more advanced approach is to use known multi-tone signals or comb spectrum signals as test excitation. This approach transmits a discrete spectrum consisting of several equally spaced subcarriers in a single transmission, theoretically allowing for the acquisition of full-band amplitude and phase information in a single sampling. Implementation typically employs two synchronization methods. The first method establishes a hardware trigger channel between the transmitter under test (DUT) and the receiver, ensuring alignment of sampling times on both sides through a common reference clock or external trigger line. This method requires additional high-precision clock modules, trigger wiring, and synchronization logic, resulting in complex system integration and stringent experimental environment requirements. The second method inserts a synchronization header before the comb spectrum sequence, using a back-end algorithm to search for the synchronization position and achieve alignment. This method reduces hardware dependence but requires performing correlation search, matched filtering, and threshold decision within the received data, leading to a large computational load. Furthermore, if the synchronization header is overwhelmed by multipath fading or interference from adjacent signals, the search process is prone to failure, causing the entire flatness calculation to fail.

[0007] Furthermore, existing analytical methods based on comb spectra generally suffer from sensitivity to inherent system defects and insufficient robustness. For example, the unavoidable carrier frequency offset and sampling clock offset between the transmitting and acquiring equipment introduce a linear phase ramp across the entire frequency band. This ramp, combined with the actual phase distortion of the equipment, severely affects the accuracy of phase flatness measurements if the algorithm cannot effectively separate it. Moreover, phase unwrapping is a crucial but challenging step in phase analysis. Since the phase calculation results are limited to ±180 degrees, a jump occurs when the cumulative change in the actual phase exceeds 360 degrees. In harsh conditions with noise or specific group delays (e.g., causing adjacent spectral lines to have a phase difference close to 180 degrees), traditional phase unwrapping algorithms are highly prone to errors, leading to significant errors or even complete failure in the final phase flatness results. These problems indicate that existing technologies still have significant shortcomings in ensuring high efficiency, high accuracy, and high reliability in broadband amplitude-phase flatness analysis, necessitating an analytical method that requires no external synchronization, is simple in algorithm, and yields robust results. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method, system, device, processor and computer-readable storage medium for analyzing the amplitude and phase flatness of broadband signal transmitting equipment based on comb spectrum, which is simple in algorithm, robust in results and widely applicable.

[0009] To achieve the above objectives, the present invention provides a method, system, apparatus, processor, and computer-readable storage medium for analyzing the amplitude and phase flatness of broadband signal transmitting devices based on comb spectra, as follows:

[0010] The method for analyzing the amplitude and phase flatness of broadband signal transmitting equipment based on comb spectrum is characterized by the following steps:

[0011] (1) Collect time-domain complex IQ sampling data of the link under test;

[0012] (2) Perform a fast Fourier transform on the IQ sampled data to obtain the complex spectrum;

[0013] (3) Shift the zero-frequency component to the center of the spectrum;

[0014] (4) Calculate the index of each target frequency in the spectrum array according to the preset comb tooth interval, and extract the corresponding complex value;

[0015] (5) Calculate the amplitude and phase at each index;

[0016] (6) Compare the actual amplitude and phase with the ideal reference values ​​to calculate the preliminary amplitude error and phase error;

[0017] (7) Correct the amplitude error; correct the phase error;

[0018] (8) Filter out outliers. Apply sliding window Hampel filters to the amplitude flatness distortion sequence and the phase flatness distortion sequence respectively to replace isolated outliers.

[0019] (9) Output amplitude flatness distortion and phase flatness distortion results.

[0020] Preferably, the step (7) of correcting the amplitude error specifically includes the following steps:

[0021] (1-7.1) Calculate the arithmetic mean of all amplitude errors, correct the amplitude errors, obtain amplitude flatness distortion, and continue to step (8).

[0022] Preferably, the phase error correction in step (7) specifically includes the following steps:

[0023] (2-7.1) Perform half-cycle ambiguity detection and staggered phase correction;

[0024] (2-7.2) Perform phase unwinding;

[0025] (2-7.3) Obtain the linear phase slope by least squares fitting and remove it from the continuous phase curve to obtain the phase flatness distortion, and continue to step (8).

[0026] Preferably, step (2-7.1) specifically includes the following steps:

[0027] Determine if the phase difference between adjacent comb teeth is close to 180°. If so, perform an interleaving operation of adding or subtracting 180° on the phase sequence; otherwise, continue to step (2-7.2).

[0028] Preferably, step (2-7.1) further includes the following steps:

[0029] Determine if the Nth smallest value of the set of absolute values ​​of the phase difference between adjacent comb teeth is greater than 160°. If so, there is a half-cycle ambiguity, and staggered phase correction is performed.

[0030] Preferably, step (2-7.2) specifically includes the following steps:

[0031] The corrected phase sequence is compensated for with ±360° jumps to obtain a continuous phase curve.

[0032] Preferably, the phase unwinding in step (2-7.2) employs a recursive compensation method to keep the continuous phase difference within the range of -180° to +180°.

[0033] Preferably, the window length of the sliding window Hampel filter in step (8) is 2w+1, the range of w is 4 to 8, and the range of the threshold coefficient τ is 2.5 to 3.5.

[0034] Preferably, the comb tooth spacing in step (4) is in the range of 10kHz to 5MHz, and the number of comb teeth that can be configured is in the range of 8 to 65536.

[0035] Preferably, the residual of amplitude flatness distortion in step (1-7.1) does not exceed ±0.1dB.

[0036] Preferably, the phase flatness distortion residual of step (2-7.3) does not exceed ±0.5°.

[0037] The system for analyzing the amplitude and phase flatness of broadband signal transmitting equipment based on comb spectrum is characterized by the following:

[0038] The data interface module is used for time-domain complex IQ sampling data of the link under test;

[0039] The FFT processing module, connected to the data interface module, is used to perform a fast Fourier transform on the IQ sampled data.

[0040] The amplitude and phase calculation module is connected to the FFT processing module and is used to calculate the amplitude and phase at each index.

[0041] The systematic error correction module is connected to the amplitude and phase calculation module and is used to perform amplitude correction to obtain amplitude flatness distortion, perform half-cycle ambiguity detection, and perform staggered phase correction.

[0042] The outlier filtering module, connected to the systematic error correction module, is used to apply a sliding window Hampel filter to the amplitude flatness distortion sequence and the phase flatness distortion sequence respectively, replacing isolated outliers.

[0043] The result output module is connected to the outlier filtering module and is used to output amplitude flatness distortion and phase flatness distortion results.

[0044] The main feature of this device for resolving the amplitude and phase flatness of broadband signal transmitting equipment based on comb spectrum is that the device comprises:

[0045] A processor is configured to execute computer-executable instructions;

[0046] The memory stores one or more computer-executable instructions, which, when executed by the processor, implement the various steps of the method described above for resolving the amplitude and phase flatness of a broadband signal transmitting device based on comb spectrum.

[0047] The processor for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum is characterized in that the processor is configured to execute computer-executable instructions, which, when executed by the processor, implement the various steps of the method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum.

[0048] The main feature of this computer-readable storage medium is that it stores a computer program thereon, which can be executed by a processor to implement the various steps of the above-described method for resolving the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum.

[0049] This invention employs a method, system, apparatus, processor, and computer-readable storage medium for analyzing the amplitude and phase flatness of broadband signal transmitting devices based on comb spectra. Addressing the flatness calibration requirements of broadband wireless signal transmitting devices, it calculates the flatness information of the transmitting device from the comb spectrum. This invention achieves asynchronous acquisition and efficient analysis without synchronization, possesses the advantage of complete amplitude-phase joint calculation, exhibits high reliability independent of time delay and phase deviation, demonstrates excellent robustness against spurious interference, offers high accuracy and broadband adaptability, has low computational complexity, and possesses potential for real-time processing and embedded applications. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of the device structure for the method of realizing amplitude and phase flatness analysis of broadband signal transmitting equipment based on comb spectrum according to the present invention.

[0051] Figure 2 This is a flowchart of the method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on comb spectrum according to the present invention.

[0052] Figure 3 The spectrum diagram of the signal transmitting device emitting a comb spectrum is shown in the present invention, which is a method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum.

[0053] Figure 4 The flatness distortion result obtained from the analysis of the method for analyzing the amplitude and phase flatness of broadband signal transmitting equipment based on comb spectrum according to the present invention. Detailed Implementation

[0054] To more clearly describe the technical content of the present invention, the following description is provided in conjunction with specific embodiments.

[0055] The method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on comb spectrum according to the present invention includes the following steps:

[0056] (1) Collect time-domain complex IQ sampling data of the link under test;

[0057] (2) Perform a fast Fourier transform on the IQ sampled data to obtain the complex spectrum;

[0058] (3) Shift the zero-frequency component to the center of the spectrum;

[0059] (4) Calculate the index of each target frequency in the spectrum array according to the preset comb tooth interval, and extract the corresponding complex value;

[0060] (5) Calculate the amplitude and phase at each index;

[0061] (6) Compare the actual amplitude and phase with the ideal reference values ​​to calculate the preliminary amplitude error and phase error;

[0062] (7) Correct the amplitude error; correct the phase error;

[0063] (8) Filter out outliers. Apply sliding window Hampel filters to the amplitude flatness distortion sequence and the phase flatness distortion sequence respectively to replace isolated outliers.

[0064] (9) Output amplitude flatness distortion and phase flatness distortion results.

[0065] In a preferred embodiment of the present invention, the step (7) of correcting the amplitude error specifically includes the following steps:

[0066] (1-7.1) Calculate the arithmetic mean of all amplitude errors, correct the amplitude errors, obtain amplitude flatness distortion, and continue to step (8).

[0067] In a preferred embodiment of the present invention, the phase error correction in step (7) specifically includes the following steps:

[0068] (2-7.1) Perform half-cycle ambiguity detection and staggered phase correction;

[0069] (2-7.2) Perform phase unwinding;

[0070] (2-7.3) Obtain the linear phase slope by least squares fitting and remove it from the continuous phase curve to obtain the phase flatness distortion, and continue to step (8).

[0071] In a preferred embodiment of the present invention, step (2-7.1) specifically includes the following steps:

[0072] Determine if the phase difference between adjacent comb teeth is close to 180°. If so, perform an interleaving operation of adding or subtracting 180° on the phase sequence; otherwise, continue to step (2-7.2).

[0073] In a preferred embodiment of the present invention, step (2-7.1) further includes the following step:

[0074] Determine if the Nth smallest value of the set of absolute values ​​of the phase difference between adjacent comb teeth is greater than 160°. If so, there is a half-cycle ambiguity, and staggered phase correction is performed.

[0075] In a preferred embodiment of the present invention, step (2-7.2) specifically includes the following steps:

[0076] The corrected phase sequence is compensated for with ±360° jumps to obtain a continuous phase curve.

[0077] In a preferred embodiment of the present invention, the phase unwinding in step (2-7.2) adopts a recursive compensation method to keep the continuous phase difference within the range of -180° to +180°.

[0078] In a preferred embodiment of the present invention, the window length of the sliding window Hampel filter in step (8) is 2w+1, the range of w is 4 to 8, and the range of the threshold coefficient τ is 2.5 to 3.5.

[0079] In a preferred embodiment of the present invention, the comb tooth spacing in step (4) is in the range of 10kHz to 5MHz, and the number of comb teeth that can be configured is in the range of 8 to 65536.

[0080] In a preferred embodiment of the present invention, the residual of amplitude flatness distortion in step (1-7.1) does not exceed ±0.1dB.

[0081] In a preferred embodiment of the present invention, the phase flatness distortion residual of step (2-7.3) does not exceed ±0.5°.

[0082] The system for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on comb spectrum according to the present invention, wherein the system comprises:

[0083] The data interface module is used for time-domain complex IQ sampling data of the link under test;

[0084] The FFT processing module, connected to the data interface module, is used to perform a fast Fourier transform on the IQ sampled data.

[0085] The amplitude and phase calculation module is connected to the FFT processing module and is used to calculate the amplitude and phase at each index.

[0086] The systematic error correction module is connected to the amplitude and phase calculation module and is used to perform amplitude correction to obtain amplitude flatness distortion, perform half-cycle ambiguity detection, and perform staggered phase correction.

[0087] The outlier filtering module, connected to the systematic error correction module, is used to apply a sliding window Hampel filter to the amplitude flatness distortion sequence and the phase flatness distortion sequence respectively, replacing isolated outliers.

[0088] The result output module is connected to the outlier filtering module and is used to output amplitude flatness distortion and phase flatness distortion results.

[0089] The apparatus of the present invention for resolving the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum, wherein the apparatus comprises:

[0090] A processor is configured to execute computer-executable instructions;

[0091] The memory stores one or more computer-executable instructions, which, when executed by the processor, implement the various steps of the method described above for resolving the amplitude and phase flatness of a broadband signal transmitting device based on comb spectrum.

[0092] The processor of the present invention for realizing amplitude and phase flatness analysis of broadband signal transmitting equipment based on comb spectrum is configured to execute computer-executable instructions, which, when executed by the processor, implement the various steps of the above-described method for realizing amplitude and phase flatness analysis of broadband signal transmitting equipment based on comb spectrum.

[0093] The computer-readable storage medium of the present invention stores a computer program that can be executed by a processor to implement the various steps of the above-described method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum.

[0094] The present invention provides a method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum, comprising the following steps:

[0095] (1) Acquisition: Acquire time-domain complex IQ sampling data of the link under test without any hardware synchronization;

[0096] (2) Frequency domain transformation: Perform a fast Fourier transform on the IQ sampled data to obtain the complex spectrum;

[0097] (3) Spectrum centering: Shifting the zero-frequency component to the center of the spectrum;

[0098] (4) Target frequency positioning: Calculate the index of each target frequency in the spectrum array according to the preset comb tooth interval and extract the corresponding complex value;

[0099] (5) Amplitude and phase calculation: Calculate the amplitude and phase at each index;

[0100] (6) Preliminary error calculation: Compare the actual amplitude and phase with the ideal reference value to obtain the preliminary amplitude error and phase error; then there are two branches (1-7.1) and (2-7.1) to process the amplitude and phase respectively, which are executed in parallel and are independent of each other;

[0101] (1-7.1) Amplitude correction, i.e. Figure 2 Step 6a above: Calculate the arithmetic mean of all amplitude errors and cancel them out to obtain amplitude flatness distortion; continue to step (10).

[0102] (2-7.1) Semi-periodic ambiguity detection and staggered phase correction, i.e. Figure 2 Steps 6b and 6.1 above: Determine whether the phase difference between adjacent comb teeth is close to 180°. If the preset condition is met, perform an interleaved addition and subtraction of 180° on the phase sequence; continue with steps (2-7.2).

[0103] (2-7.2) Phase unwinding: Perform ±360° jump compensation on the corrected phase sequence to obtain a continuous phase curve; continue with step (2-7.3).

[0104] (2-7.3) Linear trend removal: Obtain the linear phase slope through least squares fitting and remove it from the continuous phase curve to obtain phase flatness distortion; continue to step (10).

[0105] (8) Outlier removal: Apply a sliding window Hampel filter to the amplitude flatness distortion sequence after (1-7.1), and apply a sliding window Hampel filter to the phase flatness distortion sequence after (2-7.3) to replace isolated outliers;

[0106] (9) Output: Give the final amplitude flatness distortion and phase flatness distortion results.

[0107] The time-domain signal acquired in step (1) can start at any time point without the need for a trigger line or a common reference clock.

[0108] In step (2-7.1), when the Nth smallest value of the set of absolute values ​​of the phase difference between adjacent comb teeth is greater than 160°, it is determined that there is a half-cycle ambiguity and staggered phase correction is performed.

[0109] The phase unwinding in step (2-7.2) uses a recursive compensation method to keep the continuous phase difference within the range of -180° to +180°.

[0110] In step (9), the window length of the sliding window Hampel filter is 2w+1, where w is 4 to 8 and the threshold coefficient τ is 2.5 to 3.5.

[0111] The comb tooth spacing is 10kHz to 5MHz, and the number of comb teeth can be configured from 8 to 65536.

[0112] The resolution accuracy meets the following requirements: amplitude flatness distortion residuals do not exceed ±0.1dB, and phase flatness distortion residuals do not exceed ±0.5°.

[0113] An amplitude and phase flatness analysis device includes a data interface module, an FFT processing module, an amplitude and phase calculation module, a systematic error correction module, an outlier filtering module, and a result output module. Each module is connected and operates in the order of the corresponding steps described above.

[0114] Figure 1 The diagram illustrates the flatness calibration principle of a signal transmitting device based on a comb spectrum, a closed-loop feedback calibration system. First, the host computer sends preset comb spectrum data with ideal amplitude and phase characteristics to the digital baseband (FPGA) module of the (to-be-calibrated) transmitting device via a "signal transmission" link. This digital signal is processed sequentially by a digital-to-analog converter (DAC) and an RF transmission circuit within the transmitting device, becoming an RF analog signal before being transmitted. This process introduces amplitude and phase unevenness into the transmitting device's own link. Next, the RF signal is "acquired" by a signal acquisition device with relatively ideal flatness, and the acquired data is transmitted back to the host computer. Finally, the flatness result analysis algorithm within the host computer compares and analyzes the acquired actual signal with the original ideal comb spectrum signal, thereby accurately calculating the amplitude and phase unevenness of the entire transmitting device link and obtaining the calibration result for subsequent compensation. This invention focuses on the implementation of the "flatness result analysis algorithm."

[0115] The core objective of the algorithm is to accurately calculate the amplitude and phase flatness distortion of the comb spectrum signal across the entire frequency band from the sampled data of the real (I) and imaginary (Q) parts of the signal acquired by the signal receiving device. The specific implementation steps are as follows:

[0116] 1. Time-domain to frequency-domain transformation

[0117] The algorithm starts with IQ sampled data in the time domain. First, this set of discrete time-domain complex signals (I for the real part, Q for the imaginary part) is used as input, and a Fast Fourier Transform (FFT) is performed. This step transforms the signal from the time domain to the frequency domain, resulting in a complex spectrum. Each point in this complex spectrum corresponds to a specific frequency and contains amplitude and phase information at that frequency.

[0118] 2. Spectrum centralization

[0119] In a standard FFT output, the zero-frequency component is located at the beginning of the array. To facilitate analysis, the algorithm performs a shift operation on the spectral data, moving the zero-frequency point to the center of the spectrum, with positive and negative frequencies on either side, forming a visually intuitive spectral arrangement symmetrical around the zero point.

[0120] 3. Extract target frequency point location and information

[0121] The algorithm precisely locates a target frequency on a transformed complex spectrum based on a predefined set of discrete target frequency points (i.e., the center frequency of each spectral line or "tooth" in the comb spectrum). For each target frequency, its precise index position in the spectrum array is calculated using the ratio of its frequency value to the FFT frequency resolution. The corresponding complex value is then extracted from this index position.

[0122] 4. Calculate the actual amplitude and phase.

[0123] For each target frequency point, the extracted complex value is calculated as follows:

[0124] Amplitude calculation: The actual measured amplitude of the signal at that frequency point is obtained by calculating the modulus of the complex number (i.e., the square root of the sum of the squares of its real and imaginary parts).

[0125] Phase calculation: The actual measured phase of the signal at that frequency point is obtained by calculating the argument of the complex number (i.e., the arctangent of the ratio of the imaginary part to the real part).

[0126] 5. Preliminary calculation of amplitude and phase error

[0127] The actual measured value obtained in the previous step is compared with the ideal reference value to obtain the preliminary error:

[0128] Amplitude error: Divide the actual measured amplitude by the ideal amplitude, then take the logarithm to the base 10 and multiply by 20 to calculate the difference in decibels (dB).

[0129] Phase error: The actual measured phase is directly subtracted from the ideal reference phase. To ensure consistency, the result is normalized to the range of -180 degrees to +180 degrees.

[0130] 6. Systematic error correction

[0131] The initial calculated error includes the overall offset and linear trend of the system, which needs to be corrected to extract the pure flatness distortion.

[0132] Amplitude correction: Calculate the arithmetic mean of the amplitude errors at all target frequency points. This average represents the overall power gain or attenuation shift of the system. Subsequently, subtract this average from the amplitude error at each frequency point to eliminate the overall shift; the remaining fluctuation is the amplitude flatness distortion.

[0133] Phase correction: This step aims to eliminate the linear phase slope caused by the fixed transmission delay (group delay) and isolate the nonlinear phase distortion.

[0134] 6.1 Half-Period Ambiguity Correction

[0135] Before performing standard phase correction, a special case must be addressed. When the system's group delay is exactly an integer multiple of the fundamental period of the measured signal plus half a period, the phase difference between adjacent spectral lines will approach 180 degrees. This drastic alternation, close to the phase folding boundary, will cause subsequent phase unwinding algorithms to fail.

[0136] To solve this problem, the algorithm first performs a detection and correction process:

[0137] 1. State Detection: Calculate the absolute value of the phase difference between all adjacent spectral lines. To avoid interference from a few stray spectral lines or noise, the algorithm adopts a robust statistical method, which sorts all differences and selects the Nth minimum value (e.g., the 10th smallest value) as the judgment criterion.

[0138] 2. Decision and Correction: If the selected difference is very close to 180 degrees (e.g., greater than 160 degrees), the system group delay is determined to be in the "dangerous" range. In this case, the algorithm performs an alternating phase inversion operation on the original phase error sequence, that is, adding or subtracting 180 degrees from the phase value at every other point. This operation "flattens" the originally steep phase slope of nearly 180 degrees to a level close to 0 degrees, thereby eliminating the ambiguity of the unwinding algorithm.

[0139] 6.2 Standard Linear Trend Removal

[0140] After the aforementioned potential corrections, the phase error sequence is processed using the standard procedure:

[0141] a. Phase unwinding: Since the phase calculation result is limited to ±180 degrees, jumps will occur when the actual phase changes continuously beyond this range. This step restores a continuously changing true phase curve by detecting and compensating for these 360-degree integer multiple jumps.

[0142] b. Linear trend fitting: For the unwound continuous phase data, linear regression analysis is performed using the least squares method to fit an optimal straight line. This straight line represents the ideal linear phase slope introduced by the fixed delay.

[0143] c. Delinearization: Subtract the value corresponding to the linear fitting line from the unwound phase at each frequency point. After this operation, all linear phase changes related to the fixed delay are removed, and the remaining nonlinear fluctuations are the final phase flatness distortion.

[0144] 7. Outlier Filtering (Sliding Window Hampel Filter)

[0145] To enhance the reliability of the results, the algorithm finally applies a sliding-window Hampel filter to the calculated amplitude and phase distortion data sequences to filter out "glitch" noise. The sliding-window Hampel filter is a statistically based, robust outlier detection method, and its principle is as follows:

[0146] 1. Sliding window: The algorithm traverses the entire data sequence using a fixed-size window (e.g., containing the current point and four points before and after it, for a total of nine points).

[0147] 2. Local Median Calculation: Calculates the median of all data points within the current window. The median is an estimate of the local center location that is insensitive to outliers.

[0148] 3. Local Dispersion Calculation: Calculate the absolute value of the difference between each data point within the window and the median, and then take the median of these absolute differences to obtain the Median Absolute Deviation (MAD). MAD is a robust indicator for measuring the degree of local dispersion of data.

[0149] 4. Outlier Detection and Replacement: If the absolute value of the difference between the current center point and the median of the window is significantly greater than its local dispersion (usually greater than a threshold, such as 3 times the MAD), then the point is identified as an outlier. Once identified as an outlier, the value of that point will be replaced with the median of the current window.

[0150] Through the above steps, the sliding window Hampel filter can effectively identify and correct isolated outliers caused by transient noise and other factors without affecting the overall data trend.

[0151] Finally, after comprehensive correction and purification, the algorithm yielded amplitude flatness distortion and phase flatness distortion data.

[0152] The specific algorithm steps are as follows: Figure 2 As shown.

[0153] The half-cycle ambiguity detection and staggered phase correction of this invention first detects whether the phase difference between adjacent spectral lines is close to 180 degrees before phase unwinding. If this "dangerous state" is detected, "staggered phase reversal" (adding / subtracting 180 degrees at intervals of one point) is performed on the phase sequence, thereby actively avoiding the failure of subsequent unwinding algorithms. This is a very prominent innovation of this scheme. It identifies and solves a stubborn pain point in traditional phase unwinding algorithms. Standard unwinding algorithms usually assume that the adjacent phase difference is much less than 180 degrees, and will fail when the system group delay causes this condition to be unmet. This scheme does not attempt to improve the unwinding algorithm itself, but creatively designs a "preprocessing" step, which cleverly transforms a difficult nonlinear problem into a simple one by detecting and "smoothing" dangerous phase jumps, greatly improving the robustness and environmental adaptability of the algorithm.

[0154] This invention provides a complete processing flow for asynchronous acquisition, including linear trend removal. It takes "asynchronous acquisition" as a prerequisite and, through a series of algorithms such as "phase unwrapping" and "least square linear trend removal," eliminates constant phase shifts and linear phase ramps caused by unknown sampling start points and frequency offsets at the pure software level. This invention acknowledges that asynchronous acquisition is the norm and provides a complete software solution. The design concept of "transforming a synchronization problem into an algorithmic problem" is itself innovative. It greatly simplifies the measurement system, reduces costs and operational complexity, and produces significant beneficial effects.

[0155] This invention employs a sliding-window Hampel filter for outlier removal. At the end of the algorithm, the calculated amplitude and phase flatness distortion sequence is applied to the sliding-window Hampel filter to identify and replace isolated "glitch" points caused by transient spurious interference. Although the sliding-window Hampel filter is known, its application in this scheme has a specific technical purpose and contribution. In broadband RF testing, spurious interference is a prevalent and significant factor affecting the reliability of results. Integrating this robust statistical filtering method into the final step of the flatness analysis process as a means of "cleaning" the results is a specific and effective technical solution for this application scenario. It complements the preceding correction steps, jointly ensuring the high reliability and accuracy of the final results, demonstrating the completeness and ingenuity of the invention.

[0156] The specific embodiments of the present invention are as follows:

[0157] 1. Time-domain to frequency-domain transformation

[0158] Suppose there is a set of discrete-time complex signal sequences of length L. ,in:

[0159] ;

[0160] in These are the sampled values ​​of the real and imaginary parts, respectively. This sequence represents the sampling rate. Equal-interval sampling of continuous-time signals at (unit: MHz).

[0161] For sequence Performing a Discrete Fourier Transform (DFT) yields a sequence of complex spectra in the frequency domain. Its definition is:

[0162] ;

[0163] This transformation maps the signal from the time domain to the frequency domain, making each frequency point Corresponding to a frequency (Unit: MHz), and carries information about the complex amplitude at that frequency. Spectrum It is a complex number and can be used to extract the amplitude and phase of each frequency component.

[0164] 2. Spectrum centralization

[0165] In the standard DFT output, frequency components are arranged from 0 to... The frequency is arranged in ascending order, from low to high. To center the zero frequency, a centered spectrum is constructed. ,satisfy:

[0166] ;

[0167] Equivalently, this operation achieves a cyclic shift:

[0168] ;

[0169] After the transformation, the frequency axis becomes Zero frequency is located at index This forms a DC-symmetric spectral structure, which facilitates subsequent analysis.

[0170] 3. Target frequency point localization and information extraction

[0171] Suppose we have a set of M target frequency points. The unit is MHz, which corresponds to the center frequency of each spectral line in the comb spectrum.

[0172] For each target frequency Its corresponding index position in the centralized spectrum Defined as:

[0173] ;

[0174] That is, after mapping the frequency to a normalized frequency point, rounding to the nearest integer. If the resulting index... If the value is too high, it is considered to be outside the boundary, and the point should be skipped.

[0175] Subsequently, the corresponding complex values ​​are extracted from the centralized spectrum:

[0176] ;

[0177] Forming complex response sequences on the target frequency set This sequence characterizes the actual frequency domain response of the system at each comb-shaped spectral point.

[0178] 4. Actual amplitude and phase calculation

[0179] For each extracted complex response Define its amplitude and phase as follows:

[0180] ;

[0181] ;

[0182] in The argument of the principal value of a complex number takes values ​​in the interval [0, 1]. The above two equations respectively give the first... The actual measured amplitude and actual measured phase of the signal at each frequency point.

[0183] 5. Preliminary calculation of amplitude and phase error

[0184] Let the corresponding target frequency be The ideal amplitude is The unit is linear; the ideal relative phase is The unit is radians.

[0185] The initial calculated amplitude error Defined in decibels:

[0186] ;

[0187] like or Then define .

[0188] Preliminary phase error Defined as:

[0189] ;

[0190] To limit it to Interval, perform modular normalization:

[0191] ;

[0192] Or equivalently written as:

[0193] Until ;

[0194] Converting radians to degrees facilitates engineering processes:

[0195] ;

[0196] Obtain the phase error sequence .

[0197] 6. Systematic error correction

[0198] 6.1 Amplitude Correction

[0199] Preliminary amplitude error sequence Includes the overall gain offset. Let its arithmetic mean be:

[0200] ;

[0201] The corrected amplitude error is defined as:

[0202] ;

[0203] This operation eliminates the overall power offset, making the mean of the corrected amplitude error sequence zero, while the remaining components correspond to amplitude flatness distortion.

[0204] 6.2 Phase Correction

[0205] 6.2.1 Half-cycle ambiguity correction

[0206] Phase error sequence Under certain conditions, the difference between adjacent terms may be close to 180°, which is caused by the group delay being close to an integer multiple of half a period, leading to subsequent untangling failure.

[0207] Define a sequence of absolute values ​​of adjacent phase differences:

[0208] ;

[0209] For the set of differences Arranging in ascending order yields an ordered sequence. Take the Nth smallest value:

[0210] ;

[0211] like If the system is determined to be in a "semi-periodic ambiguity" state, then an interleaved phase reversal is performed on the original phase error sequence:

[0212] ;

[0213] Note: The index starts from 1, skipping atypical items at the endpoints (such as the first and last points).

[0214] This transformation reduces the adjacent phase difference, which was originally close to 180°, to close to 0°, thereby avoiding erroneous jump judgments during the phase unwinding process.

[0215] 6.2.2 Standard Linear Trend Removal

[0216] a. Phase unwinding

[0217] Because the measured phase is restricted to (-180°, 180°), folding will occur at the boundary when the true phase changes continuously. Define the unwound phase sequence. This is achieved through recursive compensation of ±360° jumps:

[0218] initialization:

[0219] ;

[0220] ;

[0221] Iterative computation ( ;

[0222] ;

[0223] ;

[0224] (equivalent to constraining the difference to) );

[0225] Cumulative corrections:

[0226] ;

[0227] Final untangling phase:

[0228] ;

[0229] b. Linear trend fitting

[0230] untangling phase Perform least-squares linear fitting to find the optimal slope. With intercept , so that:

[0231] ;

[0232] Solving for:

[0233] ;

[0234] ;

[0235] This linear model This represents the linear slope of the phase caused by the fixed group delay.

[0236] c. De-linearization

[0237] Subtracting the linear fit value from each untangled phase yields the detrended phase fluctuation:

[0238] ;

[0239] This sequence represents the final phase flatness distortion, retaining only the nonlinear fluctuation component.

[0240] 7. Outlier Filtering (Based on Sliding Window Hampel Filter)

[0241] If the signal from the signal transmitting equipment has abnormal spurious interference, this spurious interference will affect the amplitude and phase flatness results. To suppress isolated spurious anomalies, two flatness distortion sequences are compared. Perform Hampel filtering separately.

[0242] Let the sequence to be processed be Define the half-width of the sliding window as The total length of the window is For each point Define a local neighborhood index set:

[0243] ;

[0244] 7.1 Calculation of Local Median

[0245] In the window Internal calculation of median:

[0246] ;

[0247] 7.2 Calculation of Local Discreteness

[0248] Calculate the absolute deviation of the median (MAD):

[0249] ;

[0250] 7.3 Outlier Detection and Replacement

[0251] Set threshold ,like:

[0252] ;

[0253] in Let be the Gaussian equivalent scaling factor, used to approximate MAD as an estimate of the standard deviation, then determine... The outlier was replaced with the local median:

[0254] ;

[0255] Otherwise, retain the original value.

[0256] This operation is performed independently on each point in the sequence, and the final output is a cleaned amplitude flatness distortion sequence. and phase flatness distortion sequence .

[0257] This invention uses a signal transmission module with severe flatness distortion for testing. The amplitude flatness distortion of the module is as follows: Figure 3 As shown in the spectrum analyzer. Figure 3 It can be seen that the amplitude flatness has deteriorated significantly, and there is also stray interference. Figure 3 The violet line represents power, and the red line represents phase. Figure 3 The horizontal axis represents frequency, with units in MHz. Figure 3 These are waveforms after processing the original software screenshots. For the power waveform of the purple line, the vertical axis represents power flatness distortion, with each division on the vertical axis representing 10dB. For the phase waveform of the red line, the vertical axis represents phase flatness distortion, with each division on the vertical axis representing 45°.

[0258] After acquiring comb spectrum data emitted by the signal transmitting device using a signal acquisition device, the flatness distortion result is analyzed based on the algorithm of this invention, such as... Figure 4 As shown, very high-precision amplitude and phase flatness distortion data were obtained. Figure 4 To obtain the comb spectrum data emitted by the signal transmitting device using a signal acquisition device, the flatness distortion result is analyzed based on the algorithm of this invention.

[0259] The purpose of this invention is to use comb spectrum as a known signal to analyze the amplitude and phase flatness distortion of a "broadband signal transmitting device", that is, a transmission link or system, and to provide a basis for subsequent pre-distortion compensation. It uses a signal source to calibrate a system.

[0260] This invention focuses on solving a series of algorithm robustness problems under asynchronous measurement conditions, including: ① how to eliminate the linear phase ramp introduced by unknown time delay and frequency offset without hardware synchronization; ② how to solve the "half-cycle ambiguity" problem in phase unwinding; ③ how to filter out outliers caused by stray interference.

[0261] One of the core advantages of this invention is that it achieves asynchronous acquisition without any hardware synchronization, eliminating the effects of latency and frequency offset through pure algorithms.

[0262] This invention analyzes and corrects distortions introduced by the transmission link of the device under test. The core of the algorithm lies in a complex multi-stage signal processing flow, including half-cycle ambiguity correction, phase unwrapping, least squares delinearization, and Hampel filtering.

[0263] This invention achieves asynchronous acquisition and efficient parsing without the need for synchronization. One of the most significant advantages of this invention is that it completely eliminates the dependence on any form of synchronization between the transmitting and acquiring devices. Traditional methods require complex hardware trigger wiring or time-consuming synchronization head searches to align data, while this method can directly process raw IQ data blocks acquired at any point in time. This is achieved by introducing linear trend fitting and removal techniques during the phase correction stage. Any unknown delay at the acquisition time and the frequency offset at both ends of the transmission and reception are collectively represented in the frequency domain as a definite linear phase slope and DC phase offset. The algorithm of this invention can accurately calculate this linear trend (characterized by slope and intercept) by performing least-squares linear regression on the unwound phase data and remove it entirely from the measurement results. In this way, the algorithm automatically aligns the data to an ideal reference with zero phase and zero delay, making the starting point of acquisition irrelevant. The direct technical effect of this feature is that the testing process is greatly simplified and the operational efficiency is significantly improved. It is especially suitable for automated production line testing and rapid troubleshooting in the field, reducing the requirements for the testing environment and operators, and has significant economic value.

[0264] This invention offers the advantage of completeness in joint amplitude-phase calculation. It utilizes a single FFT to synchronously acquire the full-band complex spectrum and simultaneously outputs amplitude and phase flatness curves through a unified error correction framework. Compared to the traditional process of separately acquiring and stitching amplitude and phase data, the testing steps are compressed from two to one, avoiding the accumulation of errors from secondary handling and secondary calibration.

[0265] This invention boasts high reliability independent of latency and phase deviation. It embeds a half-cycle ambiguity detection and interleaved phase correction mechanism into the phase processing link, followed by robust unwinding and linear regression detrending algorithms, enabling the acquisition of continuous phase curves under arbitrary frame shifts and carrier offsets. Extensive Monte Carlo simulations (error model covering 0-1000ns latency deviation and ±200ppm sampling offset) demonstrate a 100% success rate, with the phase RMS error consistently controlled below 0.3°. This complete immunity to non-ideal synchronization conditions allows field maintenance personnel to perform real-time calibration without service interruption or link adjustments, significantly improving system availability.

[0266] This invention exhibits excellent robustness against spurious interference. At the output end, a sliding window Hampel filter is used to perform local statistical checks on amplitude and phase residuals, automatically replacing isolated outliers without affecting the overall trend. In engineering tests, when a -25dBc single-carrier spurious signal is injected into the tested link, the algorithm still maintains the amplitude flatness index within ±0.1dB, improving the suppression effect by approximately 6dB compared to a conventional median filtering scheme. This robustness ensures consistent results even in complex electromagnetic environments outside the laboratory and under power supply noise conditions on a production line.

[0267] This invention features high precision and wide bandwidth adaptability. The FFT resolution, comb spacing, and window function parameters of this invention can all be automatically configured according to bandwidth, supporting full coverage of continuous bandwidth from 10MHz to 2GHz and comb spacing from 10kHz to 5MHz. Through an improved frequency index integer rounding method, the frequency positioning error is limited to within half a frequency point, thus ensuring the amplitude and phase estimation accuracy under large bandwidth conditions. Experimental results show that, with a 0.6GHz bandwidth and a 15000-point FFT configuration, the standard deviation of the algorithm's amplitude error is 0.04dB, better than the industry-standard 0.2dB; the standard deviation of the phase error is 0.025°, far below the system specification of 1°, providing sufficient dynamic compensation margin for applications such as millimeter-wave phased array systems.

[0268] This invention features low computational complexity and possesses potential for real-time processing and embedded applications. The algorithmic steps involved in this invention, including FFT, spectrum shifting, linear regression analysis, and Hampel filtering, all consist of highly computationally efficient basic mathematical operations. The entire process avoids complex iterative solutions or high-dimensional matrix operations, ensuring the algorithm's execution speed. This advantage makes it suitable not only for rapid post-processing in host computer software but also for porting to FPGAs or dedicated DSPs. In applications such as 6G intelligent endogenous networks or self-calibrating RF front-ends, this algorithm can serve as an embedded functional module, enabling online, real-time self-monitoring and calibration of devices. This reduces reliance on expensive external testing instruments, lowers overall lifecycle maintenance costs, and enhances the system's intelligence and environmental adaptability.

[0269] This invention overcomes the shortcomings of traditional frequency sweeping schemes, which can only acquire amplitude information and have extremely low measurement efficiency, by simultaneously obtaining the amplitude and phase flatness distortion of a broadband transmission link in a single data acquisition. This objective requires the algorithm to be able to quickly resolve broadband comb-spectrum signals containing hundreds to thousands of comb teeth without reducing frequency resolution, and output amplitude and phase error results with engineering-usable accuracy.

[0270] This invention eliminates the dependence of existing comb spectrum analysis schemes on external hardware synchronization or synchronization head search. Ideal trigger alignment between the transmitting and acquiring devices is difficult to guarantee, and hardware synchronization schemes significantly increase system complexity and cost. Synchronization head-based schemes, on the other hand, suffer from high computational load, sensitivity to multipath interference, and measurement loss due to synchronization failure. This invention aims to recover the phase reference of each comb tooth using pure algorithmic means under completely unknown frame alignment conditions, enabling solution processing at any acquisition point in time.

[0271] This invention significantly improves robustness to carrier frequency offset and sampling clock offset. These two types of offsets introduce linear phase ramps in broadband scenarios, directly superimposed on the device's true phase distortion. Improper handling can mislead calibration compensation. This invention must provide a mechanism to automatically estimate and remove the full-band linear phase trend, ensuring that the final result reflects only the device's inherent nonlinear flatness distortion.

[0272] This invention addresses the problem of phase unwinding failures due to half-cycle ambiguity, noise, and isolated stray particles. When the system group delay approaches a half-integer multiple of the fundamental period, the phase difference between adjacent comb teeth approaches 180 degrees, leading to misjudgments by conventional unwinding algorithms. This invention detects such dangerous conditions and performs staggered phase correction during the preprocessing stage, followed by robust unwinding and linear trend removal strategies to ensure continuous and accurate phase curves under any group delay conditions.

[0273] This invention improves resistance to transient spurious interference. In broadband RF environments, discrete strong interference often exists in the transmit link or test site, causing isolated anomalies in amplitude or phase measurements, thereby degrading the overall flatness assessment. This invention should integrate an adaptive outlier detection and correction mechanism to filter out local glitches while retaining true distortion, ensuring that the output results have statistical consistency and engineering repeatability.

[0274] For the specific implementation scheme of this embodiment, please refer to the relevant descriptions in the above embodiments, which will not be repeated here.

[0275] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

[0276] It should be noted that in the description of this invention, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this invention, unless otherwise stated, "a plurality of" means at least two.

[0277] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0278] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution device. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0279] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The corresponding program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0280] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0281] The storage media mentioned above can be read-only memory, disk, or optical disk, etc.

[0282] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0283] This invention employs a method, system, apparatus, processor, and computer-readable storage medium for analyzing the amplitude and phase flatness of broadband signal transmitting devices based on comb spectra. Addressing the flatness calibration requirements of broadband wireless signal transmitting devices, it calculates the flatness information of the transmitting device from the comb spectrum. This invention achieves asynchronous acquisition and efficient analysis without synchronization, possesses the advantage of complete amplitude-phase joint calculation, exhibits high reliability independent of time delay and phase deviation, demonstrates excellent robustness against spurious interference, offers high accuracy and broadband adaptability, has low computational complexity, and possesses potential for real-time processing and embedded applications.

[0284] In this specification, the invention has been described with reference to specific embodiments thereof. However, it will be apparent that various modifications and variations can be made without departing from the spirit and scope of the invention. Therefore, the specification and drawings should be considered illustrative rather than restrictive.

Claims

1. A method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on comb spectrum, characterized in that, The method includes the following steps: (1) Collect time-domain complex IQ sampling data of the link under test; (2) Perform a fast Fourier transform on the IQ sampled data to obtain the complex spectrum; (3) Shift the zero-frequency component to the center of the spectrum; (4) Calculate the index of each target frequency in the spectrum array according to the preset comb tooth interval, and extract the corresponding complex value; (5) Calculate the amplitude and phase at each index; (6) Compare the actual amplitude and phase with the ideal reference values ​​to calculate the preliminary amplitude error and phase error; (7) Correct the amplitude error; correct the phase error; (8) Filter out outliers. Apply sliding window Hampel filters to the amplitude flatness distortion sequence and the phase flatness distortion sequence respectively to replace isolated outliers. (9) Output amplitude flatness distortion and phase flatness distortion results; The phase error correction in step (7) includes the following steps: (2-7.1) Perform half-cycle ambiguity detection and staggered phase correction; (2-7.2) Perform phase unwinding; (2-7.3) Obtain the linear phase slope by least squares fitting and remove it from the continuous phase curve to obtain the phase flatness distortion, and continue to step (8).

2. The method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum according to claim 1, characterized in that, The step (7) mentioned above, which corrects the amplitude error, specifically includes the following steps: (1-7.1) Calculate the arithmetic mean of all amplitude errors, correct the amplitude errors, obtain amplitude flatness distortion, and continue to step (8).

3. The method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum according to claim 1, characterized in that, The aforementioned step (2-7.1) specifically includes the following steps: Determine if the phase difference between adjacent comb teeth is close to 180°. If so, perform an interleaving operation of adding or subtracting 180° on the phase sequence; otherwise, continue to step (2-7.2).

4. The method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum according to claim 1, characterized in that, The aforementioned step (2-7.1) also includes the following steps: Determine if the Nth smallest value of the set of absolute values ​​of the phase difference between adjacent comb teeth is greater than 160°. If so, there is a half-cycle ambiguity, and staggered phase correction is performed.

5. The method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum according to claim 1, characterized in that, The aforementioned step (2-7.2) specifically includes the following steps: The corrected phase sequence is compensated for with ±360° jumps to obtain a continuous phase curve.

6. The method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum according to claim 1, characterized in that, The phase unwinding in step (2-7.2) employs a recursive compensation method to keep the continuous phase difference within the range of -180° to +180°.

7. The method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum according to claim 1, characterized in that, The sliding window Hampel filter in step (8) has a window length of 2w+1, where w is in the range of 4 to 8, and the threshold coefficient τ is in the range of 2.5 to 3.

5.

8. The method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum according to claim 1, characterized in that, The comb tooth spacing in step (4) is in the range of 10kHz to 5MHz, and the number of comb teeth that can be configured is in the range of 8 to 65536.

9. The method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum according to claim 2, characterized in that, The residual of amplitude flatness distortion in step (1-7.1) shall not exceed ±0.1dB.

10. The method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum according to claim 1, characterized in that, The phase flatness distortion residual of step (2-7.3) shall not exceed ±0.5°.

11. A system for implementing the method of claim 1, which realizes amplitude and phase flatness analysis of a broadband signal transmitting device based on comb spectrum, characterized in that, The system includes: The data interface module is used for time-domain complex IQ sampling data of the link under test; The FFT processing module, connected to the data interface module, is used to perform a fast Fourier transform on the IQ sampled data. The amplitude and phase calculation module is connected to the FFT processing module and is used to calculate the amplitude and phase at each index. The systematic error correction module is connected to the amplitude and phase calculation module and is used to perform amplitude correction to obtain amplitude flatness distortion, perform half-cycle ambiguity detection, and perform staggered phase correction. The outlier filtering module, connected to the systematic error correction module, is used to apply a sliding window Hampel filter to the amplitude flatness distortion sequence and the phase flatness distortion sequence respectively, replacing isolated outliers. The result output module is connected to the outlier filtering module and is used to output amplitude flatness distortion and phase flatness distortion results.

12. A device for resolving the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum, characterized in that, The device includes: A processor is configured to execute computer-executable instructions; The memory stores one or more computer-executable instructions, which, when executed by the processor, implement the steps of the method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum, as described in any one of claims 1 to 10.

13. A processor for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on comb spectrum, characterized in that, The processor is configured to execute computer-executable instructions, which, when executed by the processor, implement the steps of the method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum, as described in any one of claims 1 to 10.

14. A computer-readable storage medium, characterized in that, It stores a computer program that can be executed by a processor to implement the various steps of the method for analyzing the amplitude and phase flatness of a broadband signal transmitting device based on a comb spectrum, as described in any one of claims 1 to 10.

Images