Model testing method and apparatus

Abstract

The present application provides a model testing method and apparatus. The method comprises: a testing equipment (TE) sends a first reference signal to a terminal device, the first reference signal being used by the terminal device to determine first channel state information (CSI); the TE receives a first precoding matrix indicator from a simulated device; and the TE tests the performance of a first compression model on the basis of the first precoding matrix indicator. The first precoding matrix indicator is determined by the simulated device on the basis of second CSI, the second CSI is obtained by the simulated device by decompressing third CSI by using a first decompression model, the third CSI is obtained by the terminal device by compressing the first CSI by using the first compression model, and the first compression model matches the first decompression model. Therefore, the TE does not need to decompress compressed CSI by using a decompression model, and can test the performance of the first compression model with low complexity.

Description

Model testing method and device

[0001] This application claims priority to the Chinese patent application filed with the State Intellectual Property Office of China on September 26, 2023, with application number 202311274639.4 and application name “A Model Testing Method and Device”, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The present application relates to the field of communication technology, and in particular to a model testing method and device. Background Art

[0003] Artificial intelligence (AI) or machine learning (ML) can be applied to enhance channel state information (CSI) compression feedback. For example, in an AI-CSI compression feedback enhancement scenario, the terminal device compresses the CSI using an AI-CSI compression model, and the network device recovers the CSI using an AI-CSI decompression model that matches the AI-CSI compression model used by the terminal device. In this AI-CSI compression feedback enhancement scenario, both the terminal device and the network device deploy AI models, resulting in an AI dual-end model.

[0004] To ensure the performance of communication systems, the performance of AI two-end models must be tested before they are deployed. Therefore, how to test the performance of two-end models remains an open research topic.

[0005] Summary of the Invention

[0006] The embodiments of the present application provide a model testing method and apparatus, which enable a testing device to test the performance of a two-terminal model with low complexity.

[0007] In a first aspect, embodiments of the present application provide a model testing method that can be performed by a test device. The test device herein may refer to the test device itself or to a processor, module, chip, or chip system within the test device that implements the method. In this method, a test device (TE) sends a first reference signal to a terminal device. The first reference signal is used by the terminal device to determine first channel state information (CSI). The TE receives a first precoding matrix indicator from a simulation device. Based on the first precoding matrix indicator, the TE tests the performance of a first compression model.

[0008] Among them, the first precoding matrix indicator is determined by the simulation device based on the second CSI, the second CSI is obtained by the simulation device decompressing the third CSI using the first decompression model, the third CSI is obtained by the terminal device compressing the first CSI using the first compression model, and the first compression model matches the first decompression model.

[0009] In this embodiment of the present application, the first precoding matrix indicator used by the TE to test the first compression model is determined by the simulation device based on the decompressed CSI (i.e., the second CSI), and the decompressed CSI is obtained by the simulation device decompressing the compressed CSI (i.e., the third CSI) using the first decompression model. Therefore, the TE does not need to use the decompression model to decompress the compressed CSI, nor does it need to calculate the precoding matrix indicator based on the decompressed CSI, and can test the performance of the first compression model with low complexity.

[0010] In addition, the first decompression model used by the simulation device to decompress the compressed CSI (i.e., the second CSI) matches the first compression model used by the terminal device to compress the uncompressed CSI (i.e., the first CSI), which can reduce the impact of the dual-end model mismatch on the test results.

[0011] In an optional embodiment, before the TE sends the first reference signal to the terminal device, the TE further performs the following steps: receiving a first message from the terminal device, the first message including X first identifiers, one of the X first identifiers being associated with a compression model supported by the terminal device, and X being a positive integer; sending a second message to the analog device, the second message including Y second identifiers, the Y second identifiers being part or all of the X first identifiers, or being determined by the TE based on part or all of the X first identifiers, and Y being an integer less than or equal to X; receiving a third message from the analog device, the third message including M third identifiers and / or confirmation of the response. Answer ACK, the M third identifiers are part or all of the second identifiers in the Y second identifiers, or are determined by the simulation device based on part or all of the second identifiers in the Y second identifiers, the compression model associated with one of the M third identifiers matches the decompression model supported by the simulation device, and M is an integer less than or equal to Y; a fourth message is sent to the terminal device, the fourth message includes a fourth identifier, a fourth identifier is one of the M third identifiers or the Y second identifiers, or is determined by the TE based on the M third identifiers or the Y second identifiers, the fourth identifier is related to the first compression model, and the first compression model matches the first decompression model.

[0012] As can be seen, before the TE performs the dual-end model test, it also determines a fourth identifier related to the first compression model to be tested based on the X first identifiers related to one or more compression models supported by the terminal device, and the matching relationship between the decompression model supported by the simulation device and the compression model supported by the terminal device, and sends the fourth identifier to the terminal device. The first compression model is one of the compression models supported by the terminal device, and the first compression model matches the first decompression model supported by the simulation device. Therefore, during the TE test of the first compression model, the simulation device can use the first decompression model that matches the first compression model to decompress the compressed CSI, which can reduce the impact of the dual-end model mismatch on the test results.

[0013] In an optional embodiment, the TE also sends a fifth message to the analog device, the fifth message including a fifth identifier, which is one of the M third identifiers or the Y second identifiers, or is determined by the TE based on one of the M third identifiers or the Y second identifiers, the fifth identifier matches the first decompression model, and the first decompression model matches the first compression model. This method allows the analog device to determine the first decompression model associated with the fifth identifier based on the fifth identifier, that is, to determine the first decompression model that matches the first compression model. This method can ensure that the first decompression model used by the analog device matches the first compression model used by the terminal device.

[0014] In an optional embodiment, one of the X first identifiers is used to identify one or more of the following: a compression model supported by the terminal device, parameter configuration information associated with a compression model supported by the terminal device, channel condition information associated with a compression model supported by the terminal device, state condition information associated with a compression model supported by the terminal device, information associated with the current state of the terminal device, and performance indicator information associated with a compression model supported by the terminal device.

[0015] In an optional implementation, the first decompression model is provided by a terminal device, or a network device, or other devices other than the terminal device and the network device.

[0016] In an optional embodiment, before the TE receives the first precoding matrix from the analog device, the TE also receives the third CSI from the terminal device and sends the third CSI to the analog device. This method allows the TE to confirm that the third CSI is the CSI compressed by the terminal device, and can ensure that the TE tests the performance of the first compression model based on the CSI (i.e., the third CSI) obtained by the terminal device through compression using the first compression model, thereby improving the accuracy and reliability of the test of the first compression model. In addition, this method can also enable the TE to confirm that the third CSI fed back by the terminal device complies with the compressed CSI feedback format configured by the TE.

[0017] On the second aspect, an embodiment of the present application also provides a model testing method, which can be performed by a simulation device. The simulation device here can refer to the simulation device itself, or to a processor, module, chip, or chip system that implements the method in the simulation device. In this method, the simulation device receives third channel state information CSI from the test device TE or the terminal device. The third CSI is obtained by the terminal device compressing the first CSI using a first compression model. The first CSI is determined by the terminal device based on the first reference signal sent by the test device TE. The simulation device decompresses the third CSI using a first decompression model to obtain a second CSI. The first decompression model matches the first compression model. The simulation device sends a first precoding matrix indicator to the TE. The first precoding matrix indicator is determined by the simulation device based on the second CSI. The first precoding matrix indicator is used by the TE to test the performance of the first compression model.

[0018] In an embodiment of the present application, the simulation device determines a first precoding matrix indicator for the TE to test the first compression model based on the second CSI. The second CSI is obtained by the simulation device decompressing the compressed CSI (i.e., the third CSI) using the first decompression model. This facilitates the TE to directly test the performance of the first compression model using the first precoding matrix indicator provided by the simulation device without having to use the decompression model to decompress the compressed CSI. This facilitates the TE to test the performance of the first compression model with low complexity.

[0019] In addition, the first decompression model used by the simulation device to decompress the compressed CSI (i.e., the third CSI) matches the first compression model used by the terminal device to compress the uncompressed CSI (i.e., the first CSI), thereby reducing the impact of the dual-end model mismatch on the test results.

[0020] In addition, if the decompression model used by the simulation device is provided by the manufacturer of the terminal device or network device, then in this test method, the simulation device does not need to directly send the output results of the model, thereby not exposing the output results of the model and the post-processing of the output results. It only needs to feedback the precoding matrix indicator related to the output results of the model for subsequent TE testing, which can protect the private implementation of the simulation device.

[0021] In an optional embodiment, before the simulation device receives the third CSI, the simulation device further performs: receiving a second message from the TE, the second message including Y second identifiers, the Y second identifiers being part or all of the X first identifiers, or being determined by the TE based on part or all of the X first identifiers, one of the X first identifiers is associated with a compression model supported by the terminal device, X is a positive integer, and Y is an integer less than or equal to X; sending a third message to the TE, the third message including M third identifiers and / or an acknowledgment ACK, the M third identifiers being part or all of the Y second identifiers, or being determined by the simulation device based on part or all of the Y second identifiers, the compression model associated with one of the M third identifiers matches the decompression model supported by the simulation device, and M is an integer less than or equal to Y.

[0022] It can be seen that before the simulation device receives the third CSI, it also feeds back M third identifiers and / or ACK to TE based on the Y second identifiers sent by TE and the decompression model supported by itself. The M third identifiers are M identifiers related to one or more compression models that match the decompression model supported by the simulation device, so that TE determines the first compression model to be tested.

[0023] In an optional embodiment, the simulation device further receives a fifth message from the TE, where the fifth message includes a fifth identifier. The fifth identifier is one of the M third identifiers or the Y second identifiers, or is determined by the TE based on the M third identifiers or the Y second identifiers. The fifth identifier is associated with the first decompression model, and the first decompression model matches the first compression model.

[0024] It can be seen that the simulation device also receives the fifth identifier related to the first compression model to be tested. Therefore, the simulation device can determine the first decompression model that matches the first compression model based on the fifth identifier.

[0025] In an optional embodiment, one of the X first identifiers is used to identify one or more of the following: a compression model supported by the terminal device, parameter configuration information associated with a compression model supported by the terminal device, channel condition information associated with a compression model supported by the terminal device, state condition information associated with a compression model supported by the terminal device, information associated with the current state of the terminal device, and performance indicator information associated with a compression model supported by the terminal device.

[0026] In an optional implementation, the simulation device uses the first decompression model to decompress the third CSI, and before obtaining the second CSI, also obtains the first decompression model from the terminal device, or the network device, or other devices other than the terminal device and the network device.

[0027] In a third aspect, embodiments of the present application further provide a model testing method that can be performed by a test device. The test device herein may refer to the test device itself or to a processor, module, chip, or chip system within the test device that implements the method. In this method, the test device (TE) sends a second reference signal to the simulation device. The second reference signal is used by the simulation device to determine fourth channel state information (CSI). The TE receives a second precoding matrix indicator from the network device. Based on the second precoding matrix indicator, the TE tests the performance of the second decompression model.

[0028] Among them, the second precoding matrix indicator is determined by the network device based on the fifth CSI, the fifth CSI is obtained by the network device decompressing the sixth CSI using the second decompression model, the sixth CSI is obtained by the simulation device compressing the fourth CSI using the second compression model, and the second compression model matches the second decompression model.

[0029] In this embodiment of the present application, the second precoding matrix indicator used by the TE to test the second decompression model is calculated after the network device decompresses the compressed CSI (i.e., the sixth CSI) using the second decompression model. The compressed CSI is obtained by the simulation device compressing the uncompressed CSI (i.e., the fourth CSI) using a second compression model that matches the second decompression model. As a result, the TE does not need to determine the uncompressed CSI based on the reference signal, nor does it need to compress the uncompressed CSI using a compression model. The TE can test the performance of the second decompression model with low complexity.

[0030] In addition, the second compression model used by the simulation device to compress the uncompressed CSI (i.e., the fourth CSI) matches the second decompression model used by the network device to decompress the compressed CSI (i.e., the sixth CSI), which can reduce the impact of dual-end model mismatch on the test results.

[0031] In an optional embodiment, before the TE sends the second reference signal to the analog device, it also performs: receiving a sixth message from the network device, the sixth message includes S sixth identifiers, one of the S sixth identifiers is related to a decompression model supported by the network device, and S is a positive integer; sending a seventh message to the analog device, the seventh message includes T seventh identifiers, the T seventh identifiers are part or all of the S sixth identifiers, or are determined by the TE based on part or all of the S sixth identifiers, and T is an integer less than or equal to S; receiving an eighth message from the analog device, the eighth message includes N eighth identifiers and / or a confirmation response ACK, the N eighth identifiers are part or all of the seventh identifiers among the T seventh identifiers, or are determined by the simulation device based on part or all of the seventh identifiers among the T seventh identifiers, the decompression model associated with one eighth identifier among the N eighth identifiers matches the compression model supported by the simulation device, and N is an integer less than or equal to T; a ninth message is sent to the network device, the ninth message includes a ninth identifier, a ninth identifier is one of the N eighth identifiers or the T seventh identifiers, or is determined by the TE based on the N eighth identifiers or one of the T seventh identifiers, the ninth identifier is associated with the second decompression model, and the second decompression model matches the second compression model.

[0032] As can be seen, before TE tests the decompression model, it also determines a ninth identifier associated with the second decompression model to be tested based on the S sixth identifiers associated with one or more compression models supported by the network device, and the matching relationship between the compression model supported by the simulation device and the decompression model supported by the network device, and sends the ninth identifier to the network device. The second decompression model is one of the decompression models supported by the network device, and the second decompression model matches the second compression model supported by the simulation device. Thus, during TE testing of the second decompression model, the simulation device can compress the CSI using the second compression model that matches the second decompression model, thereby reducing the impact of dual-end model mismatch on the test results.

[0033] In an optional embodiment, the TE also sends a tenth message to the analog device, the tenth message including a tenth identifier, which is one of the N eighth identifiers or T seventh identifiers, or is determined by the TE based on one of the N eighth identifiers or T seventh identifiers, the tenth identifier is associated with the second compression model, and the second compression model matches the second decompression model. This method enables the analog device to determine the second compression model associated with the tenth identifier based on the tenth identifier, that is, to determine the second compression model that matches the second decompression model. This method can ensure that the second compression model used by the analog device matches the second decompression model used by the network device.

[0034] In an optional embodiment, one of the S sixth identifiers is used to identify one or more of the following: a decompression model supported by the network device, parameter configuration information associated with a decompression model supported by the network device, channel condition information associated with a decompression model supported by the network device, status condition information associated with a decompression model supported by the network device, information associated with the current state of the network device, and performance indicator information associated with a decompression model supported by the network device.

[0035] In an optional implementation, the second compression model is provided by a terminal device, or a network device, or other devices other than the terminal device and the network device.

[0036] In one optional embodiment, before receiving the second precoding matrix indicator from the network device, the TE further performs the following steps: receiving the sixth CSI from the analog device; and sending the sixth CSI to the network device. This approach enables the TE to determine that the sixth CSI is CSI compressed by the analog device, ensuring that the TE tests the performance of the second decompression model based on the CSI (i.e., the sixth CSI) obtained by compression using the second compression model by the analog device, thereby improving the accuracy or reliability of the test of the second decompression model.

[0037] In a fourth aspect, an embodiment of the present application provides a model testing method, which can be performed by a simulation device. The simulation device here can refer to the simulation device itself, or to a processor, module, chip, or chip system that implements the method in the simulation device. In this method, the simulation device receives a second reference signal from the test device TE. The simulation device compresses the fourth channel state information CSI using a second compression model to obtain a sixth CSI. The fourth CSI is determined by the simulation device based on the second reference signal, and the second compression model matches the second decompression model used by the network device. The simulation device sends the sixth CSI to the TE or the network device, and the sixth CSI is used by the TE to test the performance of the second decompression model.

[0038] In this embodiment of the present application, the sixth CSI used by the TE to test the second decompression model is obtained by compressing the fourth CSI determined based on the second reference signal using the second compression model by the simulation device. This eliminates the need for the TE to determine uncompressed CSI based on the reference signal and compress the uncompressed CSI using the compression model. This allows the TE to test the performance of the second decompression model with low complexity.

[0039] In addition, the second compression model used by the simulation device to compress the uncompressed CSI (i.e., the fourth CSI) matches the second decompression model used by the network device to decompress the compressed CSI (i.e., the sixth CSI), thereby reducing the impact of the dual-end model mismatch on the test results.

[0040] In an optional embodiment, the simulation device also performs: receiving a seventh message from the TE, the seventh message including T seventh identifiers, the T seventh identifiers being part or all of the S sixth identifiers, or being determined by the TE based on part or all of the S sixth identifiers, one of the S sixth identifiers being associated with a decompression model supported by the network device, S being a positive integer, and T being an integer less than or equal to S; sending an eighth message to the simulation device, the eighth message including N eighth identifiers and / or an acknowledgment ACK, the N eighth identifiers being part or all of the T seventh identifiers, or being determined by the simulation device based on part or all of the T seventh identifiers, the decompression model associated with one of the N eighth identifiers matching the compression model supported by the simulation device, and N being an integer less than or equal to T.

[0041] It can be seen that the simulation device also feeds back N eighth identifiers and / or ACK related to the simulation device to TE based on the T seventh identifiers sent by TE. The N eighth identifiers are N identifiers related to one or more decompression models that match the compression model supported by the simulation device, so that TE determines the second decompression model to be tested.

[0042] In an optional embodiment, the simulation device also receives a tenth message from the TE, the tenth message including a tenth identifier, the tenth identifier being one of the N eighth identifiers or the T seventh identifiers, or being determined by the TE based on one of the N eighth identifiers or the T seventh identifiers, the tenth identifier being related to the second compression model, and the second compression model matching the second decompression model.

[0043] It can be seen that the simulation device also receives the tenth identifier related to the second decompression model to be tested. Therefore, the simulation device can determine the second compression model that matches the second decompression model based on the tenth identifier.

[0044] In an optional embodiment, one of the S sixth identifiers is used to identify one or more of the following: a decompression model supported by the network device, parameter configuration information associated with a decompression model supported by the network device, channel condition information associated with a decompression model supported by the network device, status condition information associated with a decompression model supported by the network device, information associated with the current state of the network device, and performance indicator information associated with a decompression model supported by the network device.

[0045] In an optional implementation, the simulation device compresses the fourth CSI using the second compression model, and before obtaining the sixth CSI, further obtains the second compression model from the terminal device, or the network device, or other devices other than the terminal device and the network device.

[0046] In a fifth aspect, an embodiment of the present application further provides a communication device. The communication device has the function of implementing some or all of the functions of the test device described in the first or third aspect above, or implementing some or all of the functions of the simulation device described in the second or fourth aspect above. For example, the function of the communication device may have the functions of some or all of the embodiments of the test device described in the first aspect of the embodiment of the present application, or may have the function of implementing any one of the embodiments of the present application alone. The functions may be implemented by hardware, or may be implemented by hardware executing corresponding software. The hardware or software includes one or more units or modules corresponding to the above functions.

[0047] In one possible design, the communication device may include a processing unit and a communication unit. The processing unit is configured to support the communication device in performing the corresponding functions in the above method. The communication unit is configured to support communication between the communication device and other communication devices. The communication device may also include a storage unit, which is coupled to the processing unit and the communication unit and stores program instructions and data necessary for the communication device.

[0048] In one embodiment, the communication device includes: a processing unit and a communication unit, and the device is applied to a test device;

[0049] The communication unit is used to send a first reference signal to a terminal device, where the first reference signal is used by the terminal device to determine first channel state information CSI; the communication unit is also used to receive a first precoding matrix indicator from an analog device, where the first precoding matrix indicator is determined by the analog device based on a second CSI, where the second CSI is obtained by the analog device decompressing a third CSI using a first decompression model, where the third CSI is obtained by the terminal device compressing the first CSI using a first compression model, where the first compression model matches the first decompression model; and the processing unit is used to test the performance of the first compression model based on the first precoding matrix indicator.

[0050] In addition, in this aspect, other optional implementations of the communication device can refer to the relevant content of the first aspect above and will not be described in detail here.

[0051] In another embodiment, the communication device includes: a processing unit and a communication unit, and the device is applied to a simulation device;

[0052] The communication unit is used to receive third channel state information CSI from a test device TE or a terminal device, where the third CSI is obtained by the terminal device compressing the first CSI using a first compression model, and the first CSI is determined by the terminal device based on a first reference signal sent by the TE; the processing unit is used to decompress the third CSI using a first decompression model to obtain a second CSI, and the first decompression model matches the first compression model; the communication unit is also used to send a first precoding matrix indicator to the TE, where the first precoding matrix indicator is determined by the simulation device based on the second CSI, and the first precoding matrix indicator is used by the TE to test the performance of the first compression model.

[0053] In addition, in this aspect, other optional implementations of the communication device can refer to the relevant content of the second aspect above and will not be described in detail here.

[0054] In yet another embodiment, the communication device comprises: a processing unit and a communication unit, and the device is applied to a test device;

[0055] The communication unit is used to send a second reference signal to the analog device, where the second reference signal is used by the analog device to determine fourth channel state information CSI; the communication unit is also used to receive a second precoding matrix indicator from the network device, where the second precoding matrix indicator is determined by the network device based on the fifth CSI, the fifth CSI is obtained by the network device decompressing the sixth CSI using a second decompression model, and the sixth CSI is obtained by the analog device compressing the fourth CSI using a second compression model, where the second compression model matches the second decompression model; the processing unit is used to test the performance of the second decompression model based on the second precoding matrix indicator.

[0056] In addition, in this aspect, other optional implementations of the communication device can refer to the relevant content of the third aspect above and will not be described in detail here.

[0057] In yet another embodiment, the communication apparatus comprises: a processing unit and a communication unit, wherein the apparatus is applied to a simulation device;

[0058] The communication unit is used to receive a second reference signal from the test equipment TE; the processing unit is used to compress the fourth channel state information CSI using the second compression model to obtain the sixth CSI, where the fourth CSI is determined by the simulation device based on the second reference signal, and the second compression model matches the second decompression model used by the network device; the communication unit is also used to send the sixth CSI to the TE or the network device, and the sixth CSI is used by the TE to test the performance of the second decompression model.

[0059] In addition, in this aspect, other optional implementations of the communication device can refer to the relevant content of the fourth aspect above and will not be described in detail here.

[0060] As an example, the communication unit may be a transceiver or a communication interface, the storage unit may be a memory, and the processing unit may be a processor.

[0061] In one embodiment, the communication device includes: a processor and a transceiver, and the device is applied to a test device;

[0062] The transceiver is used to send a first reference signal to a terminal device, where the first reference signal is used by the terminal device to determine first channel state information CSI; the transceiver is also used to receive a first precoding matrix indicator from an analog device, where the first precoding matrix indicator is determined by the analog device based on a second CSI, where the second CSI is obtained by the analog device decompressing a third CSI using a first decompression model, where the third CSI is obtained by the terminal device compressing the first CSI using a first compression model, where the first compression model matches the first decompression model; and the processor is used to test the performance of the first compression model based on the first precoding matrix indicator.

[0063] In addition, in this aspect, other optional implementations of the communication device can refer to the relevant content of the first aspect above and will not be described in detail here.

[0064] In another embodiment, the communication device includes: a processor and a transceiver, and the device is applied to an analog device;

[0065] The transceiver is used to receive third channel state information CSI from a test device TE or a terminal device, where the third CSI is obtained by the terminal device compressing the first CSI using a first compression model, and the first CSI is determined by the terminal device based on a first reference signal sent by the TE; the processor is used to decompress the third CSI using a first decompression model to obtain a second CSI, and the first decompression model matches the first compression model; the transceiver is also used to send a first precoding matrix indicator to the TE, where the first precoding matrix indicator is determined by the simulation device based on the second CSI, and the first precoding matrix indicator is used by the TE to test the performance of the first compression model.

[0066] In addition, in this aspect, other optional implementations of the communication device can refer to the relevant content of the second aspect above and will not be described in detail here.

[0067] In yet another embodiment, the communication device comprises: a processor and a transceiver, wherein the device is applied to a test device;

[0068] The transceiver is used to send a second reference signal to the analog device, where the second reference signal is used by the analog device to determine fourth channel state information CSI; the transceiver is also used to receive a second precoding matrix indicator from the network device, where the second precoding matrix indicator is determined by the network device based on the fifth CSI, the fifth CSI is obtained by the network device decompressing the sixth CSI using a second decompression model, and the sixth CSI is obtained by the analog device compressing the fourth CSI using a second compression model, where the second compression model matches the second decompression model; the processor is used to test the performance of the second decompression model based on the second precoding matrix indicator.

[0069] In addition, in this aspect, other optional implementations of the communication device can refer to the relevant content of the third aspect above and will not be described in detail here.

[0070] In yet another embodiment, the communication device comprises: a processor and a transceiver, wherein the device is applied to an analog device;

[0071] The transceiver is used to receive a second reference signal from the test equipment TE; the processor is used to compress the fourth channel state information CSI using a second compression model to obtain sixth CSI, where the fourth CSI is obtained by the simulation device based on the second reference signal, and the second compression model matches the second decompression model used by the network device; the transceiver is also used to send the sixth CSI to the TE or the network device, and the sixth CSI is used by the TE to test the performance of the second decompression model.

[0072] In addition, in this aspect, other optional implementations of the communication device can refer to the relevant content of the fourth aspect above and will not be described in detail here.

[0073] In another embodiment, the communication device is a chip or a chip system. The processing unit may also be embodied as a processing circuit or a logic circuit; and the communication unit may be an input / output interface, an interface circuit, an output circuit, an input circuit, a pin, or related circuits on the chip or chip system.

[0074] During implementation, the processor can be used to perform, for example, but not limited to, baseband-related processing, and the transceiver can be used to perform, for example, but not limited to, radio frequency transceiver processing. The aforementioned devices can be provided on separate chips, or at least partially or entirely on the same chip. For example, the processor can be further divided into an analog baseband processor and a digital baseband processor. The analog baseband processor can be integrated with the transceiver on the same chip, while the digital baseband processor can be provided on a separate chip. With the continuous development of integrated circuit technology, more and more devices can be integrated on the same chip. For example, a digital baseband processor can be integrated with multiple application processors (such as, but not limited to, a graphics processor, a multimedia processor, etc.) on the same chip. Such a chip can be called a system on a chip (SoC). Whether each device is provided independently on different chips or integrated on one or more chips often depends on the needs of the product design. The embodiments of the present application do not limit the implementation form of the aforementioned devices.

[0075] In a sixth aspect, an embodiment of the present application further provides a processor for executing the various methods described above. In the process of executing these methods, the process of sending the above-mentioned information and receiving the above-mentioned information in the above-mentioned methods can be understood as the process of the processor outputting the above-mentioned information and the process of the processor receiving the above-mentioned information input. When outputting the above-mentioned information, the processor outputs the above-mentioned information to the transceiver so that the transceiver can transmit it. After being output by the processor, the above-mentioned information may undergo other processing before reaching the transceiver. Similarly, when the processor receives the above-mentioned information input, the transceiver receives the above-mentioned information and inputs it into the processor. Furthermore, after the transceiver receives the above-mentioned information, the above-mentioned information may undergo other processing before being input into the processor.

[0076] For the sending and receiving operations involved in the processor, unless otherwise specified, or unless they conflict with their actual functions or internal logic in the relevant descriptions, they can be more generally understood as processor output, reception, input and other operations, rather than sending and receiving operations directly performed by the RF circuit and antenna.

[0077] During implementation, the processor may be a processor specifically configured to execute these methods, or may be a processor that executes computer instructions in a memory to execute these methods, such as a general-purpose processor. The memory may be a non-transitory memory, such as a read-only memory (ROM), which may be integrated with the processor on the same chip or disposed on separate chips. The embodiments of the present application do not limit the type of memory or the configuration of the memory and the processor.

[0078] In a seventh aspect, an embodiment of the present application further provides a communication system, which may include a test device and a simulation device. In another possible design, the system may further include other devices that interact with the test device and the simulation device, such as a device to be tested.

[0079] In an eighth aspect, an embodiment of the present application provides a computer-readable storage medium for storing instructions, which, when executed on a computer, implements the method described in any one of the first to fourth aspects above.

[0080] In a ninth aspect, an embodiment of the present application further provides a computer program product comprising instructions, which, when executed on a computer, implements the method described in any one of the first to fourth aspects above.

[0081] In the tenth aspect, an embodiment of the present application provides a chip system, which includes a processor and an interface, wherein the interface is used to obtain a program or instruction, and the processor is used to call the program or instruction to implement or support the test equipment to implement the functions involved in the first or third aspect, or to implement or support the simulation equipment to implement the functions involved in the second or fourth aspect. For example, determine or process at least one of the data and information involved in the above method. In one possible design, the chip system also includes a memory, which is used to store program instructions and data necessary for the terminal. The chip system can be composed of chips, or it can include chips and other discrete devices.

[0082] In the eleventh aspect, an embodiment of the present application provides a communication device, comprising a processor for executing a computer program or executable instructions stored in a memory, so that when the computer program or executable instructions are executed, the device executes a method in any possible implementation of any one of the first to fourth aspects.

[0083] In one possible implementation, the processor and memory are integrated;

[0084] In another possible implementation, the memory is located outside the communication device.

[0085] The beneficial effects of the fifth to eleventh aspects can refer to the beneficial effects of the first to fourth aspects and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] FIG1 is a schematic diagram of a system architecture provided in an embodiment of the present application;

[0087] FIG2 is an interactive schematic diagram of a model testing method provided in an embodiment of the present application;

[0088] FIG3 is an interactive diagram of another model testing method provided in an embodiment of the present application;

[0089] FIG4 is an interactive diagram of another model testing method provided in an embodiment of the present application;

[0090] FIG5 is an interactive diagram of another model testing method provided in an embodiment of the present application;

[0091] FIG6 is a schematic structural diagram of a communication device provided in an embodiment of the present application;

[0092] FIG7 is a schematic structural diagram of another communication device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0093] The technical solutions in the embodiments of the present application are described clearly and completely below in conjunction with the drawings in the embodiments of the present application.

[0094] In order to better understand the model testing method disclosed in the embodiment of the present application, the system architecture applicable to the embodiment of the present application is described.

[0095] Please refer to Figure 1, which shows a system architecture. As shown in Figure 1, this system architecture includes test equipment (TE), a device under test (DUT), and a simulation device. The DUT can be either a terminal device or a network device, while the TE is a different device from the terminal and network devices. Transmission between the TE and the DUT can occur via a transmission medium such as radio waves or wires.

[0096] In addition, the simulation device can simulate the behavior of the terminal device during the dual-end model test. For example, the simulation device simulates the behavior of the terminal device using the compression model to compress uncompressed channel state information (CSI) during the dual-end model test. Optionally, the simulation device can also simulate the behavior of the network device during the dual-end model test. For example, the simulation device simulates the behavior of the network device using the decompression model to decompress compressed CSI during the dual-end model test. The embodiments of the present application do not limit the specific behavior of the simulation device.

[0097] It is understandable that the simulation device may be a transcoder, which has the function of simulating the behavior of the terminal device during the two-end model test and may also have the function of simulating the behavior of the network device during the two-end model test.

[0098] The system architecture shown in Figure 1 can be applied to long-term evolution (LTE) systems, fifth-generation (5G) mobile communication systems, sixth-generation (6G) mobile communication systems, and other systems evolving beyond 6G, as well as narrowband Internet of Things (NB-IoT) systems and wireless fidelity (WiFi) systems.

[0099] In an embodiment of the present application, the network device is a device with wireless transceiver functions, which is used to communicate with a terminal device. It can be an evolved base station (evolved Node B, eNB or eNodeB) in LTE, or a base station in a 5G / 6G network, or a base station in a future evolved public land mobile network (public land mobile network, PLMN), a broadband network service gateway (BNG), an aggregation switch, or a non-third generation partnership project (3GPP) access device, etc.

[0100] Optionally, the network equipment may include various forms of base stations, such as: macro base stations, micro base stations (also known as small stations), relay stations, access points, devices that realize base station functions in the future, access points (APs) in WiFi systems, transmitting and receiving points (TRPs), transmitting points (TPs), mobile switching centers, and devices that perform base station functions in device-to-device (D2D), vehicle-to-everything (V2X), and machine-to-machine (M2M) communications, devices that realize base station functions in communication systems evolved after 5G, integrated access and backhaul (IAB), and may also include centralized units (CUs) and distributed units (DUs) in cloud radio access network (C-RAN) systems, network equipment in non-terrestrial network (NTN) communication systems, that is, they can be deployed on high-altitude platforms or satellites, and may also be various devices that constitute access nodes, such as active antenna processing units (AAPs). unit (AAU), baseband unit (BBU), etc., without limitation.

[0101] Network devices can communicate and interact with core network devices to provide communication services to terminal devices. Core network devices are, for example, devices in the 5G core network (CN). As a bearer network, the core network provides an interface to the data network, providing communication connections, authentication, management, policy control, and data service delivery for terminals.

[0102] In the embodiments of the present application, the terminal device may include various handheld devices, vehicle-mounted devices, wearable devices, computing devices, or other processing devices connected to a wireless modem with wireless communication capabilities. The terminal device may also be referred to as a terminal.

[0103] Terminal equipment may also refer to user equipment (UE), access terminal, subscriber unit, user agent, cellular phone, smart phone, wireless data card, personal digital assistant (PDA), tablet computer, wireless modem, handheld device, laptop computer, smart point of sale (POS), customer-premises equipment (CPE), machine type communication (MTC) terminal, communication equipment carried by high-altitude aircraft, wearable device, drone, robot, terminal in D2D, terminal in vehicle to everything (V2X), virtual reality (VR) terminal device, augmented reality (AR) terminal device, wireless terminal in industrial control, wireless terminal in self-driving, wireless terminal in remote medical, wireless terminal in smart grid, wireless terminal in transportation security, etc. Safety), wireless terminals in smart cities, wireless terminals in smart homes, or terminal devices in future communication networks, etc., are not restricted.

[0104] The embodiments disclosed in this application can be applied to the system architecture shown in Figure 1. The embodiments of this application will present various aspects, embodiments, or features of this application around the system. It should be understood that the system architecture shown in Figure 1 can include other devices, components, modules, etc., without limitation.

[0105] The 3rd Generation Partnership Project (3GPP) has proposed applying artificial intelligence (AI) or machine learning (ML) to air interface enhancement, including CSI compression feedback enhancement. In this scenario, the AI ​​model is deployed simultaneously in both the terminal device and the network equipment, creating a dual-end AI model.

[0106] There are three training methods for AI dual-end models: the first training method is Type 1, in which one end trains the AI ​​dual-end model, that is, the terminal device or network device trains the AI ​​dual-end model; the second and third training methods are Type 2 and Type 3, in which both the terminal device and the network device participate in the training of the model, and each trains its own model. In Type 1, the terminal device and the network device need to interact to complete the intermediate results of the training in order to achieve the purpose of matching the AI ​​dual-end model. In Type 2, the terminal device and the network device exchange real-time information (including intermediate quantities such as gradients) during the training process. In Type 3, the terminal device and the network device exchange the training data of the AI ​​dual-end model, that is, the input and output when training their respective AI models. It can be seen that compared with Type 3, Type 2 requires significantly more data to be exchanged during the model training process than Type 3.

[0107] Furthermore, 3GPP Radio Access Network (RAN) 4 is currently discussing testing methods for the AI-CSI dual-end model. When testing the AI-CSI compression model of a terminal device, the TE simulates the behavior of the network device and requires the use of the AI-CSI decompression model to decompress the compressed CSI.

[0108] Currently, in non-AI CSI feedback tests, the test metric is relative throughput. That is, the TE receives the CSI fed back by the terminal device and determines the precoding matrix for downlink data transmission based on the CSI. It then performs downlink data transmission based on the precoding matrix to obtain throughput T1. The TE uses a random precoding matrix for downlink data transmission to obtain throughput T2. The TE verifies whether the increase in T1 relative to T2 meets the established indicators. If the increase in T1 relative to T2 meets the established indicators, it indicates that the CSI fed back by the UE has passed the test. For example, if the established indicator is 10%, T1 is greater than T2, and the difference between T1 and T2 is greater than 10%, the CSI fed back by the UE has passed the test.

[0109] In this non-AI CSI feedback test, the TE directly uses the CSI fed back by the terminal device to calculate the precoding matrix. However, in the AI-CSI compressed feedback test, the TE must use the AI-CSI decompression model to decompress the compressed CSI fed back by the terminal device to obtain the CSI used to calculate the precoding matrix. Therefore, the non-AI CSI feedback test method lacks a method to determine the TE's AI-CSI decompression model.

[0110] In addition, based on the different ways TE determines the decompression model, TE can also test the performance of the AI-CSI compression model of the terminal device using the following three test methods:

[0111] Test method 1: TE tests the performance of the terminal device's AI-CSI compression model based on a predefined AI-CSI decompression model.

[0112] The terminal device compresses the CSI using the AI-CSI compression model and feeds it back to the TE. The TE receives the CSI from the terminal device and decompresses it using the predefined AI-CSI decompression model. Based on the decompressed CSI, the TE calculates the precoding matrix for downlink data transmission. The TE uses the calculated precoding matrix for downlink data transmission, achieving a throughput of T1. The TE uses a random precoding matrix for downlink data transmission, achieving a throughput of T2. The TE verifies whether the increase in T1 relative to T2 meets the specified criteria. If the increase in T1 relative to T2 meets the specified criteria, the AI-CSI compression model is considered to have passed the test.

[0113] The decompression model used by the TE is predefined, for example, by a protocol, such as RAN4. In one approach, RAN4 defines all information about the AI-CSI decompression model executed by the TE, which can be understood to include the model structure and model parameters. In another approach, RAN4 defines partial information about the AI-CSI decompression model executed by the TE, which can be understood as the model structure. That is, the parameters of the AI-CSI decompression model still need to be further determined. For example, the TE determines the parameters of the AI-CSI decompression model using the AI-CSI compression model provided by the AI-CSI compression model manufacturer or the AI-CSI decompression model provided by the AI-CSI decompression model manufacturer.

[0114] However, on the one hand, there may be significant differences between the predefined decompression model and the decompression model actually deployed on the network, resulting in test results that cannot accurately evaluate the actual performance of the DUT. For example, during the test, the throughput gain achieved by the terminal device meets the performance indicators. However, the predefined AI-CSI decompression model does not match the AI-CSI decompression model actually deployed on the network equipment, that is, it is impossible to achieve a throughput gain consistent with the test in the actual network, and even a negative gain is obtained. For another example, during the test, the throughput gain achieved by the terminal device does not meet the performance indicators. However, the predefined AI-CSI decompression model matches the AI-CSI decompression model deployed on the network equipment, that is, in the actual network, a throughput gain that meets the performance indicators can be achieved, and even a throughput gain that exceeds the performance indicators can be achieved.

[0115] On the other hand, the throughput gain of the terminal device depends not only on the degree of match between the terminal device's AI-CSI compression model and the TE's AI-CSI decompression model, but is also limited by the TE's decompression model design. Predefined limited model structures may not be able to take into account the requirements of terminal devices of different specifications for the TE's decompression model performance. For example, the AI-CSI compression model of a high-specification UE is relatively complex and can be adapted to the AI-CSI decompression model of actually deployed, equally complex network equipment, that is, a higher throughput gain can be obtained in the actual network. However, if the TE's AI-CSI decompression model is relatively simple, then even if the high-specification terminal device is loaded with a more complex AI model, its throughput gain advantage cannot be highlighted in the performance comparison of the test.

[0116] Test method 2: TE tests the performance of the AI-CSI compression model of the terminal device based on the AI-CSI decompression model provided by the manufacturer of the AI-CSI compression model (i.e., the manufacturer of the terminal device).

[0117] The process for TE to test the performance of the AI-CSI compression model of a terminal device based on the AI-CSI decompression model provided by the terminal device manufacturer can be found in the above-mentioned test method 1, in which the process for TE to test the performance of the AI-CSI compression model of a terminal device based on the predefined AI-CSI decompression model is not repeated here.

[0118] However, if the terminal device manufacturer provides the AI-CSI decompression model used by the TE, the TE will need to have the AI-CSI decompression model provided by each terminal device manufacturer under test, or even different models of terminal devices from the same manufacturer, which will significantly increase the complexity of the TE equipment. Furthermore, if the terminal device provides the TE with an AI-CSI decompression model that matches its own decompression model, there is also the risk of AI model technology leakage.

[0119] Test method 3: TE tests the performance of the AI-CSI compression model on the terminal device based on the AI-CSI decompression model provided by the AI-CSI decompression model vendor (i.e., the network equipment manufacturer).

[0120] The process for TE to test the performance of the AI-CSI compression model of a terminal device based on the AI-CSI decompression model provided by the AI-CS network device manufacturer can be found in Test Method 1 above, where the process for TE to test the performance of the AI-CSI compression model of a terminal device based on the predefined AI-CSI decompression model is described here.

[0121] However, if the network equipment manufacturer provides the AI-CSI decompression model used by TE, TE will need to be able to execute the AI-CSI decompression model provided by the manufacturer of each network device to be tested, or even different models of network equipment from the same manufacturer, which will significantly increase the complexity of the TE equipment. In addition, the terminal equipment to be tested needs to pass docking tests with each network equipment manufacturer, which greatly increases the test complexity, test time, and test cost. In addition, if the network equipment provides the AI-CSI decompression model used in the actual network to TE, there is also the risk of AI model technology leakage.

[0122] In an embodiment of the present application, when the device under test is a network device, the simulation device simulates the behavior of a terminal device, such as simulating the terminal device using a compression model to compress CSI. Optionally, when the device under test is a terminal device, the simulation device simulates the behavior of a network device, such as simulating the network device using a decompression model to decompress compressed CSI, and calculating the decompressed CSI to obtain a precoding matrix indicator (PMI) for testing the compression model.

[0123] In the embodiment of the present application, the analog device may also be referred to as a first device, a second device, etc. The embodiment of the present application does not limit the naming of the analog device.

[0124] In the embodiment of the present application, the compression model matches the decompression model, which can be understood as, when the compression model matches the decompression model, the communication system can obtain better performance. For example, when a terminal device communicates with a network device, the terminal device uses compression model a to compress the CSI and feeds the compressed CSI back to the network device. The network device uses decompression model b to decompress the compressed CSI to obtain decompressed CSI. The network device then obtains a precoding matrix for downlink data transmission based on the decompressed CSI. If the compression model a matches the decompression model, the network device performs downlink data transmission based on the determined precoding matrix and can obtain better throughput. Similarly, the decompression model matches the compression model, which can also be understood as, when the decompression model matches the compression model, the communication system can obtain better performance.

[0125] The embodiments of the present application are applicable to the test scenario of the AI-CSI two-terminal model. Optionally, the embodiments of the present application are also applicable to the test scenarios of other two-terminal models. For example, the embodiments of the present application are also applicable to the test scenario of the two-terminal model for nonlinear processing. It is understandable that the embodiments of the present application are described using the test scenario of the AI-CSI two-terminal model as an example.

[0126] In the embodiments of the present application, "related" can also be replaced with "associated." For example, "one of the X first identifiers is related to a compression model supported by the terminal device" can be replaced with "one of the X first identifiers is associated with a compression model supported by the terminal device." For another example, "the fourth identifier is related to the first compression model" can be replaced with "the fourth identifier is associated with the first compression model."

[0127] An embodiment of the present application provides a model testing method 100. In the model testing method 100, a test device sends a first reference signal to a terminal device. The terminal device compresses a first CSI determined based on the first reference signal using a first compression model to obtain a third CSI, and then sends the third CSI. A simulation device decompresses the third CSI using a first decompression model to obtain a second CSI. The simulation device obtains a first precoding matrix indicator based on the second CSI and sends the first precoding matrix indicator to the test device. The test device tests the performance of the first compression model based on the first precoding matrix indicator. The first decompression model matches the first compression model. Thus, in the process of testing the first compression model by the test device, the test device does not need to use the decompression model to decompress the compressed CSI, nor does it need to calculate the PMI based on the decompressed CSI. This allows the test device to test the performance of the first compression model with low complexity. Furthermore, in this test process, the first decompression model used by the simulation device matches the first compression model used by the terminal device, which can reduce the impact of dual-end model mismatch on the test results.

[0128] The embodiment of the present application also proposes a model testing method 200 for testing a decompression model. In the model testing method 200, a test device sends a second reference signal to a simulation device. The simulation device compresses the fourth CSI determined based on the second reference signal using a second compression model to obtain a sixth CSI, and sends the sixth CSI. The network device decompresses the sixth CSI using a second decompression model to obtain a fifth CSI. The network device determines a second precoding matrix indicator based on the fifth CSI and sends the second precoding matrix indicator to the test device. The test device tests the performance of the second decompression model based on the second precoding matrix indicator. The second compression model matches the second decompression model. It can be seen that in the process of the test device testing the second decompression model, the test device does not need to use the compression model to compress the uncompressed CSI, but can directly use the second precoding matrix indicator provided by the simulation device to test the performance of the first compression model, so that the test device can test the performance of the second decompression model with low complexity. Furthermore, in this test process, the second compression model used by the simulation device matches the second decompression model used by the network device, reducing the impact of dual-end model mismatches on test results. Furthermore, if the decompression model used by the simulation device is provided by the manufacturer of the terminal device or network device, this test method eliminates the need for the simulation device to directly transmit the model's output, thereby preventing exposure to the model's output and its post-processing. Instead, the simulation device only needs to provide feedback related to the model's output, a precoding matrix indicator, for subsequent TE testing. This protects the proprietary implementation of the simulation device.

[0129] FIG2 is an interactive diagram of a model testing method 100 provided in an embodiment of the present application. The model testing method 100 is described from the perspective of the interaction between the test device, the simulation device, and the terminal device. The model testing method 100 includes but is not limited to the following steps:

[0130] S201. The test device sends a first reference signal to the terminal device. Correspondingly, the terminal device receives the first reference signal from the test device.

[0131] The first reference signal may be a channel state information reference signal (CSI-RS), a demodulation reference signal (DMRS), a sounding reference signal (SRS), or the like.

[0132] The test device sends a first reference signal to the terminal device, and the first reference signal is used by the terminal device to determine the first CSI. Therefore, this method facilitates the terminal device to determine the first CSI based on the first reference signal, and further facilitates the terminal device to compress the first CSI using a compression model to obtain compressed CSI.

[0133] S202. The terminal device determines a first CSI based on the first reference signal.

[0134] S203. The terminal device compresses the first CSI using the first compression model to obtain a third CSI.

[0135] It is understandable that the terminal device determines the first CSI based on the received first reference signal, for example, by performing channel estimation based on the first reference signal to obtain the first CSI. The terminal device compresses the first CSI using the first compression model to obtain the third CSI. Therefore, the third CSI is the CSI after the terminal device compresses the first CSI, and the first CSI is determined by the terminal device based on the first reference signal. The first compression model is the compression model of the TE test.

[0136] S204a. The terminal device sends the third CSI to the analog device.

[0137] It is understandable that the terminal device directly sends the third CSI to the analog device so that the analog device can decompress the third CSI.

[0138] S204b1. The terminal device sends the third CSI to the test device.

[0139] S204b2. The test device sends the third CSI to the simulation device.

[0140] It is understandable that the terminal device executes S204a or S204b1. When the terminal device executes S204b1, the test device also executes S204b2.

[0141] It is understandable that the terminal device sends the third CSI to the test device, and the test device then sends the third CSI to the simulation device. This method is beneficial for the test device to confirm whether the third CSI is the CSI compressed by the terminal device, or the test device to confirm whether the terminal device compresses the first CSI determined based on the first reference signal in accordance with the test requirements. This method can reduce the risk of cheating between the terminal device and the simulation device during the test. For example, in order to ensure that the performance of the first compression model can pass the test, the terminal device directly sends the uncompressed first CSI to the simulation device, resulting in the test result not being obtained based on the first compression model, affecting the accuracy and reliability of the test result. In addition, this method can also enable the test device to determine whether the third CSI fed back by the terminal device complies with the CSI feedback format configured by TE.

[0142] S205. The simulation device decompresses the third CSI using the first decompression model to obtain the second CSI, and the first decompression model matches the first compression model.

[0143] In an optional implementation, the simulation device decompresses the third CSI using the first decompression model and, before obtaining the second CSI, further receives the third CSI from the device under test. The third CSI is obtained by the terminal device compressing the first CSI using the first compression model, and the first CSI is determined by the terminal device based on the first reference signal from the test device.

[0144] After obtaining the third CSI, the simulation device decompresses the third CSI using the first decompression model to obtain the second CSI. The first decompression model used by the simulation device matches the first compression model, that is, the first decompression model used by the simulation device matches the first compression model being tested. This reduces the impact of dual-end model mismatch on test results during the test. The method for determining whether the first decompression model matches the first compression model can be found in S301 to S304 below.

[0145] In an optional implementation, the first decompression model used by the simulation device is provided by the terminal device, or the network device, or other devices other than the terminal device and the network device (ie, a third party).

[0146] In an optional implementation, the terminal device, or the network device, or the third party directly sends the first decompression model to the simulation device. When the terminal device sends the first decompression model to the simulation device, the terminal device and the simulation device belong to the same manufacturer; when the network device sends the first decompression model to the simulation device, the network device and the simulation device belong to the same manufacturer; when the third party sends the first decompression model to the simulation device, the third party and the simulation device belong to the same manufacturer. This method helps reduce the risk of the first decompression model being leaked. For example, this method helps reduce the risk of the model privacy technology being leaked when the terminal device or network device sends the first decompression model to the test device.

[0147] In another optional implementation, the manufacturer of the simulation device provides the first decompression model for the simulation device, and the manufacturer of the simulation device is the same as the manufacturer of the terminal device, or the manufacturer of the simulation device is the same as the manufacturer of the network device, or the manufacturer of the simulation device is the same as the manufacturer of a third party. Thus, the first decompression model used by the simulation device can be understood as being provided by the terminal device, the network device, or a third party. This approach can also reduce the risk of the first decompression model being leaked.

[0148] S206. The simulation device determines a first precoding matrix indicator according to the second CSI.

[0149] It is understandable that the simulation device calculates the second CSI to obtain the first precoding matrix. The indicator used to indicate the first precoding matrix is ​​the first precoding matrix indicator. Therefore, the simulation device determines the first precoding matrix indicator based on the second CSI.

[0150] S207: The simulation device sends a first precoding matrix indicator to the test device. Correspondingly, the test device receives the first precoding matrix indicator from the simulation device.

[0151] S208. The testing device tests the performance of the first compression model based on the first precoding matrix indicator.

[0152] In an optional embodiment, the testing device tests the performance of the first compression model based on the first precoding matrix indicator, including: performing downlink data transmission based on the first precoding matrix indicated by the first precoding matrix indicator to obtain a first throughput of the terminal device; performing downlink data transmission based on the random precoding matrix indicated by the random precoding matrix indicator to obtain a second throughput of the terminal device; and testing the performance of the first compression model based on the first throughput and the second throughput.

[0153] The test device performs downlink data transmission based on the first precoding matrix indicated by the first precoding matrix indicator, which means that the test device sends downlink data to the terminal device based on the first precoding matrix. Similarly, the test device performs downlink data transmission based on the random precoding matrix indicated by the random precoding matrix indicator, which means that the test device sends downlink data to the terminal device based on the random precoding matrix.

[0154] In addition, the testing device tests the performance of the first compression model based on the first throughput and the second throughput, including: determining that the performance of the first compression model passes the test when the first throughput is greater than or equal to the second throughput and the difference between the first throughput and the second throughput is greater than or equal to a first preset value; and determining that the performance of the first compression model fails the test when the first throughput is greater than or equal to the second throughput and the difference between the first throughput and the second throughput is less than the first preset value, or when the first throughput is less than the second throughput. The first preset value can be pre-set or configured.

[0155] Optionally, the testing device tests the performance of the first compression model based on the first throughput and the second throughput, including: when the ratio of the first throughput to the second throughput is greater than or equal to a second preset value, determining that the performance of the first compression model passes the test, i.e., the first model passes the test; when the ratio of the first throughput to the second throughput is less than the second preset value, determining that the performance of the first compression model fails the test, i.e., the first model fails the test. The second preset value may be pre-set or configured.

[0156] It can be seen that when the first throughput is greater than or equal to the predetermined index compared to the second throughput, the test equipment determines that the performance of the first compression model has passed the test. When the first throughput is less than the predetermined index compared to the second throughput, the test equipment determines that the performance of the first compression model has failed the test.

[0157] In another optional implementation. The test device performs downlink data transmission based on the first precoding matrix indicated by the first precoding indicator to obtain a first throughput of the terminal device. The test device obtains a third throughput of the terminal device through the above-mentioned non-AI CSI feedback test. When the first throughput is greater than or equal to the third throughput, and the difference between the first throughput and the third throughput is greater than or equal to the preset indicator, or the ratio between the first throughput and the third throughput is greater than or equal to the preset indicator, the simulation device determines that the performance of the first compression model passes the test, that is, the first model passes the test. When the first throughput is greater than or equal to the third throughput and the difference between the first throughput and the third throughput is less than the preset indicator, or the first throughput is less than the third throughput, or the ratio between the first throughput and the third throughput is less than the preset indicator, the simulation device determines that the performance of the first compression model does not pass the test, that is, the first model does not pass the test.

[0158] In another optional embodiment, the test device performs downlink data transmission based on the first precoding matrix indicated by the first precoding indicator to obtain a first throughput of the terminal device. When the first throughput is greater than or equal to a preset throughput, the test device determines that the performance of the first compression model has passed the test. When the first throughput is less than the preset throughput, the test device determines that the performance of the first compression model has failed the test.

[0159] Optionally, when the test device and multiple simulation devices test the first compression model, the test device determines whether the first compression model passes the test based on the relative throughput obtained when testing with the multiple models. For example, the simulation device is a transcoder, and the manufacturer of the transcoder is the same as the manufacturer of the network device. That is, the transcoder is provided by the network device manufacturer. To reduce the number of times the terminal device connects to various network devices, the test device selects transcoders provided by some of the multiple network device manufacturers to test the compression model of the terminal device. Therefore, the test device needs to test the first compression model of the terminal device with the multiple transcoders selected by the test device.

[0160] Optionally, the decompression models supported by each of the multiple simulation devices match the first compression model to be tested, or the decompression models supported by some of the multiple simulation devices match the first compression model to be tested, and the multiple decompression models matching the first compression model may be the same or different.

[0161] When the TE tests a first compression model to be tested with some of the multiple simulation devices, the decompression models supported by each of the simulation devices match the first compression model. For example, if the simulation devices are transcoders, and transcoder 1, transcoder 2, and transcoder 3 support decompression models 1, 2, and 3, respectively, if decompression model 1, 2, and 3 all match the first compression model, the TE tests the performance of the first compression model with transcoder 1, transcoder 2, and transcoder 3, respectively. Decompression model 1, 2, and 3 are identical, partially identical, or completely different. If decompression model 1 and decompression model 3 match the first compression model, and decompression model 2 does not match the first compression model, the TE tests the performance of the first compression model with transcoder 1 and transcoder 3, and decompression model 1 and decompression model 2 are identical or different.

[0162] Exemplarily, the test device tests the performance of the first compression model with transcoder 1, transcoder 2, and transcoder 3, respectively. When the test device tests the performance of the first compression model with transcoder 1, a first throughput #1 is obtained; when the test device tests the performance of the first compression model with transcoder 2, a first throughput #2 is obtained; when the test device tests the performance of the first compression model with transcoder 3, a first throughput #3 is obtained. When the ratios of the first throughput #1, the first throughput #2, and the first throughput #3 to the second throughput or the third throughput, respectively, are greater than or equal to a predetermined indicator, the test device determines that the performance of the first compression model has passed the test. Alternatively, the test device takes the average of the first throughput #1, the first throughput #2, and the first throughput #3 to obtain an average throughput. When the ratio of the average throughput to the second throughput or the third throughput is greater than or equal to a predetermined indicator, the test device determines that the performance of the first compression model has passed the test.

[0163] In an embodiment of the present application, a test device sends a first reference signal to a terminal device. The terminal device compresses a first CSI determined based on the first reference signal using a first compression model to obtain a third CSI. A simulation device decompresses the third CSI using a first decompression model that matches the first compression model to obtain a second CSI. The simulation device calculates the second CSI to obtain a first precoding matrix indicator corresponding to the first precoding matrix. The simulation device sends the first precoding matrix indicator to the test device. The test device tests the performance of the first compression model based on the first precoding matrix indicator.

[0164] It can be seen that the test equipment's process for testing the first compression model reuses the current non-AI CSI feedback test process. That is, the test equipment does not need to use a decompression model to decompress the compressed CSI (i.e., the third CSI), nor does it need to calculate the PMI based on the decompressed CSI. As a result, the test equipment can test the performance of the first compression model with low complexity, and the test equipment does not require firmware upgrades and updates. In addition, the first decompression model used by the simulation device to decompress the compressed CSI matches the first compression model used by the terminal device to compress the uncompressed CSI, thereby reducing the impact of dual-end model mismatch on the test results during the test process.

[0165] In an optional embodiment, before S201 in the above-mentioned model testing method 100, that is, before the test device sends the first reference signal to the terminal device, that is, before the test device executes the test process of the compression model, the terminal device, the test device, and the simulation device may also execute the interactive process shown in Figure 3. As shown in Figure 3, the interactive process includes but is not limited to:

[0166] S301. The terminal device sends a first message to the test device, the first message including X first identifiers, one of the X first identifiers corresponding to a compression model supported by the terminal device, where X is a positive integer. Accordingly, the test device receives the first message from the terminal device.

[0167] Among them, one of the X first identifiers is used to identify one or more of the following: a compression model supported by the terminal device, parameter configuration information associated with a compression model supported by the terminal device, channel condition information associated with a compression model supported by the terminal device, state condition information associated with a compression model supported by the terminal device, information associated with the current state of the terminal device, and performance indicator information associated with a compression model supported by the terminal device. The current state of the terminal device may include one or more of the terminal device's current battery level, storage capacity, computing power, movement speed, and the executable state of one or more compression models supported by the terminal device. The performance indicator information associated with a compression model supported by the terminal device may refer to the performance that the compression model is expected to achieve. For example, when the terminal device uses the compression model, after the network device decompresses the compressed CSI obtained using the compression model, the channel recovery accuracy can reach 90%, and / or the downlink throughput can reach 90% of the maximum throughput at x dB. Optionally, the performance indicator information associated with a compression model supported by the terminal device may include information related to the compression model's latency, such as one or more of the compression model activation latency and the compression model inference latency. The status condition information associated with a compression model supported by the terminal device may be one or more of the computing power, storage capacity, and power level of the terminal device.

[0168] It is understandable that when a first identifier is used to identify a compression model supported by a terminal device, the first identifier may be a model identity (model ID); when a first identifier is used to identify one or more of parameter configuration information associated with a compression model supported by the terminal device, channel condition information associated with a compression model supported by the terminal device, state condition information associated with a compression model supported by the terminal device, information associated with the current state of the terminal device, and performance indicator information associated with a compression model supported by the terminal device, the first identifier may be a function identifier. Model identifiers and function identifiers can be used as identifiers used by network devices and / or terminal devices to manage models. Model management includes but is not limited to model monitoring, model online pairing, model selection, model switching, model activation, model deactivation, and fallback, etc. Optionally, the first identifier can also explicitly identify the manufacturer of the terminal device, such as some or all fields of the first identifier directly identifying the manufacturer of the terminal device as manufacturer A. Optionally, the first identifier can also implicitly identify the manufacturer of the terminal device, such as the first identifier is used to identify information a, and other devices can obtain the manufacturer of the terminal device as manufacturer A by interpreting information a.

[0169] Optionally, the first identifier may also be a display identifier or an implicit identifier of the manufacturer of the network device, wherein the compression model supported by the manufacturer of the terminal device matches the decompression model supported by the manufacturer of the network device.

[0170] It is understandable that one of the X first identifiers is associated with a compression model supported by the terminal device. Each of the X first identifiers is associated with a different compression model supported by the terminal device, so that the X first identifiers are associated with the X compression models supported by the terminal device. Optionally, some of the X first identifiers are associated with the same compression model supported by the terminal device, so that the number of compression models associated with the X first identifiers is less than X. For example, the X first identifiers include the first identifier 1 and the first identifier 2, the first identifier 1 is associated with compression model 1, and the first identifier 2 is associated with compression model 2. For another example, the X first identifiers include the first identifier 1 and the first identifier 2, and both the first identifier 1 and the first identifier 2 are associated with compression model 1.

[0171] In an optional embodiment, after the terminal device enters the connected state from the idle state, it actively reports capability information to the test device. The capability information includes X first identifiers related to the compression models supported by the terminal device. The first message may also include other capability information supported by the terminal device in addition to the compression models.

[0172] Optionally, the test device instructs the terminal device in a connected state to report the supported compression models to the test device according to the configuration of the test device. For example, the test device instructs the terminal device to feedback some or all of the compression models supported by the terminal device under specific channel conditions and / or specific network configuration conditions. For example, the test device instructs the terminal device to feedback the compression model supported by the terminal device under the network configuration parameters of RANK=4 and / or the clustered delay line (CDL) channel conditions. After receiving the information, the terminal device reports to the test device the X first identifiers related to the compression model supported by the terminal device and applicable to the network configuration parameters of RANK=4 and / or the CDL channel conditions.

[0173] In an optional implementation, the first message is a new message in the process of TE testing a dual-end model. The new message may be reported by the terminal device to the test device according to the TE configuration.

[0174] It can be seen that before the test device tests the compression model of the terminal device, the terminal device reports X first identifiers related to one or more compression models supported by the terminal device to the test device. Therefore, the test device obtains X first identifiers related to one or more compression models supported by the terminal device.

[0175] S302. The test device sends a second message to the simulation device. The second message includes Y second identifiers, where the Y second identifiers are part or all of the X first identifiers, or are determined by the TE based on part or all of the X first identifiers, and Y is an integer less than or equal to X. In response, the simulation device receives the second message from the test device.

[0176] In one optional embodiment, after receiving X first identifiers from the terminal device, the test device selects Y first identifiers from the X first identifiers and sends the Y first identifiers to the simulation device. That is, the Y second identifiers are the Y first identifiers from the X first identifiers. Alternatively, the test device can select the Y first identifiers from the X first identifiers by selecting some or all compression models from one or more compression models reported by the terminal device and notifying the simulation device of the selected some or all compression models via the Y first identifiers. Therefore, the Y second identifiers are some or all of the first identifiers from the X first identifiers.

[0177] Exemplarily, the X first identifiers include a first identifier 1 and a first identifier 2, and the first identifier 1 and the first identifier 2 are respectively related to a compression model supported by the terminal device. The test device arbitrarily selects one from the first identifier 1 and the first identifier 2 as the second identifier, and sends the selected second identifier to the simulation device through the second information.

[0178] Exemplarily, the X first identifiers include a first identifier 1 and a first identifier 2, and the first identifier 1 and the first identifier 2 are respectively related to the compression model 1 and the compression model 2 supported by the terminal device. TE determines the second identifier based on whether the current terminal device can execute compression model 1 and / or compression model 2. Whether the terminal device can execute compression model 1 and / or compression model 2 may be reported to TE by the terminal device through other means. For example, the terminal device reports to TE that it can currently execute compression model 1 but cannot execute compression model 2, then TE sends the first identifier 1 to the simulation device through a second message, so that the simulation device determines whether the decompression model supported by itself can match the compression model 1 related to the first identifier 1.

[0179] In an optional embodiment, at least one of the X first identifiers is also used to identify one or more of the parameter configuration information associated with a compression model supported by the terminal device, the channel condition information associated with a compression model supported by the terminal device, the state condition information associated with a compression model supported by the terminal device, the information associated with the current state of the terminal device, and the performance indicator information associated with a compression model supported by the terminal device. For example, it is also used to identify the current power of the terminal device, the power condition required for the terminal device to execute the compression model associated with the first identifier, the current storage capacity of the terminal device, the storage condition required for the terminal device to execute the compression model associated with the first identifier, the current computing power capacity of the terminal device, the computing power condition required for the terminal device to execute the compression model associated with the first identifier, and whether the terminal device can currently execute one or more of the compression model associated with the first identifier. In this manner, the TE determines the compression model that the terminal device can execute based on the relevant information identified by the first identifier, and sends Y second identifiers related to the compression model that the terminal device can execute to the simulation device through the second information.

[0180] Exemplarily, the X first identifiers include a first identifier 1 and a first identifier 2, and the first identifier 1 and the first identifier 2 are respectively related to the compression model 1 and the compression model 2 supported by the terminal device. The first identifier 1 is also used to identify the conditions required to execute the compression model 1, and the first identifier 2 is also used to identify the conditions required to execute the compression model 2. The conditions required to execute the compression model may be the power conditions required to execute the compression model. For example, the first identifier 1 is also used to identify the power condition 1 required for the terminal device to execute the compression model 1, and the first identifier 2 is also used to identify the power condition 2 required for the terminal device to execute the compression model 2. TE directly determines the compression model with the least power required to execute compression model 1 and compression model 2 as the compression model executable by the terminal device. For example, if the power condition 1 is less than the power condition 2, the TE determines that the terminal device can execute the compression model 1, and sends the first identifier 1 to the simulation device through the second message.

[0181] Exemplarily, the X first identifiers include a first identifier 1 and a first identifier 2, and the first identifier 1 and the first identifier 2 are respectively associated with compression model 1 and compression model 2 supported by the terminal device. The first identifier 1 is also used to identify parameter configuration information 1 associated with compression model 1, and the first identifier 2 is also used to identify parameter configuration information 2 associated with compression model 2. The parameter configuration information associated with the compression model may be a network parameter configuration applicable to the compression model. For example, the first identifier 1 is also used to identify that the network parameter configuration applicable to the terminal device executing compression model 1 is RANK=1, and the first identifier 2 is also used to identify that the network parameter configuration applicable to the terminal device executing compression model 2 is RANK=4. If the TE determines that the test terminal device can execute the compression model under the parameter configuration of RANK=1, the TE determines that the terminal device can execute compression model 1, and sends the first identifier 1 to the simulation device through a second message.

[0182] Exemplarily, the X first identifiers include a first identifier 1 and a first identifier 2, and the first identifier 1 and the first identifier 2 are respectively related to the compression model 1 and the compression model 2 supported by the terminal device. The first identifier 1 is also used to identify the conditions required to execute the compression model 1, and the first identifier 2 is also used to identify the conditions required to execute the compression model 2. The first identifier 1 and / or the first identifier 2 are also used to identify the state of the terminal device. For example, the first identifier 1 is also used to identify the power condition 1 required for the terminal device to execute the compression model 1 and the current power of the terminal device, and the first identifier 2 is also used to identify the power condition 2 required for the terminal device to execute the compression model 2. Then, the test device can determine the compression model that the terminal device can execute based on the current power of the terminal device and the power conditions required for the terminal device to execute the compression model 1 and the compression model 2, respectively. The identifier related to the executable compression model is the second identifier.

[0183] Exemplarily, the X first identifiers include a first identifier 1 and a first identifier 2, and the first identifier 1 and the first identifier 2 are respectively related to the compression model 1 and the compression model 2 supported by the terminal device. The first identifier 1 is also used to identify the conditions required to execute the compression model 1, and the first identifier 2 is also used to identify the conditions required to execute the compression model 2. For example, the first identifier 1 is also used to identify the power condition 1 required for the terminal device to execute the compression model 1, and the first identifier 2 is also used to identify the power condition 2 required for the terminal device to execute the compression model 2. The TE determines that the terminal device can execute the compression model 1 based on the current power of the terminal device, power condition 1, and power condition 2, and then sends the first identifier 1 to the simulation device through a second message, so that the simulation device determines whether the decompression model supported by itself can match the compression model 1 associated with the first identifier 1. The current power of the terminal device may be reported by the terminal device to the TE through other means.

[0184] Exemplarily, the X first identifiers include a first identifier 1, a first identifier 2, and a first identifier 3. The first identifier 1 is used to identify compression model 1 and that the terminal device is currently capable of executing compression model 1. The first identifier 2 is used to identify compression model 2 and that the terminal device is currently capable of executing compression model 2. The first identifier 3 is used to identify compression model 3 and that the terminal device is currently incapable of executing compression model 3. The TE then determines that the terminal device is capable of executing the compression model associated with the first identifier 1 and the compression model associated with the first identifier 2. Thus, the TE sends the first identifier 1 and the first identifier 2 to the simulation device through a second message, so that the simulation device determines whether the decompression model supported by the simulation device can match the compression model in compression model 1 and compression model 2.

[0185] Exemplarily, at least one of the X first identifiers is associated with a channel condition. Based on the channel conditions associated with the X first identifiers, the test device may select Y second identifiers associated with the test channel condition from the X first identifiers and send the selected Y second identifiers to the simulation device. This approach facilitates the simulation device determining whether, among the Y second identifiers associated with the test channel condition, a decompression model supported by the simulation device matches the compression model associated with the Y second identifiers.

[0186] Exemplarily, the X first identifiers include a first identifier 1, a first identifier 2, and a first identifier 3, and the first identifier 1 and the first identifier 2 are related to channel condition 1, and the first identifier 3 is related to channel condition 2. If the test device chooses to test the performance of the compression model in a channel environment related to channel condition 1, the first identifier 1 and the first identifier 2 are sent to the simulation device, that is, the second identifiers are the first identifier 1 and the first identifier 2. Exemplarily, the X first identifiers include a first identifier 1, a first identifier 2, and a first identifier 3, and the first identifier 1, the first identifier 2, and the first identifier 3 are all related to channel condition 3. If the test device chooses to test the performance of the compression model in a channel environment of channel condition 3, the test device sends the first identifier 1, the first identifier 2, and the first identifier 3 to the simulation device, that is, the second identifiers are the first identifier 1, the first identifier 2, and the first identifier 3.

[0187] Exemplarily, the X first identifiers include first identifier 1, first identifier 2, and first identifier 3. First identifier 1, first identifier 2, and first identifier 3 are each used to identify a compression model supported by the terminal device. First identifier 1 and first identifier 2 are also respectively used to identify that compression model a and compression model b supported by the terminal device have been locally loaded on the terminal device. First identifier 3 is also used to identify that compression model c supported by the terminal device has not been locally loaded on the terminal device. Thus, the test device determines that the terminal device can use compression model a and compression model b, but cannot use compression model c. Therefore, the test device determines that the Y second identifiers are first identifier 1 and identifier 2, and sends first identifier 1 and second identifier 2 to the simulation device via a second message.

[0188] Exemplarily, the X first identifiers include a first identifier 1 and a first identifier 2, and the first identifier 1 and the first identifier 2 are respectively used to identify a compression model supported by the terminal device. The test device determines the compression model that the terminal device can execute based on the compression model loaded by the terminal device. For example, the terminal device reports to the test device through UE assistance information (UAI) or other information that the terminal device has loaded the compression model related to the first identifier 1, and the test device determines that the compression model that the terminal device can execute is the compression model related to the first identifier. Thus, the test device sends the first identifier 1 to the simulation device through the second message.

[0189] It can be seen that the test device can select Y first identifiers from the X first identifiers based on one or more of a compression model supported by the terminal device identified by a first identifier among the X first identifiers, parameter configuration information associated with a compression model supported by the terminal device, channel condition information associated with a compression model supported by the terminal device, state condition information associated with a compression model supported by the terminal device, information associated with the current state of the terminal device, and performance indicator information associated with a compression model supported by the terminal device, and inform the simulation device through a third message. It can also be understood that after obtaining the X first identifiers sent by the terminal device, the test device again confirms the Y second identifiers related to the compression model executable by the terminal device, and sends the Y second identifiers to the simulation device, so that the simulation device can determine whether there is a decompression model supported by itself that can match the compression model associated with the Y second identifiers.

[0190] In another optional embodiment, after receiving X first identifiers from the terminal device, the test device determines Y second identifiers based on some or all of the X first identifiers, and sends the Y second identifiers to the simulation device. In this approach, the Y second identifiers are determined by the test device based on some or all of the X first identifiers.

[0191] Exemplarily, the first identifier is used to identify one or more of the parameter configuration information associated with a compression model supported by the terminal device, the channel condition information associated with a compression model supported by the terminal device, the state condition information associated with a compression model supported by the terminal device, the state condition information associated with a compression model supported by the terminal device, the information associated with the current state of the terminal device, and the performance indicator information associated with a compression model supported by the terminal device, and is also used to display and identify the manufacturer of the terminal device, and / or, when the network manufacturer matches the terminal device, the test device can interpret X first identifiers and obtain the manufacturer associated with one of the X first identifiers. Then, after the test device selects Y first identifiers from the X first identifiers, it processes the Y first identifiers and obtains Y second identifiers, such as deleting the relevant information used to identify the manufacturer in the Y first identifiers to obtain Y second identifiers. For example, the second identifier is used to instruct the simulation device to activate the decompression model paired with the first compression model.

[0192] S303. The simulation device sends a third message to the test device. The third message includes M third identifiers and / or an acknowledgment (ACK). The M third identifiers are part or all of the Y second identifiers, or are determined by the simulation device based on part or all of the Y second identifiers. The compression model associated with one of the M third identifiers matches a decompression model supported by the simulation device, and M is an integer less than or equal to Y. Accordingly, the test device receives the third message from the simulation device.

[0193] In one optional embodiment, the simulation device selects, based on the decompression models supported by the simulation device, M third identifiers associated with compression models that match the decompression models supported by the simulation device from the Y second identifiers. That is, the compression model associated with one of the M third identifiers matches a decompression model supported by the simulation device. Therefore, the M third identifiers are some or all of the second identifiers in the Y second identifiers.

[0194] For example, if the Y second identifiers include second identifier 1, second identifier 2, and second identifier 3, and decompression model 1 supported by the simulation device matches compression model 1 associated with second identifier 1, and decompression model 2 supported by the simulation device matches compression model 2 associated with second identifier 2, the simulation device determines that the decompression model supported by the simulation device can match the compression model associated with second identifier 1 and the compression model associated with second identifier 2. Thus, the M third identifiers are second identifier 1 and second identifier 2.

[0195] For another example, if Y second identifiers are second identifier 1, and the decompression model 1 supported by the simulation device matches the compression model 1 related to the second identifier 1, the simulation device can feedback the second identifier 1 to the test device, or feedback an acknowledgment (ACK), or feedback the second identifier 1 and ACK.

[0196] In another optional implementation, the simulation device selects M second identifiers associated with the compression model that matches the decompression model it supports from the Y second identifiers based on the decompression model it supports, determines M third identifiers based on the M second identifiers, and feeds these back to the test device. In other words, the simulation device determines M third identifiers based on some or all of the Y second identifiers and feeds these M third identifiers back to the test device. In this approach, the M third identifiers are determined by the simulation device based on some or all of the Y second identifiers.

[0197] In an optional embodiment, when Y is an integer greater than 1 and less than X, the M third identifiers are part or all of the Y second identifiers, or are determined by the simulation device based on part or all of the Y second identifiers. In this case, the third message includes M third identifiers. When the third message includes M third identifiers, the test device defaults to the decompression model supported by the simulation device being able to match the compression model associated with the M third identifiers. Optionally, the third message includes M third identifiers and ACK, which is used to confirm that the decompression model supported by the simulation device itself is able to match the compression model associated with the M third identifiers.

[0198] In another optional embodiment, Y is equal to 1, that is, when the test device sends a second identifier to the simulation device, M is equal to 1, and a third identifier is the second identifier sent by the test device to the simulation device, or is determined by the simulation device based on the second identifier from the test device. If the decompression model supported by the simulation device can match the compression model associated with the second identifier, the third message includes the third identifier, and the test device defaults that the decompression model supported by the simulation device can match the compression model associated with the third identifier. Optionally, when Y is equal to 1, the third message includes ACK, which is used to confirm that the decompression model supported by the simulation device can match the compression model associated with the second identifier sent by the test device. Optionally, the third message includes a third identifier and ACK, and the third identifier is the second identifier sent by the test device to the model device, and the ACK is used to confirm that the decompression model supported by the simulation device can match the compression model associated with the third identifier.

[0199] In addition, the compression model associated with the second identifier sent by the test device to the simulation device is the first compression model, and the decompression model supported by the simulation device that matches the first compression model is the first decompression model, that is, the first decompression model matches the first compression model.

[0200] In another optional implementation, the test device sends a second identifier to the simulation device, and there is no decompression model that matches the compression model associated with the second identifier in the decompression models supported by the simulation device. In this case, the simulation device sends a negative response NACK to the test device, and the NACK is used to indicate that the decompression model supported by the simulation device does not match the compression model associated with the second identifier. Optionally, after receiving the NACK from the simulation device, the test device sends one or more identifiers to the simulation device so that the simulation device can reconfirm whether there is a decompression model that matches the compression model associated with the one or more identifiers in the supported decompression models. The one or more identifiers are part or all of the first identifiers in the X first identifiers, or are determined by the test device based on part or all of the first identifiers in the X first identifiers, and the one or more identifiers are different from the above-mentioned second identifiers.

[0201] It can be seen that the simulation device feeds back to the test device through the third message that among the decompression models supported by the simulation device, there is a decompression model that can match part or all of the one or more compression models associated with the Y second identifiers.

[0202] S304. The test device sends a fourth message to the terminal device. The fourth message includes a fourth identifier. The fourth identifier is one of the M third identifiers or the Y second identifiers, or is determined by the TE based on the M third identifiers or the Y second identifiers. The fourth identifier is associated with the first compression model, and the first compression model matches the first decompression model. Accordingly, the terminal device receives the fourth message from the test device.

[0203] In an optional embodiment, when Y is an integer greater than 1 and less than X, the test device learns through a third message that the simulation device supports one or more decompression models that can match one or more compression models associated with M third identifiers. Thus, the test device selects a third identifier from the M third identifiers as the fourth identifier and sends it to the terminal device, or the test device determines a fourth identifier based on one of the M third identifiers and sends the determined fourth identifier to the terminal device to activate the compression model to be tested in the terminal device. It can be seen that the fourth identifier is one of the M third identifiers, or the fourth identifier is an identifier determined by the test device based on one of the M third identifiers. The first compression model associated with the fourth identifier is the first compression model to be tested.

[0204] In another optional embodiment, when Y is equal to 1, the simulation device confirms to the test device, via a third message, that the decompression model supported by the simulation device matches the compression model associated with the second identifier sent by the test device. Therefore, the fourth identifier sent by the test device to the terminal device is the second identifier sent by the test device to the terminal device. Thus, the fourth identifier is one of the Y second identifiers, or is determined by the test device based on one of the second identifiers, and the first compression model associated with the second identifier is the first compression model to be tested.

[0205] Optionally, the test device also sends a fifth message to the simulation device, where the fifth message includes a fifth identifier, the fifth identifier is one of the M third identifiers or the Y second identifiers, or is determined by the TE based on the M third identifiers or the Y second identifiers, the fifth identifier is related to the first decompression model, and the first decompression model matches the first compression model. The fifth identifier is the same as or different from the fourth identifier. For example, the M third identifiers include the third identifier 1 and the third identifier 2, both the third identifier 1 and the third identifier 2 are related to the first compression model, the fourth identifier is the third identifier 1, and the fifth identifier is the third identifier 2. Then, the test device sends the third identifier 1 to the terminal device and sends the third identifier 2 to the simulation device, which can not only enable the terminal device to activate the first compression model to be tested, but also enable the simulation device to determine the first compression model that matches the first compression model.

[0206] For another example, the M third identifiers include a third identifier 1 and a third identifier 2, the third identifier 1 is associated with the first compression model, the third identifier 2 is not associated with the first compression model, the fourth identifier is the third identifier 1, and the fifth identifier is an identifier determined by the TE based on the third identifier 1. Then, the test device sends the fourth identifier to the terminal device and sends the fifth identifier to the simulation device, which can also cause the terminal device to activate the first compression model to be tested and cause the simulation device to determine the first decompression model that matches the first compression model.

[0207] It is understood that the test device sends the fourth identifier to the terminal device, enabling the terminal device to determine and activate the first compression model to be tested. Furthermore, the test device also sends the fifth identifier to the simulation device, enabling the simulation device to determine the first decompression model that matches the first compression model to be tested. Furthermore, during the test device's testing of the first compression model, the simulation device can utilize the first decompression model that matches the first compression model, thereby reducing the impact of dual-end model mismatch on the test results.

[0208] It can be seen that the test device, the terminal device and the simulation device can realize the first decompression model used by the simulation device for decompression processing through the above process of S301 to S304, and match it with the first compression model to be tested.

[0209] Optionally, when the test device tests the compression model of the terminal device based on multiple simulation devices, the test device and the simulation device both perform S302 to S303 above for one of the multiple simulation devices. For example, the simulation device is a transcoder, and the test device tests the compression model of the terminal device based on the multiple transcoders, including transcoder 1 and transcoder 2. The test device sends a second message to transcoder 1 and transcoder 2, the second message including Y second identifiers. Transcoder 1 returns a third message 1 to the test device, and transcoder 2 returns a third message 2 to the test device, the third message 1 including M1 third identifiers and / or ACKs, and the third message 2 including M2 third identifiers and / or ACKs. If the decompression model 1 supported by transcoder 1 matches the compression model 1 supported by the terminal device, the test device and transcoder 1 test the compression model 1 supported by the terminal device. If the decompression model 2 supported by transcoder 2 matches the compression model 2 supported by the terminal device, the test device and transcoder 2 test the compression model 2 supported by the terminal device. The compression model 1 and the compression model 2 may be the same or different.

[0210] The present embodiment also proposes a model testing method 200 as shown in FIG4 for testing the decompression model. FIG4 is an interactive diagram of the model testing method 200. The model testing method 200 is described from the perspective of the interaction between the test device, the simulation device, and the network device. The model testing method 200 includes but is not limited to the following steps:

[0211] S401. The test device sends a second reference signal to the simulation device. Correspondingly, the simulation device receives the second reference signal from the test device.

[0212] The second reference signal may refer to the first reference signal mentioned above, and will not be described in detail.

[0213] S402. The analog device determines a fourth CSI according to the second reference signal.

[0214] S403. The simulation device compresses the fourth CSI using a second compression model to obtain a sixth CSI, where the second compression model matches the second decompression model used by the network device.

[0215] It is understandable that the analog device performs channel estimation based on the received second reference signal to obtain the fourth CSI, and then compresses the fourth CSI using the second compression model to obtain the sixth CSI. Therefore, the sixth CSI is the CSI after the analog device compresses the fourth CSI.

[0216] Furthermore, the second compression model used by the simulation device matches the second decompression model used by the network device. The second decompression model used by the network device is the decompression model used for TE testing. Therefore, the second compression model used by the simulation device matches the second decompression model used for testing, thereby reducing the impact of dual-end model mismatch on test results. The implementation of matching the second compression model with the second decompression model can be found in S501 to S504 below.

[0217] In an optional implementation, the second compression model is provided by a terminal device, or a network device, or other devices other than the terminal device and the network device.

[0218] In an optional implementation, the terminal device, or the network device, or the third party directly sends the second compression model to the simulation device. When the terminal device sends the second compression model to the simulation device, the terminal device and the simulation device belong to the same manufacturer; when the network device sends the second compression model to the simulation device, the network device and the simulation device belong to the same manufacturer; when the third party sends the second compression model to the simulation device, the third party and the simulation device belong to the same manufacturer. This method helps reduce the risk of the second compression model being leaked. For example, this method helps reduce the risk of model privacy technology being leaked when the terminal device or network device sends the second compression model to the test device.

[0219] In another optional implementation, the manufacturer of the simulation device provides the second compression model for the simulation device, and the manufacturer of the simulation device is the same as the manufacturer of the terminal device, or the manufacturer of the simulation device is the same as the manufacturer of the network device, or the manufacturer of the simulation device is the same as the manufacturer of a third party. Thus, the second compression model used by the simulation device can be understood as being provided by the terminal device, the network device, or a third party. This approach can also reduce the risk of the second compression model being leaked.

[0220] S404a. The simulation device sends the sixth CSI to the network device.

[0221] It can be understood that the analog device directly sends the sixth CSI to the network device, so that the network device decompresses the sixth CSI.

[0222] S404b1. The simulation device sends the sixth CSI to the test device.

[0223] S404b2. The test device sends the sixth CSI to the network device.

[0224] It is understandable that the simulation device executes S404a or S404b1. When the simulation device executes S404b1, the test device also executes S404b2.

[0225] It is understandable that the simulation device sends the sixth CSI to the test device, and the test device then sends the sixth CSI to the network device. This method is beneficial for the test device to confirm whether the sixth CSI is the CSI compressed by the simulation device, or the test device to confirm whether the simulation device compresses the fourth CSI determined based on the second reference signal in accordance with the test requirements. This method can reduce the risk of fraud in the joint efforts of the simulation device and the network device during the test. For example, in order to enable the second decompression model to pass the test, the network device negotiates with the simulation device and allows the simulation device to directly send uncompressed CSI to the network device, such as directly sending the fourth CSI, resulting in the test result not being obtained based on the second decompression model, thereby affecting the accuracy and reliability of the test result. In addition, this method can also enable the test device to determine whether the sixth CSI fed back by the simulation device conforms to the CSI feedback format configured by TE.

[0226] S405. The network device uses the second decompression model to decompress the sixth CSI to obtain the fifth CSI.

[0227] S406. The network device determines a second precoding matrix indicator based on the fifth CSI.

[0228] The network device calculates the fifth CSI to obtain a second precoding matrix. The indicator used to indicate the second precoding matrix is ​​a second precoding matrix indicator. Thus, the network device determines the second precoding matrix indicator based on the fifth CSI.

[0229] S407: The network device sends a second precoding matrix indicator to the test device. Correspondingly, the test device receives the second precoding matrix indicator from the network device.

[0230] S408. The testing device tests the performance of the second decompression model based on the second precoding matrix indicator.

[0231] It can be understood that the implementation method of the test device testing the performance of the second decompression model based on the second precoding matrix indicator can refer to the above S208, in which the test device tests the performance of the first compression model based on the first precoding matrix indicator, and will not be repeated.

[0232] Optionally, when the testing device tests the second decompression model with multiple simulation devices, the testing device determines whether the second decompression model passes the test based on the relative throughput or average throughput obtained when testing with multiple models. The specific implementation method can be referred to as described in S208 above and will not be repeated here.

[0233] In this embodiment of the present application, the test device sends a second reference signal to the simulation device. The simulation device compresses fourth CSI obtained based on the second reference signal using a second compression model to obtain sixth CSI. The network device decompresses the sixth CSI using a second decompression model to obtain fifth CSI. The network device obtains a second precoding matrix indicator based on the fifth CSI and sends the second precoding matrix indicator to the test device. The test device tests the performance of the second decompression model based on the second precoding matrix indicator.

[0234] As can be seen, during the test device's testing of the second decompression model, the test device does not need to determine uncompressed CSI based on a reference signal, nor does it need to use a compression model to compress the uncompressed CSI. This allows the test device to test the performance of the second decompression model with low complexity, and does not require firmware upgrades or updates. Furthermore, the second compression model used by the simulation device to compress the uncompressed CSI matches the second decompression model used by the network device to decompress the compressed CSI, thereby reducing the impact of dual-end model mismatch on test results during testing.

[0235] In an optional embodiment, before S401 in the above-mentioned model testing method 200, that is, before the test device sends the second reference signal to the simulation device, that is, before the test device executes the test process of decompressing the model, the test device, the simulation device, and the network device may also execute the interaction process shown in Figure 5. As shown in Figure 5, the interaction process includes but is not limited to:

[0236] S501. The network device sends a sixth message to the test device, the sixth message including S sixth identifiers, one of the S sixth identifiers corresponding to a decompression model supported by the network device, where S is a positive integer. Accordingly, the test device receives the sixth message from the network device.

[0237] Among the S sixth identifiers, one sixth identifier is used to identify one or more of the following: a decompression model supported by the network device, parameter configuration information associated with a decompression model supported by the network device, channel condition information associated with a decompression model supported by the network device, state condition information associated with a decompression model supported by the network device, information associated with the current state of the network device, and performance indicator information associated with a decompression model supported by the network device. The current state of the network device may be one or more of the network device's current power, storage capacity, computing power, and executable state of one and / or multiple decompression models supported by the network device. The performance indicator information associated with a compression model supported by the network device may refer to the performance that the compression model is expected to achieve. Optionally, the performance indicator information associated with a decompression model supported by the network device may be information related to the latency of the decompression model, such as the latency of activating the decompression model or the latency of inferring the decompression model. The required conditions associated with a decompression model supported by the network device may be one or more of the computing power and storage capacity of the network device. The status condition information associated with a decompression model supported by the network device may be one or more of the computing power capability, storage capability, and power level of the network device.

[0238] It is understandable that when a sixth identifier is used to identify a decompression model supported by a network device, the sixth identifier may be a model identifier (model ID); when a sixth identifier is used to identify one or more of parameter configuration information associated with a decompression model supported by the network device, channel condition information associated with a decompression model supported by the network device, state condition information associated with a decompression model supported by the network device, information associated with the current state of the network device, and performance indicator information associated with a decompression model supported by the network device, the sixth identifier may be a function identifier. The model identifier and the function identifier can be referred to above and will not be repeated here.

[0239] It is understandable that one of the S sixth identifiers is associated with a decompression model supported by the network device. Each of the S sixth identifiers is associated with a different decompression model supported by the network device, and thus the S sixth identifiers are associated with the S decompression models supported by the network device. Optionally, some of the S sixth identifiers are associated with the same decompression model supported by the network device, and thus the number of decompression models associated with the S sixth identifiers is less than S.

[0240] In an optional implementation, the network device proactively reports to the test device a third identifier related to some or all of the decompression models it supports. Optionally, the test device instructs the network device to report the supported decompression models to the test device in accordance with the test device's configuration. For example, the test device instructs the network device to report back the decompression models supported by the network device under specific channel conditions. Thus, based on the received information, the network device reports to the test device a sixth identifier related to some or all of the decompression models it supports in accordance with the TE's configuration.

[0241] In an optional implementation manner, the second message is a new message in the dual-end model test, and the new message may be sent by the network device to the test device according to the TE configuration.

[0242] It can be seen that before the test device tests the decompression model of the network device, the network device reports S sixth identifiers related to one or more decompression models supported by the network device to the test device. Therefore, the test device obtains S sixth identifiers related to one or more decompression models supported by the network device.

[0243] S502. The test device sends a seventh message to the simulation device. The seventh message includes T seventh identifiers, where the T seventh identifiers are part or all of the S sixth identifiers, or are determined by the TE based on part or all of the S sixth identifiers, where T is an integer less than or equal to S. In response, the simulation device receives the seventh message from the test device.

[0244] In an optional embodiment, after receiving S sixth identifiers from the network device, the test device selects T sixth identifiers from the S sixth identifiers and sends the T sixth identifiers to the simulation device, that is, the T seventh identifiers are Y sixth identifiers from the S sixth identifiers. It can also be understood that the test device selects some or all decompression models from one or more supported decompression models reported by the network device by selecting T sixth identifiers from the S sixth identifiers, and notifies the simulation device of the selected some or all decompression models via the T sixth identifiers. Therefore, the T seventh identifiers are some or all of the sixth identifiers from the S sixth identifiers.

[0245] Optionally, the test device selects part or all of the sixth identifiers from the S sixth identifiers based on one or more of a decompression model supported by the network device identified by a sixth identifier among the S sixth identifiers, parameter configuration information associated with a decompression model supported by the network device, channel condition information associated with a decompression model supported by the network device, state condition information associated with a decompression model supported by the network device, information associated with the current state of the network device, and performance indicator information associated with a decompression model supported by the network device. The seventh identifier is part or all of the selected S sixth identifiers. The test device selects part or all of the sixth identifiers from the S sixth identifiers as the implementation method of the T seventh identifiers can refer to the above-mentioned S302 and will not be repeated here.

[0246] In another optional implementation, after receiving S sixth identifiers from the network device, the test device determines T seventh identifiers based on some or all of the S sixth identifiers, and sends the T seventh identifiers to the simulation device. In this approach, the T seventh identifiers are determined by the test device based on some or all of the S sixth identifiers.

[0247] Exemplarily, the sixth identifier is used to identify one or more of parameter configuration information associated with a decompression model supported by the network device, channel condition information associated with a decompression model supported by the network device, state condition information associated with a decompression model supported by the network device, information associated with the current state of the network device, and performance indicator information associated with a decompression model supported by the network device. In addition, it is also used to display and identify the manufacturer of the network device, and / or, when the terminal device manufacturer matches the network device, the test device can interpret the S sixth identifiers and obtain the manufacturer associated with one of the S sixth identifiers. Then, after the test device selects T sixth identifiers from the S sixth identifiers, it processes the T sixth identifiers to obtain T seventh identifiers. Thus, the T seventh identifiers are determined by the test device based on some or all of the S sixth identifiers.

[0248] S503. The simulation device sends an eighth message to the test device. The eighth message includes N eighth identifiers and / or an acknowledgment (ACK). The N eighth identifiers are some or all of the T seventh identifiers, or are determined by the simulation device based on some or all of the T seventh identifiers. The decompression model associated with one of the N eighth identifiers matches a compression model supported by the simulation device, where N is an integer less than or equal to T. Accordingly, the simulation device receives the eighth message from the test device.

[0249] In an optional embodiment, the simulation device selects N eighth identifiers related to the decompression model that matches the compression model supported by itself from T seventh identifiers based on the compression model supported by itself, that is, one eighth identifier among the N eighth identifiers matches a compression model supported by the simulation device. For example, the T seventh identifiers include seventh identifier 1, seventh identifier 2, and seventh identifier 3, the compression model 1 supported by the simulation device matches the decompression model 1 related to seventh identifier 1, and the compression model 3 supported by the simulation device matches the decompression model 3 related to seventh identifier 3, then the simulation device determines that the compression model supported by itself can match the decompression model related to seventh identifier 1 and the decompression model related to seventh identifier 3. Therefore, the N eighth identifiers are seventh identifier 1 and seventh identifier 3.

[0250] For another example, T seventh identifiers are seventh identifier 1, and the compression model 1 supported by the simulation device matches the decompression model 1 related to the seventh identifier 1. The simulation device can feedback the seventh identifier 1 to the test device, or feedback an acknowledgment response ACK, or feedback the seventh identifier 1 and ACK.

[0251] In another optional embodiment, the simulation device selects N seventh identifiers related to the decompression model that matches the compression model supported by the simulation device from the T seventh identifiers based on the compression model supported by the simulation device, and determines N eighth identifiers based on the N seventh identifiers, and feeds them back to the test device. In other words, the simulation device determines N eighth identifiers based on some or all of the T seventh identifiers, and feeds back the N eighth identifiers to the test device. In this method, the N eighth identifiers are determined by the simulation device based on some or all of the T seventh identifiers.

[0252] In an optional embodiment, when T is an integer greater than 1 and less than X, the N eighth identifiers are part or all of the seventh identifiers in the T seventh identifiers, or are determined by the simulation device based on part or all of the seventh identifiers in the T seventh identifiers. In this case, the eighth message includes N eighth identifiers. When the eighth message includes N eighth identifiers, the test device defaults to that the compression model supported by the simulation device can match the decompression model associated with the N eighth identifiers. Optionally, the third message includes N eighth identifiers and ACK, which is used to confirm that the compression model supported by the simulation device itself can match the decompression model associated with the N eighth identifiers.

[0253] In another optional implementation, T is equal to 1, that is, the test device sends a seventh identifier to the simulation device, N is equal to 1, and an eighth identifier is the seventh identifier sent by the test device to the simulation device, or is determined by the simulation device based on the seventh identifier from the test device. If the compression model supported by the simulation device can match the decompression model associated with the seventh identifier, the eighth message includes the seventh identifier, or the eighth message includes the eighth identifier determined by the test device based on the seventh identifier. At this time, the test device defaults that the compression model supported by the simulation device can match the decompression model associated with the seventh identifier. Optionally, when T is equal to 1, the eighth message includes ACK, and the ACK is used to confirm that the compression model supported by the simulation device can match the decompression model associated with the seventh identifier. Optionally, the eighth message includes the eighth identifier and ACK, and the eighth identifier is the seventh identifier sent by the test device to the simulation device, or the eighth identifier is determined by the test device based on the seventh identifier, and the ACK is used to confirm that the compression model supported by the simulation device can match the decompression model associated with the eighth identifier.

[0254] In addition, the decompression model related to the seventh identifier sent by the test device to the simulation device is the second decompression model, and the compression model supported by the simulation device that matches the second decompression model is the second compression model, that is, the second compression model matches the second decompression model.

[0255] In another optional implementation, the test device sends a seventh identifier to the simulation device, and the compression model supported by the simulation device does not match the decompression model associated with the seventh identifier. In this case, the simulation device sends a negative response NACK to the test device, and the NACK is used to indicate that the compression model supported by the simulation device does not match the compression model associated with the seventh identifier. Optionally, after receiving the NACK from the simulation device, the test device sends the one or more identifiers to the simulation device so that the simulation device can reconfirm whether there is a decompression model in the supported decompression model that matches the compression model associated with the one or more identifiers. The one or more identifiers are part or all of the S sixth identifiers, or are determined by the test device based on part or all of the S sixth identifiers, and the one or more identifiers are different from the above-mentioned seventh identifier.

[0256] It can be seen that the simulation device feeds back to the test device through the eighth message that among the decompression models supported by the simulation device, there is a compression model that can match some or all of the multiple decompression models associated with the T eighth identifiers.

[0257] S504. The test device sends a ninth message to the network device. The ninth message includes a ninth identifier. The ninth identifier is one of the N eighth identifiers or the T seventh identifiers, or is determined by the TE based on one of the N eighth identifiers or the T seventh identifiers. The ninth identifier is associated with the second decompression model, and the second decompression model matches the second compression model. In response, the network device receives the ninth message from the test device.

[0258] In an optional embodiment, when T is an integer greater than 1 and less than X, the test device learns through the eighth message that the simulated device supports multiple compression models that can match multiple decompression models associated with N eighth identifiers. Thus, the test device selects an eighth identifier from the N eighth identifiers and sends it to the network device, or determines a ninth identifier based on one of the N eighth identifiers to activate the decompression model to be tested in the network device. It can be seen that the ninth identifier is one of the N eighth identifiers, or is determined by the test device based on one of the N eighth identifiers, and the second decompression model associated with the ninth identifier is the decompression model to be tested.

[0259] In another optional embodiment, when T is equal to 1, the simulation device confirms to the test device via the eighth message that the compression model supported by the simulation device matches the decompression model associated with the seventh identifier sent by the test device. Therefore, the ninth identifier sent by the test device to the network device is the seventh identifier sent by the test device to the simulation device, that is, the ninth identifier is one of the T seventh identifiers, and the second decompression model associated with the ninth identifier is the decompression model to be tested.

[0260] Optionally, the test device further sends a tenth message to the simulation device, the tenth message including a tenth identifier, the tenth identifier being one of the N eighth identifiers or T and the seventh identifier, or being determined by the TE based on the N eighth identifiers or T and one of the seventh identifiers, the tenth identifier being associated with the second compression model, and the second compression model matching the second decompression model. The tenth identifier is the same as or different from the ninth identifier.

[0261] For example, the N eighth identifiers include the eighth identifier 1 and the eighth identifier 2, both of which are related to the second decompression model, the ninth identifier is the eighth identifier 1, and the tenth identifier is the eighth identifier 2. Then, the test device sends the eighth identifier 1 to the network device and sends the eighth identifier 2 to the simulation device, which can both enable the network device to activate the second decompression model to be tested and enable the simulation device to determine the second compression model that matches the second decompression model.

[0262] For another example, the N eighth identifiers include the eighth identifier 1 and the eighth identifier 2, the eighth identifier 1 is related to the second decompression model, the eighth identifier 2 is not related to the second decompression model, the ninth identifier is the eighth identifier 1, and the tenth identifier is an identifier determined by the TE based on the eighth identifier 1. Then, the test device sends the ninth identifier to the network device and sends the tenth identifier to the simulation device, which can also cause the network device to activate the second decompression model to be tested and cause the simulation device to determine the second compression model that matches the second decompression model.

[0263] It is understood that the test device sends the ninth identifier to the network device, enabling the network device to determine and activate the second decompression model to be tested. Furthermore, the test device also sends the tenth identifier to the simulation device, enabling the simulation device to determine a second compression model that matches the second decompression model to be tested. Furthermore, during the test device's testing of the second decompression model, the simulation device can utilize the second compression model that matches the second decompression model, thereby reducing the impact of a mismatch in the dual-end model on the test results.

[0264] It can be seen that the test device, network device and simulation device can realize the second compression model used by the simulation device for compression processing through the above process of S501 to S504, and match it with the second decompression model to be tested.

[0265] In addition, the embodiments of the present application can also be applied to a test scenario in which a test device tests the performance of a compression model of a network device, and can also be applied to a test scenario in which a test device tests the performance of a decompression model of a terminal device.

[0266] When an embodiment of the present application is applied to a test device for testing the performance of a compression model of a network device, the test device transmits a reference signal to the network device, and the network device determines the CSI based on the reference signal. The network device calculates the uplink precoding matrix (PMI) based on the determined CSI and compresses the uncompressed uplink PMI using the compression model to be tested. The network device transmits the compressed uplink PMI to a simulation device or a test device. The simulation device decompresses the compressed uplink PMI using a decompression model that matches the compression model to be tested, thereby obtaining the decompressed uplink PMI.

[0267] Based on the decompressed uplink PMI, the simulation device determines the test data required to test the compression model under test and sends the determined test data to the test device. The test device sends the received test data to the network device, which receives the test data and achieves throughput 1. The test device uses a random uplink PMI to send test data to the network device, which receives the test data and achieves throughput 2. The test device combines throughput 1 and throughput 2 to test the performance of the compression model under test.

[0268] In this compression model testing method, the test equipment does not need to use a decompression model to decompress the compressed uplink PMI, nor does it need to determine test data based on the decompressed uplink PMI. This allows the test equipment to test the compression model to be tested with low complexity. Furthermore, the decompression model used by the simulation device matches the compression model used by the network device, which can reduce the impact of dual-end model mismatch on test results. The implementation method for matching the decompression model used by the simulation device with the compression model used by the network device can refer to the implementation method for matching the compression model used by the simulation device with the decompression model used by the network device described above, and will not be repeated here.

[0269] Optionally, the simulation device sends the decompressed uplink PMI to the test device. The test device sends test data to the network device based on the decompressed uplink PMI. The network device receives the test data and obtains Throughput 1. The test device uses a random uplink PMI to send test data to the network device. The network device receives the test data and obtains Throughput 2. The test device combines Throughput 1 and Throughput 2 to test the performance of the compression model to be tested.

[0270] In the test method of this compression model, the test equipment does not need to use a decompression model to decompress the compressed uplink PMI, so that the test equipment can test the compression model to be tested with low complexity. The test equipment directly sends test data based on the uplink PMI, eliminating the impact of the simulation device sending test data on the reliability and accuracy of the test results. In addition, the decompression model used by the simulation device matches the compression model used by the network device, which can reduce the impact of the dual-end model mismatch on the test results. Among them, the implementation method of matching the decompression model used by the simulation device with the compression model used by the network device can refer to the implementation method of matching the compression model used by the simulation device with the decompression model used by the network device mentioned above, and will not be repeated here.

[0271] When an embodiment of the present application is applied to a test device to test the performance of a decompression model of a terminal device, the test device sends a reference signal to a simulation device, and the simulation device determines the CSI based on the reference signal. The simulation device calculates the uplink precoding matrix PMI based on the determined CSI, and compresses the uncompressed uplink PMI using a compression model that matches the decompression model to be tested. The simulation device sends the compressed uplink PMI to the terminal device. Optionally, the simulation device sends the compressed uplink PMI to the test device, and the test device sends the received uplink PMI to the terminal device. The terminal device decompresses the compressed uplink PMI using the decompression model to be tested to obtain the decompressed uplink PMI.

[0272] Based on the decompressed uplink PMI, the terminal device determines the test data required to test the decompression model under test and sends the determined test data to the test device. The test device sends the received test data to the simulation device, which receives the test data and achieves Throughput 1. The test device uses a random uplink PMI to send test data to the simulation device, which receives the test data and achieves Throughput 2. The test device combines Throughput 1 and Throughput 2 to test the performance of the decompression model under test.

[0273] In the testing method of this decompression model, the test equipment does not need to use the compression model to compress the uncompressed uplink PMI, nor does it need to determine the test data based on the decompressed uplink PMI, so the test equipment can test the decompression model with low complexity. In addition, the compression model used by the simulation device matches the decompression model used by the terminal device, which can reduce the impact of the dual-end model mismatch on the test results. The implementation method of matching the compression model used by the simulation device with the decompression model used by the terminal device can refer to the implementation method of matching the decompression model used by the simulation device with the compression model used by the terminal device described above, and will not be repeated here.

[0274] Optionally, the terminal device sends the decompressed uplink PMI to the test device. The test device sends test data based on the decompressed uplink PMI to the simulation device. The simulation device receives the test data and obtains throughput 1. The test device uses a random uplink PMI to send test data to the simulation device. The simulation device receives the test data and obtains throughput 2. The test device combines throughput 1 and throughput 2 to test the performance of the decompression model under test.

[0275] In the test method of this decompression model, the test equipment does not need to use a compression model to compress the uncompressed uplink PMI, so that the test equipment can test the decompression model with low complexity. The test equipment directly sends test data based on the uplink PMI, eliminating the impact of the terminal device sending test data on the reliability and accuracy of the test results. In addition, the compression model used by the simulation device matches the decompression model used by the terminal device, which can reduce the impact of the dual-end model mismatch on the test results. The implementation method of matching the compression model used by the simulation device with the decompression model used by the terminal device can refer to the implementation method of matching the decompression model used by the simulation device with the compression model used by the terminal device mentioned above, and will not be repeated here.

[0276] With respect to the technical solutions described above, the corresponding device implementation solutions are further described below.

[0277] To implement the various functions of the methods provided in the embodiments of the present application, the test equipment and simulation equipment may include hardware structures and / or software modules, and the aforementioned functions may be implemented in the form of hardware structures, software modules, or a combination of hardware structures and software modules. Whether a particular function is implemented in the form of hardware structures, software modules, or a combination of hardware structures and software modules depends on the specific application and design constraints of the technical solution.

[0278] As shown in Figure 6, an embodiment of the present application provides a communication device 600. The communication device 600 can be a component of a test device (e.g., an integrated circuit, a chip, etc.), or a component of an analog device (e.g., an integrated circuit, a chip, etc.). The communication device 600 can also be other communication units for implementing the method in the method embodiment of the present application. The communication device 600 may include: a communication unit 601 and a processing unit 602. Optionally, a storage unit 603 may also be included.

[0279] In one possible design, one or more units shown in FIG6 may be implemented by one or more processors, or by one or more processors and memories, or by one or more processors and transceivers, or by one or more processors, memories, and transceivers, although this is not limited in the present embodiment. The processors, memories, and transceivers may be provided separately or integrated.

[0280] The communication device 600 has the function of implementing the test equipment described in the embodiment of the present application, or the function of the simulation device. For example, the communication device 600 includes a test device for executing the modules or units or means corresponding to the steps involved in the test equipment in the above-mentioned method embodiments. The function or unit or means can be implemented by software, or by hardware, or by hardware executing the corresponding software implementation, or by a combination of software and hardware. For details, please refer to the corresponding description in the above-mentioned corresponding method embodiment.

[0281] In one possible design, a communication device 600 may include: a processing unit 602 and a communication unit 601 , the device being applied to a test device;

[0282] The communication unit 601 is used to send a first reference signal to a terminal device, where the first reference signal is used by the terminal device to determine first channel state information CSI; the communication unit 601 is also used to receive a first precoding matrix indicator from an analog device, where the first precoding matrix indicator is determined by the analog device based on a second CSI, where the second CSI is obtained by the analog device decompressing a third CSI using a first decompression model, where the third CSI is obtained by the terminal device compressing the first CSI using a first compression model, where the first compression model matches the first decompression model; the processing unit 602 is used to test the performance of the first compression model based on the first precoding matrix indicator.

[0283] In an optional embodiment, before the communication unit 601 sends the first reference signal to the terminal device, it is further used to perform: receiving a first message from the terminal device, the first message including X first identifiers, one of the X first identifiers is related to a compression model supported by the terminal device, and X is a positive integer; sending a second message to the analog device, the second message including Y second identifiers, the Y second identifiers being part or all of the X first identifiers, or being determined by the TE based on part or all of the X first identifiers, and Y being an integer less than or equal to X; receiving a third message from the analog device, the third message including M third identifiers and / or a confirmation response A CK, the M third identifiers are part or all of the second identifiers among the Y second identifiers, or are determined by the simulation device based on part or all of the second identifiers among the Y second identifiers, the compression model associated with one of the M third identifiers matches the decompression model supported by the simulation device, and M is an integer less than or equal to Y; a fourth message is sent to the terminal device, the fourth message including a fourth identifier, the fourth identifier being one of the M third identifiers or the Y second identifiers, or being determined by the TE based on the M third identifiers or one of the Y second identifiers, and the fourth identifier being associated with the first compression model; the first compression model matches the first decompression model.

[0284] In an optional embodiment, the communication unit 601 is also used to send a fifth message to the simulation device, and the fifth message includes a fifth identifier, and the fifth identifier is one of the M third identifiers or the Y second identifiers, or is determined by the TE based on the M third identifiers or one of the Y second identifiers, and the fifth identifier is related to the first decompression model; the first decompression model matches the first compression model.

[0285] In an optional embodiment, one of the X first identifiers is used to identify one or more of the following: a compression model supported by the terminal device, parameter configuration information associated with a compression model supported by the terminal device, channel condition information associated with a compression model supported by the terminal device, state condition information associated with a compression model supported by the terminal device, information associated with the current state of the terminal device, and performance indicator information associated with a compression model supported by the terminal device.

[0286] In an optional implementation, the first decompression model is provided by the terminal device, or the network device, or other devices other than the terminal device and the network device.

[0287] In an optional implementation, before receiving the first precoding matrix indicator from the analog device, the communication unit 601 is further configured to: receive the third CSI from the terminal device; and send the third CSI to the analog device.

[0288] The embodiments of the present application and the above-mentioned method embodiments are based on the same concept, and the technical effects they bring are also the same. For the specific principles, please refer to the description of the above-mentioned embodiments, which will not be repeated here.

[0289] In another possible design, the communication device 600 may include: a processing unit 602 and a communication unit 601, and the device is applied to an analog device.

[0290] The communication unit 601 is used to receive third channel state information CSI from a test device TE or a terminal device, where the third CSI is obtained by the terminal device compressing the first CSI using a first compression model, and the first CSI is determined by the terminal device based on a first reference signal sent by the TE; the processing unit 602 is used to decompress the third CSI using a first decompression model to obtain a second CSI, and the first decompression model matches the first compression model; the communication unit 601 is also used to send a first precoding matrix indicator to the TE, where the first precoding matrix indicator is determined by the simulation device based on the second CSI, and the first precoding matrix indicator is used by the TE to test the performance of the first compression model.

[0291] In an optional embodiment, before receiving the third CSI, the communication unit 601 is further used to perform: receiving a second message from the TE, the second message including Y second identifiers, the Y second identifiers being part or all of the X first identifiers, or being determined by the TE based on part or all of the X first identifiers, one of the X first identifiers is associated with a compression model supported by the terminal device, X is a positive integer, and Y is an integer less than or equal to X; sending a third message to the TE, the third message including M third identifiers and / or an acknowledgment ACK, the M third identifiers being part or all of the Y second identifiers, or being determined by the simulation device based on part or all of the Y second identifiers, the compression model associated with one of the M third identifiers matches the decompression model supported by the simulation device, and M is an integer less than or equal to Y.

[0292] In an optional embodiment, the communication unit 601 is also used to receive a fifth message from the TE, the fifth message including a fifth identifier, the fifth identifier being one of the M third identifiers or the Y second identifiers, or being determined by the TE based on one of the M third identifiers or the Y second identifiers, and the fifth identifier being related to the first decompression model; the first decompression model matches the first compression model.

[0293] In an optional embodiment, one of the X first identifiers is used to identify one or more of the following: a compression model supported by the terminal device, parameter configuration information associated with a compression model supported by the terminal device, channel condition information associated with a compression model supported by the terminal device, state condition information associated with a compression model supported by the terminal device, information associated with the current state of the terminal device, and performance indicator information associated with a compression model supported by the terminal device.

[0294] In an optional embodiment, the processing unit 602 uses the first decompression model to decompress the third CSI, and before obtaining the second CSI, it is also used to obtain the first decompression model from the terminal device, or the network device, or other devices other than the terminal device and the network device.

[0295] The embodiments of the present application and the above-mentioned method embodiments are based on the same concept, and the technical effects they bring are also the same. For the specific principles, please refer to the description of the above-mentioned embodiments, which will not be repeated here.

[0296] In yet another possible design, the communication device 600 may include: a processing unit 602 and a communication unit 601 , and the device is applied to a test device;

[0297] The communication unit 601 is used to send a second reference signal to the analog device, where the second reference signal is used by the analog device to determine the fourth channel state information CSI; the communication unit 601 is also used to receive a second precoding matrix indicator from the network device, where the second precoding matrix indicator is determined by the network device based on the fifth CSI, the fifth CSI is obtained by the network device decompressing the sixth CSI using a second decompression model, and the sixth CSI is obtained by the analog device compressing the fourth CSI using a second compression model, and the second compression model matches the second decompression model; the processing unit 602 is used to test the performance of the second decompression model based on the second precoding matrix indicator.

[0298] In an optional embodiment, before the communication unit 601 sends the second reference signal to the analog device, it is also used to perform: receiving a sixth message from the network device, the sixth message including S sixth identifiers, one of the S sixth identifiers is related to a decompression model supported by the network device, and S is a positive integer; sending a seventh message to the analog device, the seventh message including T seventh identifiers, the T seventh identifiers being part or all of the S sixth identifiers, or being determined by the TE based on part or all of the S sixth identifiers, and T being an integer less than or equal to S; receiving an eighth message from the analog device, the eighth message including N eighth identifiers and / or a confirmation response A CK, the N eighth identifiers are part or all of the seventh identifiers among the T seventh identifiers, or are determined by the simulation device based on part or all of the seventh identifiers among the T seventh identifiers, and the decompression model associated with one of the N eighth identifiers matches the compression model supported by the simulation device, and N is an integer less than or equal to T; a ninth message is sent to the network device, and the ninth message includes a ninth identifier, and the ninth identifier is one of the N eighth identifiers or the T seventh identifiers, or is determined by the TE based on the N eighth identifiers or one of the T seventh identifiers, and the ninth identifier is associated with the second decompression model; the second decompression model matches the second compression model.

[0299] In an optional embodiment, the communication unit 601 is also used to send a tenth message to the simulation device, the tenth message including a tenth identifier, the tenth identifier being one of the N eighth identifiers or the T seventh identifiers, or being determined by the TE based on one of the N eighth identifiers or the T seventh identifiers, the tenth identifier being related to the second compression model; the second compression model matches the second decompression model.

[0300] In an optional embodiment, one of the S sixth identifiers is used to identify one or more of the following: a decompression model supported by the network device, parameter configuration information associated with a decompression model supported by the network device, channel condition information associated with a decompression model supported by the network device, state condition information associated with a decompression model supported by the network device, information associated with the current state of the network device, and performance indicator information associated with a decompression model supported by the network device.

[0301] In an optional implementation, the second compression model is provided by the terminal device, or the network device, or other devices other than the terminal device and the network device.

[0302] In an optional implementation, before receiving the second precoding matrix indicator from the network device, the communication unit 601 is further configured to: receive the sixth CSI from the simulation device; and send the sixth CSI to the network device.

[0303] The embodiments of the present application and the above-mentioned method embodiments are based on the same concept, and the technical effects they bring are also the same. For the specific principles, please refer to the description of the above-mentioned embodiments, which will not be repeated here.

[0304] In yet another possible design, the communication apparatus 600 may include: a processing unit 602 and a communication unit 601 , wherein the apparatus is applied to a simulation device;

[0305] The communication unit 601 is used to receive a second reference signal from the test equipment TE; the processing unit 602 is used to compress the fourth channel state information CSI using the second compression model to obtain the sixth CSI, where the fourth CSI is obtained by the simulation device based on the second reference signal, and the second compression model matches the second decompression model used by the network device; the communication unit 601 is also used to send the sixth CSI to the TE or the network device, and the sixth CSI is used by the TE to test the performance of the second decompression model.

[0306] In an optional embodiment, the communication unit 601 is also used to execute: receiving a seventh message from the TE, the seventh message including T seventh identifiers, the T seventh identifiers being part or all of the S sixth identifiers, or being determined by the TE based on part or all of the S sixth identifiers, one of the S sixth identifiers being associated with a decompression model supported by the network device, S being a positive integer, and T being an integer less than or equal to S; sending an eighth message to the simulation device, the eighth message including N eighth identifiers and / or an acknowledgment ACK, the N eighth identifiers being part or all of the T seventh identifiers, or being determined by the simulation device based on part or all of the T seventh identifiers, the decompression model associated with one of the N eighth identifiers matching the compression model supported by the simulation device, and N being an integer less than or equal to T.

[0307] In an optional embodiment, the communication unit 601 is also used to perform: receiving a tenth message from the TE, the tenth message including a tenth identifier, the tenth identifier being one of the N eighth identifiers or the T seventh identifiers, or being determined by the TE based on one of the N eighth identifiers or the T seventh identifiers, the tenth identifier being related to the second compression model; the second compression model is matched with the second decompression model.

[0308] In an optional embodiment, one of the S sixth identifiers is used to identify one or more of the following: a decompression model supported by the network device, parameter configuration information associated with a decompression model supported by the network device, channel condition information associated with a decompression model supported by the network device, state condition information associated with a decompression model supported by the network device, information associated with the current state of the network device, and performance indicator information associated with a decompression model supported by the network device.

[0309] In an optional embodiment, the processing unit 602 uses the second compression model to compress the fourth CSI. Before obtaining the sixth CSI, the method further includes: obtaining the second compression model from the terminal device, or the network device, or other devices other than the terminal device and the network device.

[0310] The embodiments of the present application and the above-mentioned method embodiments are based on the same concept, and the technical effects they bring are also the same. For the specific principles, please refer to the description of the above-mentioned embodiments, which will not be repeated here.

[0311] The present application also provides a communication device 700. Figure 7 is a schematic diagram of the structure of communication device 700. Communication device 700 can be a test device, or a chip, chip system, or processor that supports the test device in implementing the above-mentioned method; or it can be a simulation device, or a chip, chip system, or processor that supports the simulation device in implementing the above-mentioned method. This device can be used to implement the method described in the above-mentioned method embodiment. For details, please refer to the description of the above-mentioned method embodiment.

[0312] The communication device 700 may include one or more processors 701. The processor 701 may be a general-purpose processor or a dedicated processor. For example, it may be a baseband processor, a digital signal processor, an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, or a central processing unit (CPU). The baseband processor may be used to process communication protocols and communication data, while the central processing unit may be used to control a communication device (e.g., a base station, a baseband chip, a terminal, a terminal chip, a distributed unit (DU), or a centralized unit (CU), etc.), execute software programs, and process data from the software programs.

[0313] Optionally, the communication device 700 may include one or more memories 702 , on which instructions 704 may be stored. The instructions may be executed on the processor 701 , so that the communication device 700 executes the method described in the above method embodiment.

[0314] Optionally, data may also be stored in the memory 702. The processor 701 and the memory 702 may be provided separately or integrated together.

[0315] Optionally, the communication device 700 may further include a transceiver 705 and an antenna 706. The transceiver 705 may be referred to as a transceiver unit, a transceiver, or a transceiver circuit, etc., and is configured to implement transceiver functions. The transceiver 705 may include a receiver and a transmitter. The receiver may be referred to as a receiver or a receiving circuit, etc., and is configured to implement a receiving function; the transmitter may be referred to as a transmitter or a transmitting circuit, etc., and is configured to implement a transmitting function.

[0316] In one possible design, the communication device 700 can be applied to a testing device. Specifically, the processor 701 is used to execute S208 in the above-mentioned model testing method 100, and to execute S408 in the above-mentioned model testing method 200; the transceiver 705 is used to execute S201 and S207 in the above-mentioned model testing method 100, and to execute S401 and S407 in the above-mentioned model testing method 200.

[0317] In another possible design, the communication device 700 can be applied to a simulation device. Specifically, the processor 701 is used to execute S205 and S206 in the above-mentioned model testing method 100, and to execute S402 and S403 in the above-mentioned model testing method 200; the transceiver 705 is used to execute S207 in the above-mentioned model testing method 100, and to execute S401, S404 and S407 in the above-mentioned model testing method 200.

[0318] Optionally, the processor 701 may store an instruction 703. The instruction 703 runs on the processor 701, which may enable the communication device 700 to perform the method described in the above method embodiment. The instruction 703 may be fixed in the processor 701. In this case, the processor 701 may be implemented by hardware.

[0319] The embodiments of the present application and the method embodiments shown in the above-mentioned model testing method 100 and model testing method 200 are based on the same concept and bring the same technical effects. For the specific principles, please refer to the description of the embodiments shown in the above-mentioned model testing method 100 and model testing method 200, which will not be repeated here.

[0320] An embodiment of the present application further provides a computer-readable storage medium for storing computer software instructions, which, when executed by a communication device, implements the functions of any of the above method embodiments.

[0321] An embodiment of the present application also provides a computer program product for storing computer software instructions, which, when executed by a communication device, implements the functions of any of the above method embodiments.

[0322] The embodiments of the present application also provide a computer program that, when executed on a computer, implements the functions of any of the above method embodiments.

[0323] The present application also provides a communication system that may include a test device and a simulation device. The test device and the simulation device may be used to perform the relevant steps in the above-described model testing methods 100 and 200, and are not described in detail herein. In another possible design, the system may further include other devices that interact with the test device and the simulation device, such as a device to be tested.

[0324] The terms "first" and "second" in the description, claims and drawings of the embodiments of this application are used to distinguish different objects, rather than to describe a specific order. "First", "second" and the like are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second" and the like may explicitly or implicitly include one or more of the features. In the description of this embodiment, unless otherwise specified, "multiple" means two or more.

[0325] Furthermore, the terms "comprise," "comprises," and "comprising," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements is not limited to the listed steps or elements, but may optionally include steps or elements not listed, or may optionally include other steps or elements inherent to the process, method, product, or apparatus.

[0326] Reference to an "embodiment" in the embodiments of this application means that a particular feature, structure, or characteristic described in conjunction with the embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it mean that each embodiment is mutually exclusive of another embodiment or an alternative embodiment. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0327] In the embodiments of the present application, "at least one (item)" refers to one or more, "more than one" refers to two or more, and "and / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can represent: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0328] In the embodiments of this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of this application should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner to facilitate understanding.

[0329] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware, or any combination thereof. When implemented using software, all or part of the embodiments may be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated therein. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a high-density digital video disc (DVD)), or a semiconductor medium (eg, an SSD).

[0330] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A model testing method, characterized in that: The method is applied to test equipment TE, comprising: Sending a first reference signal to a terminal device, where the first reference signal is used by the terminal device to determine first channel state information CSI; Receiving a first precoding matrix indicator from an analog device, where the first precoding matrix indicator is determined by the analog device according to a second CSI, where the second CSI is obtained by the analog device decompressing a third CSI using a first decompression model, where the third CSI is obtained by the terminal device compressing the first CSI using a first compression model, where the first compression model matches the first decompression model; Based on the first precoding matrix indicator, performance of the first compression model is tested.

2. The method according to claim 1, characterized in that Before sending the first reference signal to the terminal device, the method further includes: receiving a first message from the terminal device, the first message including X first identifiers, one of the X first identifiers being associated with a compression model supported by the terminal device, where X is a positive integer; Sending a second message to the simulation device, where the second message includes Y second identifiers, where the Y second identifiers are part or all of the X first identifiers, or are determined by the TE according to part or all of the X first identifiers, and Y is an integer less than or equal to X; Receive a third message from the simulation device, the third message including M third identifiers and / or an acknowledgment ACK, the M third identifiers are part or all of the Y second identifiers, or are determined by the simulation device according to part or all of the Y second identifiers, a compression model associated with a third identifier among the M third identifiers matches a decompression model supported by the simulation device, and M is an integer less than or equal to Y; Sending a fourth message to the terminal device, where the fourth message includes a fourth identifier, where the fourth identifier is one of the M third identifiers or the Y second identifiers, or is determined by the TE according to one of the M third identifiers or the Y second identifiers, and the fourth identifier is related to the first compression model; The first compression model matches the first decompression model.

3. The method according to claim 2, characterized in that The method further comprises: Sending a fifth message to the simulation device, where the fifth message includes a fifth identifier, where the fifth identifier is one of the M third identifiers or the Y second identifiers, or is determined by the TE according to one of the M third identifiers or the Y second identifiers, and the fifth identifier is related to the first decompression model; The first decompression model matches the first compression model.

4. The method according to claim 2, characterized in that: One of the X first identifiers is used to identify one or more of the following: a compression model supported by the terminal device, parameter configuration information associated with a compression model supported by the terminal device, channel condition information associated with a compression model supported by the terminal device, state condition information associated with a compression model supported by the terminal device, information associated with a current state of the terminal device, and performance indicator information associated with a compression model supported by the terminal device.

5. The method according to any one of claims 1 to 3, characterized in that: The first decompression model is provided by the terminal device, or the network device, or other devices except the terminal device and the network device.

6. The method according to claim 1, characterized in that Before receiving the first precoding matrix indicator from the analog device, the method further includes: receiving the third CSI from the terminal device; The third CSI is sent to the analog device.

7. A model testing method, characterized in that: The method is applied to a simulation device, comprising: receiving third channel state information CSI from a test device TE or a terminal device, wherein the third CSI is obtained by the terminal device compressing the first CSI using a first compression model, and the first CSI is obtained by the terminal device based on a first reference signal sent by the TE determined; Decompressing the third CSI using a first decompression model to obtain a second CSI, wherein the first decompression model matches the first compression model; A first precoding matrix indicator is sent to the TE, where the first precoding matrix indicator is determined by the simulation device according to the second CSI, and the first precoding matrix indicator is used by the TE to test the performance of the first compression model.

8. The method according to claim 7, characterized in that Before receiving the third CSI, the method further includes: receiving a second message from the TE, where the second message includes Y second identifiers, where the Y second identifiers are part or all of the X first identifiers, or are determined by the TE according to part or all of the X first identifiers, where one of the X first identifiers is related to a compression model supported by the terminal device, where X is a positive integer, and Y is an integer less than or equal to X; A third message is sent to the TE, wherein the third message includes M third identifiers and / or an acknowledgment response ACK, wherein the M third identifiers are part or all of the Y second identifiers, or are determined by the simulation device based on part or all of the Y second identifiers, and a compression model associated with a third identifier among the M third identifiers matches the decompression model supported by the simulation device, and M is an integer less than or equal to Y.

9. The method according to claim 8, characterized in that The method further comprises: receiving a fifth message from the TE, where the fifth message includes a fifth identifier, where the fifth identifier is one of the M third identifiers or the Y second identifiers, or is determined by the TE according to one of the M third identifiers or the Y second identifiers, and the fifth identifier is related to the first decompression model; The first decompression model matches the first compression model.

10. The method according to claim 8, characterized in that One of the X first identifiers is used to identify one or more of the following: a compression model supported by the terminal device, parameter configuration information associated with a compression model supported by the terminal device, channel condition information associated with a compression model supported by the terminal device, state condition information associated with a compression model supported by the terminal device, information associated with a current state of the terminal device, and performance indicator information associated with a compression model supported by the terminal device.

11. The method according to any one of claims 7 to 9, characterized in that: Before decompressing the third CSI by using the first decompression model to obtain the second CSI, the method further includes: The first decompression model is obtained from the terminal device, or the network device, or other devices except the terminal device and the network device.

12. A model testing method, characterized in that: The method is applied to test equipment TE, comprising: Sending a second reference signal to the analog device, where the second reference signal is used by the analog device to determine fourth channel state information CSI; receiving a second precoding matrix indicator from a network device, where the second precoding matrix indicator is determined by the network device according to a fifth CSI, where the fifth CSI is obtained by the network device decompressing a sixth CSI using a second decompression model, where the sixth CSI is obtained by the simulation device compressing the fourth CSI using a second compression model, where the second compression model matches the second decompression model; Based on the second precoding matrix indicator, performance of the second decompression model is tested.

13. The method according to claim 12, characterized in that Before sending the second reference signal to the analog device, the method further includes: receiving a sixth message from the network device, the sixth message including S sixth identifiers, one sixth identifier among the S sixth identifiers being associated with a decompression model supported by the network device, and S being a positive integer; Sending a seventh message to the simulation device, where the seventh message includes T seventh identifiers, where the T seventh identifiers are some or all of the S sixth identifiers, or are determined by the TE according to some or all of the S sixth identifiers, where T is an integer less than or equal to S; Receive an eighth message from the simulation device, the eighth message including N eighth identifiers and / or confirmation responses ACK, the N eighth identifiers and / or confirmation responses ACK The eighth identifier is part or all of the T seventh identifiers, or is determined by the simulation device according to part or all of the T seventh identifiers, and a decompression model associated with an eighth identifier among the N eighth identifiers matches the compression model supported by the simulation device, and N is an integer less than or equal to T; Sending a ninth message to the network device, where the ninth message includes a ninth identifier, where the ninth identifier is one of the N eighth identifiers or the T seventh identifiers, or is determined by the TE according to one of the N eighth identifiers or the T seventh identifiers, and the ninth identifier is related to the second decompression model; The second decompression model matches the second compression model.

14. The method according to claim 13, characterized in that The method further comprises: Sending a tenth message to the simulation device, where the tenth message includes a tenth identifier, where the tenth identifier is one of the N eighth identifiers or the T seventh identifiers, or is determined by the TE according to one of the N eighth identifiers or the T seventh identifiers, and the tenth identifier is related to the second compression model; The second compression model matches the second decompression model.

15. The method according to claim 13, characterized in that A sixth identifier among the S sixth identifiers is used to identify one or more of the following: a decompression model supported by the network device, parameter configuration information associated with a decompression model supported by the network device, channel condition information associated with a decompression model supported by the network device, state condition information associated with a decompression model supported by the network device, information associated with a current state of the network device, and performance indicator information associated with a decompression model supported by the network device.

16. The method according to any one of claims 12 to 14, characterized in that: The second compression model is provided by the terminal device, or the network device, or other devices except the terminal device and the network device.

17. The method according to claim 12, characterized in that Before receiving the second precoding matrix indicator from the network device, the method further includes: receiving the sixth CSI from the simulation device; The sixth CSI is sent to the network device.

18. A model testing method, characterized in that: The method is applied to a simulation device, and the method comprises: receiving a second reference signal from the test equipment TE; compressing fourth channel state information CSI using a second compression model to obtain sixth CSI, where the fourth CSI is determined by the simulation device based on the second reference signal, and the second compression model matches a second decompression model used by the network device; The sixth CSI is sent to the TE or the network device, where the sixth CSI is used by the TE to test the performance of the second decompression model.

19. The method according to claim 18, characterized in that The method further comprises: receiving a seventh message from the TE, where the seventh message includes T seventh identifiers, where the T seventh identifiers are some or all of the S sixth identifiers, or are determined by the TE according to some or all of the S sixth identifiers, where one of the S sixth identifiers is related to a decompression model supported by the network device, where S is a positive integer, and T is an integer less than or equal to S; An eighth message is sent to the simulation device, wherein the eighth message includes N eighth identifiers and / or an acknowledgment response ACK, wherein the N eighth identifiers are part or all of the T seventh identifiers, or are determined by the simulation device based on part or all of the T seventh identifiers, and a decompression model associated with an eighth identifier among the N eighth identifiers matches the compression model supported by the simulation device, and N is an integer less than or equal to T.

20. The method according to claim 19, characterized in that The method further comprises: receiving a tenth message from the TE, the tenth message including a tenth identifier, the tenth identifier being one of the N eighth identifiers or the T seventh identifiers, or being determined by the TE according to one of the N eighth identifiers or the T seventh identifiers, and the tenth identifier being related to the second compression model; The second compression model matches the second decompression model.

21. The method according to claim 19, characterized in that A sixth identifier among the S sixth identifiers is used to identify one or more of the following: a decompression model supported by the network device, parameter configuration information associated with a decompression model supported by the network device, channel condition information associated with a decompression model supported by the network device, state condition information associated with a decompression model supported by the network device, information associated with a current state of the network device, and performance indicator information associated with a decompression model supported by the network device.

22. The method according to any one of claims 18 to 20, characterized in that Before compressing the fourth CSI by using the second compression model to obtain the sixth CSI, the method further includes: The second compression model is obtained from a terminal device, or the network device, or other devices except the terminal device and the network device.

23. A communication device, characterized in that: The communication device includes a module for executing the method according to any one of claims 1 to 6, or includes a module for executing the method according to any one of claims 7 to 11, or includes a module for executing the method according to any one of claims 12 to 17, or includes a module for executing the method according to any one of claims 18 to 22.

24. A communication device, characterized in that: The communication device includes a processor, which is configured to perform the method according to any one of claims 1 to 6, or configured to perform the method according to any one of claims 7 to 11, or configured to perform the method according to any one of claims 12 to 17, or configured to perform the method according to any one of claims 18 to 22.

25. A communication system, characterized in that: The method comprises at least one of the following devices: a device for executing the method described in any one of claims 1 to 6, a device for executing the method described in any one of claims 7 to 11, a device for executing the method described in any one of claims 12 to 17, and a device for executing the method described in any one of claims 18 to 22.

26. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store instructions, which, when executed on a computer, causes the method according to any one of claims 1 to 6 to be executed, or causes the method according to any one of claims 7 to 11 to be executed, or causes the method according to any one of claims 12 to 17 to be executed, or causes the method according to any one of claims 18 to 22 to be executed.

27. A computer program product comprising instructions, characterized in that When it runs on a computer, the method according to any one of claims 1 to 6 is executed, or the method according to any one of claims 7 to 11 is executed, or the method according to any one of claims 12 to 17 is executed, or the method according to any one of claims 18 to 22 is executed.

Images