Arbitrary multi-port s-parameter calibration and de-embedding method

Abstract

The application discloses a kind of arbitrary multiport S parameter calibration and de-embedding method, belong to multiport calibration field.Method includes: using network analyzer respectively on interpolation after S parameter data base of calibration piece each single-port measurement, error term of each single-port is solved;Using network analyzer respectively on calibration piece multiport measurement, determine the arbitrary multiport S parameter measurement data of switch item and isolation item correction completion;Based on the multiport error of target through path generated by radiation, and using the multiport error of target through path, the multiport error of other path outside target through path is completed, to combine the error term of each single-port to the arbitrary multiport S parameter measurement data of switch item and isolation item correction completion is calibrated and de-embedded, and the multiport calibrated S parameter matrix is obtained.This scheme can improve calibration accuracy and calculation efficiency.

Description

Technical Field

[0001] This invention relates to the field of multiport calibration technology, and in particular to an arbitrary multiport S-parameter calibration and de-embedding method. Background Technology

[0002] In the field of radio frequency and microwave measurement, with the rapid development of cutting-edge communication technologies such as 5G / 6G, low-Earth orbit satellite communication systems, high-speed interconnected networks for artificial intelligence, terahertz integrated circuits, and large-scale multiple-input multiple-output (MIMO) antenna arrays, modern communication systems have an urgent need for highly integrated, multi-port, and multifunctional components. This has led to a surge in the number of ports on the device under test (DUT) and increasingly complex electromagnetic coupling relationships. Multi-port S-parameter calibration has become the core scientific foundation for high-precision electromagnetic characteristic characterization. Multi-port S-parameter calibration establishes an accurate error model to mathematically characterize and correct rigorous system errors in vector network analyzer systems, such as directivity, source mismatch, load mismatch, crosstalk, and frequency response, which are based on electromagnetic field theory. This allows the measurement benchmark to be traced back to international standards. Implementing multi-port S-parameter calibration can not only eliminate phase and amplitude deviations introduced by test fixtures and restore the true network characteristics of devices, but also significantly improve the consistency and repeatability of batch testing. It is a prerequisite for achieving accurate performance characterization of the entire chain from chip to system and plays a decisive role in ensuring the R&D quality and mass production reliability of key components in cutting-edge fields such as 5G / 6G, low-orbit satellite communication, and artificial intelligence.

[0003] The core problem facing current multi-port S-parameter calibration algorithms lies in the systemic challenges brought about by the non-ideal nature of high-frequency calibration components and the high complexity of multi-port systems. As the frequency increases and the port spacing decreases, crosstalk and mutual coupling effects between ports are significantly enhanced. Traditional calibration methods, represented by short-circuit, open-circuit, load, and straight-through (SOLT), become inaccurate because their calibration models ignore parasitic parameters, frequency dispersion, and losses, resulting in a significant decrease in measurement accuracy. In addition, the traditional linear system assumption is difficult to handle the harmonic distortion and intermodulation interference of nonlinear components such as high-power amplifiers and mixers. Calibration methods based on dual-port extension are prone to error accumulation due to path differences when reusing existing processes, making it difficult to balance accuracy, efficiency, and scalability.

[0004] Therefore, there is an urgent need to provide an arbitrary multi-port S-parameter calibration and de-embedding method. Summary of the Invention

[0005] To address the issues of low calibration accuracy and computational efficiency in existing multi-port S-parameter calibration algorithms, this invention provides an arbitrary multi-port S-parameter calibration and de-embedding method.

[0006] On the one hand, an arbitrary multi-port S-parameter calibration and de-embedding method is provided, the method comprising: A network analyzer was used to perform single-port measurements on the calibrator based on the interpolated S-parameter data basis. The measured reflection coefficients of the single port of the calibrator under different reflection states were input into the overdetermined equation system, and the error terms of each single port were obtained by solving the singular value decomposition method. Multi-port measurements are performed on the calibration device using a network analyzer to generate a switching term matrix after isolation term correction. The switching term matrix is ​​then corrected to obtain a switching term correction matrix. Based on the switching term correction matrix, arbitrary multi-port S-parameter measurement data collected by the network analyzer are corrected to obtain arbitrary multi-port S-parameter measurement data with completed switching term and isolation term corrections. Among them, isolation terms are added to the switching term matrix to form a 10-term error model matrix. The multi-port error of the target through path is generated, and the multi-port error of other paths outside the target through path is supplemented by depth-first search and the multi-port error of the target through path. The error terms of each single port are combined to calibrate and de-embed the arbitrary multi-port S-parameter measurement data that have completed the correction of the switching and isolation terms, so as to obtain the multi-port calibrated S-parameter matrix.

[0007] On the other hand, an arbitrary multi-port S-parameter calibration and de-embedding device based on the steps described in any method embodiment of the specification is provided, the device comprising: The solution unit is used to perform single-port measurements on the calibrator based on the interpolated S-parameter data basis using a network analyzer, so as to input the measured reflection coefficient of the single port of the calibrator under different reflection states into the overdetermined equation system, and use the singular value decomposition method to solve for the error term of each single port. The correction unit is used to perform multi-port measurements on the calibration component using a network analyzer, generate a switching term matrix after isolation term correction, correct the switching term matrix to obtain a switching term correction matrix, and correct the arbitrary multi-port S-parameter measurement data collected by the network analyzer based on the switching term correction matrix to obtain arbitrary multi-port S-parameter measurement data with completed switching term and isolation term corrections; wherein, an isolation term is added to the switching term matrix to form a 10-term error model matrix; The calibration unit is used to generate the multi-port error of the target through path, and use depth-first search and the multi-port error of the target through path to complete the multi-port error of other paths outside the target through path. In order to combine the error terms of each single port, the arbitrary multi-port S-parameter measurement data that has completed the correction of the switching term and isolation term is calibrated and de-embedded to obtain the multi-port calibrated S-parameter matrix.

[0008] On the other hand, a computer device is provided, the computer device including a memory and a processor, the memory for storing a computer program, and the processor for executing the computer program stored in the memory to implement the steps of the method described above.

[0009] On the other hand, a computer-readable storage medium is provided, wherein a computer program is stored therein, and when the computer program is executed by a processor, it implements the steps of the method described above.

[0010] On the other hand, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the method described above.

[0011] The technical solution provided by this invention can bring at least the following beneficial effects: This paper replaces the traditional polynomial model with measured data from calibration components based on S-parameter data, addressing the issues of non-ideal high-frequency models and calibration failures caused by long-term wear. It fully considers the error terms introduced by inter-port crosstalk and mutual coupling effects, as well as the non-ideal characteristics of the RF switches inside the vector network analyzer. Corresponding isolation and switching terms are introduced to correct these effects and improve data accuracy, and the 8-term error model is rewritten into 10 terms. For multi-port calibration, the paper optimizes the through-connection method by using n-1 connections to reduce the number of connections, i.e., the target through-path. A depth-first search method is used to automatically solve for the through-path between the remaining two ports, reducing the computational dimensionality. Therefore, this scheme can improve calibration accuracy and computational efficiency. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart of an arbitrary multi-port S-parameter calibration and de-embedding method provided by an embodiment of the present invention; Figure 2 This is a schematic diagram of a radial port connection path provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of a chain port connection path provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of a hybrid port connection path provided in an embodiment of the present invention; Figure 5 This is a structural diagram of an arbitrary multi-port S-parameter calibration and de-embedding device provided in an embodiment of the present invention; Figure 6 This is a hardware architecture diagram of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0015] The following describes the specific implementation of the above concept.

[0016] Please refer to Figure 1 This invention provides an arbitrary multi-port S-parameter calibration and de-embedding method, which includes: Step 100: Use a network analyzer to perform single-port measurements on the calibration piece based on the interpolated S-parameter data basis. Input the measured reflection coefficients of the single port of the calibration piece under different reflection states into the overdetermined equation system and use the singular value decomposition method to solve for the error terms of each single port. Step 102: Use a network analyzer to perform multi-port measurements on the calibration component to generate a switching term matrix after isolation term correction. Correct the switching term matrix to obtain a switching term correction matrix. Based on the switching term correction matrix, correct the arbitrary multi-port S-parameter measurement data collected by the network analyzer to obtain arbitrary multi-port S-parameter measurement data with completed switching term and isolation term corrections. Among them, an isolation term is added to the switching term matrix to form a 10-term error model matrix. Step 104: Generate the multi-port error of the target through path, and use the depth-first search and the multi-port error of the target through path to complete the multi-port error of other paths besides the target through path. Combine the error terms of each single port to calibrate and de-embed the arbitrary multi-port S-parameter measurement data that have completed the correction of the switching and isolation terms, and obtain the multi-port calibrated S-parameter matrix.

[0017] In this embodiment of the invention, measured data from calibration components based on S-parameter data are used to replace the traditional polynomial model, solving the problems of non-ideal high-frequency models and calibration failure caused by long-term wear. The error terms introduced by port crosstalk and mutual coupling effects, as well as the non-ideal characteristics of the RF switches inside the vector network analyzer, are fully considered. Isolation and switching terms are introduced to correct these effects and improve data accuracy, and the 8-term error model is rewritten into 10 terms. For multi-port calibration, the direct connection method is optimized by using n-1 connections to reduce the number of connections, i.e., the target direct path. A depth-first search method is used to automatically solve for the direct connection path between the remaining two ports, reducing the computational dimensionality. Therefore, this solution can improve calibration accuracy and computational efficiency.

[0018] The following description Figure 1 The execution method of each step is shown.

[0019] For step 100: Traditional vector network analyzers using polynomial model calibration kits neglect parasitic parameters, frequency dispersion, and losses, leading to inaccuracies and a significant decrease in measurement accuracy. Furthermore, they struggle to accurately characterize the non-ideal characteristics of calibration kits caused by parasitic capacitance / inductance, contact losses, and mechanical wear. This solution replaces these with calibration kits based on S-parameter data. Its core advantage lies in abandoning the reliance of traditional polynomial models on lumped parameter approximations or finite-order analytical expressions. Instead, it directly utilizes the measured complex S-parameters (including amplitude and phase) at all frequencies uniquely corresponding to the physical standard to characterize its electrical characteristics. This method not only fully preserves the true non-ideal response of the calibration kit across a wide bandwidth caused by parasitic resonance, frequency dispersion, and port discontinuities, but also fundamentally avoids the interpolation errors introduced by polynomial fitting at high frequencies due to order limitations, model assumption biases, and extrapolation divergences. This allows for more accurate extraction of systematic error terms, significantly improving measurement consistency, uncertainty quantification capabilities, and traceability accuracy in high-frequency measurement scenarios such as millimeter waves and terahertz, better meeting the requirements of high-precision RF and microwave measurements.

[0020] When discussing calibration-related issues, it is generally assumed that the vector network analyzer is set to the target frequency before calibration, and then measurements and calibrations are performed at these frequencies. However, it is impractical to extract data before each use of the data-based calibration component. Therefore, this solution employs a data interpolation algorithm for frequency adaptive scaling. Since the S-parameters of the calibration component change rapidly with frequency, it is difficult to interpolate between data points. The simplest approach is to interpolate the real and imaginary parts separately for complex numbers, but a better method is to interpolate the amplitude and phase separately. However, traditional Lagrange interpolation (polynomial) and cubic spline interpolation are inconvenient to use. Therefore, this solution flexibly selects circular interpolation and vector fitting interpolation methods for data interpolation scaling based on data characteristics to obtain the interpolated S-parameter data basis. This embodiment changes the mathematical description of the calibration component from the traditional polynomial model to one based on measured data, solving the problems of non-ideal high-frequency models and calibration failure caused by long-term wear. The introduction of a frequency adaptive interpolation algorithm improves the testing flexibility and ease of use of the calibration component.

[0021] In this embodiment, in order to improve the numerical stability of the calibration algorithm and reduce the dependence on the quality of the calibration components, this patent selects more (>3) different reflection coefficients to construct overdetermined equation (3), and uses singular value decomposition to solve it numerically, further reducing the random error of the single-port error term. Setting redundant terms to construct an overdetermined system of equations has the core scientific value of breaking the singularity of the system of equations by introducing redundant observation information, transforming the ill-conditioned problem that is extremely sensitive to noise and has non-unique or unstable solutions into a well-posed problem with statistically optimal solutions; with the help of the singular value decomposition optimization algorithm, this mechanism can effectively average random measurement errors, suppress outlier interference and significantly reduce the condition number of the solution, thereby greatly improving the robustness of numerical calculation and the accuracy and reliability of the results when there is uncertainty in the data.

[0022] The overdetermined equations are: in, Indicates the reflection coefficient of the x-port of the calibration component itself. The reflection coefficient of the single port of the network splitter terminal calibration component x under different reflection states is represented by i, where i corresponds to different reflection states. This represents the error term for each individual port, where x represents any port and n is the number of ports.

[0023] It should be noted that the numerical solution can be obtained. , and ,So It can be determined that the multi-port calibration used in step 104 is employed. This embodiment increases the number of single-port calibration components, adds redundant information to construct an overdetermined system of equations, and uses singular value decomposition to solve the overdetermined system of equations. The algorithm is stable, highly accurate, and less susceptible to the influence of ill-conditioned matrices.

[0024] Regarding step 102: Generally, before actually running the calibration algorithm, the multi-port raw data obtained in the "error acquisition" step will be corrected according to the test requirements and the selected error model. This mainly includes two steps: switch term correction and isolation term correction.

[0025] The "switching error term" in the error model represents the matching variation caused by the source switch. This error term represents the ratio of the incident wave to the reflected wave at the same measurement port. These switching terms are independent of external components or connection methods and are therefore entirely internal parameters of the vector network analyzer. In most cases, these error terms are very stable, and once their values ​​are determined by some method, they can be applied to other calculations using the 8-term model without having to measure the incident wave at the termination port again.

[0026] When measuring high-isolation devices (such as band-stop filters), an isolation error term is added for measurement and data correction to improve data accuracy in the "high insertion loss" region. Therefore, this embodiment adds an isolation term to the 8-term error model, correcting the model to a 10-term model. Taking ports 1 and 2 as an example, the isolation error term is the S-parameter measured when each port under test is connected to a load, i.e. and Based on this definition, the isolation term can be extended to multi-port networks. Typically, multiple averaging steps are required to accurately obtain the isolation error term; otherwise, the measurement result is merely the noise level of the vector network analyzer.

[0027] The correction process for the switching and isolation terms in arbitrary multi-port S-parameter measurement data is as follows: in, In the formula, This is the switching term matrix after the isolation term correction. This represents the switching term at port i as measured by the vector network analyzer. , Let j be the incident and reflected waves measured at this port, and j be the source port at this time. The S-parameters, i.e., the isolation term, are measured when the load is connected to both ports. Here are the S-parameter measurements for each tested port pair under the direct connection condition, and [STC] is the switching term correction matrix. The S-parameters of port ij collected by the network analyzer, i.e., S-parameter measurement data of any number of ports. Element; The switch term matrix after isolation term correction The element value; Arbitrary multiport S-parameter measurement data for completing the correction of switching and isolation terms.

[0028] This embodiment fully considers the crosstalk and mutual coupling effects between ports and the error terms introduced by the non-ideal characteristics of the internal RF switches of the vector network analyzer. Correspondingly, isolation terms and switching terms are introduced to correct these effects and improve data accuracy, and the 8 error models are rewritten into 10 terms.

[0029] Regarding step 104: It should be noted that it can be done according to Figures 2-4 Select the target direct path as shown. Figures 2-4 These include radial, chain, and hybrid types. Depending on the specific requirements, you can choose either radial, chain, or a hybrid form. However, considering transmission accuracy and error accumulation, radial is the preferred option.

[0030] Therefore, in some implementations, the step "generating the multiport error of the target straight path and using depth-first search and the multiport error of the target straight path to complete the multiport errors of other paths besides the target straight path" includes: Determine the number of ports n of the calibration piece, and designate one of the ports as the common port; Based on the radial approach, the path from the common port to each other port is taken as the target straight-through path. Straight-through measurement is performed on the target straight-through path to determine the multi-port error of each target straight-through path; the number of target straight-through paths is n-1. Automatic paths are generated using depth-first search; the automatic path is from other port x to another other port y via common port c. For each automatic path, the following is performed: Based on the multiport error of the corresponding target direct path, calculate the multiport error between port x and port y in the automatic path; Continue until the multi-port errors of all paths except the target direct path are completed, thus obtaining the multi-port errors between all port paths.

[0031] The data acquisition process for existing multiport calibration error models requires a pass-through connection for measurement. This results in n×(n-1) / 2 direct paths. Taking n as the number of ports, for example with 20 ports, this would create 190 direct paths, which is very time-consuming.

[0032] In this embodiment, reference Figure 2 The diagram illustrates a radial connection path. Port 1 is taken as the common port, and the direct connection between port 1 and other ports is taken as the target direct path. Assuming the number of ports n is 16, then only 15 connections are needed, that is, port 1 is connected to ports 2-16 respectively, resulting in 15 target direct paths.

[0033] The multiport error elements of the target through path in the subsequent transition matrix, which are axisymmetric about the axis, satisfy the following relationship: In the formula, Multi-port error for the target direct path. and For single-port error terms, By simplifying the numerator and denominator, we can see that the multi-port errors of the axisymmetric multi-ports are equal. Therefore, without repeated measurements, the multi-port errors of the direct path of each target can be measured.

[0034] For paths other than the target direct path, such as the paths between ports 2 and 3, the automatic path generated by depth-first search is port 2-port 1-port 3; the multi-port error elements of these automatic paths in the subsequent transition matrix satisfy the following relationship: In the formula, x is 2, y is 3, and c is common port 1. It can be seen that... and The multi-port error of the measured target straight-through path. The single-port error term obtained in step 100, This represents the simplification of the numerator and denominator. Therefore, the automatic path can be automatically calculated based on the multi-port error of the target through path and the single-port error of the common port, without the need for connection measurement using a network analyzer.

[0035] Therefore, after completing n-1 straight-through measurements in multi-port calibration, it is necessary to complete the transmission paths between ports in the network. This embodiment uses a depth-first search method for automatic path generation to directly calculate the multi-port errors of paths other than the target straight-through path. This reduces the number of straight-through connections from n×(n-1) / 2 to n-1, significantly reducing the collection of error terms, simplifying the operation, and improving computational efficiency.

[0036] In some implementations, the step "combining the error terms of each single port to calibrate and de-embed the arbitrary multi-port S-parameter measurement data after completing the switching and isolation term corrections, to obtain the multi-port calibrated S-parameter matrix" includes steps A1-A2: A1, substitute the error terms of each single port, the multi-port error between all port paths, and the arbitrary multi-port S-parameter measurement data after completing the correction of the switching and isolation terms into the transition matrix.

[0037] In some implementations, the transition matrix is ​​determined by the following formula: in, In the formula, This is the transition matrix. This is the diagonalized error matrix corresponding to position pq. This represents the element value at the corresponding position of the nth port. To complete the measurement data of arbitrary multiport S-parameters for switching and isolation terms correction, To complete the correction of switching and isolation terms, the corresponding position element values ​​of arbitrary multiport S-parameter measurement data are obtained. and For each single port, The multi-port error between port paths is represented by x and y, where x and y represent port pairs composed of any two ports.

[0038] A2, based on the transition matrix, calculate the S-parameter matrix after multi-port calibration.

[0039] In this step, the S-parameter matrix after multiport calibration is determined by the following formula: In the formula, The S-parameter matrix after multi-port calibration. It is the identity matrix. is the diagonalization error matrix at the corresponding position, and -1 is the inverse.

[0040] The above process enables full matrix processing in the calibration algorithm of a multiport vector network analyzer, abstracting the complex physical error network into rigorous linear algebraic operators, thus achieving a strictly closed decoupling capability mathematically. By integrating the coupling error terms, crosstalk effects, and mismatch disturbances between all measurement ports into a complete error coefficient matrix or cascaded transmission matrix, the calibration process is transformed into a problem of inverting and decomposing a large matrix. This not only completely breaks the dependence of the traditional cascaded method on the port topology order and recursive accumulation error by utilizing unified mathematical tools such as matrix partitioning and inversion, but also enables numerical early warning of ill-conditioned calibration states through matrix condition number analysis, and separates random errors and systematic errors in one step using the rank property of the matrix. More importantly, this full matrix form makes the uncertainty propagation of the multiport network conform to strict multivariate statistical laws, significantly improving computational complexity efficiency and numerical stability, thereby establishing a universal metrological benchmark from physical measurement to mathematical reconstruction in multiport vector network analysis.

[0041] In some implementations, after obtaining the S-parameter matrix after multiport calibration, the following steps are also included: Based on the isolation term correction matrix, the isolation term is corrected twice on the S-parameter matrix after multi-port calibration to obtain the final S-parameter matrix after multi-port calibration.

[0042] Specifically, the final S-parameter matrix after multiport calibration is determined as follows: Obtain the isolation term correction matrix: The S-parameter matrix after multi-port calibration is then subjected to a second correction using isolation terms to obtain the final S-parameter matrix after multi-port calibration: In the formula, The final S-parameter matrix after multiport calibration. The S-parameter matrix after multi-port calibration. The correction matrix for the isolation term, These are the S-parameters measured when the load is connected to two ports.

[0043] In this embodiment, after obtaining the S-parameter matrix after multi-port calibration, a secondary isolation term correction can be performed on the matrix to finally obtain a multi-port calibrated S-parameter matrix with higher calibration accuracy.

[0044] Please refer to Figure 5 This invention provides an arbitrary multi-port S-parameter calibration and de-embedding device for implementing the steps of any method embodiment in the specification. The device includes: Solver 501 is used to perform single-port measurements on the calibrator based on the interpolated S-parameter data basis using a network analyzer, so as to input the measured reflection coefficient of the single port of the calibrator under different reflection states into the overdetermined equation system, and use the singular value decomposition method to solve for the error term of each single port. The correction unit 502 is used to perform multi-port measurements on the calibration component using a network analyzer to generate a switching term matrix after isolation term correction. The switching term matrix is ​​then corrected to obtain a switching term correction matrix. Based on the switching term correction matrix, the arbitrary multi-port S-parameter measurement data collected by the network analyzer is corrected to obtain arbitrary multi-port S-parameter measurement data with both switching and isolation term corrections completed. Among these, an isolation term is added to the switching term matrix to form a 10-term error model matrix. The calibration unit 503 is used to generate multi-port errors of the target through path radially, and to supplement the multi-port errors of other paths outside the target through path by using depth-first search and the multi-port errors of the target through path. In order to combine the error terms of each single port, the arbitrary multi-port S-parameter measurement data that has completed the correction of the switching term and the isolation term are calibrated and de-embedded to obtain the multi-port calibrated S-parameter matrix.

[0045] It should be noted that the above device embodiments and method embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0046] Embodiments of this application also provide a computer device, please refer to... Figure 6 The computer device includes a processor and a memory, the memory storing at least one instruction, at least one program, code set or instruction set, the at least one instruction, at least one program, code set or instruction set being loaded and executed by the processor to implement the arbitrary multi-port S-parameter calibration and de-embedding method provided in the above method embodiments.

[0047] Embodiments of this application also provide a computer-readable storage medium storing at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, at least one program, code set, or instruction set is loaded and executed by a processor to implement the arbitrary multi-port S-parameter calibration and de-embedding method provided in the above-described method embodiments.

[0048] Embodiments of this application also provide a computer program product, which includes a computer program. A processor of a computer device reads the computer program from a computer-readable storage medium and executes the computer program, causing the computer device to perform any of the arbitrary multi-port S-parameter calibration and de-embedding methods described in the above embodiments.

[0049] For ease of description, the above devices or apparatuses are described separately according to their functions, divided into various modules or units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0050] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of the embodiments of this application.

[0051] Finally, it should be noted that in this document, relational terms such as first, second, third, and fourth are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0052] The above are merely preferred embodiments of this application. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for arbitrary multi-port S-parameter calibration and de-embedding, characterized in that, include: A network analyzer was used to perform single-port measurements on the calibrator based on the interpolated S-parameter data basis. The measured reflection coefficients of the single port of the calibrator under different reflection states were input into the overdetermined equation system, and the error terms of each single port were obtained by solving the singular value decomposition method. Multi-port measurements are performed on the calibration device using a network analyzer to generate a switching term matrix after isolation term correction. The switching term matrix is ​​then corrected to obtain a switching term correction matrix. Based on the switching term correction matrix, arbitrary multi-port S-parameter measurement data collected by the network analyzer are corrected to obtain arbitrary multi-port S-parameter measurement data with complete switching term and isolation term corrections. Among them, isolation terms are added to the switching term matrix to form a 10-term error model matrix. The multi-port error of the target through path is generated, and the multi-port error of other paths outside the target through path is supplemented by depth-first search and the multi-port error of the target through path. The error terms of each single port are combined to calibrate and de-embed the arbitrary multi-port S-parameter measurement data that have completed the correction of the switching and isolation terms, so as to obtain the multi-port calibrated S-parameter matrix. By combining the error terms of each single port, the S-parameter measurement data of any multi-port that has completed the correction of the switching and isolation terms are calibrated and de-embedded to obtain the multi-port calibrated S-parameter matrix, including: Substitute the error terms of each single port, the multi-port error between all port paths, and the arbitrary multi-port S-parameter measurement data after completing the correction of the switching and isolation terms into the transition matrix; Based on the transition matrix, calculate the S-parameter matrix after multi-port calibration; The transition matrix is ​​determined by the following formula: in, In the formula, This is the transition matrix. This is the diagonalized error matrix corresponding to position pq. This represents the element value at the corresponding position of the nth port. To complete the measurement data of arbitrary multiport S-parameters for switching and isolation terms correction, To complete the correction of switching and isolation terms, the corresponding position element values ​​of arbitrary multiport S-parameter measurement data are obtained. and For each single port, The multiport error between port paths is represented by x and y, where x and y represent port pairs composed of any two ports. The S-parameter matrix after multi-port calibration is determined by the following formula: In the formula, The S-parameter matrix after multi-port calibration. It is the identity matrix. is the diagonalization error matrix at the corresponding position, and -1 is the inverse.

2. The method as described in claim 1, characterized in that, After obtaining the S-parameter matrix after multi-port calibration, the following is also included: Based on the isolation term correction matrix, the isolation term is corrected twice on the multi-port calibrated S-parameter matrix to obtain the final multi-port calibrated S-parameter matrix.

3. The method as described in claim 1, characterized in that, The process of generating the multi-port error of the target direct path, and using depth-first search and the multi-port error of the target direct path to complete the multi-port errors of other paths besides the target direct path, includes: Determine the number of ports n of the calibration component, and designate one of the ports as a common port; Based on the radial approach, the path between the common port and each other port is taken as the target straight-through path. Straight-through measurement is performed on the target straight-through path to determine the multi-port error of each target straight-through path; the number of target straight-through paths is n-1. An automatic path is generated using depth-first search; the automatic path is from other port x to another other port y via the common port c. For each of the automatic paths, the following is performed: based on the multiport error of the corresponding target direct path, calculate the multiport error between port x and port y in the automatic path; Continue until the multi-port errors of all paths except the target direct path are completed, thus obtaining the multi-port errors between all port paths.

4. The method as described in claim 2, characterized in that, The final S-parameter matrix after multiport calibration is determined as follows: Obtain the isolation term correction matrix: The S-parameter matrix after multi-port calibration is then subjected to a second correction using isolation terms to obtain the final S-parameter matrix after multi-port calibration: In the formula, The final S-parameter matrix after multiport calibration. The S-parameter matrix after multi-port calibration. The correction matrix for the isolation term, These are the S-parameters measured when the load is connected to two ports.

5. An arbitrary multi-port S-parameter calibration and de-embedding device, used to implement the steps of the method according to any one of claims 1-4, characterized in that, include: The solution unit is used to perform single-port measurements on the calibrator based on the interpolated S-parameter data basis using a network analyzer, so as to input the measured reflection coefficient of the single port of the calibrator under different reflection states into the overdetermined equation system, and use the singular value decomposition method to solve for the error term of each single port. The correction unit is used to perform multi-port measurements on the calibration component using a network analyzer, generate a switching term matrix after isolation term correction, correct the switching term matrix to obtain a switching term correction matrix, and correct the arbitrary multi-port S-parameter measurement data collected by the network analyzer based on the switching term correction matrix to obtain arbitrary multi-port S-parameter measurement data with completed switching term and isolation term corrections; wherein, an isolation term is added to the switching term matrix to form a 10-term error model matrix; The calibration unit is used to generate the multi-port error of the target through path, and use depth-first search and the multi-port error of the target through path to complete the multi-port error of other paths outside the target through path. In order to combine the error terms of each single port, the arbitrary multi-port S-parameter measurement data that has completed the correction of the switching term and isolation term is calibrated and de-embedded to obtain the multi-port calibrated S-parameter matrix.

6. A computer device, characterized in that, The computer device includes a memory and a processor. The memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory to implement the steps of the method according to any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the method described in any one of claims 1-4.

8. A computer program product, characterized in that, Includes a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1-4.

Images