Mobile terminal test apparatus and mobile terminal test method

Abstract

The present invention provides a mobile terminal testing device and a mobile terminal testing method that can reduce the number of measurements and set the output level of a test signal to a testable level in a shorter time, and can efficiently perform a reception sensitivity test of a mobile terminal. A reception sensitivity test control section (18) provided in a measurement device (1) and repeatedly performing transmission and reception of a test signal and performing a reception sensitivity test of a mobile terminal is configured to have a test condition setting section (18a) that initially sets test conditions including an initial step level (SL0) of a test signal at the start of a reception sensitivity test and a start output level (OL0), and an AI prediction test condition change setting section (18b) that sets an AI predicted output level (OL01) that is AI predicted in advance from an AI prediction model based on the test conditions set by the test condition setting section (18a) as a primary output level instead of the output level (OL0).

Description

Technical Field

[0001] This invention relates to a mobile terminal testing apparatus and method that performs a receiving sensitivity test by repeatedly causing the test subject to receive test signals and measuring the receiving sensitivity, and tests the test subject, i.e., a mobile terminal. Background Technology

[0002] For wireless terminals that are being developed in recent years to transmit and receive broadband signals in the millimeter wave band, such as IEEE 802.11ad and 5G cellular, the performance test is conducted to determine whether the wireless communication antennas of the wireless terminals meet the predetermined benchmarks by measuring the output level of the transmitted radio waves and the receiving sensitivity set for each communication standard.

[0003] For example, in performance testing of 5G NR (New Radio System) devices using wireless terminals (hereinafter referred to as 5G wireless terminals) as the device under test (DUT), OTA tests are conducted using an anechoic chamber (OTA chamber) called a compact antenna test range (CATR) that is unaffected by the surrounding radio wave environment.

[0004] As an example of OTA testing, a DUT receiver sensitivity test can be described as repeatedly performing actions to make the DUT (e.g., a 5G wireless terminal) receive test signals sent from a signal generator to measure receiver sensitivity, and then statistically analyzing the measurement results. In the DUT receiver sensitivity test, as described in 7.3.2 of Chapter 38.521-2 of Non-Patent Document 1, there are specifications such as a minimum permissible output level [dB] for the test signal.

[0005] As a method for implementing a receiver sensitivity test, which involves repeatedly reducing the output level of the test signal to the minimum permissible output level while measuring the receiver sensitivity of the DUT, there are known methods such as linearly changing the output level of the test signal from the value set in the initial receiver sensitivity test to a certain level each time the number of measurements increases, and non-linearly changing the output level.

[0006] Furthermore, it is known in the past that there are devices that set conditions (CP conditions) as check points, and when the throughput measured in the test meets the CP conditions, the time to obtain the measurement result is shortened by increasing the step level of the test signal (devices employing CP determination processing) (for example, see Patent Document 1).

[0007] Non-patent literature 1: 3GPP (registered trademark) Technical Specification 38.521-2, Chapter 7.3.2

[0008] Patent Document 1: Japanese Patent Application Publication No. 2022-54744

[0009] exist Figure 15 The diagram illustrates an example of the output level variation pattern (EIS search path) up to the end of the receiver sensitivity test, in both methods of linearly varying the output level of the test signal and methods of nonlinearly varying it, within the aforementioned conventional mobile terminal test apparatus. Here, EIS is an abbreviation for Equivalent Isotropic Sensitivity. Figure 15 As shown, in both the method of linearly changing the output level of the test signal (reference curves L1, L2) and the method of nonlinearly changing it (reference curves NL1, NL2), the point at which the EIS search begins from a fixed value, i.e., the initial measurement of the OL (output) level, is the same. Furthermore, in both methods, there is a large difference between the initial measurement of the OL level and the output level at the end of the receiver sensitivity test (EIS point 1, 2).

[0010] Therefore, in conventional mobile terminal measurement devices that apply any of the above methods, since a number of EIS searches need to be performed after the test signal is set to a fixed value, i.e., the initial output level, to start the EIS search, it is difficult to significantly shorten the EIS search time.

[0011] On the other hand, in conventional devices employing the CP determination processing described in Patent Document 1, the variation range of the output level is increased with CP as the boundary. Therefore, compared with methods that change the output level linearly or non-linearly, the EIS search time can be shortened. In Patent Document 1, the relationship between the output level and the number of measurements up to the end of the receiver sensitivity test is shown in both the method of non-linearly changing the output level of the test signal and the method of changing it via CP determination processing. Figure 11 In the curve graph. Figure 11 In the diagram, the curve corresponding to nonlinear changes is represented by the symbol C3, while the curve corresponding to changes caused by CP determination is represented by the symbol C1. For example... Figure 11 As shown, by employing a method that changes the output level of the test signal via CP decision processing, the number of measurements can be reduced (up to 4 in this example) compared to a nonlinear variation method, thereby shortening the corresponding EIS search time. However, in conventional devices employing CP decision processing, it is difficult to reduce the number of measurements (EIS search times) to even fewer, and there is a limit to the reduction in EIS search time. Summary of the Invention

[0012] The present invention was made to solve this previous problem, and its purpose is to provide a mobile terminal testing apparatus and a mobile terminal testing method that can set the output level of the test signal to a testable level in a shorter time by minimizing the number of measurements and can effectively perform mobile terminal receiving sensitivity tests.

[0013] To address the aforementioned issues, the mobile terminal testing apparatus according to the first aspect of the present invention includes: a signal generator 20 that generates a test signal; and a receiver sensitivity test execution unit 18 that performs a receiver sensitivity test on the mobile terminal by repeatedly transmitting and receiving the test signal between the signal generator and the mobile terminal 100. The mobile terminal testing apparatus 1, which tests the test object, i.e., the mobile terminal, is characterized by: a setting unit 18a that initially sets the test conditions, including the step level SL0 and output level OL0 of the test signal at the start of the receiver sensitivity test; and a pass / fail judgment unit 18f that performs a pass / fail judgment by transmitting and receiving the test signal under the test conditions. Throughput measurement is performed, and the throughput measurement result is compared with a judgment threshold to determine whether it is qualified or not; output level setting unit 18e sets the output level according to the previous throughput measurement qualification or failure judgment result when the qualification or failure judgment result is determined to be no; and AI prediction test condition change setting unit 18b holds an AI prediction model generated by machine learning based on the correlation between parameter data and measurement result data of the measurement log file of the past receiver sensitivity test, and sets the AI ​​prediction output level OL01, which is predicted in advance by the AI ​​prediction model according to the test conditions set by the setting unit, as the initial output level instead of the output level OL0.

[0014] According to this structure, the mobile terminal testing apparatus of the first aspect of the present invention can quickly set the initial output level of the test signal to a value close to the end value of the measurement under the original nonlinear control by AI prediction. Compared with the conventional methods such as nonlinear control and CP control, it can significantly reduce the number of measurements and set the output level of the test signal to a testable level in a very short time, and can effectively conduct mobile terminal receiving sensitivity tests.

[0015] Furthermore, in the mobile terminal testing apparatus according to the second aspect of the present invention, the structure may be as follows: the AI ​​prediction test condition change setting unit sets the step level for subsequent measurements to the AI ​​prediction step level SL01 predicted in advance by the AI ​​prediction model, based on the measurement result of the initial throughput measurement performed by transmitting and receiving the test signal with the AI ​​prediction output level, instead of the step level SL0.

[0016] According to this structure, the mobile terminal testing device of the second aspect of the present invention not only dynamically controls the initial output level of the test signal through AI prediction, but also dynamically controls the step level. Therefore, compared with the situation where measurements are continued in the conventional way such as nonlinear control and CP control, the test signal can be set to a testable level in a shorter time, thereby further improving the efficiency of mobile terminal receiving sensitivity testing.

[0017] Furthermore, in the mobile terminal testing apparatus according to the third aspect of the present invention, the structure may be as follows: when the result of the pass / fail judgment on the measurement result of the initial throughput measurement is judged as no, then, limited to the second throughput measurement, the AI ​​prediction test condition change setting unit uses the AI ​​prediction step level SL01 to perform the output level update processing for the next measurement.

[0018] According to this structure, the mobile terminal testing apparatus of the third aspect of the present invention dynamically controls the initial output level and step level of the test signal predicted by AI until the second throughput is measured, thereby avoiding the situation where the above-mentioned dynamic control is still repeating in vain when the measurement is deliberately started from the value at the end of the measurement under near nonlinear control, etc.

[0019] Furthermore, in the mobile terminal testing apparatus according to the fourth aspect of the present invention, the structure may be as follows: the AI ​​prediction test condition change setting unit specifies the loss of the external connection cable, the azimuth angle (θ, φ) of the mobile terminal, and the frequency of the measurement object as the parameter data, and holds the AI ​​prediction model generated by machine learning through the relationship between the parameter data and the measurement result data.

[0020] According to this structure, the mobile terminal testing device involved in the fourth aspect of the present invention can still use the AI ​​prediction model to derive accurate AI prediction results that take into account the various parameter data, even in environments where the loss of external connection cables, the orientation angle (θ, φ) of the mobile terminal, and the frequency of the measured object affect the measurement accuracy. This ensures that the measurement accuracy is not reduced and the test signal can be reliably set to a testable level in a very short time.

[0021] Furthermore, the mobile terminal testing device involved in the fifth aspect of the present invention can be structured as follows: when the result of the pass / fail judgment on the measurement result of the second throughput measurement is judged as no, the receiving sensitivity test execution unit controls the output level setting unit to perform level reduction processing A, level reduction processing B, level boosting processing A, or level boosting processing B.

[0022] According to this structure, when the output level of the test signal fails to reach a testable level due to the dynamic control of the initial output level and step level of the test signal predicted by AI, the mobile terminal testing device of the fifth aspect of the present invention can immediately switch to the normal measurement program to reliably set the output level of the test signal to a testable level.

[0023] Furthermore, the mobile terminal testing apparatus according to the sixth aspect of the present invention may have the following structure, namely: an anechoic chamber 50 having an internal space 51; and scanning units 16 and 56, which drive the scanning of the mobile terminal in a manner that allows the orientation of the mobile terminal to change continuously within the internal space, and the receiving sensitivity test is performed in all directions within the internal space on the object being scanned by the scanning unit under an OTA (Over The Air) measurement environment.

[0024] According to this structure, the mobile terminal testing device involved in the sixth aspect of the present invention is applicable to the dynamic control of the initial output level and step level of the test signal predicted by AI, even when it is necessary to measure the receiving sensitivity in all directions under the OTA environment, and can set the output level of the test signal to a testable level in a short time.

[0025] Furthermore, in order to solve the above-mentioned problems, the mobile terminal testing method according to the seventh aspect of the present invention uses a mobile terminal testing device 1 for testing the test object, i.e., the mobile terminal. The device performs a receiver sensitivity test on the mobile terminal by repeatedly transmitting and receiving the test signal between a signal generator and the mobile terminal. The mobile terminal testing device 1 includes: a signal generator 20 that generates a test signal; and a receiver sensitivity test execution unit 18 that performs a receiver sensitivity test on the mobile terminal by repeatedly transmitting and receiving the test signal between the signal generator and the mobile terminal 100. The mobile terminal testing method is characterized by including a setting step S1, which initially sets the test conditions, including the step level SL of the test signal at the start of the receiver sensitivity test. 0. Output level OL0; AI prediction test condition change setting step S1a, holding an AI prediction model generated by machine learning based on the correlation between parameter data and measurement result data of the measurement log file of the past receiving sensitivity test, sets the AI ​​prediction output level OL01, which is predicted by the AI ​​prediction model in advance according to the test conditions set in the setting step, as the initial output level instead of the output level OL0; pass / fail judgment steps S4 and S5, performing the initial throughput measurement by transmitting and receiving the test signal with the AI ​​prediction output level, and comparing the initial throughput measurement result with the judgment threshold to determine whether it is pass / fail; and step level change setting step, setting the AI ​​prediction step level SL01, which is predicted by the AI ​​prediction model in advance according to the measurement result of the initial throughput measurement, as the step level for the second and subsequent measurements instead of the step level SL0.

[0026] According to this structure, the mobile terminal testing method of the seventh aspect of the present invention, by using the mobile terminal testing device described in any one of the first to sixth aspects and applying the present mobile terminal testing method, can quickly set the initial output level of the test signal to a value close to the end value of the measurement under the original nonlinear control by AI prediction. Compared with conventional methods such as nonlinear control and CP control, it can significantly reduce the number of measurements and set the output level of the test signal to a testable level in a very short time, thereby enabling effective testing of the receiving sensitivity of the mobile terminal.

[0027] Invention Effects

[0028] The present invention provides a mobile terminal testing apparatus and method that can set the output level of the test signal to a testable level in a shorter time by minimizing the number of measurements and effectively conducting mobile terminal receiver sensitivity tests. Attached Figure Description

[0029] Figure 1 This is a diagram showing the general structure of the measuring device according to one embodiment of the present invention.

[0030] Figure 2 This is a block diagram illustrating the functional structure of a measuring device according to an embodiment of the present invention.

[0031] Figure 3 This is a block diagram illustrating the functional structure of a comprehensive control device for a measuring apparatus according to an embodiment of the present invention, and the controlled components thereof.

[0032] Figure 4 This is a block diagram illustrating the functional structure of the NR system simulator in a measuring device according to an embodiment of the present invention.

[0033] Figure 5 This is a diagram showing a global surface scan pattern of the test object within the OTA chamber of a measuring device according to an embodiment of the present invention. Figure 5 (a) indicates the configuration of the test subject relative to the center of the spherical coordinate system. Figure 5 (b) indicates the distribution of the angle specimen point PS in the spherical coordinate system.

[0034] Figure 6 It is for use Figure 5 The diagram shown illustrates the configuration of the test antenna 5 within the OTA chamber of the measuring device according to one embodiment of the present invention, using a spherical coordinate system (r, θ, φ).

[0035] Figure 7 This is a conceptual diagram illustrating the generation and processing steps of the AI ​​prediction model used in a measuring device according to an embodiment of the present invention.

[0036] Figure 8 This is a schematic diagram illustrating the difference in the measurement start position involved in the receiver sensitivity test of a DUT using AI prediction in a measurement device according to an embodiment of the present invention and the receiver sensitivity test of a DUT in a conventional device.

[0037] Figure 9 This is a flowchart illustrating the variable setting control operation of the output level of the test signal for which AI prediction is applied in the receiver sensitivity test of the DUT in the measuring device according to an embodiment of the present invention.

[0038] Figure 10 It means Figure 9 A flowchart detailing the processing steps in step S8. Figure 10 (a) indicates the level reduction process (A). Figure 10 (b) indicates the level reduction process (B).

[0039] Figure 11 This is a graph showing the relationship between the number of measurements and the good / bad judgment results of the output level and throughput measurement results of the DUT in the measurement device according to an embodiment of the present invention and the DUT in the conventional device.

[0040] Figure 12 It means Figure 11 The graph shows the relationship between the number of measurements and the judgment results of the output level and throughput measurement results of the test signal in the measurement device with AI prediction applied to the DUT in an embodiment of the present invention and the DUT in the conventional device.

[0041] Figure 13 This is a graph illustrating the difference in EIS search paths involved in the receiver sensitivity test of a DUT using AI prediction in a measurement device according to an embodiment of the present invention and the receiver sensitivity test of a DUT in a conventional device.

[0042] Figure 14 This is a graph showing the relationship between the measurement time of EIS-CDF measurement using AI prediction in a measurement device according to an embodiment of the present invention and the number of measurements used for machine learning.

[0043] Figure 15 It is a graph representing the arrangement of various EIS search paths for linear and nonlinear control of the test signal involved in the receiver sensitivity test of the DUT in the original device. Detailed Implementation

[0044] The embodiments of the mobile terminal testing apparatus and mobile terminal testing method of the present invention will be described below with reference to the accompanying drawings.

[0045] First, refer to Figures 1-4 The structure of the measuring device 1 according to one embodiment of the present invention will be described. The measuring device 1 constitutes the mobile terminal testing apparatus of the present invention. The measuring device 1 according to this embodiment has as a whole the following characteristics: Figure 1 The appearance structure shown is composed of, as Figure 2 The functional blocks shown are configured as follows. Among them, in Figure 1 , Figure 2 The diagram shows the configuration of each component of the OTA chamber 50 as viewed from its side.

[0046] Measuring device 1, for example, in having Figure 1The aforementioned constituent components are mounted on each frame 90a of the frame structure 90 shown in the diagram. Figure 1 The diagram shows an example in which an integrated control unit 10, an NR system simulator 20, and an OTA chamber 50 are respectively mounted on each rack 90a of the rack structure 90.

[0047] like Figure 2 As shown, the measuring device 1 involved in this embodiment includes an integrated control device 10, an NR system simulator 20, a signal processing unit 25, and an OTA chamber 50.

[0048] Regarding these structures, for ease of understanding, we will first explain the OTA chamber 50. For example... Figure 1 , Figure 2 As shown, the OTA chamber 50 is, for example, composed of a metal frame body 52 with an internal space 51 in the shape of a cuboid, and the internal space 51 houses the DUT 100 with an antenna 110, the test antenna 5, the reflector 7 and the DUT scanning mechanism 56.

[0049] An electromagnetic wave absorber 55 is attached to the entire inner surface area of ​​the OTA chamber 50, namely the bottom surface 52a, side surface 52b, and top surface 52c of the main body 52. ​​This enhances the OTA chamber 50's ability to restrict the intrusion of external electromagnetic waves and the radiation of electromagnetic waves to the outside of the components (DUT 100, test antenna 5, reflector 7, DUT scanning mechanism 56) located within the internal space 51. Thus, the OTA chamber 50 achieves an anechoic chamber with an internal space 51 unaffected by the surrounding electromagnetic environment. The anechoic chamber used in this embodiment is, for example, of the Anechoic type.

[0050] Within the unit housed in the internal space 51 of the OTA chamber 50, the DUT100 is, for example, a wireless terminal such as a smartphone. Examples of communication standards for the DUT100 include cellular (LTE, LTE-A, W-CDMA, GSM, CDMA2000, 1xEV-DO, TD-SCDMA, etc.), wireless LAN (IEEE 802.11b / g / a / n / ac / ad, etc.), Bluetooth, GNSS (GPS, Galileo, GLONASS, BeiDou, etc.), FM, and digital broadcasting (DVB-H, ISDB-T, etc.). Furthermore, the DUT100 can also be a wireless terminal capable of transmitting and receiving wireless signals in millimeter-wave bands corresponding to IEEE 802.11ad or 5G cellular networks.

[0051] In this embodiment, the antenna 110 of the DUT100 uses, for example, a wireless signal in a specified frequency band conforming to the 5G NR communication standard. The DUT100 constitutes the test object, or mobile terminal, in this invention.

[0052] Within the internal space 51 of the OTA chamber 50, the DUT 100 is held by a portion of the DUT scanning mechanism 56. The DUT scanning mechanism 56 extends vertically from the bottom surface 52a of the main body 52 within the internal space 51 of the OTA chamber 50. The DUT scanning mechanism 56 holds the DUT 100 for performance testing and performs a global surface scan on the DUT 100 (see reference ). Figure 5 , Figure 6 ).

[0053] like Figure 1 As shown, the DUT scanning mechanism 56 includes a turntable 56a, a support member 56b, a DUT mounting section 56c, and a drive section 56e. The turntable 56a is composed of a plate member with a disk shape and has a structure that rotates around an azimuth axis (vertical rotation axis) (see reference). Figure 3 The support member 56b is composed of columnar members arranged on the plate surface of the turntable 56a in a vertical direction.

[0054] The DUT mounting section 56c is arranged parallel to the turntable 56a near the upper end of the support member 56b, and has a mounting plate 56d for mounting the DUT 100. The DUT mounting section 56c has a structure capable of rotating about a roller axis (a horizontal rotation axis) (see reference). Figure 3 ).

[0055] like Figure 3 As shown, the drive unit 56e includes, for example, a drive motor 56f for rotating the orientation axis and a drive motor 56g for rotating the roller shaft. The drive unit 56e is configured as a dual-axis positioner having a mechanism that rotates the orientation axis and the roller shaft in their respective rotational directions via the drive motors 56f and 56g. Thus, the drive unit 56e can rotate the DUT 100 mounted on each mounting plate 56d along two axes (the orientation axis and the roller shaft) for each mounting plate 56d. Hereinafter, the DUT scanning mechanism 56, including the drive unit 56e, will sometimes be referred to as a dual-axis positioner (see reference). Figure 3 The drive unit 56e, drive motors 56f and 56g respectively constitute the drive unit, the first rotary drive unit, and the second rotary drive unit in this invention. The mounting plate 56d constitutes the test object mounting unit in this invention.

[0056] Assuming that the DUT100, which is mounted (held) on the mounting plate 56d, is arranged, for example, on a sphere (see reference) in the DUT scanning mechanism 56. Figure 5At the center O1 of sphere B), a global surface scan is performed, sequentially changing the orientation of DUT 100 with antenna 110 facing all directions of the sphere's surface. The DUT scan in the DUT scanning mechanism 56 is controlled by the DUT scan control unit 16, described later. The DUT scanning mechanism 56 and the DUT scan control unit 16 constitute the scanning unit of this invention.

[0057] The test antenna 5 is mounted at a desired position on the bottom surface 52a of the main body 52 of the OTA chamber 50 using a suitable retainer (not shown). The mounting position of the test antenna 5 ensures visibility from the reflector 7 via the opening 67a provided in the bottom surface 52a. The test antenna 5 uses the same specified (NR standard) frequency band as the antenna 110 of the DUT 100.

[0058] When performing NR-related measurements of the DUT100 within the OTA chamber 50, the test antenna 5 transmits a test signal to the DUT100 from the NR system simulator 20 and receives the measured signal transmitted from the DUT100, which has received the test signal. The test antenna 5 is configured such that its light-receiving surface is at the focal point F of the reflector 7. However, the reflector 7 is not essential when the test antenna 5 can be configured such that its light-receiving surface faces the DUT100 and can appropriately receive light.

[0059] The reflector 7 is mounted at the desired position on the side 52b of the OTA chamber 50 using the reflector holder 58. The reflector 7 realizes the radio wave path that reflects the wireless signals (test signals and measured signals) transmitted and received by the antenna 110 of the DUT 100 back to the light-receiving surface of the test antenna 5.

[0060] Next, the structure of the integrated control device 10 and the NR system simulator 20 will be explained.

[0061] like Figure 2 As shown, the integrated control unit 10 is connected to the NR system simulator 20, for example, via a network 19 such as Ethernet (registered trademark), enabling communication with each other. Furthermore, the integrated control unit 10 is also connected via the network 19 to controlled system components such as the DUT scanning control unit 16 in the OTA chamber 50.

[0062] The integrated control unit 10 centrally controls the NR system simulator 20 and the DUT scanning control unit 16 via a network 19, and is, for example, composed of a personal computer (PC). Additionally, the DUT scanning control unit 16 is attached to the OTA chamber 50 (see reference). Figure 2 Besides independent settings, such as Figure 3 As shown, it can also be installed in the integrated control device 10. Hereinafter, the integrated control device 10 will be described as having... Figure 3The structure shown is for illustrative purposes only.

[0063] like Figure 3 As shown, the integrated control device 10 includes a control unit 11, an operation unit 12, and a display unit 13. The control unit 11 is, for example, a computer device. This computer device includes a CPU (Central Processing Unit) 11a that performs predetermined information processing for realizing the functions of the measuring device 1 and centrally controls the NR system simulator 20 and the signal processing unit 25; a ROM (Read Only Memory) 11b that stores the OS (Operating System) used to start the CPU 11a, other programs, and control parameters; a RAM (Random Access Memory) 11c that stores the OS, application program execution code, and data used in the operation of the CPU 11a; an external I / F unit 11d; and input / output ports (not shown).

[0064] The external I / F unit 11d is connected to the NR system simulator 20, the signal processing unit 25, and the drive unit 56e of the DUT scanning mechanism (dual-axis positioner) 56 via the network 19, respectively, enabling communication between them. An operation unit 12 and a display unit 13 are connected to the input / output ports. The operation unit 12 is a functional unit for inputting various information such as commands, and the display unit 13 is a functional unit for displaying the input screen containing the aforementioned information and various information such as measurement results.

[0065] The aforementioned computer device functions as a control unit 11 by using the CPU 11a to execute programs stored in the ROM 11b, with the RAM 11c as the working area. For example... Figure 3 As shown, the control unit 11 includes a call connection control unit 14, a signal transceiver control unit 15, a DUT scanning control unit 16, a signal analysis control unit 17, and a receiver sensitivity test control unit 18. The call connection control unit 14, the signal transceiver control unit 15, the DUT scanning control unit 16, the signal analysis control unit 17, and the receiver sensitivity test control unit 18 are also implemented by the CPU 11a executing a predetermined program stored in the ROM 11b in the working area of ​​the RAM 11c.

[0066] The call connection control unit 14 drives the test antenna 5 via the NR system simulator 20 and the signal processing unit 25 to transmit and receive control signals (wireless signals) between the NR system simulator 20 and the DUT 100, thereby controlling the establishment of a call (the state of being able to transmit and receive wireless signals) between the NR system simulator 20 and the DUT 100.

[0067] The signal transceiver control unit 15 performs the following control: it monitors the user operation in the operation unit 12, and takes the start of operation of the predetermined measurement involved in the measurement of the transmission and reception characteristics of the DUT 100 by the user as the starting point, sends a signal transmission command to the NR system simulator 20 after the call is established by the call connection control, and controls the transmission of test signals from the NR system simulator 20 via the test antenna 5 and the transmission and reception of signal commands, and controls the reception of the measured signal via the test antenna 5.

[0068] The DUT scan control unit 16 performs a global surface scan of the DUT 100 mounted on the mounting plate 56d of the DUT mounting unit 56c by driving the drive motors 56f and 56g of the DUT scan mechanism 56. To achieve this control, for example, a DUT scan control table 16a is pre-installed in the ROM 11b. The DUT scan control table 16a stores, for example, the spherical coordinate system involved in the global surface scan of the DUT 100 (see reference). Figure 5 (a) Specimen points PS at various angles (reference) Figure 5 (b) The coordinates of the drive motors 56f and 56g are associated with the coordinates of the specimen points PS at each angle, and the control data associated with the stop time (measurement time) at each specimen point PS at each angle are also associated. When the drive motors 56f and 56g are, for example, step motors, the number of drive pulses is stored as the aforementioned drive data.

[0069] The DUT scan control unit 16 expands the DUT scan control table 16a into the working area of ​​RAM 11c, and drives the drive motors 56f and 56g of the DUT scan mechanism 56 according to the control data stored in the DUT scan control table 16a. This performs a global surface scan of the DUT 100 mounted on the DUT mounting unit 56c. During the global surface scan, for each angle specimen point PS, the antenna surface of the DUT 100's antenna 110 is sequentially moved towards each angle specimen point PS in spherical coordinates, stopping for a predetermined time (the aforementioned stopping time) before moving to the next angle specimen point PS (the scan of the DUT 100).

[0070] When performing a global surface scan on the DUT100, the signal analysis and control unit 17 reads the radio signals related to NR and LTE received by the test antenna 5 via the NR system simulator 20 and the signal processing unit 25, and analyzes and processes them as signals for specified measurement items (measurement processing).

[0071] The receiver sensitivity test control unit 18 performs control by executing multiple receiver sensitivity tests to measure the receiver sensitivity of the DUT100 by causing it to receive test signals sent from the signal generation unit 21a of the NR system simulator 20, and statistically analyzing the measurement results of the multiple receiver sensitivity tests as the test result. The receiver sensitivity test control unit 18 constitutes the receiver sensitivity test execution unit and pass / fail determination unit of the present invention.

[0072] like Figure 3 As shown, the receiver sensitivity test control unit 18 includes a test condition setting unit 18a, an AI prediction test condition change setting unit 18b, a throughput measurement unit 18c, a degradation state determination unit 18d, an output level variable setting unit 18e, and a measurement result output unit 18f. The test condition setting unit 18a, the AI ​​prediction test condition change setting unit 18b, and the output level variable setting unit 18e respectively constitute the setting unit, the test condition change setting unit, and the output level setting unit of the present invention.

[0073] The test condition setting unit 18a is a functional unit for setting the test conditions for the receiver sensitivity test. Examples of test conditions (setting parameters) set by the test condition setting unit 18a include the initial step level SL0, the starting output level OL0, the error tolerance level EL (error tolerance of boundary level), and the connection drop threshold DT (connection drop threshold). The initial step level SL0 represents the initial value of the step variation amplitude of the output level of the test signal, which varies periodically during the receiver sensitivity test. The starting output level OL0 represents the output level of the DUT100 when the receiver sensitivity test begins (the output level during the first transmission and reception). The error tolerance level EL represents the variation amplitude (predetermined value) of the output level of the test signal between the previous and current tests, and is used to determine whether to continue the next receiver sensitivity test. The connection drop threshold DT is a minimum setting value; that is, if set to a value lower than this, the call connection will be dropped (resulting in a call drop). Figure 9 In the algorithm (alg orithm), the output level is gradually reduced with large step sizes. Therefore, a threshold needs to be set to ensure that the level decreases to a certain point and then stops, thus preventing call drops. Users can preset this value.

[0074] The AI ​​prediction test condition change setting unit 18b is a functional unit that uses an AI prediction model (Dai) obtained by performing machine learning (ML) on the correlation between parameter data and measurement result data based on the measurement log file of the receiver sensitivity test of the DUT100 previously performed in the measurement device 1, using artificial intelligence (AI). Figure 7 Instead of the initial set value, the setting is changed to a specific set parameter value predicted in advance by the AI ​​prediction model Dai based on the test conditions (setting parameters) set by the test condition setting unit 18a at the start of the test (initial setting). The specific set parameter, which is the object of the AI ​​prediction test condition change setting unit 18b changing the setting based on the AI ​​prediction using the AI ​​prediction model Dai, can be exemplified by, for example, the output level (starting output level OL0) and the initial step level SL0 of the test signal initially set. (See below for reference) Figure 9 The example given is that the output levels OL0, OL1 and the initial step level SL0 of the test signal are changed based on AI predictions during the period up to the second measurement.

[0075] The change setting process performed by the AI ​​prediction test condition change setting unit 18b on the initial set start output level OL0 is implemented, for example, in the following process: Figure 9 In step S1a of the flowchart shown, the AI ​​prediction output level OL01, predicted in advance by the AI ​​prediction model Dai (AI prediction model A1) based on the test conditions (setting parameters) initially set in the previous step S1, is used to replace the initial output level OL0 set in step S1 and set as the initial output level. Furthermore, the change setting process for the initial step level SL0 is implemented as follows: Figure 9 In step S26 of the flowchart shown, instead of the initial step level SL0 initially set in step S1, the AI ​​predicted step level SL01 is set to be pre-predicted by the AI ​​prediction model Dai (AI prediction model A2) based on the processing results of previous steps S4, S5, and S25 (the measurement results of the initial throughput measurement when the test signal set to the initial output level in step S1a is changed). This will be referred to later. Figure 7 The generation and processing steps of the AI ​​prediction models Dai (A1, A2) are explained in detail.

[0076] The throughput measurement unit 18c is a functional unit that measures the throughput related to the receiving capability of the DUT100 each time a receive sensitivity test is performed. For example, the throughput measurement unit 18c may be structured such that, along with the transmission rate of the test signal, it also transmits the transmission rate to the DUT100, and then, together with the DUT100, it notifies the NR system simulator 20 of the receiving result (receive transmission rate) of the test signal, and measures the throughput based on the receive transmission rate.

[0077] The degradation state determination unit 18d has the function of determining whether the throughput measurement result measured by the throughput measurement unit 18c has decreased to a predetermined proportion within a rapidly decreasing region related to the rapidly decreasing characteristic. To achieve this function, for example, a determination condition is preset by the test condition setting unit 18a to determine whether the throughput measurement result has decreased to a certain proportion within the rapidly decreasing region. The degradation state determination unit 18d determines whether the state is degraded based on whether the measured throughput is within the rapidly decreasing region indicated by the determination condition. For example, as a determination condition for determining whether the state is degraded, when the throughput value measured by the throughput measurement unit 18c through the transmission and reception of the first test signal is set to 100%, a range exceeding 95% and below 99% can be set. Therefore, when the measured throughput is within the set range, a degraded state can be determined; when it is a proportion exceeding the above range, a non-degraded state can be determined (see reference). Figure 9 Steps S7 and S9a). Here, the determination criteria are not limited to setting the throughput value measured by transmitting and receiving the first test signal to 100% (reference value) as described above, and setting it to a range of proportions exceeding 95% and below 99% relative to the reference value. Other proportion ranges relative to the reference value may also be set.

[0078] The variable output level setting unit 18e is a functional unit that, based on the comparison between the throughput measurement result measured by the throughput measurement unit 18c and a preset predetermined threshold (throughput threshold), variably sets the output level of the test signal in the next receiver sensitivity test in an upward (level increase) or downward (level decrease) direction, such that the output level of the test signal is different between the previous and subsequent receiver sensitivity tests.

[0079] The measurement result output unit 18f has the following function: when the variation range between the test result (throughput measurement result) of this receiver sensitivity test based on the test signal with a variable output level and the test result of the previous receiver sensitivity test exceeds the variation range (EL) set by the test condition setting unit 18a, it proceeds to the next receiver sensitivity test (throughput measurement); when the variation range of the test result falls within the variation range (EL), it outputs the test result. The measurement result output unit 18f, together with the receiver sensitivity test control unit 18, constitutes the pass / fail determination unit of the present invention.

[0080] like Figure 4 As shown, the NR system simulator 20 includes a signal generation unit 21a, a transceiver unit 21f, a signal measurement unit 21b, a control unit 21c, an operation unit 21d, and a display unit 21e. The NR system simulator 20 constitutes the signal generator of the present invention.

[0081] The signal generation unit 21a generates a signal (baseband signal) to serve as a test signal source. The transceiver unit 21f generates test signals corresponding to the frequencies of each communication standard based on the signal generated by the signal generation unit 21a and outputs them to the signal processing unit 25. It also functions as an RF unit to reconstruct the baseband signal based on the measured signal transmitted from the signal processing unit 25. The signal measurement unit 21b performs measurement processing on the measured signal based on the baseband signal reconstructed by the transceiver unit 21f.

[0082] The control unit 21c centrally controls the functional units of the signal generation unit 21a, signal measurement unit 21b, transceiver unit 21f, operation unit 21d, and display unit 21e. The operation unit 21d is a functional unit for inputting various information such as commands, and the display unit 21e is a functional unit for displaying various information such as input screens and measurement results.

[0083] In the measuring device 1 with the above-described structure, a DUT 100 can be placed on the mounting plate 56d of the DUT scanning mechanism 56 (biaxial positioner) within the internal space 51 of the OTA chamber 50. While rotating the DUT 100 in the biaxial (azimuth axis and roller axis) direction along each mounting plate 56d (changing the angle of the positioner), measurements of wireless signal-related EIRP (Equivalent Isotropic Radiated Power)-CDF (Cumulative Distribution Function), EIS (Equivalent Isotropic Sensitivity)-CDF, TRP (Total Radiated Power) and other measurement items of the DUT 100 are performed.

[0084] Here, for reference Figure 5 , Figure 6 The angle control (global surface scan) of the DUT100 required for measuring the above-mentioned measurement items by changing the angle of the biaxial positioner is explained.

[0085] Generally, for radiated power measurements on a DUT100, methods for measuring equivalent isotropic radiated power (EIRP) and total radiated power (TRP) are known. EIRP, for example, is... Figure 5 (a) shows the electrical power values ​​measured at each measurement point (θ, φ) in the spherical coordinate system (r, θ, φ). In contrast, TRP refers to a predetermined number of angled specimen points PS (reference points) located at equal distances on a sphere from the center O1 (hereinafter referred to as the reference point) of the global surface scan of the DUT100 in all orientations of the aforementioned spherical coordinate system (r, θ, φ). Figure 5 The value is obtained by summing the EIRP values ​​at (b)).

[0086] In this embodiment, the number of segments Nθ and Nφ used to calculate the total radiated power (TRP) are set to 12, for example. Therefore, in this embodiment, the number of angle specimens (N) is calculated as N = 132 (= (12-1) × 12). If these 132 angle specimen points PS are represented on the surface of sphere B, they become as follows: Figure 5 The location shown in (b)

[0087] In the measuring device 1 according to this embodiment, such as Figure 5As shown in (b), EIRP is measured at 132 points equidistant from the reference point in the spherical coordinate system (r, θ, φ), and the EIRP at all points is summed. Then, based on the sum of the above EIRPs, that is, the sum of the EIRPs at all angle specimen points PS of the 132 points, the total radiated power (TRP) of DUT100 is calculated.

[0088] During TRP measurement, the integrated control unit 10 performs a global surface scan of the DUT 100 by driving and controlling the DUT scanning mechanism 56. During the global surface scan of the DUT 100, the integrated control unit 10 repeatedly drives the motor 56f (driving / non-driving) and rotates the drive turntable 56a around the azimuth axis. Simultaneously, it repeatedly drives the motor 56g (driving / non-driving) and rotates the mounting plate 56d around the roller axis. At this time, each time the antenna surface of the antenna 110 is aligned with an angular specimen point PS, the integrated control unit 10 controls the drive motors 56f and 56g to be non-driven. Through this global surface scan control of the DUT 100, the DUT 100, mounted on the mounting plate 56d, is kept in a state where the antenna 110 is positioned at the center of the sphere B in the specified spherical coordinate system (r, θ, φ), i.e., the reference point. The antenna 110 is rotated around the reference point in such a manner that its antenna surface sequentially faces (points to) all angular specimen points PS of the sphere B.

[0089] like Figure 6 As shown, the test antenna 5 is positioned at a specific angle specimen point PS (point 1) in the aforementioned spherical coordinate system (r, θ, φ). During the global surface scan described above, the DUT 100 is driven (scanned) with the antenna surface of the antenna 110 sequentially facing the light-receiving surface of the test antenna 5. Thus, the test antenna 5 can transmit and receive signals for TRP measurements between itself and the antenna 110 of the DUT 100 performing the global surface scan. The transmitted and received signals are the test signal transmitted from the NR system simulator 20 via the test antenna 5, and the measured signal received by the DUT 100 via the antenna 110 after receiving the test signal.

[0090] In the integrated control device 10, Figure 5 In the spherical coordinate system (r, θ, φ) shown in (b), the DUT100 is scanned by means of angular specimen points PS in the φ direction while maintaining a certain angle θ. In conjunction with this, the NR system simulator 20 is driven to generate the aforementioned test signal via the signal generation unit 21a and the transceiver unit 21f, and the test signal is transmitted from the test antenna 5 via the signal processing unit 25. If the test signal is received by the antenna 110, the DUT100 transmits a response signal corresponding to the reception of the test signal.

[0091] The integrated control unit 10 further drives the NR system simulator 20, and the DUT 100 transmits in response to the reception of the aforementioned test signal. The signal measurement unit 21b receives the signal received by the test antenna 5 from the signal processing unit 25 via the transceiver unit 21f as the measured signal. Furthermore, the integrated control unit 10 drives the control signal measurement unit 21b to perform signal processing related to the EIRP measurement based on the received measured signal. This EIRP measurement control is implemented by coordinating a global scan of the DUT 100 through all angle specimen points PS by changing the θ angle. In the NR system simulator 20, the EIRP can be measured for all angle specimen points PS in the spherical coordinate system (r, θ, φ) corresponding to the NR. Furthermore, the integrated control unit 10 can calculate the sum of the EIRP measurements for all angle specimen points PS, i.e., the TRP.

[0092] Furthermore, the integrated control device 10 has an output level control function that, before performing performance tests on the DUT100 while changing the angle of the dual-axis positioner (DUT scanning mechanism 56) within the OTA chamber 50, specifically, before performing measurements such as EIRP-CDF, EIS-CDF, and TRP, the output level (power level) of the test signal in the NR system simulator 20 is adjusted to an appropriate level as specified in the 3GPP standard. Based on this output level control function, when measuring the aforementioned items, a search is performed for the output level of the test signal that allows the DUT100 to perform at its maximum capacity (EIS search). Therefore, the output level control function performed by the NR system simulator 20 can also be understood as the control function involved in the receiver sensitivity test for searching the receiver sensitivity of the DUT100. The control function involved in the receiver sensitivity test is implemented by the receiver sensitivity test control unit 18 provided in the control unit 11 of the integrated control device 10.

[0093] (Convergence control of the output level of the test signal in the receiver sensitivity test)

[0094] In the integrated control unit 10, the receiver sensitivity test control unit 18 controls the receiver sensitivity test of the DUT100 by repeatedly transmitting and receiving test signals between the signal generator, i.e., the NR system simulator 20, and the DUT100. In this control, the throughput is measured in conjunction with the transmission and reception of each test signal in the receiver sensitivity test. While reducing or increasing the output level of the test signal based on the comparison result of the measured throughput value and the throughput threshold, the output level is gradually converged to an output level at which an appropriate throughput value (measurement result) can be obtained.

[0095] As an example of convergence control of the output level of the test signal in the aforementioned receiver sensitivity test, in the measurement device 1 of this embodiment, the receiver sensitivity test control unit 18 of the integrated control device 10 is provided with an initial measurement parameter change setting function that changes specific setting parameters (e.g., initial output level OL0, initial step level SL0) in the test conditions set when the DUT 100 starts measurement from the initial setting value according to AI prediction technology. This dynamically controls the output level of the test signal so that the start location of the EIS search is as close as possible to the end location of the original EIS search. By setting the aforementioned initial measurement parameter change setting function, in the measurement device 1 of this embodiment, the output level changes sequentially (linearly) from the initially set output level with a certain level each time the number of measurements increases (see reference). Figure 13 The method of curves L1 and L2) and the method of controlling the output level by repeatedly decreasing or increasing the output level of the test signal while making the output level change non-linearly (see reference). Figure 13 Compared to existing methods that use CP determination processing to reduce the number of measurements (such as curves NL1 and NL2), this method can quickly converge the output level of the test signal to the target level, thereby significantly reducing the EIS search time.

[0096] As an initial measurement parameter change setting function, the AI ​​prediction test condition change setting unit 18b of the receiver sensitivity test control unit 18 holds an AI prediction model Dai obtained by machine learning based on the measurement log file of the receiver sensitivity test of the DUT100 of the measurement device 1 in the past. Using the AI ​​prediction model Dai, the starting output level (OL0) and the initial step level (SL0) are changed to values ​​for the first measurement or the second measurement based on the AI ​​prediction.

[0097] refer to Figure 7 The generation and processing steps of the AI ​​prediction model Dai are explained in detail. For example... Figure 7 As shown, in order to generate the AI ​​prediction model Dai, measurement log files collected from multiple past measurements of the receiver sensitivity of the DUT 100 in the measurement apparatus 1 according to this embodiment are prepared (step S31). Examples of measurement log files to be prepared include EIS measurement log files or EIS-CDF measurement log files. Preferably, the number of these measurement log files collected is equivalent to the number of measurements (e.g., 10 or more) that allow the AI ​​prediction model Dai, as the target model, to maintain sufficient accuracy.

[0098] Next, in this generation process step, EIS measurement result data (the EIS-CDF log file) and parameter data used (set) during each measurement are extracted from the EIS measurement log file (or EIS-CDF log file) collected from the n measurements prepared in step S31 (step S32). Here, as the parameter data to be extracted, at least the loss of the external connection cable (external path loss), the angle of the DUT scanning mechanism 56, i.e., the azimuth angle (θ, φ) of the DUT100, and the frequency of the measured object can be cited as examples.

[0099] Furthermore, in this generation process, AI is used to perform machine learning on the relationship between the parameter data extracted in step S32 and the measurement result data (step S33). The machine learning results are then summarized in the AI ​​prediction model Dai and stored in a predetermined storage area (step S34). Here, various AI prediction models Dai that reflect the categories, etc., of parameter data extracted from past measurement log files are generated. In this embodiment, the AI ​​prediction model Dai is, for example, stored in an AI prediction model storage unit 18b1 provided in the RAM 11c of the control unit 11 of the integrated control device 10.

[0100] from Figure 7 The generation process shown is understandable, as the AI ​​prediction model Dai generated in this process is a model composed of data obtained by statistically analyzing the correlation between parameter data and measurement result data from multiple past DUT100 receiver sensitivity tests. In summary, the AI ​​prediction model Dai represents statistical data indicating what measurement results can be obtained for what parameter data values.

[0101] Therefore, in the measurement device 1 according to this embodiment, based on the initial setting parameters set in the receiver sensitivity test of the DUT100, and referring to the AI ​​prediction model Dai, the output level or step level of the test signal at EIS search points within a predetermined number of search times starting from the last EIS search point on the EIS search path in the original measurement method can be determined by AI prediction. In other words, in the measurement device 1 according to this embodiment, the EIS search up to the aforementioned last EIS search point using the initially set setting parameters can be skipped, and the EIS search can start from that last EIS search point, thereby significantly reducing the number of EIS searches.

[0102] Furthermore, in this embodiment, external path loss, the azimuth angle (θ, φ) of the DUT100, and the frequency of the measured object are used as parameter data for generating the AI ​​prediction model Dai. Therefore, the AI ​​prediction model can be used to derive accurate AI prediction results that take into account these parameter data, and the test signal can be reliably set to a testable level in a very short time without reducing the measurement accuracy.

[0103] A comparison example is shown of the measurement start position of the receiver sensitivity test of the DUT using AI prediction in the measurement device 1 according to this embodiment and the receiver sensitivity test of the DUT in a conventional device. Figure 8 In the middle. For example Figure 8 As shown, in the existing device, regardless of whether the output level of the test signal is linearly or nonlinearly controlled, the measurement starts from a fixed value initially set at the beginning of the EIS measurement (e.g., OL0 = -80dBm) ("Previous Measurements" in the figure). Therefore, assuming that an "EIS1" measurement that meets the termination condition at -90dBm is performed in the "Previous Measurement" step, the number of measurements is, for example, n. On the other hand, assuming that an "EIS2" measurement that meets the termination condition at -96dBm is performed in the "Previous Measurement" step, the number of measurements becomes more, for example, N (N > n).

[0104] In contrast, in the measurement apparatus 1 according to this embodiment, which uses AI prediction (“AI measurement” in the figure), the EIS search begins with the value (e.g., OL01:(predicted value)) predicted by the AI ​​prediction model Dai based on the initial settings parameters described above. Here, when performing EIS1 measurement, the AI-predicted value OL01 is, for example, the value corresponding to EIS search points within a predetermined range (in this example, before the first measurement) starting from the last EIS search point in the “previous measurement” (the point corresponding to the “nth measurement”). Thus, in the AI-predicted EIS1 measurement, the output level of the test signal is set to, for example, -89.8dBm and the first measurement is performed. Then, the output level is set to, for example, -90dBm and the second measurement is performed. Here, it is determined that the termination condition is met, and the EIS search ends with the second measurement.

[0105] Furthermore, during the EIS2 measurement, the AI-predicted value OL01 is, for example, the value corresponding to an EIS search point within a predetermined range (in this example, 1 time ago) starting from the last EIS search point in the "previous measurements" (the point corresponding to the "Nth measurement"). Thus, in the AI-predicted EIS2 measurement, the output level of the test signal is set, for example, to -95.8 dBm for the first measurement, and then the output level is set, for example, to -96 dBm for the second measurement. This is deemed to meet the termination condition, and the EIS search ends with the second measurement.

[0106] like Figure 8 As shown, in the measuring device 1 of this embodiment, by dynamically and adaptively controlling the output level of the test signal that starts the measurement according to the set parameters, compared with the "previous measurement" that usually starts the measurement from a fixed value, the EIS search path can be shortened significantly, and the number of EIS searches can be reduced significantly, so that the output level of the test signal can be set to a testable level in a shorter time.

[0107] Based on the above-described method of shortening the receiver sensitivity test time of DUT100 using AI prediction model, the following references... Figure 9 , Figure 10 The variable setting control operation of the output level of the test signal involved in the receiver sensitivity test of the DUT100 performed by the integrated control device 10 of the measuring device 1 according to this embodiment will be described.

[0108] Figure 9 This is a flowchart illustrating the variable setting control operation of the output level of the test signal, which utilizes AI prediction, during the receiver sensitivity test of the DUT100 in the measuring apparatus 1 according to this embodiment. Figure 9 In this process, steps S20, S1a, S25, S26, and S27 constitute the output level change setting process for which the AI ​​prediction described above applies. As a method for storing the control data used to implement the output level change setting process for which the AI ​​prediction applies, one approach is to store data (prediction result data) of the results predicted in advance by the AI ​​prediction model Dai based on the setting parameters, and another approach is to store the AI ​​prediction model Dai and perform AI prediction based on the AI ​​prediction model Dai each time the setting parameters are initially set at the start of the receiver sensitivity test of the DUT100. The following description assumes the former approach is applicable. The storage location for the control data is, for example, the AI ​​prediction model storage unit 18b1.

[0109] In the measuring device 1 according to this embodiment, in order to... Figure 9The flowchart shown begins the receiver sensitivity test of DU T100. First, the receiver sensitivity test control unit 18 in the control unit 11 of the integrated control device 10 sets the test conditions (setting parameters) (initial setting) (step S1). Specifically, the test condition setting unit 18a sets the values ​​of the initial step level SL0, the start output level OL0, the fault tolerance level EL, the connection disconnection determination threshold DT, and the determination condition for the sharp decrease in throughput (refer to steps S7 and S9a) through the operation input in the receiver operation unit 12.

[0110] The initial output level OL0 and initial step level SL0 set in step S1 are assumed to be, for example, -75dBm and 10dB, respectively. The fault tolerance level EL is assumed to be, for example, 0.2dB. The connection disconnection determination threshold DT is assumed to be, for example, -90dBm. Furthermore, as a condition for determining a region of sharp decrease in throughput, the condition is assumed to be, for example, a range in which the throughput measurement value exceeds 95% but is less than 99% of the aforementioned reference value.

[0111] After completing the initial setting of the test conditions in step S1, the receiver sensitivity test control unit 18 performs a process of changing and setting the initial set starting output level OL0 based on AI prediction (step S1a). Specifically, in the receiver sensitivity test control unit 18, the AI ​​prediction test condition change setting unit 18b searches for AI prediction result data predicted by the AI ​​prediction model Dai stored in the AI ​​prediction model storage unit 18b1 based on the starting output level OL0 set in step S1 (step S20), and changes the value of the set starting output level OL0 to the AI ​​prediction output level (OL01) predicted in advance by the AI ​​corresponding to the starting output level OL0 (step S1a). That is, the initially set starting output level OL0 is changed to the initial output level OL01.

[0112] Next, the receiver sensitivity test control unit 18, based on an increment of +1 to the number of measurements n (step S2), sets the parameters involved in the Nth measurement and reads the value of the output level OL set in steps such as step S8 (including steps S8a and S8b) for OL level reduction processing or step S9 for OL level enhancement processing (step S3). Then, the receiver sensitivity test control unit 18 controls the measurement by sending a test signal based on the parameters set (or read) in step S3 and performing the nth measurement related to the throughput of the DUT 100 (step S4).

[0113] As a specific example of the control (throughput measurement control) in steps S3 and S4, the receiver sensitivity test control unit 18 reads the initial output level OL01 set in the test condition settings in step S1 and the change settings based on AI prediction in step S20 as parameters related to the first measurement, and uses the initial output level OL01 to drive and control the DUT100 while performing throughput measurement.

[0114] Next, the receiver sensitivity test control unit 18 checks the interval of the step level involved in this throughput measurement relative to the previous throughput measurement, that is, it checks whether the step level SL(n) is greater than the fault tolerance level EL set in step S1 (step S5). Here, if it is determined that the step level SL(n) is greater than the fault tolerance level EL ("Yes" in step S5), the receiver sensitivity test control unit 18 proceeds to step S25 and continues to perform throughput measurement and step level search control. In addition, during the first throughput measurement, as described above, it starts from the transmission of the test signal of the initial output level OL01, and there is no change in SL relative to the previous measurement, so the processing of step S5 is skipped and the process proceeds to step S25.

[0115] In step S25, the receiver sensitivity test control unit 18 checks whether the number of measurements N is 1. Here, if it is determined that the number of measurements N is not 1 (i.e., it is 2 or more) ("No" in step S25), the receiver sensitivity test control unit 18 proceeds to the processing after step S6, which is the measurement processing based on the original measurement method.

[0116] In contrast, when it is determined that the number of measurements N is 1 ("Yes" in step S25), the receiver sensitivity test control unit 18 acquires the AI ​​prediction step level SL01, which is predicted in advance based on the initial step level SL0 set in step S1, and performs processing to set the output level OL1 for the second measurement based on the acquired AI prediction step level SL01 (steps S26 and S27). Specifically, in the receiver sensitivity test control unit 18, the AI ​​prediction test condition change setting unit 18b searches the AI ​​prediction result data stored in the AI ​​prediction model storage unit 18b1 and acquires the AI ​​prediction step level SL01, which is predicted in advance based on the initial step level SL0 set in step S1 (step S26). Next, the AI ​​prediction test condition change setting unit 18b performs a process of setting the output level OL1 (=OL01+SL01) of the test signal for the second measurement based on the acquired AI prediction step level SL01 and the initial output level OL01 changed in step S1a (step S27).

[0117] Then, the receiver sensitivity test control unit 18 continues to perform steps S2, S3, and S4. That is, the receiver sensitivity test control unit 18 controls the process in the following manner: it increments the number of measurements n by +1 (step S2), sets the parameters involved in the second measurement, and, based on the previously set (or read) value of the output level OL1 set according to AI prediction in step S27 (step S3), sends a test signal according to the set (or read) parameters involved in the measurement and performs a second measurement related to the throughput of the DUT100 (step S4).

[0118] Specifically, in the control (throughput measurement control) of steps S3 and S4, the receiver sensitivity test control unit 18 reads the output level OL1 (=OL01+SL01) set in step S27, and uses the output level OL1 to drive and control the DUT100 and perform the second throughput measurement.

[0119] Furthermore, the receiver sensitivity test control unit 18 checks the interval of the step level involved in this throughput measurement relative to the previous throughput measurement, that is, checks whether the step level SL(n) is greater than the fault tolerance level EL set in step S1 (step S5).

[0120] Here, when it is determined that the step level SL(n) is below the fault tolerance level EL (No in step S5), the receiver sensitivity test control unit 18 stops the throughput measurement and the step level search, outputs the measurement result at this time (step S10), and ends a series of measurement actions. Correspondingly, the receiver sensitivity test control unit 18 feeds back the throughput measurement result output in step S10 and the parameter data set by the AI ​​prediction test condition change setting unit 18b, and performs the process of updating the AI ​​prediction model Dai (AI prediction model A1) used in step S20 (step S11).

[0121] In contrast, when it is determined that the step level SL(n) is greater than the fault tolerance level EL ("Yes" in step S5), the receiver sensitivity test control unit 18 checks whether the number of measurements N is 1 (step S25). Here, when it is determined that the number of measurements N is not 1 (i.e., is 2 or more) ("No" in step S25), the receiver sensitivity test control unit 18 performs the processing after step S6, that is, the measurement processing based on the original measurement method.

[0122] Thus, in Figure 9In the variable setting control operation of the output level of the test signal involved in the receiver sensitivity test of the DUT100 in the measuring device 1 of this embodiment, the following example is given: the setting change based on AI prediction, starting from the initial output level OL0 initially set in step S1, is limited to 2 times (steps S1a and S26). If no "yes" determination result is obtained in step S5 during this period, the measurement processing based on the original measurement method is performed from step S6 onwards. Here, the number of setting changes based on AI prediction is not limited to 2 times, and can be set to other numbers as well.

[0123] The measurement processing based on the original measurement method after step S6 is similar to that in Patent Document 1. Figure 7 The processing after step 6 in the flowchart is the same, roughly as follows.

[0124] First, in step S6, the receiver sensitivity test control unit 18 compares the throughput measurement value in step S4 with the preset throughput threshold. If it is determined that the throughput measurement value is above the throughput threshold (the "passed" state in step S6), then it proceeds to step S7 and determines whether the determination condition that the throughput measurement value is within the range of 95% to 99% of the above-mentioned benchmark value is met.

[0125] Here, when it is determined that the measured throughput value exceeds 99% of the aforementioned reference value and does not meet the aforementioned determination condition ("No" in step S7), the receiver sensitivity test control unit 18 performs an OL level (output level) reduction process (A) to reduce the output level of the test signal (step S8a). On the other hand, when the reduction state determination unit 18d determines that the measured throughput value is within the range of more than 95% and less than 99% of the aforementioned reference value and meets the aforementioned determination condition ("Yes" in step S7), the receiver sensitivity test control unit 18 performs an OL level reduction process (B) (step S8b). The OL level reduction process (A) is as follows: Figure 10 (a) shows the OL level reduction process (B) as shown. Figure 10 As shown in (b).

[0126] On the other hand, if it is determined in step S6 that the throughput (measured value) is below the throughput threshold (the state of "failure" in step S6), the receiver sensitivity test control unit 18 then performs the determination process in step S9a. Here, when it is determined that the measured value of the throughput exceeds 80% of the above-mentioned reference value and meets the above-mentioned determination condition ("yes" in step S9a), the OL level boosting process (A) for boosting the output level of the test signal is performed. On the other hand, when it is determined that the measured value of the throughput is below 80% of the above-mentioned reference value and does not meet the above-mentioned determination condition ("no" in step S9a), the OL level boosting process (B) or (C) is performed according to the determination result in step S9c.

[0127] After performing the above-mentioned OL level reduction processing (A) and (B) or OL level enhancement processing (A), (B), and (C), the receiver sensitivity test control unit 18 continues to perform steps S2 to S4. Then, in step S5, if it is determined that the step level SL(n) is below the fault tolerance level EL (No in step S5), the throughput measurement and step level search are stopped (step S10), and then the series of measurement operations are ended. Here, the receiver sensitivity test control unit 18 also feeds back the throughput measurement result and performs a process to update the AI ​​prediction model Dai (AI prediction model A1) used in step S20 (step S11).

[0128] according to Figure 9 The series of measurement controls shown, for the target object, will change the initial output level OL0 and initial step level SL0 initially set in step S1 according to AI prediction (refer to steps S20, S1a, S25, S26, S27) for example, until the second measurement.

[0129] Based on the AI ​​prediction and the change setting of the initial output level OL0, for example, the measurement can be completed with only two EIS path searches, compared to... Figure 9 Compared to the nonlinear measurement based on the original measurement method performed after step S6 and employing the CP decision condition, this method can further reduce the number of EIS searches. Specific examples will be provided below to illustrate this point.

[0130] Figure 11 This is a graph showing the relationship between the number of measurements involved in the receiver sensitivity test of the DUT100 with AI prediction in the measurement device 1 according to this embodiment and the receiver sensitivity test of the DUT100 in the conventional device, and the good or bad judgment results of the output level and throughput measurement results of the test signal. Figure 12 It means Figure 11The graph shows the relationship between the number of measurements and the results of the good or bad judgment of the output level and throughput measurement of the DUT100 in the measurement device 1 with AI prediction in this embodiment and the DUT100 in the conventional device.

[0131] exist Figure 11 In the chart shown, columns 2 and 3 from the left represent data examples involved in the receiver sensitivity test of the DUT100 based on nonlinear output level control in the original device. Columns 4 and 5 from the left represent the relationship between the number of measurements involved in the receiver sensitivity test of the DUT100 using AI prediction-based output level control in the measurement device 1 according to this embodiment and the good / bad judgment results of the output level and throughput measurement results of the test signal. Similarly, in Figure 12 The curve shown will be related to Figure 11 In this embodiment, the curve corresponding to the data example of output level control of the experimental signal based on AI prediction is represented by the symbol C1 (a curve representing the relationship between the number of measurements and the output level), and will be compared with... Figure 11 The curve corresponding to the output level control data of the nonlinear test signal in the original device is represented by the symbol C2 (a curve representing the relationship between the number of measurements and the output level).

[0132] like Figure 11 (Refer to columns 1, 2, and 3 from the left) and Figure 12 As shown in characteristic C2, in the receiver sensitivity test of DUT100 based on nonlinear output level control in the original device, the DUT100 was tested for the first time at the measurement point P21 of characteristic C2 with an output level OL(0) = -80dBm. At this time, the measured throughput was determined to be within the allowable range (pass), and the OL level was reduced.

[0133] As a result of the OL level reduction process, the DUT100 was tested for the second time at the measurement point P22 of characteristic C2 with an output level OL(1) = -90dBm. The measured throughput was determined to be within the allowable range (passed), and the OL level reduction process was further implemented.

[0134] As a result of the OL level reduction process, the DUT100 was tested for the third time at the measurement point P23 of characteristic C2 with an output level OL(2) = -95dBm. The measured throughput was determined to be within the allowable range (passed), and the OL level reduction process was further implemented.

[0135] As a result of the OL level reduction process, the DUT100 was tested for the fourth time at the measurement point P24 of characteristic C2 with an output level OL(3) = -97.5dBm. At this time, the measured throughput was determined to be outside the allowable range (failure), and the OL level was increased.

[0136] As a result of the OL level boosting process, the DUT100 was tested for the fifth time at the measurement point P25 of characteristic C2 with an output level of OL(4) = -96.2dBm. At this time, the measured throughput was determined to be outside the allowable range (failure), and the OL level boosting process was further implemented.

[0137] As a result of the OL level boosting process, the DUT100 was tested for the sixth time at the measurement point P26 of characteristic C2 with an output level OL(5) = -95.6dBm. At this time, the measured throughput was determined to be within the allowable range (pass), and the OL level was reduced.

[0138] As a result of the OL level reduction process, the DUT100 was tested for the 7th time at the measurement point P27 of characteristic C2 with an output level OL(6) = -95.9dBm. At this time, the measured throughput was determined to be outside the allowable range (failure), and the OL level was increased.

[0139] As a result of the OL level boosting process, the DUT100 was tested for the 8th time at the measurement point P28 of characteristic C2 with an output level OL(7) = -95.8dBm. At this time, the measured throughput was determined to be outside the allowable range (failure), but the step level SLn between the throughput and the throughput measured in the previous (7th) became below the fault-tolerant level EL, and the series of measurements ended.

[0140] On the other hand, such as Figure 11 (Refer to columns 1, 4-5 from the left) and Figure 12 As shown in characteristic C1, according to the measurement device 1 of this embodiment, the receiving sensitivity test of the DUT100 with AI prediction-based output level control was conducted. The first time, at measurement point P11 of characteristic C1, the signal was obtained through AI prediction (reference...). Figure 9 In steps S20 and S1a), the initial output level OL(01) is changed to -95.9dBm to start the test on DUT100. At this time, the measured throughput is determined to be outside the allowable range (failure) and OL level boosting is implemented.

[0141] The result of this OL level boosting process, the second time at measurement point P12 of characteristic C1, is predicted by AI (reference). Figure 9In steps S26 and S27), the output level OL(1) is set to -95.8dBm to start the test on DUT100. At this time, the measured throughput is determined to be within the allowable range (pass), but the step level SLn between the throughput and the throughput measured in the previous (first) measurement becomes the fault-tolerant level EL and the measurement ends.

[0142] like Figure 12 As shown in characteristic C2, even with nonlinear output level control implemented in the receiver sensitivity test of the DUT100 in the original device, eight measurements are still required on P21, P22, P23, P24, P25, P26, P27, and P28.

[0143] In contrast, in the receiver sensitivity test of the DUT100 in the measuring device 1 according to this embodiment, the receiver sensitivity test can be completed by applying dynamic control of the initial output level of the test signal predicted by AI, so that two measurements on P11 and P12 are sufficient.

[0144] The main reason why the receiver sensitivity test of the DUT100 in the measuring device 1 according to this embodiment can be completed with fewer measurements is that the initial measurement starts from a value close to the end of a past measurement, using AI prediction technology based on the AI ​​prediction model Dai generated from past measurement log files. To achieve this, in the measuring device 1 according to this embodiment, the initially set starting output level OL0 (e.g., -75dBm) is changed to an AI predicted output level OL01 (e.g., -95.9dBm: reference OL01) that is equivalent to the value two measurements prior to the end of a past measurement, using AI prediction. Figure 11 Here, the closer the value of the changed setting parameter (AI predictive output level (initial output level) OL01) is to the value near the end of the measurement of this setting parameter in past measurements, the shorter the convergence time to the output level that yields the target throughput value (measurement result).

[0145] refer to Figure 13 The time reduction effect of the AI-predicted DUT100 in the measuring device 1 according to this embodiment on the receiving sensitivity test will be explained. Figure 13In the graph shown, curves L1 and L2 represent the search paths of the EIS search based on linear output level control in the receiver sensitivity test of the DUT100 in the original device, while curves NL1 and NL2 represent the search paths of the EIS search based on nonlinear output level control in the receiver sensitivity test of the DUT100 in the original device. As shown by curves L2 and NL2, the EIS searches based on both linear and nonlinear output level control in the receiver sensitivity test of the DUT100 in the original device ended at EIS point 1.

[0146] In this measurement situation, when performing measurements using AI prediction in the measurement apparatus 1 according to this embodiment, for example, instead of the initial measurement OL level (starting output level) set during the measurement based on the aforementioned nonlinear output level control, AI prediction using the AI ​​prediction model Dai can be used. For example, the value (AI prediction result) corresponding to the AI ​​prediction point Pai0 within the AI ​​prediction range Rai within a predetermined range from EIS point 1 is set to the aforementioned initial output level OL01 (reference). Figure 9 Step S1a). Thus, as Figure 13 As shown by curve Lai on the graph, the EIS search path in the measuring device 1 when AI prediction is applied includes two measurements within the AI ​​prediction range Rai, starting from AI prediction point Pai0, passing through the next AI prediction point Pai1 (measurement result), and ending at EIS point 1.

[0147] Available Figure 11 , Figure 12 The description is based on the receiver sensitivity test of the DUT100 when the measurement results data are shown. In the measuring device 1 according to this embodiment, in Figure 9 In step S1a, based on AI prediction, the initial output level OL0 set in step S1 is changed to be the same as... Figure 13 The value corresponding to the AI ​​prediction point Pai0 within the AI ​​prediction range Rai shown (initial output level OL01). Then, during the second measurement, in Figure 9 In step S27, the settings are changed to match the AI ​​prediction. Figure 13 The value OL1 = (OL01 + SL01) corresponding to the AI ​​prediction point Pai2 within the AI ​​prediction range Rai shown is obtained after the second throughput measurement. The "yes" judgment in step S5 is obtained and the measurement ends.

[0148] Figure 13 It was made public. Figure 5 (b) shows a graph of the AI ​​prediction results for the output level of each angled specimen point PS during EIS measurement. More time is required when measuring EIS-CDF. See here for reference. Figure 14The relationship between the measurement time achievable using AI-predicted EIS-CDF measurement in the measurement apparatus 1 according to this embodiment and the number of measurements used for machine learning was verified. Regarding the measurement conditions in the measurement apparatus 1 during this verification, for example, Meas.Metric (measurement object) was set to EIS-CDF, Grid Type was set to Constant Step, Pont (number of angle specimen points PS) was set to 168, Range (measurement range) was set to a hemisphere, Step (angle to move to the next angle specimen point PS) was set to 15 degrees, and Error Control (same as the "error tolerance level EL" mentioned above) was set to 0.2 dB.

[0149] Under the aforementioned measurement conditions, an EIS-CD F measurement was performed in the measuring apparatus 1 used for verification preparation. The results, based on the number of measurements, yielded [data / information]. Figure 14 The verification results are shown in the graph. Figure 14 In the graph, the measurement corresponding to "Measurement 0" corresponds to the measurement result when measured using a conventional device without AI prediction. The measurement corresponding to "Measurement 1" corresponds to the measurement result when measured using the default prepared AI prediction model. Furthermore, measurements corresponding to "Measurement 2" and later (in this example, the measurement corresponding to "Measurement 2" and the measurement corresponding to "Measurement 3") correspond to measurement results obtained using an AI prediction model acquired through machine learning based on the preceding measurements, i.e., the measurements corresponding to "Measurement 1" and "Measurement 2". Figure 14 As shown in the graph, in the measurement corresponding to "Measurement 0", it took 173 minutes to complete. On the other hand, in the measurement corresponding to "Measurement 1", machine learning was applied, and the measurement started from the AI-predicted output level OL01, resulting in a shorter measurement time of 126 minutes compared to "Measurement 0". Furthermore, in the measurements corresponding to "Measurement 2" and "Measurement 3", the times to complete were 72 minutes and 67 minutes respectively, showing a trend of decreasing measurement time as the number of machine learning iterations increases. Figure 14 The graph shown illustrates the machine learning effect that leads to a reduction in measurement time starting around the 2nd or 3rd iteration. Therefore, it can be inferred that further improvements in measurement speed can be expected by using measurement device 1 to store measurement information in the future.

[0150] Furthermore, the operation of using measuring device 1 to store measurement information is as described above, for example, it is possible to... Figure 7The steps shown are performed. Here, regarding the measurement conditions and number of measurements for which an improvement in measurement speed can be expected, when the measurement conditions are set to "automatically update by performing only EIS / EIS-CDF measurements", the inventors of the present invention were able to confirm that an improvement in measurement speed begins to appear from about 10 machine learning iterations.

[0151] Furthermore, while the above embodiment exemplifies specialized receiver sensitivity test control for measurements on a single surface (EIS measurement), this embodiment can also be applied to measurements on a global surface (TRP measurement: reference). Figure 5 In the relevant receiver sensitivity test.

[0152] Furthermore, in the above embodiments, a system structure example in which an integrated control device 10 is provided outside the measuring device 1 is disclosed. However, the present invention can also be a structure in which the control function of the integrated control device 10 is provided in the measuring device 1.

[0153] As described above, the measuring device 1 according to this embodiment has the following structure: it includes an NR system simulator 20 that generates test signals; and a receiver sensitivity test execution unit 18 that performs a receiver sensitivity test on the mobile terminal by sending and receiving test signals multiple times between the NR system simulator 20 and the mobile terminal (DUT100). The measuring device 1 performs tests on the test object, i.e., the DUT100, and includes a test condition setting unit 18a that initially sets the test conditions, including the step level SL0 and the output level OL0 of the test signal at the start of the receiver sensitivity test; and a measurement result output unit 18f that performs throughput by sending and receiving test signals under the above test conditions. The system measures throughput and compares the results with a threshold to determine whether the throughput is qualified or not. The output level variable setting unit 18e sets the output level based on the previous throughput measurement results when the result of the qualification or failure determination is negative. The AI ​​prediction test condition change setting unit 18b holds an AI prediction model Dai generated by machine learning based on the correlation between parameter data and measurement result data from past receiver sensitivity test measurement log files. The AI ​​prediction output level OL01, which is predicted by the AI ​​prediction model Dai in advance according to the test conditions set by the test condition setting unit 18a, is set as the initial output level instead of the output level OL0.

[0154] According to this structure, the measuring device 1 involved in this embodiment can quickly set the initial output level of the test signal to a value close to the end value of the measurement under the original nonlinear control by AI prediction. Compared with conventional methods such as nonlinear control and CP control, it can significantly reduce the number of measurements and set the output level of the test signal to a testable level in a very short time, and can effectively perform mobile terminal receiving sensitivity tests.

[0155] Furthermore, in the measuring device 1 according to this embodiment, it can be structured such that the AI ​​prediction test condition change setting unit 18b sets the AI ​​prediction step level SL01, which is predicted in advance by the AI ​​prediction model Dai, to the step level for subsequent measurements after the second measurement, based on the measurement result of the first throughput measurement performed by transmitting and receiving test signals with AI prediction output level.

[0156] According to this structure, the measuring device 1 involved in this embodiment not only dynamically controls the initial output level of the test signal through AI prediction, but also dynamically controls the initial step level. As a result, compared with the case of continuing measurement in the conventional way such as nonlinear control and CP control, the test signal can be set to a testable level in a shorter time, thereby further improving the efficiency of mobile terminal receiving sensitivity testing.

[0157] Furthermore, in the measuring device 1 according to this embodiment, it can be structured such that when the result of judging whether the initial throughput measurement result is qualified or not is judged as not being qualified, then, limited to the second throughput measurement, the AI ​​prediction test condition change setting unit 18b uses the AI ​​prediction step level SL01 to perform output level update processing for the next measurement.

[0158] According to this structure, the measuring device 1 of this embodiment can dynamically control the initial output level and step level of the test signal predicted by AI until the second throughput is measured, thereby avoiding the situation where the above-mentioned dynamic control is still repeating in vain when the measurement is deliberately started from the value close to the end of the measurement under nonlinear control, etc.

[0159] Furthermore, in the measuring device 1 according to this embodiment, it can be structured such that the AI ​​prediction test condition change setting unit 18b specifies the loss of the external connection cable, the azimuth angle (θ, φ) of the DUT100 and the frequency of the measured object as the above-mentioned parameter data, and holds an AI prediction model generated by machine learning through the relationship between the parameter data and the measurement result data.

[0160] According to this structure, the measuring device 1 involved in this embodiment can use the AI ​​prediction model to derive accurate AI prediction results that take into account the various parameter data of the external connection cable loss, the azimuth angle (θ, φ) of the DUT100 and the frequency of the measured object, even in environments where the measurement accuracy is affected by the loss of the external connection cable, the azimuth angle (θ, φ) of the DUT100 and the frequency of the measured object. This allows the test signal to be reliably set to a testable level in a very short time without reducing the measurement accuracy.

[0161] Furthermore, in the measuring device 1 according to this embodiment, it can be structured such that when the result of judging whether the second throughput measurement result is qualified or not is judged as not being qualified, the receiving sensitivity test control unit 18 controls the output level variable setting unit 18e to perform level reduction processing A, level reduction processing B, level increase processing A, or level increase processing B.

[0162] According to this structure, when the output level of the test signal fails to reach a testable level due to dynamic control of the initial output level and initial step level of the test signal predicted by AI, the measuring device 1 of this embodiment can immediately switch to a normal measurement program to reliably set the output level of the test signal to a testable level.

[0163] Furthermore, the measuring device 1 involved in this embodiment may have the following structure: an OTA chamber 50 having an internal space 51; and a DUT scanning control unit 16 and a DUT scanning mechanism 56, which drive the scanning of the DUT 100 in a manner that allows the orientation of the DUT 100 to change continuously within the internal space 51, and perform a sensitivity test on the OTA (Over The Air) measurement environment within the internal space 51, covering all directions of the object scanned by the DUT scanning control unit 16 and the DUT scanning mechanism 56.

[0164] According to this structure, the measuring device 1 involved in this embodiment is applicable to the dynamic control of the initial output level and step level of the test signal predicted by AI, even when it is necessary to measure the receiving sensitivity in all directions under the OTA environment. It can set the output level of the test signal to a testable level in a short time.

[0165] Furthermore, to address the aforementioned issues, the mobile terminal testing method of this embodiment uses a measurement device 1 with the aforementioned structure and performs a mobile terminal reception sensitivity test by repeatedly transmitting and receiving test signals between the NR system simulator 20 and the mobile terminal. The mobile terminal testing method is characterized by including: a setting step S1, which initially sets the test conditions, including the step level SL0 and output level OL0 of the test signal at the start of the reception sensitivity test; and an AI prediction test condition change setting step S1a, which uses artificial intelligence to perform correlation analysis between parameter data and measurement result data based on measurement log files from past reception sensitivity tests. The AI ​​prediction model generated by machine learning will set the initial output level as OL0 by replacing the output level OL0 with the AI ​​prediction output level OL01 predicted in advance by the AI ​​prediction model according to the test conditions set in the setting step; in the pass / fail judgment steps S4 and S5, the test signal is transmitted and received using the AI ​​prediction output level to perform the initial throughput measurement, and the pass / fail judgment is made by comparing the initial throughput measurement result with the judgment threshold; and in the step level change setting step, the step level SL01 predicted in advance by the AI ​​prediction model will be replaced with the step level SL0 and set as the step level for subsequent measurements based on the measurement result of the initial throughput measurement.

[0166] According to this structure, the mobile terminal testing method of this embodiment, by using the measuring device 1 with the above structure and applying this mobile terminal testing method, can quickly set the initial output level of the test signal to a value close to the end value of the measurement under the original nonlinear control by AI prediction. Compared with conventional methods such as nonlinear control and CP control, it can significantly reduce the number of measurements and set the output level of the test signal to a testable level in a very short time, thereby enabling effective testing of the receiving sensitivity of the mobile terminal.

[0167] Industrial availability

[0168] As described above, the mobile terminal testing apparatus and mobile terminal testing method of the present invention can minimize the number of measurements and set the output level of the test signal to a testable level in a shorter time, and can effectively conduct mobile terminal receiving sensitivity tests. It is effective in mobile terminal testing apparatus and measurement methods for conducting receiving sensitivity tests on mobile terminals with high-speed communication capabilities, such as 5G wireless terminals.

[0169] Symbol Explanation

[0170] 1-Measuring device (mobile terminal testing device), 16-DUT scanning control unit (scanning unit), 18-Receiver sensitivity test control unit (receiver sensitivity test execution unit, pass / fail judgment unit), 18a-Test condition setting unit (setting unit), 18b-AI prediction test condition change setting unit (AI prediction test condition change setting unit), 18e-Output level variable setting unit (output level setting unit), 51-Internal space, 56-DUT scanning mechanism (scanning unit), 100-DUT (test object, mobile terminal).

Claims

1. A mobile terminal testing device, comprising: Signal generator (20) generates test signals; and The receiver sensitivity test execution unit (18) performs a receiver sensitivity test on the mobile terminal by sending and receiving the test signal multiple times between the signal generator and the mobile terminal (100). The mobile terminal testing apparatus (1) for testing the test object, namely the mobile terminal, is characterized in that it has: The setting unit (18a) initially sets the test conditions, which include the step level SL0 and the output level OL0 of the test signal at the start of the receiver sensitivity test. The pass / fail judgment unit (18f) performs throughput measurement by transmitting and receiving the test signal under the test conditions, and judges whether the pass / fail status is determined by comparing the throughput measurement result with the judgment threshold. The output level setting unit (18e) sets the output level based on the pass / fail judgment result of a previous throughput measurement when the pass / fail judgment result is negative; and The AI ​​prediction test condition change setting unit (18b) holds an AI prediction model generated by performing machine learning on the correlation between parameter data and measurement result data based on the measurement log file of the past receiver sensitivity test. The AI ​​prediction output level OL01, which is predicted in advance by the AI ​​prediction model according to the test conditions set by the setting unit, is set as the initial output level instead of the output level OL0.

2. The mobile terminal testing apparatus according to claim 1, characterized in that, The AI ​​prediction test condition change setting unit will set the step level for subsequent measurements by replacing the step level SL0 with the AI ​​prediction step level SL01, which was predicted in advance by the AI ​​prediction model, based on the measurement result of the initial throughput measurement performed by transmitting and receiving the test signal with the AI ​​prediction output level.

3. The mobile terminal testing apparatus according to claim 2, characterized in that, When the result of the pass / fail judgment on the initial throughput measurement is determined to be no, then, limited to the second throughput measurement, the AI ​​prediction test condition change setting unit uses the AI ​​prediction step level SL01 to perform the output level update processing for the next measurement.

4. The mobile terminal testing apparatus according to claim 1, characterized in that, The AI ​​prediction test condition change setting unit specifies the loss of the external connection cable, the azimuth angle (θ, φ) of the mobile terminal, and the frequency of the measurement object as the parameter data, and holds the AI ​​prediction model generated by machine learning through the relationship between the parameter data and the measurement result data.

5. The mobile terminal testing apparatus according to claim 2, characterized in that, When the result of the second throughput measurement is determined to be unqualified, the receiver sensitivity test execution unit controls the output level setting unit to perform level reduction processing A, level reduction processing B, level boosting processing A, or level boosting processing B.

6. The mobile terminal testing apparatus according to claim 1, characterized in that, have: An anechoic chamber (50) having an internal space (51); and The scanning units (16, 56) drive the scanning of the mobile terminal within the internal space in a manner that allows the orientation of the mobile terminal to change continuously. The receiver sensitivity test is conducted in the internal space on all sides of the object being scanned by the scanning unit under the OTA (Over The Air) measurement environment.

7. A mobile terminal testing method, wherein the method uses a mobile terminal testing device (1) for testing the test object, i.e., the mobile terminal, and performs a receiving sensitivity test on the mobile terminal by repeatedly transmitting and receiving test signals between a signal generator and the mobile terminal, wherein the mobile terminal testing device (1) has: The signal generator (20) generates a test signal; and The receiver sensitivity test execution unit (18) performs a receiver sensitivity test on the mobile terminal by repeatedly transmitting and receiving the test signal between the signal generator and the mobile terminal (100). The mobile terminal testing method is characterized by including: The setting step (S1) involves initially setting the test conditions, which include the step level SL0 and output level OL0 of the test signal at the start of the receiver sensitivity test. In the AI ​​prediction test condition change setting step (S1a), an AI prediction model generated by performing machine learning on the correlation between parameter data and measurement result data based on the measurement log file of the past receiver sensitivity test is held. The AI ​​prediction output level OL01, which is predicted in advance by the AI ​​prediction model according to the test conditions set in the setting step, is set as the initial output level instead of the output level OL0. The pass / fail determination steps (S4, S5) involve performing an initial throughput measurement by transmitting and receiving the test signal using the AI ​​predicted output level, and comparing the initial throughput measurement result with a judgment threshold to determine whether the signal is pass / fail; and The step level change setting step is to replace the step level SL0 with the AI ​​predicted step level SL01, which is predicted in advance by the AI ​​prediction model based on the measurement results of the first throughput measurement, and set it as the step level for subsequent measurements.

Images