A digital modulated signal hierarchical screening electromagnetic interference prediction system

Abstract

The present application relates to a kind of digital modulation signal hierarchical screening electromagnetic interference prediction system, system includes library building stage including signal / spectrum acquisition module, reference PS data extraction module, ASK / PSK envelope modeling module, FSK envelope modeling module, parameter optimization module, smoothing processing module and database access module, responsible for extracting reference PSE from actual measurement or simulation PS, complete model fitting, parameter optimization, smoothing processing and database access;Application stage includes GS / IM application module, responsible for calling the PSE entry of having been in the library in actual EMI prediction task and participating in GS frequency screening and IM determination.The present application forms reusable PSE database: through the unified storage of parameterized model, optimization parameter, applicable scope and access index, the same kind of digital modulation signal can be quickly called in multiple EMI prediction tasks.

Description

Technical Field

[0001] This invention relates to the field of electromagnetic compatibility and interference, and in particular to a digital modulation signal hierarchical screening electromagnetic interference prediction system. Background Technology

[0002] As electronic information systems evolve towards broadband, multi-standardization, and high integration, spectrum resources within and between systems are becoming increasingly congested. Unintended coupling can occur between different transmitters, receivers, antennas, and RF front-ends, leading to increased bit error rates, degraded receiver sensitivity, decreased link quality, and even communication failures in interfered receiving equipment. Therefore, establishing a stable, reusable, and engineering-compliant spectrum description of interference sources during the system design phase, and embedding it into a tiered screening and interference margin (IM) assessment process, has direct engineering value in reducing subsequent rectification costs.

[0003] Traditional tiered screening processes typically use the measured power spectrum (PS) as the frequency domain input and make sensitivity judgments based on frequency thresholds, amplitude thresholds, or receiver sensitivity thresholds. This approach is relatively mature in analog modulation or quasi-narrowband interference scenarios. However, in digital modulation signals, the power spectrum exhibits significant ripple and quasi-periodic fluctuations caused by sinc roll-off, symbol pulses, finite observation length, window functions, resolution bandwidth (RBW), and carrier frequency offset. Directly using the raw power spectrum for threshold screening can lead to unstable judgment results near the critical frequency band.

[0004] Existing technologies directly use the measured power spectrum of digital modulated signals as the frequency domain descriptor of the interference source in the GS (Graded Screening) / IM electromagnetic interference prediction process, and use the amplitude of the power spectrum at the receiver operating frequency or passband for interference margin calculation. This type of method treats a snapshot of the power spectrum obtained from a measurement as input for subsequent screening, but lacks a unified modeling mechanism for digital modulated signal spectral ripple, out-of-band roll-off, and measurement setting sensitivity. Furthermore, existing technologies also employ methods to "envelope" or simplify the power spectrum, such as peak detection, maximum hold, polynomial fitting, spline fitting, moving average smoothing, average power spectral density curves, discrete spectral line sets, equivalent bandwidth representation, or static transmit masks in communication standards. While these methods can suppress ripple or provide compliance upper bounds to some extent, they still have the following shortcomings: 1. Peak detection and maximum hold still depend on specific measurement conditions, and the resulting curves are not normalized, parameterized, or reusable models, making them difficult to directly use as entries in the GS / EMIP database. 2. Data-driven fitting methods such as polynomials, splines, and moving averages easily inherit measurement dependencies from sampling conditions like RBW, window functions, and observation durations, lacking physical parameters corresponding to the modulation mechanism. 3. Static emission masks primarily serve for standard compliance judgments, typically only providing upper bound constraints and failing to characterize cross-components between FSK tones, the slope of the central transition region, and out-of-band energy accumulation characteristics. 4. Existing methods typically lack quantitative database admission criteria, making it difficult to determine whether a particular spectral model is sufficiently stable or reusable across analytical conditions. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a digital modulation signal hierarchical screening electromagnetic interference prediction system, which solves the deficiencies of the prior art.

[0006] The objective of this invention is achieved through the following technical solution: a digital modulation signal hierarchical screening electromagnetic interference prediction system, the system comprising:

[0007] Signal / Spectrum Acquisition Module: Input digital modulated signal, power spectrum data and basic parameters, establish relative frequency axis f, and map it to the actual radio frequency;

[0008] Reference PS data extraction module: Perform local peak search based on PS data, connect peak points in frequency order to obtain reference PSE, where PS represents power spectrum and PSE represents power spectrum envelope;

[0009] ASK / PSK envelope modeling module: Selects the ASK / PSK analytical model based on the modulation method;

[0010] FSK envelope modeling module: Builds a 2FSK model PSE based on the pitch offset set, symbol duration, and initial parameters;

[0011] Parameter optimization module: Based on the reference PSE, model PSE, and parameter boundaries, construct the genetic algorithm objective function and optimize the parameters;

[0012] Smoothing module: Smooths the optimized model PSE using Savitzky-Golay to obtain the calibrated PSE;

[0013] Database admission module: Calculates database admission criteria and makes admission decisions based on the calibration PSE and reference PSE;

[0014] GS / IM Application Module: Calculates the interference power and IM at the receiver input at the GS frequency selection point, and outputs the electromagnetic interference sensitivity prediction result. GS represents graded selection, and IM represents interference margin.

[0015] The signal / spectrum acquisition module specifically includes:

[0016] Obtain the modulation scheme, modulation order, and symbol duration of the signal to be analyzed. carrier frequency Frequency offset set Transmit power, sampling rate, spectrum estimation parameters, and power spectrum data. This represents the offset of the i-th FSK tone on the relative frequency axis;

[0017] The input time-domain waveform generated by simulation or the measured power spectrum output by a spectrum analyzer satisfies the relationship between the relative frequency and the absolute RF frequency. The relationship, among which, This represents the frequency offset relative to the frequency axis. This refers to the actual radio frequency.

[0018] The reference PS data extraction module specifically includes:

[0019] If the input is a time-domain waveform, the power spectrum is obtained using the Welch method. If the input is a power spectrum, then a local peak search is performed directly;

[0020] Local peak points are detected on the power spectrum curve to obtain the set of peak frequencies. And take the peak amplitude as the reference upper envelope sample. The reference PSE is formed by connecting these peak samples in frequency order, where, This represents the total number of local peak points extracted from the power spectrum curve.

[0021] The ASK / PSK envelope modeling module specifically includes:

[0022] For 2ASK and 2PSK, under the bipolar NRZ rectangular pulse abstraction, their relative power spectral envelope is represented by two segments: the main lobe plateau and the out-of-band roll-off. The model is as follows: This model characterizes the approximate plateau of the main lobe region and the decreasing trend of the out-of-band region with increasing frequency offset in ASK / PSK. For M-ASK and M-PSK, under the premise that changes in amplitude or phase state number do not alter the relative envelope shape, a segmented relative envelope structure of main lobe plateau + out-of-band roll-off is adopted. Amplitude and frequency mapping are performed using actual transmit power, bandwidth parameters, or database entries. This is a bandwidth parameter related to the symbol rate.

[0023] The FSK envelope modeling module specifically includes:

[0024] For 2FSK, the power spectrum consists of a structured superposition of two carrier frequency shift components. Then the parameterized model PSE for 2FSK calibration is ,in, and These are the carrier frequency shift components at two frequencies during 2FSK modulation. for and The center frequency, Indicates the first The base band roll-off of each FSK component This indicates the relative strength coupling between two carrier frequency shift components. For correction items;

[0025] For higher-order FSKs, a hierarchical pyramid structure is adopted, with 4FSK consisting of two calibrated 2FSK envelopes. and Composed of two calibrated 4FSK envelopes, 8FSK is formed by combining two calibrated 4FSK envelopes. and Composed of various elements.

[0026] The parameter optimization module specifically includes:

[0027] A population of multiple candidate parameter individuals is initialized. For each individual, the predicted sample of the model PSE at the peak frequency point is calculated and compared with the reference PSE peak sample to obtain the objective function value.

[0028] Selection, crossover, and mutation are performed based on the objective function value to generate the next generation of parameter population;

[0029] Repeat the above process until the convergence threshold is met or the maximum number of iterations is reached, and finally output the parameter vector with the optimal objective function value.

[0030] The smoothing module specifically includes:

[0031] Substitute the optimal parameter vector into the corresponding ASK / PSK analytical envelope model or FSK parameterized envelope model to generate the PSE model on the complete frequency axis.

[0032] Savitzky-Golay smoothing was performed on the model PSE to obtain the calibrated PSE.

[0033] The database access control module includes:

[0034] For each candidate PSE, the peak sample error is calculated separately in the main lobe region, roll-off region, and out-of-band region. Modulation type Measurement of PS, To predict PSE, the peak error is: , and These represent the predicted peak PSE sample and the reference peak PSE sample, respectively.

[0035] Using main lobe bandwidth parameters Divide the peak index into , and ,in, The modulation method is At that time, the first power spectrum extracted from the original power spectrum Local peak frequencies, The modulation method is The set of peak indexes for the main lobe region at that time. The modulation method is The set of peak indexes in the roll-off region at that time. The modulation method is The set of out-of-band peak indexes at that time;

[0036] Obtain the main lobe region index and Roll-off zone indicators and Indicators from outside the zone and ,in, The modulation method is indicated as Time control of the average absolute error of the main lobe region, The modulation method is indicated as Time control of the overall root mean square error of the main lobe region The modulation method is Peak sample size in the main lobe region The modulation method is Peak number of samples in the roll-off region The modulation method is Peak sample size in the outer band The modulation method is Average absolute deviation of the roll-off zone The modulation method is The out-of-band predicted integral energy is calculated from the peak samples of the PSE model. The modulation method is The out-of-band reference integral energy is obtained by converting the reference PSE or the peak sample of the original power spectrum. Indicates an indicator function, The modulation method is Peak coverage at that time The modulation method is The relative error of the out-of-band energy at that time;

[0037] When a candidate PSE meets multiple set conditions, it is written into the PSE database and saved as a PSE database entry.

[0038] The GS / IM application module specifically includes:

[0039] Based on the modulation scheme, modulation order, symbol time, tone shift, carrier frequency, and transmit power of the interference source to be analyzed, matching entries are retrieved from the PSE database;

[0040] If a perfectly matching entry exists in the PSE database, the entry is used directly. If an approximately matching entry exists in the PSE database, the PSE is used according to the frequency normalization rule, transmit power scaling rule, and bandwidth parameter interpolation rule. If no entry that meets the conditions exists in the PSE database, the model is remodeled or the model is indicated as unavailable.

[0041] Map the normalized PSE to the actual RF frequency axis By combining transmit power, path loss, antenna gain, coupling path, or receiver front-end transfer function, the interference signal power at the receiver input is calculated. And calculate the interference margin. , This refers to the receiver sensitivity threshold or performance degradation threshold.

[0042] when A value greater than 0 dB indicates that the interference power has reached or exceeded the receiver's tolerance, classifying it as sensitive or critically sensitive. When the value is less than 0 dB, it is considered insensitive or low risk.

[0043] The present invention has the following advantages:

[0044] 1. Improve frequency domain input stability: By replacing the original PS with PSE, local fluctuations caused by sinc ripple, RBW, window function, observation length and carrier frequency offset are reduced, so that GS frequency screening no longer depends on a single measurement spectrum snapshot.

[0045] 2. Improve the consistency of critical IM determination: In the sensitive / insensitive boundary region where IM is close to 0 dB, PSE can reduce misjudgments caused by local spectral valleys or peaks, thereby improving the repeatability of receiver sensitivity prediction.

[0046] 3. Form a reusable PSE database: Through the unified storage of parameterized models, optimized parameters, applicable scope and admission criteria, the same type of digital modulation signal can be quickly called up in multiple EMI prediction tasks.

[0047] 4. Compatible with different digital modulation types: ASK / PSK can be modeled by parsing the envelope, and FSK can be modeled by parameterization and hierarchical pyramid structure, covering typical multi-tone scenarios such as 2FSK, 4FSK, and 8FSK.

[0048] 5. Reduce engineering modeling costs: After the database is built, there is no need to perform a complete spectrum fitting every time in the application stage. You only need to search the database and perform frequency / power mapping to participate in GS / IM calculation. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the process of the present invention;

[0050] Figure 2 A schematic diagram of the construction of a recursive pyramid.

[0051] Figure 3 A flowchart illustrating the process of calibrating GA parameters.

[0052] Figure 4 This is a flowchart of the PSE database admission decision process. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of this application provided below with reference to the accompanying drawings is not intended to limit the scope of protection of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application. The present invention will be further described below with reference to the accompanying drawings.

[0054] like Figure 1 As shown, this invention specifically relates to a digital modulation signal hierarchical screening electromagnetic interference prediction system, realizing a closed-loop design for GS / IM electromagnetic interference prediction, encompassing "power spectrum envelope modeling - parameter calibration - database admission - hierarchical screening application". The system includes a database construction phase and an application phase. The database construction phase includes a signal / spectrum acquisition module, a reference PS data extraction module, an ASK / PSK envelope modeling module, an FSK envelope modeling module, a parameter optimization module, a smoothing module, and a database admission module. This module is responsible for extracting reference PSEs from measured or simulated PSs, completing model fitting, parameter optimization, smoothing, and database admission. The application phase includes a GS / IM application module, responsible for calling the already-included PSE entries in the database to participate in GS frequency screening and IM determination in actual EMI prediction tasks.

[0055] Among them, the signal / spectrum acquisition module: inputs digital modulation signal, power spectrum data and basic parameters, establishes relative frequency axis f, and maps it to the actual radio frequency;

[0056] Reference PS data extraction module: Perform local peak search based on PS data, connect peak points in frequency order to obtain reference PSE, where PS represents power spectrum and PSE represents power spectrum envelope;

[0057] ASK / PSK envelope modeling module: Selects the ASK / PSK analytical model based on the modulation method;

[0058] FSK envelope modeling module: Builds a 2FSK model PSE based on the pitch offset set, symbol duration, and initial parameters;

[0059] Parameter optimization module: Based on the reference PSE, model PSE, and parameter boundaries, construct the genetic algorithm objective function and optimize the parameters;

[0060] Smoothing module: Smooths the optimized model PSE using Savitzky-Golay to obtain the calibrated PSE;

[0061] Database admission module: Calculates database admission criteria and makes admission decisions based on the calibration PSE and reference PSE;

[0062] GS / IM Application Module: Calculates the interference power and IM at the receiver input at the GS frequency selection point, and outputs the electromagnetic interference sensitivity prediction result. GS represents graded selection, and IM represents interference margin.

[0063] Furthermore, the signal / spectrum acquisition module specifically includes the following:

[0064] The signal / spectrum acquisition module obtains the modulation scheme, modulation order, and symbol duration of the signal to be analyzed. carrier frequency Frequency offset set Transmit power, sampling rate, spectrum estimation parameters, and power spectrum data. This represents the offset of the i-th FSK tone on the relative frequency axis;

[0065] The input time-domain waveform generated by simulation or the measured power spectrum output by a spectrum analyzer satisfies the relationship between the relative frequency and the absolute RF frequency. The relationship, among which, This represents the frequency offset relative to the frequency axis. This refers to the actual radio frequency.

[0066] Furthermore, the PS data extraction module specifically includes the following:

[0067] If the input is a time-domain waveform, the power spectrum is obtained using a spectrum estimation method or other spectrum estimation methods. If the input is a power spectrum, then a local peak search is performed directly;

[0068] The system detects local peak points on the power spectrum curve to obtain a set of peak frequencies. And take the peak amplitude as the reference upper envelope sample. , This represents the total number of local peak points extracted from the power spectrum curve. The reference PSE is formed by connecting these peak samples in frequency order. Compared with directly fitting all power spectrum sampling points, peak samples are more suitable for expressing the upper envelope trend and can reduce the influence of spectral valleys and local ripples on model fitting.

[0069] Furthermore, the ASK / PSK envelope modeling module specifically includes the following:

[0070] For 2ASK (second-order ASK) and 2PSK (second-order PSK), under the bipolar NRZ rectangular pulse abstraction, their relative power spectral envelope is represented by two segments: the main lobe plateau and the out-of-band roll-off. The model is as follows: This model characterizes the approximate plateau of the main lobe region and the decreasing trend of the out-of-band region with increasing frequency offset in ASK / PSK. For M-ASK (M-order ASK, where M is the modulation order) and M-PSK (M-order PSK), under the premise that changes in amplitude or phase state number do not change the relative envelope shape, a segmented relative envelope structure of main lobe plateau + out-of-band roll-off is adopted. Amplitude and frequency mapping is performed using actual transmit power, bandwidth parameters, or database entries. This is a bandwidth parameter related to the symbol rate.

[0071] Furthermore, the FSK envelope modeling module specifically includes the following:

[0072] For 2FSK (second-order FSK), the power spectrum consists of a structured superposition of two carrier frequency shift components. Due to the cross-components and transition region coupling between the two tones, it is difficult to stably represent this spectrum using only a single closed roll-off formula. Therefore, this invention introduces a parameterized power spectrum envelope model for 2FSK.

[0073] set up Then the 2FSK calibration envelope is ,in, and These are the carrier frequency shift components at two frequencies during 2FSK modulation. for and The center frequency, Indicates the first The base band roll-off of each FSK component This indicates the relative strength coupling between two carrier frequency shift components. This is a correction term used to correct residual deviations near the pitch.

[0074] like Figure 2 As shown, hierarchical construction of 4FSK, 8FSK and... FSK PSE;

[0075] For higher-order FSK, this invention employs a hierarchical pyramid structure. 4FSK consists of two calibrated 2FSK envelopes. and Combining; 8FSK consists of two calibrated 4FSK envelopes. and Composed of various elements.

[0076] The original combinatorial envelope of 4FSK is:

[0077] ,

[0078] in, , , Controlling the nonlinear fusion of two 2FSK components, Used to inhibit , Nearby cross items; This represents the two tones in the second group of the 4FSK construction. and The center frequency, This represents the central transition frequency between two 2FSK sub-envelopes in a 4FSK hierarchical combination.

[0079] Further introduce a central correction term: ,in, Used for correction Nearby residual error and transition zone slope, This indicates the 4FSK after adding center correction and optimization. This indicates the original 4FSK that has not yet been centered.

[0080] The original combined envelope of 8FSK is:

[0081] ,

[0082] in, , , , This represents the first group of four tones in the 8FSK construction. , , and The center frequency, This represents the second group of four tones in the 8FSK construction. , , and The center frequency, This represents the center transition frequency between two 4FSK sub-envelopes in an 8FSK hierarchical combination. , and These represent the nonlinear fusion intensity coefficient, envelope ratio index, and cross-term suppression intensity coefficient, respectively.

[0083] Further introduce a central correction term: .

[0084] in, , , and These represent the central residual error correction strength, the locality index of the correction term, the slope correction coefficient of the transition zone, and the slope correction index, respectively.

[0085] for FSK of order, which can be repeatedly performed on two calibrated units. The recursive process of FSK envelope combination, central region cross term correction, GA parameter calibration, and smoothing is as follows:

[0086] The above hierarchical combination yields a high-order FSK model PSE containing parameters to be optimized. Its nonlinear fusion parameters, cross-term suppression parameters, and center correction parameters are uniformly calibrated by the subsequent parameter optimization module using GA based on the reference PSE. The calibrated model PSE then enters the smoothing module to obtain a calibrated PSE that can be used for database admission judgment.

[0087] Furthermore, the parameter optimization module specifically includes the following:

[0088] In this invention, GA refers to the Genetic Algorithm, a global optimization algorithm based on population search, selection, crossover, and mutation mechanisms. Its role in this invention is as a model parameter calibration operator, rather than directly smoothing the original power spectrum.

[0089] like Figure 3 As shown in the figure, the input and initialization correspond to the input of the reference PSE peak sample, candidate envelope model, parameter boundary and GA hyperparameter; "GA iterative optimization" corresponds to objective function evaluation, tournament selection, intermediate crossover, uniform mutation and new population generation; output and post-processing correspond to the optimal parameter vector output, Savitzky-Golay smoothing and candidate PSE submission to the database admission judgment, where the candidate PSE is the calibrated PSE after parameter optimization and smoothing, which is used as the candidate PSE for database entry.

[0090] During execution, the parameter optimization module first initializes multiple candidate parameter individuals to form a population, calculates the predicted sample of the model PSE at the peak frequency point for each individual, and compares it with the reference PSE peak sample to obtain the objective function value;

[0091] Selection, crossover, and mutation are performed based on the objective function value to generate the next generation of parameter population;

[0092] Repeat the above process until the convergence threshold is met or the maximum number of iterations is reached, and finally output the parameter vector with the minimum (optimal) objective function value.

[0093] Furthermore, the parameter optimization module compares the model PSE with the reference PSE at the same peak frequency point, and the model predicted sample is defined as... ,in, These are the parameters to be optimized.

[0094] Genetic algorithms minimize the following objective function with a boundary penalty:

[0095] ,

[0096] in, Indicates the reference peak sample. and The first Lower and upper bounds of each parameter; For boundary penalty weights, the optimal choice is... .

[0097] For 2FSK, the preferred parameter boundary is: .

[0098] The optimal hyperparameters for the genetic algorithm are: tournament selection. Intersection in the middle Uniform mutation, population size 2000, maximum number of iterations 50, convergence threshold The role of GA (Genetic Algorithm) is to calibrate the parameters of the analytical PSE model, rather than directly smoothing the measurement spectrum.

[0099] In the parameter optimization module, when the model is 2FSK, the parameters to be optimized include the basic out-of-band roll-off, relative strength coupling, and correction term parameters. When the model is 4FSK, 8FSK, or M-FSK (M-order modulated FSK), the parameters to be optimized further include nonlinear fusion parameters introduced during the hierarchical combination process, cross-term suppression parameters, and center transition region correction parameters. For the corresponding order FSK model PSE, the parameter optimization module calculates model prediction samples at the peak frequency point of the reference PSE, and minimizes the error between the model prediction samples and the reference peak samples using a genetic algorithm, finally outputting the optimal parameter vector for that order FSK envelope model.

[0100] Furthermore, the smoothing module specifically includes the following:

[0101] After the genetic algorithm outputs the optimal parameter vector, the system substitutes this parameter vector into the corresponding ASK / PSK analytical envelope model or FSK parameterized envelope model to generate the model PSE on the complete frequency axis. Since the parameterized model may still have local high-frequency fluctuations near local peaks caused by numerical fitting, frequency sampling, or cross-term correction, the system further performs Savitzky-Golay smoothing on the model PSE to obtain the calibrated PSE.

[0102] This smoothing process is only a post-processing step after model parameter calibration, used to suppress residual ripple and improve curve continuity; it does not change the optimization effect of the genetic algorithm on model structural parameters. The smoothed calibration PSE will be sent to the database admission module for further calculation of main lobe error, roll-off error, out-of-band peak coverage, and out-of-band energy error. If the admission threshold is met, it is saved as a PSE database entry; otherwise, it is rejected or model optimization is re-executed.

[0103] Furthermore, such as Figure 4 As shown, the database admission module includes the following:

[0104] The database admission module calculates the peak sample error for each candidate PSE in the main lobe region, roll-off region, and out-of-band region, respectively. Modulation type Measurement of PS, To predict PSE, the peak error is: ,in, and These are the predicted PSE peak sample and the reference PSE peak sample, respectively.

[0105] Using main lobe bandwidth parameters Divide the peak index into , and ,in, The modulation method is At that time, the first power spectrum extracted from the original power spectrum Local peak frequencies, The modulation method is The set of peak indexes for the main lobe region at that time. The modulation method is The set of peak indexes in the roll-off region at that time. The modulation method is The set of out-of-band peak indexes at that time;

[0106] Obtain the main lobe region index and Roll-off zone indicators and Indicators from outside the zone and ,in, The modulation method is indicated as Time control of the average absolute error of the main lobe region, The modulation method is indicated as Time control of the overall root mean square error of the main lobe region The modulation method is The peak number of samples in the main lobe region, i.e. , The modulation method is The peak number of samples in the roll-off region, i.e. , The modulation method is The number of peak samples in the time-band region, i.e. , The modulation method is The mean absolute deviation of the roll-off region is used to measure the average error between the peak sample of the roll-off region model and the reference PSE. The modulation method is The out-of-band predicted integral energy is calculated from the peak samples of the PSE model. The modulation method is The out-of-band reference integral energy is obtained by converting the reference PSE or the peak sample of the original power spectrum. Indicates an indicator function, The modulation method is Peak coverage at that time The modulation method is The relative error of the out-of-band energy at that time;

[0107] and The dB peak sample can be converted into linear power, and then obtained by trapezoidal integration over the peak frequency grid. Candidate PSEs simultaneously satisfy... , , , , At that time, it is written to the PSE database and saved as a PSE database entry, where, This indicates the average absolute error of the main lobe region. This indicates the overall root mean square error of the main lobe region. This indicates the maximum deviation in the control transition roll-off region. Indicates peak coverage. This indicates the relative error of the out-of-band energy.

[0108] Furthermore, PSE models that meet the admission requirements are saved as PSE database entries, as shown in Table 1 below:

[0109] Table 1. Database Entry Fields Table

[0110]

[0111] Furthermore, the GS / IM application module specifically includes the following:

[0112] In actual EMI prediction tasks, the system retrieves matching entries from the PSE database based on the modulation method, modulation order, symbol time, tone shift, carrier frequency, and transmit power of the interference source to be analyzed.

[0113] If a perfectly matching entry exists in the PSE database, the entry is used directly. If an approximately matching entry exists in the PSE database, the PSE is used according to the frequency normalization rule, transmit power scaling rule, and bandwidth parameter interpolation rule. If no entry that meets the conditions exists in the PSE database, the model is remodeled or the model is indicated as unavailable.

[0114] When applying this method, the normalized PSE is mapped to the actual RF frequency axis. By combining transmit power, path loss, antenna gain, coupling path, or receiver front-end transfer function, the interference signal power at the receiver input is calculated. And calculate the interference margin. , This refers to the receiver sensitivity threshold or performance degradation threshold.

[0115] when When the interference power is greater than or close to 0 dB, it indicates that the interference power has reached or exceeded the receiver's tolerance, and the receiver is judged to be sensitive or critically sensitive. When the value is significantly less than 0 dB, it is considered insensitive or low-risk.

[0116] Therefore, the output of this step is an electromagnetic interference prediction result: the system calls the PSE entry in the database to predict the interference power at the receiver, and determines whether the receiving device is sensitive, critically sensitive, or insensitive at the corresponding frequency point and power combination based on the predicted IM. In verification scenarios with measured or simulated reference data, the system can also compare the predicted IM with the reference IM or the reference judgment result obtained based on BER and sensitivity degradation, and output the prediction accuracy within the error range of ±3 dB, ±6 dB, etc.

[0117] This invention uses a stable PSE-corrected frequency domain input, which reduces screening jumps caused by single-measurement ripple, and is particularly suitable for... Critical boundary regions with small absolute values.

[0118] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and improvements, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A digital modulation signal hierarchical screening electromagnetic interference prediction system, characterized in that: The system includes: Signal / Spectrum Acquisition Module: Input digital modulated signal, power spectrum data and basic parameters, establish relative frequency axis f, and map it to the actual radio frequency; Reference PS data extraction module: Perform local peak search based on PS data, connect peak points in frequency order to obtain reference PSE, where PS represents power spectrum and PSE represents power spectrum envelope; ASK / PSK envelope modeling module: Selects the ASK / PSK analytical model based on the modulation method; FSK envelope modeling module: Builds a 2FSK model PSE based on the pitch offset set, symbol duration, and initial parameters; Parameter optimization module: Based on the reference PSE, model PSE, and parameter boundaries, construct the genetic algorithm objective function and optimize the parameters; Smoothing module: Smooths the optimized model PSE using Savitzky-Golay to obtain the calibrated PSE; Database admission module: Calculates database admission criteria and makes admission decisions based on the calibration PSE and reference PSE; GS / IM Application Module: Calculates the interference power and IM at the receiver input at the GS frequency selection point, and outputs the electromagnetic interference sensitivity prediction result. GS represents graded selection, and IM represents interference margin.

2. The digital modulation signal hierarchical screening electromagnetic interference prediction system according to claim 1, characterized in that: The signal / spectrum acquisition module specifically includes: Obtain the modulation scheme, modulation order, and symbol duration of the signal to be analyzed. carrier frequency Frequency offset set Transmit power, sampling rate, spectrum estimation parameters, and power spectrum data. This represents the offset of the i-th FSK tone on the relative frequency axis; The input time-domain waveform generated by simulation or the measured power spectrum output by a spectrum analyzer satisfies the relationship between the relative frequency and the absolute RF frequency. The relationship, among which, This represents the frequency offset relative to the frequency axis. This refers to the actual radio frequency.

3. The digital modulation signal hierarchical screening electromagnetic interference prediction system according to claim 2, characterized in that: The reference PS data extraction module specifically includes: If the input is a time-domain waveform, the power spectrum is obtained using the Welch method. If the input is a power spectrum, then a local peak search is performed directly; Local peak points are detected on the power spectrum curve to obtain the set of peak frequencies. And take the peak amplitude as the reference upper envelope sample. The reference PSE is formed by connecting these peak samples in frequency order, where, This represents the total number of local peak points extracted from the power spectrum curve.

4. The digital modulation signal hierarchical screening electromagnetic interference prediction system according to claim 1, characterized in that: The ASK / PSK envelope modeling module specifically includes: For 2ASK and 2PSK, under the bipolar NRZ rectangular pulse abstraction, their relative power spectral envelope is represented by two segments: the main lobe plateau and the out-of-band roll-off. The model is as follows: This model characterizes the approximate plateau of the main lobe region and the decreasing trend of the out-of-band region with increasing frequency offset in ASK / PSK. For M-ASK and M-PSK, under the premise that changes in amplitude or phase state number do not alter the relative envelope shape, a segmented relative envelope structure of main lobe plateau + out-of-band roll-off is adopted. Amplitude and frequency mapping are performed using actual transmit power, bandwidth parameters, or database entries. This is a bandwidth parameter related to the symbol rate.

5. The digital modulation signal hierarchical screening electromagnetic interference prediction system according to claim 1, characterized in that: The FSK envelope modeling module specifically includes: For 2FSK, the power spectrum consists of a structured superposition of two carrier frequency shift components. Then the parameterized model PSE for 2FSK calibration is ,in, and These are the carrier frequency shift components at two frequencies during 2FSK modulation. for and The center frequency, Indicates the first The base band roll-off of each FSK component This indicates the relative strength coupling between two carrier frequency shift components. For correction items; For higher-order FSKs, a hierarchical pyramid structure is adopted, with 4FSK consisting of two calibrated 2FSK envelopes. and Composed of two calibrated 4FSK envelopes, 8FSK is formed by combining two calibrated 4FSK envelopes. and Composed of various elements.

6. The digital modulation signal hierarchical screening electromagnetic interference prediction system according to claim 1, characterized in that: The parameter optimization module specifically includes: A population of multiple candidate parameter individuals is initialized. For each individual, the predicted sample of the model PSE at the peak frequency point is calculated and compared with the reference PSE peak sample to obtain the objective function value. Selection, crossover, and mutation are performed based on the objective function value to generate the next generation of parameter population; Repeat the above process until the convergence threshold is met or the maximum number of iterations is reached, and finally output the parameter vector with the optimal objective function value.

7. The digital modulation signal hierarchical screening electromagnetic interference prediction system according to claim 6, characterized in that: The smoothing module specifically includes: Substitute the optimal parameter vector into the corresponding ASK / PSK analytical envelope model or FSK parameterized envelope model to generate the PSE model on the complete frequency axis. Savitzky-Golay smoothing was performed on the model PSE to obtain the calibrated PSE.

8. The digital modulation signal hierarchical screening electromagnetic interference prediction system according to claim 1, characterized in that: The database access control module includes: For each candidate PSE, the peak sample error is calculated separately in the main lobe region, roll-off region, and out-of-band region. Modulation type Measurement of PS, To predict PSE, the peak error is: , and These represent the predicted peak PSE sample and the reference peak PSE sample, respectively. Using main lobe bandwidth parameters Divide the peak index into , and ,in, The modulation method is At that time, the first power spectrum extracted from the original power spectrum Local peak frequencies, The modulation method is The set of peak indexes for the main lobe region at that time. The modulation method is The set of peak indexes in the roll-off region at that time. The modulation method is The set of out-of-band peak indexes at that time; Obtain the main lobe region index and Roll-off zone indicators and Indicators from outside the zone and ,in, The modulation method is indicated as Time control of the average absolute error of the main lobe region, The modulation method is indicated as Time control of the overall root mean square error of the main lobe region The modulation method is Peak sample size in the main lobe region The modulation method is Peak number of samples in the roll-off region The modulation method is Peak sample size in the outer band The modulation method is Average absolute deviation of the roll-off zone The modulation method is The out-of-band predicted integral energy is calculated from the peak samples of the PSE model. The modulation method is The out-of-band reference integral energy is obtained by converting the reference PSE or the peak sample of the original power spectrum. Indicates an indicator function, The modulation method is Peak coverage at that time The modulation method is The relative error of the out-of-band energy at that time; When a candidate PSE meets multiple set conditions, it is written into the PSE database and saved as a PSE database entry.

9. The digital modulation signal hierarchical screening electromagnetic interference prediction system according to claim 1, characterized in that: The GS / IM application module specifically includes: Based on the modulation scheme, modulation order, symbol time, tone shift, carrier frequency, and transmit power of the interference source to be analyzed, matching entries are retrieved from the PSE database; If a perfectly matching entry exists in the PSE database, the entry is used directly. If an approximately matching entry exists in the PSE database, the PSE is used according to the frequency normalization rule, transmit power scaling rule, and bandwidth parameter interpolation rule. If no entry that meets the conditions exists in the PSE database, the model is remodeled or the model is indicated as unavailable. Map the normalized PSE to the actual RF frequency axis By combining transmit power, path loss, antenna gain, coupling path, or receiver front-end transfer function, the interference signal power at the receiver input is calculated. And calculate the interference margin. , This refers to the receiver sensitivity threshold or performance degradation threshold. when A value greater than 0 dB indicates that the interference power has reached or exceeded the receiver's tolerance, classifying it as sensitive or critically sensitive. When the value is less than 0 dB, it is considered insensitive or low risk.

Images