A variable input and output frequency V-Q band test transponder and test method

Abstract

This invention discloses a V-Q band test repeater and test method with variable input and output frequencies. The test repeater is located between the uplink and downlink and includes a down-conversion mixer circuit, a variable high local oscillator circuit, an intermediate frequency (IF) processing circuit, an up-conversion mixer circuit, and a variable low local oscillator circuit. The uplink can transmit the V-band input signal to the down-conversion mixer circuit for down-conversion mixing with the variable high IF signal. The generated IF signal can be sent to the up-conversion mixer circuit for down-conversion mixing with the variable low IF signal, and output a Q-band output signal. Through two frequency conversions and the independent configuration of the local oscillator signal by the variable high and variable low IF circuits, the input and output frequencies within the band can be arbitrarily configured. This enables the conversion of any uplink V-band signal into any downlink Q-band signal. Furthermore, the performance indicators of all frequency point combinations across the entire band can be predicted using measured data from a small number of characteristic frequency points, thereby improving test efficiency.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and in particular to a VQ band test transponder and test method with variable input and output frequencies. Background Technology

[0002] Existing satellite communication ground stations include uplink and downlink, with different frequencies, both significantly higher than baseband signal frequencies, especially the V-band uplink and Q-band downlink. To perform loopback testing of the uplink and downlink radio frequencies on the ground, the uplink frequency needs to be converted to the Q-band. Existing upconverters or downconverters have the same input and output frequency bandwidth, or a fixed intermediate frequency signal frequency, with the radio frequency signal frequency changed via a local oscillator. This only allows upconverting the baseband signal to a very high frequency or downconverting the received high-frequency satellite signal to a very low baseband signal. When anomalies occur in the satellite signal received by the satellite or the satellite signal received by the ground, it is impossible to determine whether the problem lies with the ground equipment or the satellite itself. Therefore, a frequency relay device is needed to perform ground loopback testing to pinpoint whether the anomaly originates from the satellite itself. Summary of the Invention

[0003] In view of this, the present invention proposes a VQ band test transponder and test method with variable input and output frequencies. By using two frequency conversions and two independent configurable local oscillator signals, the input and output frequencies within the band can be arbitrarily configured, realizing the conversion of any uplink frequency into any downlink band signal.

[0004] The technical solution of this invention is implemented as follows: A VQ band test repeater with variable input and output frequencies is configured between the uplink and downlink. It includes a down-conversion mixer circuit, a variable high local oscillator circuit, an intermediate frequency (IF) processing circuit, an up-conversion mixer circuit, and a variable low local oscillator circuit. The variable high IF circuit generates a variable high IF signal. The down-conversion mixer circuit down-converts the input signal from the uplink and the variable high IF signal to generate an IF signal. The IF processing circuit processes the IF signal. The variable low IF circuit generates a variable low IF signal. The up-conversion mixer circuit up-converts the processed IF signal and the variable low IF signal to generate an output signal that is sent to the downlink.

[0005] Preferably, it further includes an input filter and an output filter. The input filter filters the input signal of the uplink and then sends it to the downconverter mixer circuit. The output filter filters the output signal and then sends it to the downlink.

[0006] Preferably, it also includes an internal and external parameter reference circuit, which is electrically connected to the variable high local oscillator circuit and the variable low local oscillator circuit, respectively.

[0007] Preferably, the variable high local oscillator circuit includes a high local oscillator phase-locked loop, a high local oscillator frequency multiplier, and a high local oscillator filter. The high local oscillator phase-locked loop is electrically connected to the high local oscillator frequency multiplier and the internal and external parameter reference circuits, respectively. The high local oscillator frequency multiplier is electrically connected to the down-conversion mixer circuit, and the high local oscillator filter is electrically connected between the high local oscillator phase-locked loop and the high local oscillator frequency multiplier.

[0008] Preferably, the variable low local oscillator circuit includes a low local oscillator phase-locked loop, a low local oscillator frequency multiplier, and a low local oscillator filter. The low local oscillator phase-locked loop is electrically connected to the low local oscillator frequency multiplier and the internal and external parameter reference circuits, respectively. The low local oscillator frequency multiplier is electrically connected to the down-conversion mixer circuit, and the low local oscillator filter is electrically connected between the low local oscillator phase-locked loop and the low local oscillator frequency multiplier.

[0009] Preferably, the internal and external parameter reference circuit provides a reference power divider signal to the high local oscillator phase-locked loop and the low local oscillator phase-locked loop via a built-in temperature-controlled crystal oscillator or by an externally provided reference power divider signal.

[0010] A test method for a VQ band transponder using variable input and output frequencies includes the following steps: Step S1: Based on the input bandwidth of the V-band input signal and the output bandwidth of the Q-band output signal of the test transponder, construct a full-band test frequency point grid and select a sparse sampling frequency point set from it; Step S2: Control the test repeater to switch to each group of frequency points in the sparse sampling frequency point set in sequence, and measure the link performance parameters corresponding to each group of frequency points. Step S3: Input the link performance parameters into the pre-trained dual mapping prediction model, use the mid-frequency axis dual symmetry as a physical constraint, and output the performance parameter prediction results corresponding to all frequency point combinations in the full frequency band. Step S4: Using the V-band frequency points as the horizontal axis and the Q-band frequency points as the vertical axis, map the performance parameters corresponding to all frequency point combinations in the entire frequency band into a two-dimensional heat map, and combine it with the preset qualified threshold to determine the performance qualification.

[0011] Preferably, step S1 includes the following steps: The input bandwidth of the V-band input signal and the output bandwidth of the Q-band output signal of the test transponder are divided in an equally spaced step manner to obtain N sets of input frequency points and M sets of output frequency points; N sets of input frequency points and M sets of output frequency points form N×M sets of frequency point combinations, and a full-band test frequency point grid is constructed. The initial sampling frequency point combination is selected from the full-band test frequency point grid according to the dual mapping constraint formula, where the dual mapping constraint formula is: , Let be the absolute value of the offset of the i-th input frequency point relative to the center frequency of the intermediate frequency signal. This represents the ratio of the multiplication order of the high-local-oscillator frequency multiplier to that of the low-local-oscillator frequency multiplier. It is the absolute value of the offset of the j-th output frequency point relative to the center frequency of the intermediate frequency signal; Based on the initial sampling frequency point combination, the upper and lower edge frequency points of the V band and the upper and lower edge frequency points of the Q band are added to form a sparse sampling frequency point set.

[0012] Preferably, step S2 includes the following specific steps: Synchronously adjust the output frequencies of the variable high local oscillator circuit and the variable low local oscillator circuit to match the input and output frequencies of the test transponder with the frequency combination selected from the sparse sampling frequency points; After the link status stabilizes, the link gain, phase noise, spurious suppression, output 1dB compression point, and in-band flatness corresponding to the selected frequency point combination are collected as link performance parameters. The link performance parameters are fixed and the deviations are eliminated and stored.

[0013] Preferably, step S3 includes the following specific steps: A few-sample graph neural network with embedded physical constraints is constructed as a dual mapping prediction model, with the frequency coordinates and link performance parameters corresponding to the frequency point combination as inputs; The dual mapping prediction model introduces a dual symmetric loss constraint term during inference and optimization, imposing a mid-frequency axisymmetric constraint on the prediction results. The expression for the dual symmetric loss constraint term is as follows:

[0014] in For dual symmetric loss values, For the i-th input frequency point in band V Q-band output frequency point j The corresponding model prediction performance value, , These are the Q-band and V-band mirror frequencies after intermediate frequency axis mirror mapping. These are the model prediction performance values ​​corresponding to the Q-band mirror frequency and the V-band mirror frequency. The dual mapping prediction model is iteratively optimized with the goal of minimizing the total loss function, which includes the mean square error loss of the measured and predicted values ​​and the dual symmetric loss constraint term, and finally outputs the performance parameter prediction results corresponding to all frequency point combinations in the full frequency band.

[0015] Compared with the prior art, the beneficial effects of the present invention are: This invention discloses a VQ band test transponder with variable input and output frequencies. After receiving the V-band input signal transmitted from the uplink, the transponder performs down-conversion mixing with a variable high local oscillator signal through a down-conversion mixing circuit to obtain a lower intermediate frequency (IF) signal. The IF signal is then processed and sent to an up-conversion mixing circuit, which performs up-conversion mixing with the IF signal and the variable low IF signal to output a higher Q-frequency signal. By performing two frequency conversions, the V-band is shifted to the Q-band, thereby enabling loopback testing of ground equipment.

[0016] The present invention provides a testing method based on the dual mapping rule of dual local oscillators of the same source. By combining the measured link performance parameters of a small number of sparsely sampled frequency points with a dual mapping prediction model with physical constraints, the performance deduction and qualification determination of all frequency point combinations in the entire frequency band can be completed. Under the premise of ensuring comprehensive test coverage and reliable results, the testing workload and time consumption in the large bandwidth scenario of VQ band are significantly reduced, and the testing efficiency and operation and maintenance response speed are effectively improved. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only preferred embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of a VQ band test transponder with variable input and output frequencies according to the present invention. Figure 2 This is a schematic diagram illustrating the application of a VQ band test transponder with variable input and output frequencies according to the present invention. Figure 3 A flowchart illustrating a test method for a VQ band transponder with variable input and output frequencies according to the present invention; In the diagram, 101 is the input filter; 102 is the down-converter mixer circuit; 20 is the variable high local oscillator circuit; 201 is the high local oscillator frequency multiplier; 202 is the high local oscillator filter; 203 is the high local oscillator phase-locked loop; 301 is the internal and external parameter reference circuit; 40 is the variable low local oscillator circuit; 401 is the low local oscillator frequency multiplier; 402 is the low local oscillator filter; 403 is the low local oscillator phase-locked loop; 501 is the intermediate frequency processing circuit; 601 is the up-converter mixer circuit; and 701 is the output filter. Detailed Implementation

[0019] To better understand the technical content of this invention, a specific embodiment is provided below, and the invention will be further described in conjunction with the accompanying drawings.

[0020] See Figures 1 to 2 This invention provides a VQ band test repeater with variable input and output frequencies, positioned between the uplink and downlink. It includes a down-conversion mixer circuit 102, a variable high local oscillator circuit 20, an intermediate frequency (IF) processing circuit 501, an up-conversion mixer circuit 601, and a variable low local oscillator circuit 40. The variable high IF circuit 20 generates a variable high IF signal. The down-conversion mixer circuit 102 down-converts the input signal from the uplink and the variable high IF signal to generate an IF signal. The IF processing circuit 501 processes the IF signal. The variable low IF circuit 40 generates a variable low IF signal. The up-conversion mixer circuit 601 up-converts the processed IF signal and the variable low IF signal to generate an output signal that is sent to the downlink.

[0021] This invention provides a VQ band test transponder with variable input and output frequencies, connected to the uplink and downlink, enabling loopback testing of ground equipment. The loopback test converts the coupled output signal of the uplink to the coupled input signal of the downlink. The coupled output signal of the uplink is the uplink transmitted V-band signal, and the coupled input signal of the downlink is the downlink received Q-band signal. The test transponder uses the coupled output signal as its input signal, performs two frequency conversions to shift the V-band signal to the Q-band signal, and uses the final output signal as the coupled input signal of the downlink. This achieves loopback testing of the uplink and downlink of the satellite ground station, enabling full-link self-testing, indicator verification, fault location, and routine maintenance without consuming satellite resources, thus improving system commissioning efficiency, maintenance convenience, and operational security.

[0022] Specifically, the V-band signal, as the input signal, is sent to the down-conversion mixer circuit 102 for down-conversion mixing. The signal mixed with the input signal is the variable high local oscillator signal generated by the variable high local oscillator circuit 20. Through down-conversion mixing, a lower intermediate frequency signal can be obtained and enters the intermediate frequency processing circuit 501. The intermediate frequency processing circuit 501 performs noise suppression, signal amplification, and attenuation on the intermediate frequency signal. The adjusted signal is input to the up-conversion mixer circuit 601. The up-conversion mixer circuit 601 simultaneously receives the variable low local oscillator signal sent by the variable low local oscillator circuit 40 and performs up-conversion mixing to obtain a higher radio frequency signal that is different from the transmitted signal, namely the required Q-band signal. Thus, loopback testing is realized. By independently setting the variable high local oscillator circuit 20 and the variable low local oscillator circuit 40, the input frequency and output frequency within the band can be arbitrarily configured, realizing the conversion of any uplink frequency band signal into any downlink frequency band signal.

[0023] Preferably, it also includes an input filter 101 and an output filter 701. The input filter 101 filters the input signal of the uplink and then sends it to the downconverter mixer circuit 102. The output filter 701 filters the output signal and then sends it to the downlink.

[0024] The input filter 101 can filter the input signal, removing the image frequency and other out-of-band spurious signals before it enters the down-conversion mixer circuit 102. The output signal generated by the up-conversion mixer circuit 601 carries out-of-band spurious signals such as local oscillator leakage. The output filter 701 can suppress out-of-band spurious signals, remove local oscillator and other spurious signals, and output a purer radio frequency signal.

[0025] Preferably, it also includes an internal and external parameter reference circuit 301, which is electrically connected to the variable high local oscillator circuit 20 and the variable low local oscillator circuit 40, respectively.

[0026] The internal and external parameter reference circuit 301 can provide a reliable reference signal for the variable high local oscillator circuit 20 and the variable low local oscillator circuit 40, so that the variable high local oscillator circuit 20 and the variable low local oscillator circuit 40 can generate variable high local oscillator signals and variable low local oscillator signals that are synchronized with the reference.

[0027] Preferably, the variable high local oscillator circuit 20 includes a high local oscillator phase-locked loop 203, a high local oscillator frequency multiplier 201, and a high local oscillator filter 202. The high local oscillator phase-locked loop 203 is electrically connected to the high local oscillator frequency multiplier 201 and the internal and external parameter reference circuit 301, respectively. The high local oscillator frequency multiplier 201 is electrically connected to the down-conversion mixer circuit 102, and the high local oscillator filter 202 is electrically connected between the high local oscillator phase-locked loop 203 and the high local oscillator frequency multiplier 201.

[0028] The intermediate frequency (IF) signal is a fixed-frequency output. When different input V-band signals are configured, the high local oscillator (LO) signal can be calculated using the known IF signal frequency. Then, the LO fundamental frequency signal to be output by the LO phase-locked loop (PLL) 203 can be calculated. Therefore, the LO fundamental frequency can be automatically calculated based on the input signal frequency and the predefined IF signal frequency. The LO PLL 203 then locks the LO fundamental frequency signal onto the reference signal. After that, the LO filter 202 suppresses out-of-band spurious signals and filters out the redundant components generated by the LO PLL 203. Then, the LO frequency multiplier 201 multiplies the frequency of the LO PLL 203 to a local oscillator signal that can drive the down-conversion mixer. Thus, the down-conversion mixer circuit 102 can down-convert the input V-band signal to the IF signal.

[0029] Preferably, the variable low local oscillator circuit 40 includes a low local oscillator phase-locked loop 403, a low local oscillator frequency multiplier 401, and a low local oscillator filter 402. The low local oscillator phase-locked loop 403 is electrically connected to the low local oscillator frequency multiplier 401 and the internal and external parameter reference circuit 301, respectively. The low local oscillator frequency multiplier 401 is electrically connected to the down-conversion mixer circuit 102, and the low local oscillator filter 402 is electrically connected between the low local oscillator phase-locked loop 403 and the low local oscillator frequency multiplier 401.

[0030] Similarly, the low local oscillator fundamental signal can be calculated based on the configured output signal frequency and intermediate frequency signal. Then, the low local oscillator phase-locked loop 403 locks the low local oscillator fundamental signal onto the reference signal. After being filtered by the low local oscillator filter 402, the low local oscillator frequency multiplier 401 multiplies the calculated low local oscillator phase-locked loop 403 frequency to the local oscillator signal that drives the up-conversion mixer. Thus, the up-conversion mixer circuit 601 can perform up-conversion mixing to generate the Q-band output signal.

[0031] Preferably, the internal and external parameter reference circuit 301 provides a reference power divider signal to the high local oscillator phase-locked loop 203 and the low local oscillator phase-locked loop 403 via its own temperature-controlled crystal oscillator or by an externally provided reference power divider signal.

[0032] The reference signals for both the high-LO phase-locked loop 203 and the low-LO phase-locked loop 403 come from the internal and external reference circuit 301. The internal and external reference circuit 301 can have its own temperature-controlled crystal oscillator or can be externally provided with reference power divided to the high-LO phase-locked loop 203 and the low-LO phase-locked loop 403. When the internal temperature-controlled crystal oscillator is present and there is no external reference signal input, the signal of this temperature-controlled crystal oscillator is divided into two paths by the power divider circuit of the reference circuit and given to the high-LO phase-locked loop 203 and the low-LO phase-locked loop 403. When there is an external reference signal input, the phase-locked loop circuit inside this reference circuit performs phase-locking of the internal and external reference signals. After locking, it stably outputs the reference signal required by the high-LO phase-locked loop 203 and the low-LO phase-locked loop 403 to be synchronized with the external reference.

[0033] Reference Figure 3 As shown, a test method for a VQ band transponder using variable input and output frequencies includes the following steps: Step S1: Based on the input bandwidth of the V-band input signal and the output bandwidth of the Q-band output signal of the test transponder, construct a full-band test frequency point grid and select a sparse sampling frequency point set from it; Step S2: Control the test repeater to switch to each group of frequency points in the sparse sampling frequency point set in sequence, and measure the link performance parameters corresponding to each group of frequency points. Step S3: Input the link performance parameters into the pre-trained dual mapping prediction model, use the mid-frequency axis dual symmetry as a physical constraint, and output the performance parameter prediction results corresponding to all frequency point combinations in the full frequency band. Step S4: Using the V-band frequency points as the horizontal axis and the Q-band frequency points as the vertical axis, map the performance parameters corresponding to all frequency point combinations in the entire frequency band into a two-dimensional heat map, and combine it with the preset qualified threshold to determine the performance qualification.

[0034] As the core device for realizing loopback self-testing of terrestrial radio frequency links, the test transponder supports flexible configuration of any input and output frequency points within the full operating bandwidth. The number of frequency point combinations that can be covered increases exponentially with the increase of bandwidth. However, traditional testing methods generally adopt a full-frequency point traversal test scheme, and the test time increases linearly with the number of frequency points. In the case of tens of thousands of frequency point combinations in the millimeter wave high-bandwidth scenario, the full performance test of a single device can take several hours, which seriously restricts the efficiency of factory calibration and the response speed of on-site operation and maintenance. If ordinary random sampling or simple interpolation is used to reduce the amount of testing, the prediction error cannot be effectively constrained because the inherent architecture characteristics of the test transponder, which has dual local oscillators from the same source and shares an intermediate frequency, are not taken into account. This can easily lead to the omission of performance testing at key locations such as frequency band edges and harmonic points, making it difficult to guarantee the comprehensiveness and reliability of test results.

[0035] To address this, the present invention provides a testing method that constructs a full-band test frequency grid based on the input and output bandwidths. This grid covers all configurable input and output frequencies of the test transponder, ensuring no frequency blind spots in the final test results. Since the uplink and downlink of the test transponder share the same intermediate frequency (IF) signal and reference signal, link performance distortion exhibits a natural dual correlation. Therefore, only a sparse sampling frequency set needs to be selected from the full-band test frequency grid for testing to preserve the full-band performance distribution characteristics, thereby significantly reducing the workload of actual measurements. During testing, the test transponder is controlled to sequentially switch to each frequency combination in the selected sparse sampling frequency set, and the link performance parameters are measured. The measured link performance parameters are then fed into a dual mapping prediction model, which embeds IF axisymmetric physical parameters. Constraints can significantly improve prediction accuracy with few samples, allowing reliable extrapolation of performance parameters for all frequency combinations across the entire frequency band with only a small number of sampling points. This achieves a leap from localized measurements to full-domain performance reconstruction. Finally, a two-dimensional heatmap is constructed, using different color levels to represent the numerical values ​​of performance parameters. Simultaneously, the predicted performance parameters for each frequency combination can be compared with preset acceptable thresholds, automatically marking performance deviation points and statistically analyzing their distribution areas and proportions. Based on the statistical results of deviation points, a full-band performance acceptance judgment is output, along with a link performance degradation trend analysis result. This transforms full-domain performance data into intuitive and actionable test conclusions, achieving a several-fold increase in test efficiency while ensuring the reliability and comprehensiveness of test results. It fully adapts to the testing requirements of VQ band transponders with large bandwidth and multiple frequency combinations.

[0036] Preferably, step S1 includes the following steps: The input bandwidth of the V-band input signal and the output bandwidth of the Q-band output signal of the test transponder are divided in an equally spaced step manner to obtain N sets of input frequency points and M sets of output frequency points; N sets of input frequency points and M sets of output frequency points form N×M sets of frequency point combinations, and a full-band test frequency point grid is constructed. The initial sampling frequency point combination is selected from the full-band test frequency point grid according to the dual mapping constraint formula, where the dual mapping constraint formula is: , Let be the absolute value of the offset of the i-th input frequency point relative to the center frequency of the intermediate frequency signal. This represents the ratio of the multiplication factor of the high-local-oscillator frequency multiplier 201 to that of the low-local-oscillator frequency multiplier 401. It is the absolute value of the offset of the j-th output frequency point relative to the center frequency of the intermediate frequency signal; Based on the initial sampling frequency point combination, the upper and lower edge frequency points of the V band and the upper and lower edge frequency points of the Q band are added to form a sparse sampling frequency point set.

[0037] The input and output frequency step values ​​are both no greater than the minimum frequency tuning step of the test transponder. The final full-band test frequency grid completely covers all configurable operating frequencies of the test transponder. Then, the dual mapping rule of the same source dual local oscillator is used for screening. Before screening, the center frequency of the intermediate frequency signal is determined, and then the offset of the input and output frequencies relative to the center frequency of the intermediate frequency signal is determined. Finally, the dual mapping constraint formula can be constructed by combining the fixed ratio of the frequency harmonics. Based on the dual mapping constraint formula, preliminary screening can be performed to obtain the initial sampling frequency combination. Finally, the initial sampling frequency combination is supplemented to obtain a sparse sampling frequency set accounting for 1%-5% of the total number of combinations.

[0038] The dual mapping constraint formula uses the center frequency of the intermediate frequency signal as the reference axis and establishes a proportional relationship between the input frequency offset and the output frequency offset based on the fixed ratio of the high and low local oscillator frequency harmonics. Essentially, it anchors the symmetrical distribution law of amplitude / phase frequency distortion of the two local oscillator harmonic links. Since the two local oscillators are from the same source and have the same architecture, the sampling frequency points selected along this proportional trajectory correspond exactly to the positions where the distortion characteristics of the two local oscillator links are consistent. It can capture the core distortion law of the entire frequency band with the fewest sampling points and avoid the omission of features caused by random sampling.

[0039] Preferably, step S2 includes the following specific steps: Synchronously adjust the output frequencies of the variable high local oscillator circuit 20 and the variable low local oscillator circuit 40 so that the input and output frequencies of the test transponder are matched with the frequency combination selected from the sparse sampling frequency points. After the link status stabilizes, the link gain, phase noise, spurious suppression, output 1dB compression point, and in-band flatness corresponding to the selected frequency point combination are collected as link performance parameters. The link performance parameters are fixed and the deviations are eliminated and stored.

[0040] After sending a frequency control command to the test transponder, the test transponder can adjust the output frequency of the local oscillator circuit so that the input and output frequencies of the test transponder match the currently selected frequency combination. After the phase-locked loop is locked and the link status is stable, the link performance parameters can be collected. At the same time, the pre-stored test instrument system error calibration coefficients are called to perform fixed deviation elimination processing on each set of measured link performance parameters. Finally, the calibrated valid data is stored in the sampling dataset.

[0041] Preferably, step S3 includes the following specific steps: A few-sample graph neural network with embedded physical constraints is constructed as a dual mapping prediction model, with the frequency coordinates and link performance parameters corresponding to the frequency point combination as inputs; The dual mapping prediction model introduces a dual symmetric loss constraint term during inference and optimization, imposing a mid-frequency axisymmetric constraint on the prediction results. The expression for the dual symmetric loss constraint term is as follows:

[0042] in The dual symmetric loss value, For the i-th input frequency point in band V Q-band output frequency point j The corresponding model prediction performance value, , These are the Q-band and V-band mirror frequencies after intermediate frequency axis mirror mapping. These are the model prediction performance values ​​corresponding to the Q-band mirror frequency and the V-band mirror frequency. The dual mapping prediction model is iteratively optimized with the goal of minimizing the total loss function. The total loss function includes the mean square error loss of the measured and predicted values ​​and the dual symmetric loss constraint term, and finally outputs the performance parameter prediction results corresponding to all frequency point combinations in the full frequency band.

[0043] The dual mapping prediction model is pre-trained using transfer learning. First, the basic weights of the model are trained using historical full-band measured datasets of test transponders with the same architecture. Then, it is fine-tuned and adapted using sparse sampling data of the current test transponder. After adaptation, it is used for full-band performance inference. Its input includes link performance parameters and the corresponding frequency point combination coordinates. The dual mapping prediction model introduces dual symmetric loss constraints during inference and optimization. It uses the hardware characteristics of the test transponder itself as a strong constraint to improve the prediction accuracy under few samples, and solves the problem of high prediction error of ordinary interpolation algorithms in millimeter-wave high-bandwidth scenarios. Finally, the dual mapping prediction model is iteratively optimized with the goal of minimizing its total loss function, which can obtain the performance parameter prediction results corresponding to all frequency point combinations.

[0044] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A VQ band test transponder with variable input and output frequencies, configured between the uplink and downlink, characterized in that, The system includes a down-conversion mixer circuit, a variable high local oscillator circuit, an intermediate frequency (IF) processing circuit, an up-conversion mixer circuit, and a variable low local oscillator circuit. The variable high IF circuit generates a variable high IF signal. The down-conversion mixer circuit down-converts and mixes the input signal from the uplink with the variable high IF signal to generate an IF signal. The IF processing circuit processes the IF signal. The variable low IF circuit generates a variable low IF signal. The up-conversion mixer circuit up-converts and mixes the processed IF signal with the variable low IF signal to generate an output signal that is sent to the downlink.

2. A VQ band test transponder with variable input and output frequencies according to claim 1, characterized in that, It also includes an input filter and an output filter. The input filter filters the input signal of the uplink and then sends it to the downconverter mixer circuit. The output filter filters the output signal and then sends it to the downlink.

3. A VQ band test transponder with variable input and output frequencies according to claim 1, characterized in that, It also includes an intrinsic and extrinsic parameter reference circuit, which is electrically connected to the variable high local oscillator circuit and the variable low local oscillator circuit, respectively.

4. A VQ band test transponder with variable input and output frequencies according to claim 3, characterized in that, The variable high local oscillator circuit includes a high local oscillator phase-locked loop, a high local oscillator frequency multiplier, and a high local oscillator filter. The high local oscillator phase-locked loop is electrically connected to the high local oscillator frequency multiplier and the internal and external parameter reference circuits, respectively. The high local oscillator frequency multiplier is electrically connected to the down-conversion mixer circuit, and the high local oscillator filter is electrically connected between the high local oscillator phase-locked loop and the high local oscillator frequency multiplier.

5. A VQ band test transponder with variable input and output frequencies according to claim 4, characterized in that, The variable low local oscillator circuit includes a low local oscillator phase-locked loop, a low local oscillator frequency multiplier, and a low local oscillator filter. The low local oscillator phase-locked loop is electrically connected to the low local oscillator frequency multiplier and the internal and external parameter reference circuits, respectively. The low local oscillator frequency multiplier is electrically connected to the down-conversion mixer circuit, and the low local oscillator filter is electrically connected between the low local oscillator phase-locked loop and the low local oscillator frequency multiplier.

6. A VQ band test transponder with variable input and output frequencies according to claim 5, characterized in that, The internal and external parameter reference circuits provide reference power division signals to the high local oscillator phase-locked loop and the low local oscillator phase-locked loop via a built-in temperature-controlled crystal oscillator or externally.

7. A test method for a VQ band transponder with variable input and output frequencies as described in any one of claims 1-6, characterized in that, Includes the following steps: Step S1: Based on the input bandwidth of the V-band input signal and the output bandwidth of the Q-band output signal of the test transponder, construct a full-band test frequency point grid and select a sparse sampling frequency point set from it; Step S2: Control the test repeater to switch to each group of frequency points in the sparse sampling frequency point set in sequence, and measure the link performance parameters corresponding to each group of frequency points. Step S3: Input the link performance parameters into the pre-trained dual mapping prediction model, use the mid-frequency axis dual symmetry as a physical constraint, and output the performance parameter prediction results corresponding to all frequency point combinations in the full frequency band. Step S4: Using the V-band frequency points as the horizontal axis and the Q-band frequency points as the vertical axis, map the performance parameters corresponding to all frequency point combinations in the entire frequency band into a two-dimensional heat map, and combine it with the preset qualified threshold to determine the performance qualification.

8. The test method according to claim 7, characterized in that, The specific steps of step S1 include: The input bandwidth of the V-band input signal and the output bandwidth of the Q-band output signal of the test transponder are divided in an equally spaced step manner to obtain N sets of input frequency points and M sets of output frequency points; N sets of input frequency points and M sets of output frequency points form N×M sets of frequency point combinations, and a full-band test frequency point grid is constructed. The initial sampling frequency point combination is selected from the full-band test frequency point grid according to the dual mapping constraint formula, where the dual mapping constraint formula is: , Let be the absolute value of the offset of the i-th input frequency point relative to the center frequency of the intermediate frequency signal. This represents the ratio of the multiplication order of the high-local-oscillator frequency multiplier to that of the low-local-oscillator frequency multiplier. It is the absolute value of the offset of the j-th output frequency point relative to the center frequency of the intermediate frequency signal; Based on the initial sampling frequency point combination, the upper and lower edge frequency points of the V band and the upper and lower edge frequency points of the Q band are added to form a sparse sampling frequency point set.

9. The test method according to claim 7, characterized in that, The specific steps of step S2 include: Synchronously adjust the output frequencies of the variable high local oscillator circuit and the variable low local oscillator circuit to match the input and output frequencies of the test transponder with the frequency combination selected from the sparse sampling frequency points; After the link status stabilizes, the link gain, phase noise, spurious suppression, output 1dB compression point, and in-band flatness corresponding to the selected frequency point combination are collected as link performance parameters. The link performance parameters are fixed and the deviations are eliminated and stored.

10. The test method according to claim 7, characterized in that, The specific steps of step S3 include: A few-sample graph neural network with embedded physical constraints is constructed as a dual mapping prediction model, with the frequency coordinates and link performance parameters corresponding to the frequency point combination as inputs; The dual mapping prediction model introduces a dual symmetric loss constraint term during inference and optimization, imposing a mid-frequency axisymmetric constraint on the prediction results. The expression for the dual symmetric loss constraint term is as follows: in The dual symmetric loss value, For the i-th input frequency point in band V Q-band output frequency point j The corresponding model prediction performance value, , These are the Q-band and V-band mirror frequencies after intermediate frequency axis mirror mapping. These are the model prediction performance values ​​corresponding to the Q-band mirror frequency and the V-band mirror frequency. The dual mapping prediction model is iteratively optimized with the goal of minimizing the total loss function. The total loss function includes the mean square error loss of the measured and predicted values ​​and the dual symmetric loss constraint term, and finally outputs the performance parameter prediction results corresponding to all frequency point combinations in the full frequency band.

Images