A simple test method based on MIMO-OTA

Abstract

The application relates to a simple test method based on MIMO-OTA, and belongs to the technical field of wireless communication. The test method shields the air interface influence of OTA, and for multi-base station test, multi-terminal test or pull-out test, only the multi-base station or multi-terminal needs to be placed in an actual darkroom, then a corresponding topological structure is established according to actual test needs by using control software of a simulation instrument, a channel impulse response is generated in channel modeling software, and channel coefficients are loaded into a channel simulation instrument, so that OTA test can be realized, and the influence of the darkroom air interface on the test system does not need to be considered.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of wireless communication, and particularly relates to a simple test method based on MIMO-OTA. BACKGROUND

[0002] According to the 3GPP-38901 protocol, a conventional geometric-based modeling formula is as follows:

[0003]

[0004] wherein represents a non-line-of-sight (NLOS) diameter, represents a line-of-sight (LOS) diameter, K R is a Rician K-factor, representing the proportion of the LOS path in all paths. Each NLOS path is superimposed by 20 sub-paths. and The modeling method of and is as follows:

[0005]

[0006]

[0007] wherein:

[0008] u represents an antenna index of a receiving antenna; s represents an antenna index of a transmitting antenna; n represents a cluster index; m represents a sub-path index; P n represents a normalized cluster power; M represents a sub-path number; theta represents an elevation angle; represents an azimuth angle F rx,u,θ represents a vertical direction pattern of the receiving antenna; represents a horizontal direction pattern of the receiving antenna; kappa represents a cross-polarization ratio; Phi represents a random phase; represents a coordinate of a receiving terminal in a spherical coordinate system; represents a coordinate of a transmitting base station in the spherical coordinate system; represents a coordinate vector of the u-th receiving antenna; represents a coordinate vector of the s-th transmitting antenna; lambda0 represents a wavelength; represents a velocity vector of the terminal.

[0009] The channel modeling method in the 3GPP-38901 protocol is generally used for the path modeling of the 5G wireless communication performance test system. For base station testing, in the Sub6G frequency band, the fading process experienced by the terminal can be simulated by loading the wireless channel impulse response through the conductive connection channel simulator. In the millimeter wave frequency band, due to the large number of base station arrays and the characteristics of the millimeter wave frequency band, the base station may no longer provide a port for conductive testing. Therefore, the multi-probe anechoic chamber is currently the most suitable OTA test scheme. However, the multi-probe anechoic chamber needs to optimize the probe position according to the wireless channel model to match the fading channel test, and needs to be strictly calibrated to meet the fading channel performance test requirements.

[0010] Therefore, at the present stage, a simple test method based on MIMO-OTA needs to be designed to solve the above problems. SUMMARY

[0011] The present application aims to provide a simple test method based on MIMO-OTA for solving the technical problems existing in the prior art, in order to take advantage of the convenience of millimeter wave anechoic chamber OTA testing, and to avoid the complex calibration and verification process and test process of multi-probe anechoic chamber OTA. The present application has great advantages in millimeter wave base station remote testing, multi-base station switching testing and multi-terminal access testing.

[0012] To achieve the above purpose, the technical scheme of the present application is:

[0013] A simple test method based on MIMO-OTA, comprising the following steps:

[0014] Assuming that the terminal side uses a terminal anechoic chamber, the terminal uses an omnidirectional antenna, the number of antennas is N, and the number of base station antennas is M; the channel impulse response of the wireless link between the terminal and the base station is described as:

[0015] H = H u H c H b (4)

[0016] Wherein, H u represents the channel impulse response between the terminal antenna and the terminal anechoic chamber probe, which is an N x 2 channel matrix; H c is a 2 x 2 channel matrix simulated by a channel simulator; H b is a channel impulse response between the base station anechoic chamber probe and the base station antenna, which is a 2 x M channel matrix;

[0017] In formula (4), the form is similar to the singular value decomposition of a matrix; in an actual outdoor wireless channel, H is an N x M channel matrix, and singular value decomposition of H can be obtained:

[0018] HN×M = U N×N S N×M V M×M (5)

[0019] wherein U and V are N x N and M x M unitary matrices respectively, and S is a singular value matrix; in the singular value matrix, singular values are arranged in descending order, and in many cases, the sum of the first 1%-10% singular values accounts for more than 99% of the sum of all singular values; in machine learning, dimensionality reduction operation using singular value decomposition is used for data compression and denoising, and dimensionality reduction operation is performed on the matrix in formula (5), that is, formula (5) is simplified as:

[0020] H N×M ≈ U N×2 S 2×2 V 2×M (6).

[0021] Further, considering that the position of the terminal in the dark room will affect the dimensionality reduction operation of the singular value decomposition, the terminal needs to find a suitable position in the terminal dark room, as follows:

[0022] Configure the channel simulator to generate H c = 2 x 2 channel orthogonal matrix;

[0023] Reasonably adjust the power attenuation of the frequency converter and the channel simulator so that the power of the system transceiver is at a reasonable value;

[0024] Configure the test system to perform service testing, and adjust the terminal position until the system service reaches 2 flow peaks;

[0025] If the service of the system reaches the maximum flow peak, it means that H in formula (6) approximates to an orthogonal matrix; after dimensionality reduction operation, if the rank of the original channel matrix H N×M is less than or equal to 2, the dimensionality reduction operation does not affect the system flow test; if the rank of the original channel matrix H N×M is greater than 2, the dimensionality reduction operation is equivalent to taking only the first two eigenvalues for channel simulation, which will result in a decrease in the total flow of the test system.

[0026] Further, after the dimensionality reduction operation of singular value decomposition is completed;

[0027] H c in formula (4) can be realized by using the traditional channel modeling method, i.e., formula (1), formula (2) and formula (3) to realize OTA system channel modeling and testing;

[0028] Assuming that the half-wavelength sampling density is p, representing the number of sampling points on the half-wavelength length in the modeling process, the interval between two sampling points is Lambda represents wavelength, for traditional cellular modeling theory, it is assumed that terminal moves from point a to point b, and its propagation delay is Transformed into Wherein c represents the speed of light, the control of propagation delay is controlled by the wireless channel simulator hardware by delaying the input signal, and the channel modeling algorithm does not involve the control of delay;

[0029] According to the Doppler formula:

[0030]

[0031] Wherein f center Is the center frequency, v represents the terminal running speed, omega is the incoming angle, which is determined according to the geometric relationship between the transmitting and receiving ends; according to T0 represents the sampling time between two sample points, and v is brought into formula (7) to obtain

[0032]

[0033] Let Represent the baseband sampling rate, then formula (8) can be expressed as:

[0034]

[0035] Wherein Represents the maximum Doppler shift; when the terminal speed v is constant, f d,max Is a constant value; formula (9) shows that the wireless channel simulator can control the Doppler shift as long as t0 is strictly controlled, and the spatial angle theta is calculated in real time according to the geometric relationship; at the same time, the angle information introduced by the outside field can be brought into formula (2) or (3) to calculate the projection of the antenna directional diagram, the cross polarization ratio matrix On the antenna panel, so as to realize the generation and simulation of the wireless channel impulse response.

[0036] Compared with the prior art, the present application has the beneficial effects that:

[0037] The innovation of the present application lies in that the test method shields the air interface influence of OTA, for multi-base station test, multi-terminal test or pull-out test, only need to place the multi-base station or multi-terminal in the actual darkroom, then according to the actual test needs, the control software of the simulator is used to establish the corresponding topology structure, the channel impulse response is generated in the channel modeling software, and the channel coefficient is loaded into the channel simulator to realize the OTA test, without considering the influence of the darkroom air interface on the test system. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1Schematic diagram of simple OTA test principle of MIMO-OTA base station of the embodiment of the application.

[0039] Figure 2 Schematic diagram of simple OTA test device of MIMO-OTA base station of the embodiment of the application.

[0040] Figure 3 Schematic diagram of base station anechoic chamber probe placement position of the embodiment of the application, top view (left) and side view (right).

[0041] Figure 4 Schematic diagram of singular value decomposition dimension reduction operation of the embodiment of the application.

[0042] Figure 5 Schematic diagram of MIMO-OTA millimeter wave base station test equivalent transmission device of the embodiment of the application.

[0043] Figure 6 Schematic diagram of trajectory sampling point of the embodiment of the application.

[0044] Figure 7 Schematic diagram of millimeter wave base station high-speed rail OTA test scene of the embodiment of the application.

[0045] Figure 8 Schematic diagram of millimeter wave base station high-speed rail OTA test scene trajectory of the embodiment of the application.

[0046] Figure 9 Schematic diagram of 6 millimeter wave base stations impacting terminal antenna 1 high-speed rail OTA channel impulse response of the embodiment of the application. DETAILED DESCRIPTION

[0047] The technical solutions in the embodiments of the application will be described below in detail with reference to the drawings of the application. Figures 1-9 The technical solutions in the embodiments of the application will be described below in detail with reference to the drawings of the application.

[0048] Embodiment:

[0049] In order to utilize the convenience of millimeter wave anechoic chamber OTA test, and in order to avoid the complex calibration verification process and test process of multi-probe full-wave anechoic chamber OTA, a simple test device and method based on MIMO-OTA are proposed in this paper, which has great advantages in millimeter wave base station remote test, multi-base station switching test and multi-terminal access test.

[0050] According to the 3GPP-38901 protocol, the traditional geometric-based modeling formula is as follows:

[0051]

[0052] wherein denotes non-line-of-sight (NLOS) path, denotes line-of-sight (LOS) path, K R is the Rician K-factor, representing the proportion of LOS path in all paths. Each NLOS path is composed of 20 sub-paths. and The modeling method is as follows:

[0053]

[0054]

[0055] wherein:

[0056] u denotes the antenna index of the receiving antenna;

[0057] s denotes the antenna index of the transmitting antenna;

[0058] n denotes the cluster index;

[0059] m denotes the sub-path index;

[0060] P n denotes the normalized cluster power;

[0061] M denotes the number of sub-paths;

[0062] θ denotes the elevation angle;

[0063] denotes the azimuth angle

[0064] F rx,u,θ denotes the vertical pattern of the receiving antenna;

[0065] denotes the horizontal pattern of the receiving antenna;

[0066] κ denotes the cross-polarization ratio;

[0067] Φ denotes the random phase;

[0068] denotes the coordinates of the receiving terminal in the spherical coordinate system;

[0069] denotes the coordinates of the transmitting base station in the spherical coordinate system;

[0070] denotes the coordinate vector of the u-th receiving antenna;

[0071] denotes the coordinate vector of the s-th transmitting antenna;

[0072] λ0 represents wavelength;

[0073] denotes the velocity vector of the terminal;

[0074] The schematic diagram of the simple test principle based on MIMO-OTA proposed in this document is shown in Figure 1 The terminal darkroom can also use a terminal simulator instead. Since the commercial channel simulator generally supports a frequency range up to 6 GHz, the purpose of the frequency converter is to shift the signal in the millimeter wave band to the frequency band supported by the wireless channel simulator. The purpose of the terminal darkroom and the base station darkroom is to receive the radio signals radiated by the terminal and the base station in an OTA (Over The Air) manner, and the channel simulator is used to add a wireless fading channel in the test system.

[0075] The schematic diagram of the test device proposed in this document is shown in Figure 2 The terminal darkroom in Figure 2 can use a terminal simulator instead. There is only one dual-polarized probe in the base station darkroom. Figure 3 The "cross" in represents the base station antenna panel, and the "triangle" represents the probe position, which is located on the central normal line of the base station panel. The straight-line distance between the base station panel and the probe should meet the Fraunhofer far-field distance, so that the base station can subsequently perform beamforming tests. The turntable in the base station darkroom is used to simulate the influence of the base station rotation on the system test.

[0076] Assuming that the terminal side uses a terminal darkroom, the terminal uses an omnidirectional antenna, and the number of antennas is N, and the number of base station antennas is M. The channel impulse response of the wireless link between the terminal and the base station can be described as:

[0077] u H c H b (4)

[0078] wherein H u denotes the channel impulse response between the terminal antenna and the terminal darkroom probe, which is an N×2 channel matrix; H c is a 2×2 channel matrix simulated by the channel simulator; and H b is the channel impulse response between the base station darkroom probe and the base station antenna, which is a 2×M channel matrix.

[0079] Examining formula (4), its form is similar to the singular value decomposition of a matrix. In an actual outdoor wireless channel, H is an N×M channel matrix, and singular value decomposition of H can be obtained as:

[0080] H N×M = U N×N S N×M VM×M (5)

[0081] Where U and V are N×N and M×M unitary matrices, respectively, and S is the singular value matrix. In the singular value matrix, the singular values ​​are arranged in descending order, and the singular values ​​decrease very rapidly. In many cases, the sum of the first 10% or even 1% of the singular values ​​accounts for more than 99% of the total sum of the singular values. In machine learning, singular value decomposition is often used for dimensionality reduction to compress data and denoise. This scheme follows this approach. Since the dimension of a single dual-polarization probe is 2, the device in this scheme is equivalent to performing a dimensionality reduction operation on the matrix in formula (5), that is, formula (5) is simplified to:

[0082] H N×M ≈U N×2 S 2×2 V 2×M (6)

[0083] Schematic diagram as follows Figure 4 As shown.

[0084] Since commercial terminals generally use omnidirectional antennas, and the antenna positions vary between manufacturers, the terminal's location in the anechoic chamber will affect the aforementioned dimensionality reduction decomposition. To find a suitable location for the terminal in the anechoic chamber, the following method is used:

[0085] 1) Configure the channel simulator to generate H c = 2×2 channel orthogonality matrix;

[0086] 2) Adjust the power attenuation of the frequency converter and channel simulator appropriately to ensure that the power of the system's transmission and reception is at a reasonable value;

[0087] 3) Configure the test system to conduct business tests, adjust the terminal positions, until the system business reaches the peak of 2 flows.

[0088] If the system's traffic reaches the maximum flow (2 flows) peak, it means that H in formula (6) approximately reaches an orthogonal matrix. After the dimensionality reduction operation shown above, if H of the original channel matrix... N×M If the rank is less than or equal to 2, this operation does not affect the system traffic test; if the original channel matrix H N×M If the rank is greater than 2, the dimensionality reduction operation is equivalent to taking only the two largest eigenvalues ​​for channel simulation. This device will reduce the total traffic that the test system can simulate.

[0089] Although the dimensionality reduction operation of the aforementioned OTA testing device imposes certain limitations on the total system traffic test, it transforms the aforementioned OTA testing device into something like... Figure 5 The equivalent conduction test apparatus shown is as follows:

[0090] It can be seen that the device not only completes the import and export of air interface data between the base station and the terminal via OTA, but also shields the air interface environment of the base station anechoic chamber from the impact on system testing. Therefore, the channel modeling process does not need to consider the influence of the base station anechoic chamber air interface, i.e., H in formula (4) c Traditional channel modeling methods, namely Equations (1), (2) and (3), can be used to realize the system channel modeling and testing of OTA.

[0091] Assuming the half-wavelength sampling density is ρ, representing the number of sampling points along half a wavelength during modeling, and the interval between two sampling points is... λ represents wavelength, such as Figure 6 As shown. In traditional cellular modeling theory, assuming a terminal moves from point a to point b, its propagation delay is... Transform into Where c represents the speed of light, the propagation delay is controlled by the wireless channel simulator hardware through delay control of the input signal, and the channel modeling algorithm does not involve delay control.

[0092] According to the Doppler formula:

[0093]

[0094] Among them, f center The center frequency is Ω, v represents the terminal operating speed, and Ω is the angle of arrival, which is determined based on the geometric relationship between the transmitting and receiving ends. t0 represents the sampling time between two sample points. Substituting v into formula (7) yields...

[0095]

[0096] make Let the baseband sampling rate be represented, then formula (8) can be expressed as:

[0097]

[0098] in This represents the maximum Doppler frequency shift. When the terminal velocity v is constant, f d,max It is a definite value. Formula (9) shows that as long as the wireless channel simulator strictly controls t0 and calculates the spatial angle θ in real time according to the geometric relationship (which can be represented by the azimuth and elevation angles), it can achieve the control of the Doppler frequency shift. At the same time, the angle information imported from the external field can be substituted into formula (2) or (3) to calculate the antenna pattern (F). tx and F rx Cross-polarization ratio matrix (NLOS scenario), projection of the diameter onto the antenna panel ( and ) the impact on the channel impulse response, thereby achieving the generation and simulation of the impulse response of the wireless channel.

[0099] Since the test device shields the air interface impact of OTA, for multi-base station test, multi-terminal test or pull-out test, only need to place the multi-base station or multi-terminal in the actual darkroom, then according to the actual test needs, use the control software of the simulator to establish the corresponding topology, generate the channel impulse response in the channel modeling software, and load the channel coefficient into the channel simulator to realize the OTA test, without considering the impact of the darkroom air interface on the test system. Taking the millimeter wave high-speed rail OTA test as an example, its test scheme is shown in Figure 7

[0100] The terminal trajectory configuration is shown in Figure 8 , wherein the station spacing is 1000m, the station track spacing is 150m, and the terminal makes a round trip trajectory, wherein the outbound movement is 3000m, the return movement is 3000m, and the round trip movement is 6000m.

[0101] According to the generated impulse response, the time domain impulse response of the first LOS path is drawn as shown in Figure 9 .

[0102] It can be seen that the generated channel impulse response can completely describe the path loss power change between the transmitter and the receiver with the distance. The channel simulator can perform joint simulation of large-scale fading and small-scale fading according to the above specified attenuation, thereby realizing high-speed rail channel simulation. Due to the round trip movement of the terminal, the overall impulse response of the LOS path is axisymmetric. With the movement of the terminal, the switching sequence of the base station is BS0-BS1-BS2-BS3-BS4-BS5-BS4-BS3-BS2-BS1-BS0.

[0103] The above is the preferred embodiment of the present application, any changes made according to the technical solutions of the present application, as long as the generated function does not exceed the scope of the technical solutions of the present application, belongs to the protection scope of the present application.​

Claims

1. A simple test method based on MIMO-OTA, characterized in that, Comprising the following steps: Assuming that the terminal side uses a terminal darkroom, the terminal uses an omnidirectional antenna, the number of antennas is N, and the number of base station antennas is M; the channel impulse response of the wireless link between the terminal and the base station is described as: H = H u H c H b (4) where H u is the channel impulse response from the terminal antenna to the terminal darkroom probe, which is an N x 2 channel matrix; H c is the channel impulse response from the terminal antenna to the terminal darkroom probe, which is an N x 2 channel matrix; H b is the channel impulse response from the terminal antenna to the terminal darkroom probe, which is an N x 2 channel matrix; H In formula (4), the form is similar to the singular value decomposition of a matrix; in the actual field of wireless channels, H is an N x M channel matrix, and singular value decomposition of H can be obtained: H N×M = U N×N S N×M V M×M (5) Where U and V are N x N and M x M unitary matrices, and S is a singular value matrix; in the singular value matrix, the singular values are arranged in descending order, and in many cases, the sum of the first 1%-10% of the singular values accounts for more than 99% of the total singular value; in machine learning, the dimensionality reduction operation using singular value decomposition is used for data compression and denoising, and the matrix in formula (5) is subjected to dimensionality reduction operation, that is, formula (5) is simplified to: H N×M ≈U N×2 S 2×2 V 2×M (6); Considering that the position of the terminal in the darkroom will affect the dimensionality reduction operation of the singular value decomposition, the appropriate position of the terminal in the terminal darkroom needs to be found, as follows: The configuration channel simulator generates H c = 2 x 2 channel orthogonal matrix; Reasonably adjust the power attenuation of the frequency converter and the channel simulator to make the power of the system transceiver reasonable; Configure the test system to test the service, adjust the terminal position, and until the system service reaches 2 flow peak value; If the system traffic reaches the maximum flow peak, it means that H in formula (6) is approximately a orthogonal matrix; after dimension reduction operation, if the rank of H of the original channel matrix is less than or equal to 2, the dimension reduction operation does not affect the system flow test; if the rank of H of the original channel matrix is greater than 2, the dimension reduction operation is equivalent to taking only the first two eigenvalues for channel simulation, which will cause the total flow of the total simulatable test of the test system to decrease. N×M If the system traffic reaches the maximum flow peak, it means that H in formula (6) is approximately a orthogonal matrix; after dimension reduction operation, if the rank of H of the original channel matrix is less than or equal to 2, the dimension reduction operation does not affect the system flow test; if the rank of H of the original channel matrix is greater than 2, the dimension reduction operation is equivalent to taking only the first two eigenvalues for channel simulation, which will cause the total flow of the total simulatable test of the test system to decrease. N×M If the system traffic reaches the 2. The simple test method based on MIMO-OTA according to claim 1, characterized in that, After the dimensionality reduction operation of singular value decomposition is completed; H in equation (4) c The system channel modeling and testing of the OTA can be implemented by using the conventional channel modeling method, i.e., shown in equation (1), equation (2) and equation (3). Where the geometric-based modeling formula is as follows: wherein denotes non-line-of-sight (NLOS) path, denotes line-of-sight (LOS) path, K R is the Rician K-factor, representing the proportion of the LOS path in all paths; each NLOS path is composed of 20 sub-paths; and The modeling method of is as follows: Where: u denotes the antenna index of the receive antenna; s denotes the antenna index of the transmit antenna; n denotes the cluster index; m denotes the sub-radiation index; P n denotes the normalized cluster power; M denotes the number of sub-radiations; θ denotes the elevation angle; denotes the azimuth angle F rx,u,θ denotes the vertical pattern of the receive antenna; denotes the horizontal pattern of the receive antenna; κ denotes the cross-polarization ratio; Φ denotes the random phase; denotes the coordinates of the receive terminal in the spherical coordinate system; denotes the coordinates of the transmit base station in the spherical coordinate system; denotes the coordinate vector of the u-th receive antenna; denotes the coordinate vector of the s-th transmit antenna; λ0denotes the wavelength; denotes the velocity vector of the terminal; Assuming the half-wavelength sampling density is p, representing the number of sampling points in the half wavelength length in the modeling process, the interval between two sampling points is λ represents the wavelength, and for the traditional cellular modeling theory, it is assumed that the terminal moves from point a to point b, and the propagation delay is converted into where c represents the speed of light, and the control of the propagation delay is controlled by the wireless channel simulator hardware by delaying the input signal, and the channel modeling algorithm does not involve the control of the delay; According to the Doppler formula: where f center is the center frequency, v represents the terminal running speed, and Ω is the incoming wave angle, which is determined according to the geometric relationship between the transmitting and receiving ends; according to t0 represents the sampling time between two samples, and v is brought into formula (7) to obtain Let denote the baseband sampling rate, then equation (8) can be expressed as: wherein represents the maximum Doppler shift; when the terminal speed v is constant, f d,max is a fixed value; formula (9) shows that the wireless channel simulator can control the Doppler shift as long as t0 is strictly controlled and the spatial angle θ is calculated in real time according to the geometric relationship; at the same time, the angle information introduced from the field can be brought into formula (2) or (3) to calculate the antenna pattern and the cross-polarization ratio matrix the projection of the radial on the antenna panel affects the channel impulse response, thereby realizing the generation and simulation of the wireless channel impulse response.

Images