Aerospace radio frequency signal quality multi-domain evaluation and compensation method based on model library

Abstract

The invention discloses a model library-based spaceflight radio frequency signal quality multi-domain evaluation and compensation method, and belongs to the technical field of wireless communication. The implementation method comprises the steps of determining signal system parameters based on various CCSDS standards and a spaceflight protocol system, extracting related parameters of spaceflight standards in a time-frequency domain, generating standard time-frequency domain radio frequency signals, converting the standard time-frequency domain radio frequency signals into a related domain, a modulation domain, an energy domain and a polarization domain, and establishing spaceflight radio frequency signal sub-model libraries of the domains; obtaining a time-frequency domain measured signal through a receiver, integrating time domain information and frequency domain information, and constructing a time-frequency domain sub-evaluation model; constructing a multi-domain comprehensive evaluation model by adopting a parallel data transmission mode and a weight data processing method; calculating the weight of the evaluation result of each domain through an entropy weight method, and obtaining a comprehensive quality evaluation result of the spacecraft radio frequency signal; and based on the distortion coefficient, a robust signal equalization algorithm is adopted to process the signal, the influence of the distortion effect on the signal quality is compensated, and the robustness and the anti-interference capability of signal transmission are improved.

Description

Technical Field

[0001] The invention belongs to the technical field of wireless communications and relates to a multi-domain evaluation and compensation method for multi-system aerospace radio frequency signal quality based on a model library. Background Art

[0002] Ensuring the high quality of spacecraft RF signals is crucial to maintaining the stability of space communications, navigation, and intersatellite measurement and control systems. By comprehensively and reasonably evaluating the quality of spacecraft RF signals, reliable communication between spacecraft and the ground can be ensured, supporting its positioning, navigation, and measurement and control functions, thereby ensuring the reliability and safety of the system.

[0003] In recent years, with the development of modern aerospace signal systems, the complexity of characterizing the quality of spacecraft RF signals has continued to increase, and the corresponding evaluation requirements have become more complex and diverse. Early evaluation methods focused on the availability and reliability of signals, relying on the strength or signal-to-noise ratio (SNR) of received signals to judge their quality. For example, NASA's Spacecraft Rendezvous Assessment and Collision Avoidance Best Practices Manual mentions that signal strength and signal-to-noise ratio are important factors in evaluating signal quality. However, a single-indicator signal evaluation system cannot fully describe the quality of RF signals in a complex multi-system environment.

[0004] On the other hand, in actual communication systems, the space environment in which spacecraft are located is harsh, and the hardware of the spacecraft's RF signal transmitter and receiver often has non-ideal characteristics, including hardware distortion, amplitude and phase noise, nonlinear effects, frequency drift, etc. These factors will increase the complexity and uncertainty of signal processing and signal quality assessment. However, most existing studies assume ideal transceiver hardware, which has a significant impact on subsequent signal processing and quality assessment.

[0005] In response to the above problems, some studies on signal quality assessment methods have considered more potential characterization dimensions, including signal integrity, clock accuracy, and multipath effects. In addition, some related studies have considered non-ideal hardware characteristics and compensated for them in communication system design. However, most of these studies focus on single-dimensional analysis and are unable to fully cope with complex aerospace communication scenarios.

[0006] In high-dynamic scenarios, the rapid movement and frequent attitude changes of spacecraft will cause significant Doppler shifts, affecting the frequency stability, positioning accuracy, reception and demodulation of the signal. At the same time, the distortion effect and environmental interference have an increasingly significant impact on signal quality during signal transmission. In order to meet the needs of comprehensive and accurate evaluation of aerospace RF signals with high representation complexity, multiple systems, non-ideal hardware, and high dynamics, it is urgent to propose a new evaluation method that can not only perform real-time signal quality evaluation, but also effectively identify distortion effects and perform signal compensation, providing strong support for the normal operation of spacecraft in complex space environments. Summary of the invention

[0007] In order to meet the high real-time, high dynamic and reliability requirements of spacecraft radio frequency signals that are widely used in different satellite communication systems, the purpose of the present invention is to provide a multi-domain evaluation and compensation method for aerospace radio frequency signal quality based on a model library, which can perform multi-dimensional hierarchical comprehensive evaluation of spacecraft radio frequency signals, realize fast, reliable and efficient radio frequency signal evaluation, and can compensate for noise interference and distortion effects in the signal transmission process based on the mapping relationship between quality evaluation indicators and compensation coefficients, restore the original characteristics of the signal, and significantly improve the signal transmission rate and anti-interference ability.

[0008] The objective of the present invention is achieved through the following technical solutions:

[0009] The present invention discloses a method for multi-domain evaluation and compensation of aerospace radio frequency signal quality based on a model library, comprising the following steps:

[0010] Step 1: Determine the signal system parameters based on various CCSDS standards and aerospace protocol systems, extract the relevant parameters of aerospace standards in the time and frequency domain, generate standard time and frequency domain RF signals, and convert them to the relevant domain, modulation domain, energy domain, and polarization domain, and establish a sub-model library of aerospace RF signals in each domain. Based on the orthogonal time-frequency-space OTFS transformation method, transform the standard time and frequency domain signals generated by the model library into the delay-Doppler domain, and establish a delay-Doppler domain signal sub-model library. The signal system parameters include carrier frequency, coding mode, modulation mode, frame length, code rate, spread spectrum mode, spread spectrum code type, spread spectrum code length, spread spectrum code rate, Doppler frequency deviation range, and Doppler frequency deviation change rate.

[0011] Step 2: Obtain the measured signal in the time and frequency domain through the receiver, and extract the noise size n and noise tolerance ε of the time domain parameter spread spectrum signal N , pseudo code zero-crossing distortion v parameter, extract frequency domain parameter synthesis power spectrum deviation |ΔPSD(f)|, carrier leakage L carrier , integrating time domain information and frequency domain information, constructing a time-frequency domain sub-evaluation model; converting the received time-frequency domain measured signals into correlation domain, modulation domain, energy domain, polarization domain and delay Doppler domain, and constructing independent evaluation models in the six domains respectively, and using parallel data transmission method and weighted data processing method to integrate the six independent evaluation models, and constructing a multi-domain comprehensive evaluation model.

[0012] Step 2.1: Obtain the measured signal in the time and frequency domain through the receiver.

[0013] Step 2.1.1: Connect the receiving antenna to the SDR receiver.

[0014] Step 2.1.2: Amplify and filter the analog signal, and then convert it into a digital signal through a high-speed analog-to-digital converter ADC.

[0015] Step 2.1.3: Use the digital signal processing DSP unit inside or outside the receiver to demodulate and process the signal, and finally obtain the time domain waveform and spectrum information of the signal.

[0016] Step 2.2: Extract the time domain parameters of the measured signal obtained in step 2.1, including the spread spectrum signal noise size n and the noise margin ε N , pseudo-code zero-crossing distortion v parameters, and then build a time domain sub-evaluation model.

[0017] Step 2.2.1: Perform coherent cumulative averaging processing on the measured signal obtained in step 2.1 to obtain superimposed code chips after coherent cumulative averaging processing;

[0018] Step 2.2.2: Analyze the waveform of the superimposed code chips in step 2.2.1 to check whether there is obvious distortion. Count the duration of each positive and negative code chip of the signal within the code period, and compare the duration with the ideal signal code chip generated by the model library in step 1, and calculate the difference sequence between the superimposed code chips in step 2.2.1 and the ideal signal code chips in step 1.

[0019] Step 2.2.3: Perform statistical analysis on the difference sequence obtained in step 2.2.2, and calculate its maximum value, minimum value, peak value, standard deviation and mean. Based on the statistical results, determine the time domain parameters, which include the spread spectrum signal noise size n, the noise tolerance ε N And the pseudo code zero-crossing distortion v.

[0020] Step 2.3: Extract the frequency domain parameters of the measured signal obtained in step 2.1, including the synthetic power spectrum deviation |ΔPSD(f)|, carrier leakage L carrier , to construct the frequency domain sub-evaluation model.

[0021] Step 2.3.1: Derive the signal power spectrum curve from the measured RF signal obtained in step 2.1, and perform fitting analysis with the ideal signal power spectrum generated by the model library in step 1, and test and compare the signal bandwidth of the fitting curve.

[0022] The power spectrum curve PSD of the measured signal is obtained by Fourier transforming the time domain measured RF signal in step 2.1 meas (f): That is in, represents the Fourier transform of the signal x(t).

[0023] Step 2.3.2: Import the measured RF signal in step 2.1 into a spectrum analyzer or phase noise analyzer for detailed power spectrum analysis. Then, compare it with the standard power spectrum PSD of the ideal signal generated by the model library in step 1. ideal (f) is compared to calculate the frequency domain synthetic power spectrum deviation |ΔPSD(f)| at the main energy distribution point of the signal. The calculation formula can be expressed as:

[0024] |ΔPSD(f)|=|PSD meas (f)-PSD ideal (f)|

[0025] Step 2.3.3: Detect the carrier leakage L of the RF signal measured in step 2.1 through power spectrum analysis carrier situation, evaluate the degree of spectral distortion of the signal.

[0026] Carrier leakage ratio L carrier The measured carrier leakage power P carrier and the total received signal power P total The ratio represents: Carrier leakage power P carrier It can be directly read from the power spectrum of the measured signal: P carrier =PSD meas (f carrier ), where f carrier is the carrier frequency.

[0027] Step 2.4: Extract the correlation domain parameter cross-correlation function CCF(τ) and correlation loss of the measured signal obtained in step 2.1 Zero drift ε b (δ), construct the relevant domain sub-evaluation model.

[0028] Step 2.4.1: Perform carrier stripping on the measured signal obtained in step 2.1 to obtain a baseband waveform of the measured signal;

[0029] Step 2.4.2: Calculate the normalized cross-correlation function CCF(τ) between the signal obtained in step 2.4.1 and the local reference code in step 1, and calculate the difference between the ideal power of the local reference code in step 1 and the actual power of the signal obtained in step 2.4.1, i.e., the correlation loss

[0030] Among them, the correlation function CCF(τ) refers to the baseband signal s to be evaluated rec1 (t) The ideal band-limited signal s reproduced with the local signal model library 0 The normalized cross-correlation function of (t), the correlation function CCF(τ) expression is:

[0031]

[0032] In the formula, the integration time T p The length of the main code cycle.

[0033] Related loss The time domain expression is:

[0034]

[0035] In the formula, For rec1 (t) and s 0 (t), For rec1 The autocorrelation function of (t), For 0 (t) is the autocorrelation function of

[0036] Step 2.4.2: Draw the phase-comparison curve of the measured signal obtained in step 2.1, and the zero-crossing drift ε b (δ) Curve of change with the lead-lag interval δ.

[0037] Among them, the expression of zero-point drift of the discrimination function curve is:

[0038]

[0039] In the formula, ε 1 and ε 2 are two adjacent time-delayed sampling points, δ is the lead-lag spacing of the correlator, and D(ε,δ) is the expression of the discrimination function curve, that is,

[0040] Step 2.4.3: Within the transmission bandwidth of the measured signal obtained in step 2.1, the correlator interval is within the range of 0 to 1 code chip, according to ε b (δ), further analyze the changes of the zero-crossing point deviation SCB and its slope in the data segments of several code periods, and construct a related domain sub-evaluation model.

[0041] Step 2.5: Extract the modulation domain parameters I / Q branch amplitude ratio Δa, I / Q branch phase orthogonality deviation Δθ, and constellation error vector amplitude ε from the measured signal obtained in step 2.1 EVM , signal component amplitude difference Δp, and construct the modulation domain sub-evaluation model.

[0042] Step 2.5.1: The measured signal obtained in step 2.1 is passed through an ideal FIR cutoff filter with a bandwidth equal to the transmission bandwidth, and a baseband signal is obtained after demodulation.

[0043] Step 2.5.2: Sample the baseband signal obtained in step 2.5.1, draw the constellation diagram, and calculate the I / Q branch amplitude ratio Where |I (t) |and|Q (t) |represent the absolute amplitude average values ​​of the I branch and Q branch signals respectively.

[0044] Step 2.5.3: Calculate the phase orthogonality deviation of the I / Q branch based on the constellation diagram obtained in step 2.5.2

[0045] Step 2.5.4: Based on the constellation diagram obtained in step 2.5.2, calculate the deviation between the ideal signal point generated by the signal model library in step 1 and the actual received signal point in step 2.5.1, and calculate the error vector amplitude of the constellation diagram

[0046] Step 2.5.5: Calculate the component amplitude difference between the ideal signal generated by the signal model library in step 1 and the measured baseband signal obtained in step 2.5.1

[0047] Step 2.6: Use a power meter or vector network analyzer (VNA) to measure the power of the downlink measured signal obtained in step 2.1. Obtain a series of power readings through multiple measurements, perform statistical analysis on the power reading data, and extract the power variation range ∈ in the energy domain parameter P and stability σ P . Among them, the power variation range ∈ P Indicates the maximum fluctuation amplitude of the signal power during the measurement process, while the stability σ P The standard deviation of the power reading is calculated to reflect the fluctuation of the signal power. P and stability σ P , construct the energy domain sub-evaluation model.

[0048] Among them, the power variation range ∈ P =P max -P min , where P max and P min are the maximum and minimum values ​​of the measured signal power respectively; the measured signal power stability

[0049] Step 2.7: For the measured signal obtained in step 2.1 and the ideal signal generated by the signal model library in step 1, use professional spectrum test tools such as vector network analyzer VNA or spectrum analyzer to measure these signals respectively. By comparing the response ε of the measured signal obtained in step 2.1 in the frequency domain Polar The frequency domain response ε′ of the ideal signal generated by the signal model library in step 1 Polar , preliminarily determine the polarization mode of the signal. Then, use different polarization measurement settings such as linear polarization and circular polarization antennas to test the signal. By gradually adjusting the polarization direction of the receiving antenna (such as from horizontal polarization to vertical polarization, or from right-hand circular polarization to left-hand circular polarization), observe and record the change ΔP of the measured signal power in step 2.1 and the change ΔP′ of the ideal signal power generated by the model library in step 1. ε Polar ΔP-ε′ Polar ΔP′ is a quantitative indicator of the polarization domain, thus constructing a polarization domain sub-evaluation model.

[0050] Step 2.8: Transform the measured signal obtained in step 2.1 and the ideal signal generated by the model library in step 1 into the delay-Doppler domain, and extract the mean square error (MSE), peak signal-to-noise ratio (PSNR), and cross-correlation coefficient (COR) according to the delay-Doppler signal matrix of the measured signal in step 2.1 and the ideal signal in step 1. Dopper , construct the delay-Doppler domain sub-evaluation model.

[0051] Step 2.8.1: Perform frame processing on the measured time domain signal obtained in step 2.1 and the ideal time domain signal in step 1 respectively, and divide the continuous time domain signal into multiple frames (or blocks).

[0052] Step 2.8.2: Perform a discrete time Fourier transform (DTFT) on each frame (block) to convert the time domain signal into a frequency domain signal.

[0053] Step 2.8.3: Perform a sigmoid Fourier transform (SFT) on the converted frequency domain signal to convert it into a delay-Doppler domain signal.

[0054] Step 2.8.4: Arrange the obtained delay-Doppler domain signals into a matrix form, where each matrix element corresponds to a specific delay and Doppler frequency shift.

[0055] Step 2.8.5: Extract the mean square error (MSE), peak signal-to-noise ratio (PSNR), and cross-correlation coefficient (COR) based on the delay-Doppler signal matrix of the measured signal in step 2.1 and the ideal signal in step 1. Dopper , construct the delay-Doppler domain sub-evaluation model.

[0056] Specifically, the mean square error is expressed as The peak signal-to-noise ratio is expressed as The mutual correlation coefficient is expressed as Among them, X dd represents the delay-Doppler domain matrix form of the standard signal in step 1; Y dd Represents the delay-Doppler domain matrix form of the received signal in step 2.1.

[0057] Step 3: Compare the measured spacecraft RF signal with the standard signal generated by the model library, and calculate the normalized results of each quantitative index in the six domains: time-frequency domain index S tf , related domain index S r , modulation domain index S m , Energy domain index S e , polarization domain index S p and delay-Doppler domain index S d ; According to the normalized results of the six domain sub-evaluation models under the measured data, the weight of the evaluation results of each domain is calculated by the entropy weight method, and then the weighted average of the evaluation results of each domain is calculated to obtain the comprehensive quality evaluation result S of the spacecraft RF signal.

[0058] Step 3.1: Calculate the calculation indicators in the six domains respectively.

[0059] Time-frequency domain indicators The weight w tf1 ,w tf2 ,w tf3 ,w tf4 ,w tf5 All need to be adjusted according to the actual measurement situation and w tf1 +w tf2 +w tf3 +w tf4 +w tf5 =1,

[0060] Related domain indicators The weight w r1 ,w r2 ,w r3 All need to be adjusted according to the actual measurement situation through expert experience method and w r1 +w r2 +w r3 =1,

[0061] Modulation Domain Indicators The weight w m1 ,w m2 ,w m3 ,w m4 All need to be adjusted according to the actual measurement situation through expert experience method and w m1 +w m2 +w m3 +wm4 =1,

[0062] Energy Domain Index

[0063] Polarization Domain Indicators

[0064] Delay-Doppler domain index S dd =MSE 1 / PSNR+1 / PSNR 1 / COR Dopper +MSE·1 / COR Dopper .

[0065] Step 3.2: Obtain the respective weights of the six domain sub-assessment results {w tf ,w r ,w m ,w e ,w p ,w dd},

[0066] Step 3.2.1: Standardize the relevant raw data involved in the sub-assessment results of the six domains (i.e., the six indicators) to eliminate the impact of the dimension. ij To standardize:

[0067]

[0068] Among them, the j-th index value of the i-th sample is j=1,2,3,4,5,6.

[0069] Step 3.2.2: Calculate the entropy value of each of the six indicators:

[0070]

[0071] Among them, p ij is the weight of the i-th sample under the j-th index: n is the sample size.

[0072] Step 3.2.3: Calculate the coefficient of variation for each indicator:

[0073] Calculate the coefficient of difference d of the jth indicator j :

[0074] d j =1-E j

[0075] Step 3.2.4: Calculate the weight w of each indicator based on the coefficient of variation j :

[0076]

[0077] Step 3.3: Calculate the comprehensive quality assessment result S = w tf S tf +w r S r +w m S m +w e S e +w p S p +w dd S dd Among them, {w tf ,w r ,w m ,w e ,w p ,w dd}={w 1 ,w 2 ,w 3 ,w 4 ,w 5 ,w 6}.

[0078] Step 4: Based on the real-time quality assessment results of the spacecraft RF signal obtained in step 3, further analyze the noise interference and distortion effects of the signal during transmission, and calculate and obtain the distortion effect coefficient β of the transmitter and its related links. T and the distortion effect coefficient β of the receiving end and its related links R Based on the distortion coefficient, a robust signal equalization algorithm is used to process the signal, compensate for the influence of the distortion effect on the signal quality, restore the original characteristics of the signal, and improve the robustness and anti-interference ability of signal transmission.

[0079] Distortion effect coefficient β of the transmitter and its related links T and the distortion effect coefficient β of the receiving end and its related links R Expressed as

[0080] β T =β R =S=w rf S tf +w r S r +w m S m +w e S e +w p S p +w dd S dd

[0081] Furthermore, according to the obtained distortion effect coefficient β of the transmitter and its related links T, the distortion effect coefficient β of the receiving end and its related links R , based on the minimum mean square error criterion, the equalization processing matrix G is designed to cope with the influence of distortion and noise. By performing robust equalization on the received signal, the influence of the non-ideal characteristics of the hardware at the transmitting and receiving ends and the distortion on the signal evaluation results is reduced.

[0082] Specifically, in a multi-antenna system, the receiving end robust equalization matrix G based on the minimum mean square error criterion is expressed as:

[0083] G=W H H H (HWW H H H +Z) -1 ,

[0084] in,

[0085]

[0086] Furthermore, W represents a transmitting end precoding matrix, and H represents an equivalent baseband channel matrix between the transmitting and receiving ends obtained through channel estimation.

[0087] Beneficial effects:

[0088] 1. The present invention discloses a method for multi-domain evaluation and compensation of aerospace radio frequency signal quality based on a model library. The method determines signal system parameters based on various CCSDS standards and aerospace protocol systems, extracts relevant parameters of aerospace standards in the time and frequency domains, generates standard time and frequency domain radio frequency signals, and converts them into correlation domains, modulation domains, energy domains, and polarization domains. A sub-model library of aerospace radio frequency signals in each domain is established to provide an accurate and comprehensive benchmark model for signal evaluation, which helps to improve the accuracy and efficiency of multi-domain evaluation and compensation of aerospace radio frequency signal quality.

[0089] 2. The present invention discloses a multi-domain evaluation and compensation method for aerospace radio frequency signal quality based on a model library. Based on the orthogonal time-frequency space modulation OTFS method, the standard time-frequency domain signal generated by the model library is transformed into the delay-Doppler domain, and a delay-Doppler domain signal sub-model library is established to determine whether the signal still has high frequency stability and receiving demodulation accuracy when the frequency changes drastically. This method can effectively deal with the Doppler frequency shift problem of spacecraft in high dynamic scenarios.

[0090] 3. The present invention discloses a multi-domain evaluation and compensation method for aerospace radio frequency signal quality based on a model library, which obtains the measured signal in the time-frequency domain through a receiver and converts it into the correlation domain, modulation domain, energy domain, polarization domain and delay Doppler domain. Based on the comparison of the standard signal generated by the multi-system signal model library with the measured signal in the time-frequency domain, correlation domain, modulation domain, energy domain, polarization domain and delay Doppler domain, a multi-dimensional signal quality evaluation model is established, and the evaluation indicators of each domain are integrated through the entropy weight method. The comprehensive quality evaluation result of the spacecraft radio frequency signal is obtained by parallel data transmission, so as to realize the multi-dimensional, fast, efficient and reliable evaluation of the comprehensive quality of the radio frequency signal.

[0091] 4. The present invention discloses a multi-domain evaluation and compensation method for aerospace radio frequency signal quality based on a model library. The method analyzes the noise interference and distortion effects of the signal during transmission according to the real-time evaluation results, calculates and obtains the distortion effect coefficients of the transmitter, receiver and related links, and performs robust equalization and efficient compensation to restore the original characteristics of the signal, improve the robustness and anti-interference ability of signal transmission, and provide important guarantee for the stability and reliability of signal transmission in aerospace systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] Figure 1 A flow chart of a method for multi-domain evaluation and compensation of aerospace radio frequency signal quality based on a model library of the present invention;

[0093] Figure 2 is the mean square error of the system received signal as the system distortion effect coefficient changes. DETAILED DESCRIPTION

[0094] Embodiments of the present invention are described in detail below. The embodiments are exemplary and intended to be used to explain the present invention, but should not be construed as limiting the present invention.

[0095] This embodiment determines the signal system parameters based on various CCSDS standards and aerospace protocol systems, extracts the relevant parameters of aerospace standards in the time-frequency domain, generates standard time-frequency domain radio frequency signals, and converts them to the correlation domain, modulation domain, energy domain, and polarization domain, and establishes a sub-model library of aerospace radio frequency signals in each domain. Based on the orthogonal time-frequency space modulation (OTFS) method, the standard time-frequency domain signal generated by the model library is transformed into the delay Doppler domain, and a delay Doppler domain signal sub-model library is established. Next, the present invention obtains the measured signal in the time-frequency domain through a receiver, and converts it to the correlation domain, modulation domain, energy domain, polarization domain and delay Doppler domain. Based on the comparison of the standard signal generated by the multi-system signal model library with the measured signal in the time-frequency domain, correlation domain, modulation domain, energy domain, polarization domain and delay Doppler domain, a multi-dimensional signal quality evaluation model is established, and the evaluation indicators of each domain are integrated through the entropy weight method, and the comprehensive quality evaluation result of the spacecraft radio frequency signal is obtained through parallel data transmission. The specific structure is shown in Table 1:

[0096] Table 1 Contents of the multi-domain evaluation framework for multi-system aerospace signals

[0097]

[0098] According to the real-time evaluation results, the noise interference and distortion effects of the signal during transmission are analyzed, the distortion effect coefficients of the transmitter, receiver and related links are calculated and obtained, and robust equalization and efficient compensation are performed to restore the original characteristics of the signal and improve the robustness and anti-interference ability of signal transmission.

[0099] like Figure 1 As shown, this embodiment discloses a multi-domain evaluation and compensation method for aerospace radio frequency signal quality based on a model library, and the specific implementation steps are as follows:

[0100] Step 1: Based on the CCSDS131.0-B-3 standard (the standard provided for low-density parity-check code LDPC) and the CCSDS 413.0-G-1 standard (recommended practice for data compression) and other standards, combined with the aerospace protocol system (such as ISO 22665:2009 Radio frequency and modulation selection protocol for satellite communications), determine the signal system parameters, extract the key parameters in the time-frequency domain, generate a standard time-frequency domain radio frequency signal, and convert it to the correlation domain, modulation domain, energy domain and polarization domain, and establish a space radio frequency signal sub-model library for each domain. Based on the orthogonal time-frequency-space OTFS transformation method, the standard time-frequency domain signal generated by the model library is transformed into the delay-Doppler domain, and a delay-Doppler domain signal sub-model library is established. The signal system parameters include:

[0101] Carrier frequency: S-band (2GHz-4GHz), X-band (7GHz-8.5GHz),

[0102] Coding method: LDPC coding (based on CCSDS131.0-B-3 standard),

[0103] Modulation mode: QPSK (quadrature phase shift keying), 8PSK (8-phase phase shift keying),

[0104] Frame length: 2048 bits, 4096 bits,

[0105] Bit rate: 10Mbps~100Mbps,

[0106] Spread spectrum method: DSSS (direct sequence spread spectrum),

[0107] Spread spectrum code type: PN code (pseudo-random noise code),

[0108] Spread spectrum code length: 1023 bits,

[0109] Spread spectrum code rate: 10MHz~100MHz,

[0110] Doppler frequency deviation range: ±40kHz (determined by satellite orbit speed),

[0111] Doppler frequency deviation change rate: 1kHz / s~20kHz / s (affected by the motion state of the spacecraft).”

[0112] Step 2: Obtain the measured signal in the time and frequency domain through the receiver, and extract the noise size n and noise tolerance ε of the time domain parameter spread spectrum signal N , pseudo code zero-crossing distortion v parameter, extract frequency domain parameter synthesis power spectrum deviation |ΔPSD(f)|, carrier leakage L carrier , integrating time domain information and frequency domain information, constructing a time-frequency domain sub-evaluation model; converting the received time-frequency domain measured signals into correlation domain, modulation domain, energy domain, polarization domain and delay Doppler domain, and constructing independent evaluation models in the six domains respectively, and using parallel data transmission method and weighted data processing method to integrate the six independent evaluation models, and constructing a multi-domain comprehensive evaluation model.

[0113] Step 2.1: Obtain the measured signal in the time and frequency domain through the receiver.

[0114] Step 2.1.1: Connect the receiving antenna to the SDR receiver.

[0115] Step 2.1.2: Amplify and filter the analog signal, and then convert it into a digital signal through a high-speed analog-to-digital converter ADC.

[0116] Step 2.1.3: demodulate and process the signal using a digital signal processing DSP unit inside or outside the receiver, and finally obtain the time domain waveform and spectrum information of the signal. Specifically in the embodiment, by connecting to a software defined radio receiver, receiving a satellite communication signal with a bandwidth of 5 MHz, the sampling rate is set to 10 MHz, and collecting signal data with a duration of 1 ms.

[0117] Step 2.2: Extract the time domain parameters of the measured signal obtained in step 2.1, including the spread spectrum signal noise size n and the noise margin ε N , pseudo-code zero-crossing distortion v parameters, and then build a time domain sub-evaluation model.

[0118] Step 2.2.1: Perform coherent cumulative averaging processing on the measured signal obtained in step 2.1 to obtain superimposed code chips after coherent cumulative averaging processing;

[0119] Step 2.2.2: Analyze the waveform of the superimposed code chips in step 2.2.1 to check whether there is obvious distortion. Count the duration of each positive and negative code chip of the signal within the code period, and compare the duration with the ideal signal code chip generated by the model library in step 1, and calculate the difference sequence between the superimposed code chips in step 2.2.1 and the ideal signal code chips in step 1.

[0120] Step 2.2.3: Perform statistical analysis on the difference sequence obtained in step 2.2.2, and calculate its maximum value, minimum value, peak value, standard deviation and mean. Based on the statistical results, determine the time domain parameters, which include the spread spectrum signal noise size n, the noise tolerance ε N And the pseudo code zero-crossing distortion v.

[0121] Specifically in the embodiment, the collected spread spectrum signal is processed by coherent accumulation and averaging, and the obtained superimposed code chip has a noise level of -95dBm, a noise margin of 2dB, and a zero-crossing distortion v of 0.01 compared with the ideal code chip. A time domain evaluation model is established through these parameters to determine the quality of the signal.

[0122] Step 2.3: Extract the frequency domain parameters of the measured signal obtained in step 2.1, including the synthetic power spectrum deviation |ΔPSD(f)|, carrier leakage L carrier , to construct the frequency domain sub-evaluation model.

[0123] Step 2.3.1: Derive the signal power spectrum curve from the measured RF signal obtained in step 2.1, and perform fitting analysis with the ideal signal power spectrum generated by the model library in step 1, and test and compare the signal bandwidth of the fitting curve.

[0124] The power spectrum curve PSD of the measured signal is obtained by Fourier transforming the time domain measured RF signal in step 2.1 meas (f): That is in, represents the Fourier transform of the signal x(t).

[0125] Step 2.3.2: Import the measured RF signal in step 2.1 into a spectrum analyzer or phase noise analyzer for detailed power spectrum analysis. Then, compare it with the standard power spectrum PSD of the ideal signal generated by the model library in step 1. ideal (f) is compared to calculate the frequency domain synthetic power spectrum deviation |ΔPSD(f)| at the main energy distribution point of the signal. The calculation formula can be expressed as:

[0126] |ΔPSD(f)|=|PSD meas (f)-PSD ideal (f)|

[0127] Step 2.3.3: Detect the carrier leakage L of the RF signal measured in step 2.1 through power spectrum analysis carrier situation, evaluate the degree of spectral distortion of the signal.

[0128] Carrier leakage ratio L carrier The measured carrier leakage power P carrier and the total received signal power P total The ratio represents: Carrier leakage power P carrier It can be directly read from the power spectrum of the measured signal: P carrier =PSD meas (f carrier ), where f carrier is the carrier frequency.

[0129] Specifically in the embodiment, the signal is analyzed by a spectrum analyzer, and the power spectrum deviation obtained is 0.1 dB, and the carrier leakage is -40 dBc. No serious carrier leakage is detected in the spectrum analysis, and the signal quality is good.

[0130] Step 2.4: Extract the correlation domain parameter cross-correlation function CCF(τ) and correlation loss of the measured signal obtained in step 2.1 Zero drift ε b (δ), construct the relevant domain sub-evaluation model.

[0131] Step 2.4.1: Perform carrier stripping on the measured signal obtained in step 2.1 to obtain a baseband waveform of the measured signal;

[0132] Step 2.4.2: Calculate the normalized cross-correlation function CCF(τ) between the signal obtained in step 2.4.1 and the local reference code in step 1, and calculate the difference between the ideal power of the local reference code in step 1 and the actual power of the signal obtained in step 2.4.1, i.e., the correlation loss

[0133] Among them, the correlation function CCF(τ) refers to the baseband signal s to be evaluated rec1 (t) The ideal band-limited signal s reproduced with the local signal model library 0 The normalized cross-correlation function of (t), the correlation function CCF(τ) expression is:

[0134]

[0135] In the formula, the integration time T p The length of the main code cycle.

[0136] Related loss The time domain expression is:

[0137]

[0138] In the formula, For rec1 (t) and s 0 (t), For rec1 The autocorrelation function of (t), For 0 (t) is the autocorrelation function of

[0139] Step 2.4.2: Draw the phase-comparison curve of the measured signal obtained in step 2.1, and the zero-crossing drift ε b (δ) Curve of change with the lead-lag interval δ.

[0140] Among them, the expression of zero-point drift of the discrimination function curve is:

[0141]

[0142] In the formula, ε 1 and ε 2 are two adjacent time-delayed sampling points, δ is the lead-lag spacing of the correlator, and D(ε,δ) is the expression of the discrimination function curve, that is,

[0143] Step 2.4.3: Within the transmission bandwidth of the measured signal obtained in step 2.1, the correlator interval is within the range of 0 to 1 code chip, according to ε b (δ), further analyze the changes of the zero-crossing point deviation SCB and its slope in the data segments of several code periods, and construct a related domain sub-evaluation model.

[0144] Specifically in the embodiment, the cross-correlation function is used to calculate and the CCF (τ=0) is 0.95, indicating that the correlation is high. The correlation loss is -3dB and the zero-crossing drift is 0.02, indicating that the pseudo code drift is within the allowable range.

[0145] Step 2.5: Extract the modulation domain parameters I / Q branch amplitude ratio Δa, I / Q branch phase orthogonality deviation Δθ, and constellation error vector amplitude ε from the measured signal obtained in step 2.1 EVM , signal component amplitude difference Δp, and construct the modulation domain sub-evaluation model.

[0146] Step 2.5.1: The measured signal obtained in step 2.1 is passed through an ideal FIR cutoff filter with a bandwidth equal to the transmission bandwidth, and a baseband signal is obtained after demodulation.

[0147] Step 2.5.2: Sample the baseband signal obtained in step 2.5.1, draw the constellation diagram, and calculate the I / Q branch amplitude ratio Where |I(t) |and|Q (t) |represent the absolute amplitude average values ​​of the I branch and Q branch signals respectively.

[0148] Step 2.5.3: Calculate the phase orthogonality deviation of the I / Q branch based on the constellation diagram obtained in step 2.5.2

[0149] Step 2.5.4: Based on the constellation diagram obtained in step 2.5.2, calculate the deviation between the ideal signal point generated by the signal model library in step 1 and the actual received signal point in step 2.5.1, and calculate the error vector amplitude of the constellation diagram

[0150] Step 2.5.5: Calculate the component amplitude difference between the ideal signal generated by the signal model library in step 1 and the measured baseband signal obtained in step 2.5.1

[0151] Specifically in the embodiment, a constellation diagram is drawn for the demodulated signal, and the amplitude ratio of the I / Q branch is calculated to be 1.02, the phase orthogonality deviation is 0.03 rad, and the error vector amplitude is 2.5%. The modulation quality is good and the signal is relatively stable.

[0152] Step 2.6: Use a power meter or vector network analyzer (VNA) to measure the power of the downlink measured signal obtained in step 2.1. Obtain a series of power readings through multiple measurements, perform statistical analysis on the power reading data, and extract the power variation range ∈ in the energy domain parameter P and stability σ P . Among them, the power variation range ∈ P Indicates the maximum fluctuation amplitude of the signal power during the measurement process, while the stability σ P The standard deviation of the power reading is calculated to reflect the fluctuation of the signal power. P and stability σ P , construct the energy domain sub-evaluation model.

[0153] Among them, the power variation range ∈ P =P max -P min , where P max and P min are the maximum and minimum values ​​of the measured signal power respectively; the measured signal power stability

[0154] Specifically in the embodiment, the signal power is measured multiple times, and the maximum power fluctuation amplitude is 0.5dB and the stability is 0.1dB. The signal fluctuation is small, indicating that the energy of signal transmission is relatively stable.

[0155] Step 2.7: For the measured signal obtained in step 2.1 and the ideal signal generated by the signal model library in step 1, use professional spectrum test tools such as vector network analyzer VNA or spectrum analyzer to measure these signals respectively. By comparing the response ε of the measured signal obtained in step 2.1 in the frequency domain Polar The frequency domain response ε′ of the ideal signal generated by the signal model library in step 1 Polar , preliminarily determine the polarization mode of the signal. Then, use different polarization measurement settings such as linear polarization and circular polarization antennas to test the signal. By gradually adjusting the polarization direction of the receiving antenna (such as from horizontal polarization to vertical polarization, or from right-hand circular polarization to left-hand circular polarization), observe and record the change ΔP of the measured signal power in step 2.1 and the change ΔP′ of the ideal signal power generated by the model library in step 1. ε Polar ΔP-ε′ Polar ΔP′ is a quantitative indicator of the polarization domain, thus constructing a polarization domain sub-evaluation model.

[0156] Specifically in the embodiment, by comparing the polarization response of the ideal signal, the polarization difference of the measured signal is 0.02 dB, and the change of the received signal power is 0.5 dB, which indicates that the polarization characteristics are basically matched and the polarization domain performs well.

[0157] Step 2.8: Transform the measured signal obtained in step 2.1 and the ideal signal generated by the model library in step 1 into the delay-Doppler domain, and extract the mean square error (MSE), peak signal-to-noise ratio (PSNR), and cross-correlation coefficient (COR) according to the delay-Doppler signal matrix of the measured signal in step 2.1 and the ideal signal in step 1. Dopper , construct the delay-Doppler domain sub-evaluation model.

[0158] Step 2.8.1: Perform frame processing on the measured time domain signal obtained in step 2.1 and the ideal time domain signal obtained in step 1 respectively, and divide the continuous time domain signal into multiple frames (or blocks), each frame containing 1024 sampling points.

[0159] Step 2.8.2: Perform a 2048-point discrete time Fourier transform (DTFT) on each frame (block) to convert the time domain signal into a frequency domain signal.

[0160] Step 2.8.3: Perform a sigmoid Fourier transform (SFT) on the converted frequency domain signal to convert it into a delay-Doppler domain signal, where the resolution of the sigmoid Fourier transform is determined by the resolution of the delay and Doppler frequency shift, the delay resolution is 0.5 nanoseconds, and the Doppler resolution is 5 Hz.

[0161] Step 2.8.4: Arrange the obtained delay-Doppler domain signal into a matrix form, where each matrix element corresponds to a specific delay and Doppler shift. The rows and columns of the matrix represent the delay and Doppler shift, respectively, which determines the dimension of the matrix, denoted as M (delay dimension) and N (Doppler dimension). M = 128, N = 256 means that the delay dimension of the matrix is ​​128 and the Doppler dimension is 256.

[0162] Step 2.8.5: Extract the mean square error (MSE), peak signal-to-noise ratio (PSNR), and cross-correlation coefficient (COR) based on the delay-Doppler signal matrix of the measured signal in step 2.1 and the ideal signal in step 1. Dopper , construct the delay-Doppler domain sub-evaluation model.

[0163] Specifically, the mean square error is expressed as The peak signal-to-noise ratio is expressed as The mutual correlation coefficient is expressed as Among them, X dd represents the delay-Doppler domain matrix form of the standard signal in step 1; Y dd Represents the delay-Doppler domain matrix form of the received signal in step 2.1.

[0164] In the embodiment, by comparing the delay-Doppler matrix of the ideal signal and the measured signal, the calculated MSE=0.002, and by the ratio of the maximum signal to the mean square error, the calculated PSNR=35dB. Based on the cross-correlation calculation of the measured signal and the ideal signal, the calculated COR Dopper =0.98.

[0165] Step 3: Compare the measured spacecraft RF signal with the standard signal generated by the model library, and calculate the normalized results of each quantitative index in the six domains: time-frequency domain index S tf , related domain index S r , modulation domain index S m , Energy domain index S e , polarization domain index S p and delay-Doppler domain index S d ; According to the normalized results of the six domain sub-evaluation models under the measured data, the weight of the evaluation results of each domain is calculated by the entropy weight method, and then the weighted average of the evaluation results of each domain is calculated to obtain the comprehensive quality evaluation result S of the spacecraft RF signal.

[0166] Step 3.1: Calculate the calculation indicators in the six domains respectively.

[0167] Time-frequency domain indicators The weight w tf1 ,w tf2 ,w tf3 ,wtf4 ,w tf5 All need to be adjusted according to the actual measurement situation and w tf1 +w tf2 +w tf3 +w tf4 +w tf5 =1,

[0168] Related domain indicators The weight w r1 ,w r2 ,w r3 All need to be adjusted according to the actual measurement situation through expert experience method and w r1 +w r2 +w r3 =1,

[0169] Modulation Domain Indicators The weight w m1 ,w m2 ,w m3 ,w m4 All need to be adjusted according to the actual measurement situation through expert experience method and w m1 +w m2 +w m3 +w m4 =1,

[0170] Energy Domain Index

[0171] Polarization Domain Indicators

[0172] Delay-Doppler domain index S dd =MSE 1 / PSNR+1 / PSNR1 / COR Dopper +MSE·1 / COR Dopper .

[0173] Step 3.2: Obtain the respective weights of the six domain sub-assessment results {w tf ,w r ,w m ,w e ,w p ,w dd},

[0174] Step 3.2.1: Standardize the relevant raw data involved in the sub-assessment results of the six domains (i.e., the six indicators) to eliminate the impact of the dimension. ij To standardize:

[0175]

[0176] Among them, the j-th index value of the i-th sample is j=1,2,3,4,5,6.

[0177] Step 3.2.2: Calculate the entropy value of each of the six indicators:

[0178]

[0179] Among them, p ij is the weight of the i-th sample under the j-th index: n is the sample size.

[0180] Step 3.2.3: Calculate the coefficient of variation for each indicator:

[0181] Calculate the coefficient of difference d of the jth indicator j :

[0182] d j =1-E j

[0183] Step 3.2.4: Calculate the weight w of each indicator based on the coefficient of variation j :

[0184]

[0185] Step 3.3: Calculate the comprehensive quality assessment result S = w tf S tf +w r S r +w m S m +w e S e +w p S p +w dd S dd Among them, {w tf ,w r ,w m ,w e ,w p ,w dd}={w 1 ,w 2 ,w 3 ,w 4 ,w 5 ,w 6}.

[0186] Step 4: Based on the real-time quality assessment results of the spacecraft RF signal obtained in step 3, further analyze the noise interference and distortion effects of the signal during transmission, and calculate and obtain the distortion effect coefficient β of the transmitter and its related links. T and the distortion effect coefficient β of the receiving end and its related links RBased on the distortion coefficient, a robust signal equalization algorithm is used to process the signal, compensate for the influence of the distortion effect on the signal quality, restore the original characteristics of the signal, and improve the robustness and anti-interference ability of signal transmission.

[0187] Distortion effect coefficient β of the transmitter and its related links T and the distortion effect coefficient β of the receiving end and its related links R Expressed as

[0188] β T =β R =S=w tf S tf +w r S r +w m S m +w e S e +w p S p +w dd S dd

[0189] According to the obtained distortion effect coefficient β of the transmitter and its related links T , the distortion effect coefficient β of the receiving end and its related links R , based on the minimum mean square error criterion, the equalization processing matrix G is designed to cope with the influence of distortion and noise. By performing robust equalization on the received signal, the influence of the non-ideal characteristics of the hardware at the transmitting and receiving ends and the distortion on the signal evaluation results is reduced.

[0190] In a multi-antenna system, the receiving end robust equalization matrix G based on the minimum mean square error criterion is expressed as:

[0191] G=W H H H (HWW H H H +Z) -1 ,

[0192] in,

[0193]

[0194] Furthermore, W represents a transmitting end precoding matrix, and H represents an equivalent baseband channel matrix between the transmitting and receiving ends obtained through channel estimation.

[0195] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention without departing from the principles and purpose of the present invention.

Claims

1. A multi-domain evaluation and compensation method for aerospace radio frequency signal quality based on a model library, characterized by: The following steps are included: Step 1: Determine the signal system parameters based on various CCSDS standards and aerospace protocol systems, extract the relevant parameters of aerospace standards in the time and frequency domain, generate standard time and frequency domain RF signals, and convert them to the correlation domain, modulation domain, energy domain, and polarization domain, and establish a sub-model library of aerospace RF signals in each domain; Based on the orthogonal time-frequency-space (OTFS) transformation method, the standard time-frequency domain signal generated by the model library is transformed into the delay-Doppler domain, and a delay-Doppler domain signal sub-model library is established; the signal system parameters include carrier frequency, coding mode, modulation mode, frame length, code rate, spread spectrum mode, spread spectrum code type, spread spectrum code length, spread spectrum code rate, Doppler frequency deviation range and Doppler frequency deviation change rate; Step 2: Obtain the measured signal in the time and frequency domain through the receiver, and extract the noise size n and noise tolerance ε of the time domain parameter spread spectrum signal N , pseudo code zero-crossing distortion v parameter, extract frequency domain parameter synthesis power spectrum deviation |ΔPSD(f)|, carrier leakage L carrier , integrate time domain information and frequency domain information to build a time-frequency domain sub-evaluation model; convert the received time-frequency domain measured signals into the correlation domain, modulation domain, energy domain, polarization domain and delay Doppler domain, and build independent evaluation models in the six domains respectively, and use parallel data transmission and weighted data processing methods to integrate the six independent evaluation models to build a multi-domain comprehensive evaluation model; Step 3: Compare the measured spacecraft RF signal with the standard signal generated by the model library, and calculate the normalized results of each quantitative index in the six domains: time-frequency domain index S tf , related domain index S r , modulation domain index S m , Energy domain index S e , polarization domain index S p and delay-Doppler domain index S d ; According to the normalized results of the six domain sub-evaluation models under the measured data, the weight of each domain evaluation result is calculated by the entropy weight method, and then the weighted average of the evaluation results of each domain is calculated to obtain the comprehensive quality evaluation result S of the spacecraft RF signal; Step 4: Based on the real-time quality assessment results of the spacecraft RF signal obtained in step 3, further analyze the noise interference and distortion effects of the signal during transmission, calculate and obtain the distortion effect coefficient βT of the transmitter and its related links and the distortion effect coefficient β of the receiver and its related links R Based on the distortion coefficient, a robust signal equalization algorithm is used to process the signal, compensate for the influence of the distortion effect on the signal quality, restore the original characteristics of the signal, and improve the robustness and anti-interference ability of signal transmission.

2. The method for multi-domain evaluation and compensation of aerospace radio frequency signal quality based on a model library according to claim 1, characterized in that: Step 2 is implemented as follows: Step 2.1: Obtain the measured signal in the time and frequency domain through the receiver; Step 2.1.1: Connect the receiving antenna to the software defined radio receiver; Step 2.1.2: Amplify and filter the analog signal, and then convert the analog signal into a digital signal through a high-speed analog-to-digital converter ADC; Step 2.1.3: Use the digital signal processing DSP unit inside or outside the receiver to demodulate and process the signal, and finally obtain the time domain waveform and spectrum information of the signal; Step 2.2: For the measured signal obtained in step 2.1, extract the time domain parameters, including the spread spectrum signal noise size n, the noise margin εN, and the pseudo code zero-crossing distortion v parameter, and then construct a time domain sub-evaluation model; Step 2.2.1: Perform coherent cumulative averaging processing on the measured signal obtained in step 2.1 to obtain superimposed code chips after coherent cumulative averaging processing; Step 2.2.2: Analyze the waveform of the superimposed code chips in step 2.2.1 to check whether there is obvious distortion; count the duration of each positive and negative code chip of the signal within the code period, and compare the duration with the ideal signal code chip generated by the model library in step 1, and calculate the difference sequence between the superimposed code chip in step 2.2.1 and the ideal signal code chip in step 1; Step 2.2.3: Perform statistical analysis on the difference sequence obtained in step 2.2.2, and calculate its maximum value, minimum value, peak value, standard deviation and mean; Based on the statistical results, the time domain parameters are determined, and the time domain parameters include the spread spectrum signal noise size n, the noise margin εN and the pseudo code zero-crossing distortion v; Step 2.3: Extract the frequency domain parameters of the measured signal obtained in step 2.1, including the synthetic power spectrum deviation |ΔPSD(f)|, carrier leakage L carrier , to construct the frequency domain sub-assessment model; Step 2.3.1: Derive the signal power spectrum curve from the measured RF signal obtained in step 2.1, and perform fitting analysis with the ideal signal power spectrum generated by the model library in step 1, and test and compare the signal bandwidth of the fitting curve; The power spectrum curve PSD of the measured signal is obtained by Fourier transforming the time domain measured RF signal in step 2.1 meas (f): That is in, represents the Fourier transform of the signal x(t); Step 2.3.2: Import the measured RF signal in step 2.1 into a spectrum analyzer or phase noise analyzer for detailed power spectrum analysis; then, compare it with the standard power spectrum PSD of the ideal signal generated by the model library in step 1. ideal (f) is compared to calculate the frequency domain synthetic power spectrum deviation |ΔPSD(f)| at the main energy distribution point of the signal. The calculation formula is expressed as: |ΔPSD(f)|=|PSD meas (f)-PSD ideal (f)| Step 2.3.3: Detect the carrier leakage L of the RF signal measured in step 2.1 through power spectrum analysis carrier Situation, evaluate the degree of spectrum distortion of the signal; Carrier leakage ratio L carrier The measured carrier leakage power P carrier and the total received signal power P total The ratio represents: Carrier leakage power P carrier It can be directly read from the power spectrum of the measured signal: P carrier =PSD meas (f carrier ), where f carrier is the carrier frequency; Step 2.4: Extract the correlation domain parameter cross-correlation function CCF(τ) and correlation loss of the measured signal obtained in step 2.1 Zero drift ε b (δ), construct the sub-evaluation model of the relevant domain; Step 2.4.1: Perform carrier stripping on the measured signal obtained in step 2.1 to obtain a baseband waveform of the measured signal; Step 2.4.2: Calculate the normalized cross-correlation function CCF(τ) between the signal obtained in step 2.4.1 and the local reference code in step 1, and calculate the difference between the ideal power of the local reference code in step 1 and the actual power of the signal obtained in step 2.4.1, i.e., the correlation loss Among them, the correlation function CCF(τ) refers to the baseband signal S to be evaluated rec1 The normalized cross-correlation function of the ideal band-limited signal s0(t) reproduced by the local signal model library is expressed as: In the formula, the integration time T p The length of the main code cycle; Related loss The time domain expression is: In the formula, For rec1 The cross-correlation function of s(t) and s0(t), For rec1 The autocorrelation function of (t), is the autocorrelation function of s0(t); Step 2.4.2: Draw the phase-comparison curve of the measured signal obtained in step 2.1, and the zero-crossing drift ε b (δ) Curve of variation with lead-lag interval δ; Among them, the expression of zero-point drift of the discrimination function curve is: In the formula, ε1 and ε2 are two adjacent delayed sampling points, δ is the lead-lag interval of the correlator, and D(ε, δ) is the expression of the discrimination function curve, that is, Step 2.4.3: Within the transmission bandwidth of the measured signal obtained in step 2.1, the correlator interval is within the range of 0 to 1 code chip, according to ε b (δ), further analyze the changes in the zero-crossing point deviation SCB and its slope of the S curve in the data segments of several code periods, and construct a related domain sub-evaluation model; Step 2.5: Extract the modulation domain parameters I / Q branch amplitude ratio Δa, I / Q branch phase orthogonality deviation Δθ, and constellation error vector amplitude ε from the measured signal obtained in step 2.1 EVM , signal component amplitude difference Δp, and construct a modulation domain sub-evaluation model; Step 2.5.1: The measured signal obtained in step 2.1 is passed through an ideal FIR cutoff filter with a bandwidth equal to the transmission bandwidth, and then demodulated to obtain a baseband signal; Step 2.5.2: Sample the baseband signal obtained in step 2.5.1, draw the constellation diagram, and calculate the I / Q branch amplitude ratio Where |I (t) |and|Q (t) |represents the absolute amplitude averages of the I branch and Q branch signals respectively; Step 2.5.3: Calculate the phase orthogonality deviation of the I / Q branch based on the constellation diagram obtained in step 2.5.2 Step 2.5.4: Based on the constellation diagram obtained in step 2.5.2, calculate the deviation between the ideal signal point generated by the signal model library in step 1 and the actual received signal point in step 2.5.1, and calculate the error vector amplitude of the constellation diagram Step 2.5.5: Calculate the component amplitude difference between the ideal signal generated by the signal model library in step 1 and the measured baseband signal obtained in step 2.5.1 Step 2.6: Use a power meter or vector network analyzer (VNA) to measure the power of the downlink measured signal obtained in step 2.1; obtain a series of power readings through multiple measurements, perform statistical analysis on the power reading data, and extract the power variation range ∈ in the energy domain parameter P and stability σ P ; Among them, the power variation range ∈ P Indicates the maximum fluctuation amplitude of the signal power during the measurement process, while the stability σ P The standard deviation of the power reading is calculated to reflect the fluctuation of the signal power. P and stability σ P , construct the energy domain sub-assessment model; Among them, the power variation range ∈ P =P max -P min , where P max and P min are the maximum and minimum values ​​of the measured signal power respectively; the measured signal power stability Step 2.7: For the measured signal obtained in step 2.1 and the ideal signal generated by the signal model library in step 1, use a professional spectrum test tool such as a vector network analyzer VNA or a spectrum analyzer to measure these signals respectively; by comparing the response ε of the measured signal obtained in step 2.1 in the frequency domain Polar The frequency domain response ε′ of the ideal signal generated by the signal model library in step 1 Polar , preliminarily determine the polarization mode of the signal; then, use different polarization measurement settings such as linear polarization and circular polarization antennas to test the signal; by gradually adjusting the polarization direction of the receiving antenna, observe and record the change AP of the measured signal power in step 2.1 and the change ΔP′ of the ideal signal power generated by the model library in step 1; ε Polar ΔP-ε′ Polar ΔP′ is a quantitative index of polarization domain, and a polarization domain sub-evaluation model is constructed; Step 2.8: Transform the measured signal obtained in step 2.1 and the ideal signal generated by the model library in step 1 into the delay-Doppler domain, and extract the mean square error (MSE), peak signal-to-noise ratio (PSNR), and cross-correlation coefficient (COR) according to the delay-Doppler signal matrix of the measured signal in step 2.1 and the ideal signal in step 1. Dopper , construct the delay-Doppler domain sub-evaluation model; Step 2.8.1: Perform frame processing on the measured time domain signal obtained in step 2.1 and the ideal time domain signal obtained in step 1, respectively, and divide the continuous time domain signal into multiple frames; Step 2.8.2: Perform discrete time Fourier transform (DTFT) on each frame to convert the time domain signal into a frequency domain signal; Step 2.8.3: Perform sigmoid Fourier transform (SFT) on the converted frequency domain signal to convert it into a delay-Doppler domain signal; Step 2.8.4: Arrange the obtained delay-Doppler domain signals into a matrix form, where each matrix element corresponds to a specific delay and Doppler frequency shift; Step 2.8.5: Extract the mean square error (MSE), peak signal-to-noise ratio (PSNR), and cross-correlation coefficient (COR) based on the delay-Doppler signal matrix of the measured signal in step 2.1 and the ideal signal in step 1. Dopper , construct the delay-Doppler domain sub-evaluation model; The mean square error is expressed as The peak signal-to-noise ratio is expressed as The mutual correlation coefficient is expressed as Among them, X dd represents the delay-Doppler domain matrix form of the standard signal in step 1; Y dd Represents the delay-Doppler domain matrix form of the received signal in step 2.

1.

3. The method for multi-domain evaluation and compensation of aerospace radio frequency signal quality based on a model library as claimed in claim 2, characterized in that: Step 3 is implemented as follows: Step 3.1: Calculate the calculation indicators in the six domains respectively; Time-frequency domain indicators The weight w tf1 , w tf2 , w tf3 , w tf4 , w tf5 All need to be adjusted according to the actual measurement situation and w tf1 +w tf2 +w tf3 +w tf4 +w tf5 =1, Related domain indicators The weight w r1 , w r2 , w r3 All need to be adjusted according to the actual measurement situation through expert experience method and w r1 +w r2 +w r3 =1, Modulation Domain Indicators The weight w m1 , w m2 , w m3 , w m4 All need to be adjusted according to the actual measurement situation through expert experience method and w m1 +w m2 +w m3 +w m4 =1, Energy Domain Index Polarization Domain Indicators Delay-Doppler domain index S dd =MSE 1 / PSNR+1 / PSNR 1 / COR Dopper +MSE·1 / COR Dopper ; Step 3.2: Obtain the respective weights of the six domain sub-assessment results {w tf , w r , w m , w e , w p , w dd }, Step 3.2.1: Standardize the relevant raw data involved in the sub-assessment results of the six domains to eliminate the impact of dimensions; ij To standardize: Among them, the jth index value of the i-th sample, j = 1, 2, 3, 4, 5, 6; Step 3.2.2: Calculate the entropy value of each of the six indicators: Among them, p ij is the weight of the i-th sample under the j-th index: n is the sample size; Step 3.2.3: Calculate the coefficient of variation for each indicator: Calculate the coefficient of difference d of the jth indicator j : d j =1-E j Step 3.2.4: Calculate the weight w of each indicator based on the coefficient of variation j : Step 3.3: Calculate the comprehensive quality assessment result S = w tf S tf +w r S r +w m S m +w e S e +w p S p +w dd S dd ; Among them, {w tf , w r , w m , w e , w p , w dd }={w1, w2, w3, w4, w5, w6}.

4. The method for multi-domain evaluation and compensation of aerospace radio frequency signal quality based on a model library as claimed in claim 3, characterized in that: In step 4, The distortion effect coefficient βT of the transmitter and its related links and the distortion effect coefficient βR of the receiver and its related links are expressed as β T =β R =S=w tf S tf +w r S r +w m S m +w e S e +w p S p +w dd S dd According to the obtained distortion effect coefficient βT of the transmitter and its related links, and the distortion effect coefficient βR of the receiver and its related links, the equalization processing matrix G is designed based on the minimum mean square error criterion to cope with the influence of distortion and noise; by performing robust equalization processing on the received signal, the influence of non-ideal characteristics of the hardware of the transmitter and receiver and distortion on the signal evaluation results is reduced.

5. The method for multi-domain evaluation and compensation of aerospace radio frequency signal quality based on a model library as claimed in claim 4, characterized in that: In a multi-antenna system, the receiving end robust equalization matrix G based on the minimum mean square error criterion is expressed as: G=W H H H (HWW H H H +Z) -1 , in, Furthermore, W represents a transmitting end precoding matrix, and H represents an equivalent baseband channel matrix between the transmitting and receiving ends obtained through channel estimation.

Images