Data transmission method and communication device

Abstract

The present application relates to a data transmission method and a communication device. The method comprises: a first communication device sends processing data, the processing data being obtained by processing a plurality of pieces of information to be processed. According to embodiments of the present application, the plurality of pieces of information to be processed are processed, and then the processing data is transmitted, so that the amount of data needing to be transmitted can be reduced, thereby reducing the signaling overhead of a system.

Description

Data transmission method and communication device Technical Field

[0001] The present application relates to the field of communications, and more specifically, to a data transmission method and a communication device. Background Art

[0002] Artificial Intelligence (AI)-based solutions are increasingly being used in wireless communication systems. For example, AI can be used to implement channel state information (CSI) feedback and CSI prediction. However, information such as the performance evaluation results of AI solutions can result in significant data transmission and signaling overhead.

[0003] Summary of the Invention

[0004] The embodiments of the present application provide a data transmission method and a communication device, which can reduce the amount of data that needs to be transmitted.

[0005] An embodiment of the present application provides a data transmission method, including: a first communication device sends processed data, where the processed data is obtained by processing a plurality of information to be processed.

[0006] An embodiment of the present application provides a data transmission method, including: a second communication device receives processed data, where the processed data is obtained after a first communication device processes a plurality of information to be processed.

[0007] An embodiment of the present application provides a first communication device, including: a sending unit, configured to send processed data, where the processed data is obtained by processing a plurality of information to be processed.

[0008] An embodiment of the present application provides a second communication device, including: a receiving unit, configured to receive processing data, where the processing data is obtained after the first communication device processes a plurality of information to be processed.

[0009] An embodiment of the present application provides a communication device, comprising a processor and a memory, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory, so that the communication device executes the above-mentioned data transmission method.

[0010] The embodiment of the present application provides a chip for implementing the above-mentioned data transmission method. Specifically, the chip includes: a processor for calling and running a computer program from a memory, so that a device equipped with the chip executes the above-mentioned data transmission method.

[0011] An embodiment of the present application provides a computer-readable storage medium for storing a computer program. When the computer program is executed by a device, the device executes the above-mentioned data transmission method.

[0012] An embodiment of the present application provides a computer program product, including computer program instructions, which enable a computer to execute the above-mentioned data transmission method.

[0013] An embodiment of the present application provides a computer program, which, when executed on a computer, enables the computer to execute the above-mentioned data transmission method.

[0014] In the embodiment of the present application, after processing a plurality of information to be processed, the processed data is transmitted, which can reduce the amount of data to be transmitted and thus reduce the system signaling overhead. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] FIG1 is a schematic diagram of an application scenario according to an embodiment of the present application.

[0016] FIG2A is a schematic flow chart of the CSI feedback problem.

[0017] FIG2B is a schematic flow chart of the CSI prediction problem.

[0018] FIG3 is a schematic flowchart of a data transmission method according to an embodiment of the present application.

[0019] FIG4 is a schematic flowchart of a data transmission method according to an embodiment of the present application.

[0020] FIG5A is a schematic diagram of a UE transmitting first data to a network.

[0021] 5B to 5E are diagrams illustrating examples of a UE transmitting first data to a network.

[0022] FIG6A is a schematic diagram of a network transmitting second data to a UE.

[0023] 6B to 6E are diagrams illustrating examples of the network transmitting second data to the UE.

[0024] FIG7 is a schematic block diagram of a first communication device according to an embodiment of the present application.

[0025] FIG8 is a schematic block diagram of a second communication device according to an embodiment of the present application.

[0026] FIG9 is a schematic block diagram of a communication device according to an embodiment of the present application.

[0027] FIG10 is a schematic block diagram of a chip according to an embodiment of the present application.

[0028] FIG11 is a schematic block diagram of a communication system according to an embodiment of the present application. DETAILED DESCRIPTION

[0029] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

[0030] The technical solutions of the embodiments of the present application can be applied to various communication systems, such as: Global System of Mobile communication (GSM) system, Code Division Multiple Access (CDMA) system, Wideband Code Division Multiple Access (WCDMA) system, General Packet Radio Service (GPRS), Long Term Evolution (LTE) system, Advanced Long Term Evolution (LTE-A) system, New Radio (NR) system, NR system evolution system, LTE on unlicensed spectrum (LTE-U) system, NR on unlicensed spectrum (NR-U) system, Non-Terrestrial Networks (NTN) system, Universal Mobile Telecommunication System (UMTS), Wireless Local Area Networks (WLAN), Wireless Fidelity (Wireless Fidelity) system. Fidelity, WiFi), fifth-generation communication (5th-Generation, 5G) system or other communication systems, etc.

[0031] Generally speaking, traditional communication systems support a limited number of connections and are easy to implement. However, with the development of communication technology, mobile communication systems will not only support traditional communications, but will also support, for example, device-to-device (D2D) communication, machine-to-machine (M2M) communication, machine-type communication (MTC), vehicle-to-vehicle (V2V) communication, or vehicle-to-everything (V2X) communication, etc. The embodiments of the present application can also be applied to these communication systems.

[0032] In one embodiment, the communication system in the embodiment of the present application can be applied to a carrier aggregation (CA) scenario, a dual connectivity (DC) scenario, and a standalone (SA) networking scenario.

[0033] In one embodiment, the communication system in the embodiment of the present application can be applied to an unlicensed spectrum, wherein the unlicensed spectrum can also be considered as a shared spectrum; or, the communication system in the embodiment of the present application can also be applied to an authorized spectrum, wherein the authorized spectrum can also be considered as an unshared spectrum.

[0034] The embodiments of the present application describe various embodiments in conjunction with network devices and terminal devices, wherein the terminal device may also be referred to as user equipment (UE), access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication device, user agent or user device, etc.

[0035] The terminal device can be a station (STAION, ST) in a WLAN, a cellular phone, a cordless phone, a Session Initiation Protocol (SIP) phone, a Wireless Local Loop (WLL) station, a Personal Digital Assistant (PDA) device, a handheld device with wireless communication capabilities, a computing device or other processing device connected to a wireless modem, a vehicle-mounted device, a wearable device, a terminal device in a next-generation communication system such as an NR network, or a terminal device in a future evolved Public Land Mobile Network (PLMN) network, etc.

[0036] In an embodiment of the present application, the terminal device can be deployed on land, including indoors or outdoors, handheld, wearable or vehicle-mounted; it can also be deployed on the water surface (such as ships, etc.); it can also be deployed in the air (such as airplanes, balloons and satellites, etc.).

[0037] In an embodiment of the present application, the terminal device may be a mobile phone, a tablet computer, a computer with wireless transceiver function, a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal device in industrial control, a wireless terminal device in self-driving, a wireless terminal device in remote medical, a wireless terminal device in a smart grid, a wireless terminal device in transportation safety, a wireless terminal device in a smart city, or a wireless terminal device in a smart home, etc.

[0038] As an example and not a limitation, in the embodiment of the present application, the terminal device may also be a wearable device. Wearable devices may also be called wearable smart devices, which are a general term for wearable devices that are intelligently designed and developed using wearable technology for daily wear, such as glasses, gloves, watches, clothing, and shoes. A wearable device is a portable device that is worn directly on the body or integrated into the user's clothes or accessories. Wearable devices are not only hardware devices, but also achieve powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include those that are fully functional, large in size, and can achieve complete or partial functions without relying on smartphones, such as smart watches or smart glasses, as well as those that only focus on a certain type of application function and need to be used in conjunction with other devices such as smartphones, such as various smart bracelets and smart jewelry for vital sign monitoring.

[0039] In an embodiment of the present application, the network device may be a device for communicating with a mobile device. The network device may be an access point (AP) in WLAN, a base station (BTS) in GSM or CDMA, a base station (NodeB, NB) in WCDMA, an evolved base station (eNB or eNodeB) in LTE, or a relay station or access point, or a vehicle-mounted device, a wearable device, and a network device (gNB) in an NR network, or a network device in a future evolved PLMN network or a network device in an NTN network, etc.

[0040] As an example and not a limitation, in an embodiment of the present application, the network device may have a mobile feature, for example, the network device may be a mobile device. Alternatively, the network device may be a satellite or a balloon station. For example, the satellite may be a low earth orbit (LEO) satellite, a medium earth orbit (MEO) satellite, a geostationary earth orbit (GEO) satellite, a high elliptical orbit (HEO) satellite, etc. Optionally, the network device may also be a base station set up in a location such as land or water.

[0041] In an embodiment of the present application, a network device may provide services for a cell, and a terminal device may communicate with the network device through the transmission resources (e.g., frequency domain resources, or spectrum resources) used by the cell. The cell may be a cell corresponding to a network device (e.g., a base station). The cell may belong to a macro base station or a base station corresponding to a small cell. The small cells here may include: a metro cell, a micro cell, a pico cell, a femto cell, etc. These small cells have the characteristics of small coverage and low transmission power, and are suitable for providing high-speed data transmission services.

[0042] FIG1 exemplarily illustrates a communication system 100. The communication system includes a network device 110 and two terminal devices 120. In one embodiment, the communication system 100 may include multiple network devices 110, and each network device 110 may include a different number of terminal devices 120 within its coverage area, which is not limited in this embodiment of the present application.

[0043] In one embodiment, the communication system 100 may further include other network entities such as a Mobility Management Entity (MME) and an Access and Mobility Management Function (AMF), which is not limited in this embodiment of the present application.

[0044] Among them, the network equipment may include access network equipment and core network equipment. That is, the wireless communication system also includes multiple core networks for communicating with the access network equipment. The access network equipment can be an evolutionary base station (evolutional node B, abbreviated as eNB or e-NodeB) macro base station, micro base station (also called "small base station"), pico base station, access point (AP), transmission point (TP) or new generation base station (new generation Node B, gNodeB), etc. in a long-term evolution (LTE) system, a next-generation (mobile communication system) (next radio, NR) system or an authorized auxiliary access long-term evolution (LAA-LTE) system.

[0045] It should be understood that in the embodiments of the present application, a device having a communication function in a network / system may be referred to as a communication device. Taking the communication system shown in Figure 1 as an example, the communication device may include a network device and a terminal device having a communication function. The network device and the terminal device may be specific devices in the embodiments of the present application and will not be described in detail here. The communication device may also include other devices in the communication system, such as a network controller, a mobility management entity, and other network entities, which are not limited in the embodiments of the present application.

[0046] It should be understood that the terms "system" and "network" are often used interchangeably herein. The term "and / or" is simply a description of an association between related objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " generally indicates that the related objects are in an "or" relationship.

[0047] It should be understood that the "indication" mentioned in the embodiments of this application can be a direct indication, an indirect indication, or an indication of an association. For example, "A indicates B" can mean that A directly indicates B, for example, B can be obtained through A; it can also mean that A indirectly indicates B, for example, A indicates C, and B can be obtained through C; it can also mean that there is an association between A and B.

[0048] In the description of the embodiments of the present application, the term "corresponding" may indicate a direct or indirect correspondence between the two, or an association relationship between the two, or a relationship between indication and being indicated, configuration and being configured, etc.

[0049] To facilitate understanding of the technical solutions of the embodiments of the present application, the relevant technologies of the embodiments of the present application are described below. The following relevant technologies can be arbitrarily combined with the technical solutions of the embodiments of the present application as optional solutions, and they all fall within the protection scope of the embodiments of the present application.

[0050] The CSI feedback problem, as shown in Figure 2A, can achieve AI-based CSI information compression and feedback through the AI ​​encoder and AI decoder. The CSI prediction problem, as shown in Figure 2B. The measured CSI (Meastured CSI) is input into the CSI prediction model (CSI prediction model) to obtain the predicted CSI (Predicted CSI). The subscript W represents the time, for example, W t-10 Indicates the CSI information at time (t-10).

[0051] The performance of AI / machine learning (ML) solutions can be affected by scenarios, data, and application conditions, and these effects can fluctuate, perturb, and erratically change. For example, AI / ML-based solutions can include AI / ML-based CSI compression and recovery, AI / ML-based prediction, or other AI / ML-based solutions. In actual systems, if an AI / ML solution is deemed to have failed due to a brief performance dip, this can lead to unnecessary failure determinations and unnecessary solution switching. This can also result in unnecessary model and solution updates, reconfiguration, and additional signaling overhead. For example, if a CSI compression and recovery solution resumes normal operation after a brief period of ineffectiveness, there's no need to determine that the solution has failed and trigger a new alternative, or to first determine that the solution has failed, then trigger a new alternative, and then switch back to the original solution. The current basic solution involves performing AI / ML performance evaluations, such as through multiple evaluations (including multiple transmissions of evaluation data and multiple performance assessments) to obtain relatively stable evaluation results.

[0052] While multiple performance evaluations can yield relatively stable AI / ML solution performance evaluation results, they can also incur significant evaluation-related data transmission and signaling overhead. For example, in CSI compression, recovery, or CSI prediction solutions, if the difference between original CSI information and compressed and recovered CSI information is used as the evaluation target, multiple evaluations require the UE to transmit the original CSI information to the network multiple times, or the network to transmit compressed and recovered CSI information to the UE multiple times. The combined air interface overhead caused by these frequent transmissions and the amount of information transmitted each time can result in significant evaluation overhead.

[0053] FIG3 is a schematic flow chart of a data transmission method 300 according to an embodiment of the present application. The method can optionally be applied to the system shown in FIG1 , but is not limited thereto. The method includes at least part of the following contents.

[0054] S310: The first communication device sends processed data, where the processed data is obtained by processing a plurality of pieces of information to be processed.

[0055] In an embodiment of the present application, a first communication device may send processed data to a second communication device. For example, the first communication device is a terminal device, and the second communication device is a network device. In another example, the first communication device is a network device, and the second communication device is a terminal device. In another example, the first communication device is a first terminal device, and the second communication device is a second terminal device. The multiple pieces of information to be processed may include multiple pieces of information that the first communication device needs to send to the second communication device. The first communication device may first process the multiple pieces of information to be processed to obtain processed data, which may then be sent to the second communication device. This can reduce the amount of data sent, thereby reducing signaling overhead.

[0056] In one embodiment, the plurality of information to be processed includes data related to model performance monitoring, training, testing, or use in the first communication device. In this embodiment of the present application, the first communication device and / or the second communication device may have a model for communication scenario problems. For example, a model related to CSI. The plurality of information to be processed can be various types of data used in the performance monitoring, training, testing, or use phases of the model, such as model input data, output data, label data, evaluation data, etc.

[0057] In one embodiment, one piece of information to be processed among the plurality of information to be processed includes at least one of the following: a channel; a feature vector representing channel state information (CSI); input information of a model; and output information of a model.

[0058] In the embodiments of the present application, the multiple pieces of information to be processed may include multiple channels. For example, channel H1 is obtained by the UE through channel estimation. Another example is channel H2 obtained by the network through a channel information recovery scheme or prediction scheme. The multiple pieces of information to be processed may include multiple eigenvectors representing CSI. For example, a eigenvector representing CSI information is obtained by the UE through SVD decomposition of channel H1. Another example is a eigenvector representing CSI information is obtained by the network through a CSI information recovery scheme or prediction scheme.

[0059] In one embodiment, the channel includes at least one of the following:

[0060] An initial channel composed of time domain information, frequency domain information, angle domain information, and antenna domain information;

[0061] For the initial channel, information extraction is performed in the time domain, and a channel is obtained by selecting a subset of the time domain information;

[0062] For the initial channel, information extraction is performed in the frequency domain, and a channel is obtained by selecting a subset of the frequency domain information;

[0063] For the initial channel, information extraction is performed in the angle domain, and a channel is obtained by selecting a subset of the angle domain information;

[0064] For the initial channel, information extraction is performed in the antenna domain, and a channel is obtained by selecting a subset of antenna domain information;

[0065] The channel obtained by quantizing the initial channel. For example, the channel obtained by the terminal device by performing K-bit quantization on channel H1. Another example is the channel obtained by the network device by performing K-bit quantization on channel H2.

[0066] For example, the channels formed by time domain information may include channels across multiple symbols and multiple time slots in the time domain. The channels formed by frequency domain information may include channels across multiple subcarriers and multiple RBs in the frequency domain. The channels formed by angle domain information may include channels across multiple transmission angles and multiple reception angles in the angle domain. The channels formed by antenna domain information may include channels across multiple transmission antennas and transmission ports in the antenna domain.

[0067] In one embodiment, the feature vector representing the CSI includes at least one of the following:

[0068] Performing singular value decomposition (SVD) on the channel to obtain the eigenvector representing the CSI;

[0069] A feature vector representing the CSI information obtained by the CSI information recovery scheme;

[0070] The feature vector representing the CSI information is obtained through the CSI information prediction scheme.

[0071] For example, the first communications device is a terminal device, which can perform SVD on channel H1 to obtain a feature vector representing the CSI. For another example, the first communications device is a network device. During an AI / ML-based CSI compression recovery process, the network device can obtain a feature vector representing the CSI information using a CSI information recovery solution. During an AI / ML-based CSI prediction process, the network device can obtain a feature vector representing the CSI information using a CSI information prediction solution.

[0072] In one embodiment, the input information of the model includes the channel and / or the feature vector representing the CSI. For example, the input information of the model may also be referred to as label data, input data, initial data, expected data, etc.

[0073] In one embodiment, the output information of the model includes the channel and / or the feature vector representing the CSI. For example, the output information of the model may also be referred to as evaluation data, output data, prediction data, etc.

[0074] In one embodiment, the processing method of the plurality of information to be processed includes:

[0075] Obtaining the average value of the plurality of information to be processed;

[0076] The average value of the processed elements of the multiple pieces of information to be processed is obtained.

[0077] In one embodiment, element processing of the plurality of information to be processed includes at least one of the following:

[0078] Taking absolute values ​​of elements in the plurality of information to be processed;

[0079] Modulo values ​​of elements in the plurality of information to be processed;

[0080] Taking the square of the absolute value of the elements in the plurality of information to be processed;

[0081] Square the modulus value of the elements in the plurality of information to be processed;

[0082] An element in the plurality of pieces of information to be processed is multiplied by the conjugate of the element.

[0083] In an embodiment of the present application, the elements of multiple information to be processed can be directly averaged, or the elements of multiple information to be processed can be first processed by taking absolute values, taking modulos, square operations, conjugate operations, etc., and then the average is obtained. The above-mentioned method of element processing is only an example and not a limitation, and other processing can also be performed on the elements as needed. In an embodiment of the present application, performing average processing on multiple information to be processed is beneficial to reducing the amount of data. After performing element processing on multiple information to be processed, the average processing can be better reflected by taking absolute values, taking modulos, square operations, conjugate operations, etc. For example, the original data may be positive or negative. Taking the absolute value first and then taking the average error will not cancel out the positive and negative errors, but will accumulate.

[0084] For example, the elements of a to-be-processed information D1 include {d11, d12, d13, d14}. After taking the absolute value of the to-be-processed information, the absolute values ​​of the elements of D1 can be obtained to include {|d11|, |d12|, |d13|, |d14|}. After taking the absolute value of the to-be-processed information and then squaring it, the square of the absolute values ​​of the elements of D1 can be obtained to include {|d11| 2 ,|d12| 2 ,|d13| 2 ,|d14| 2 After taking the modulus of the information to be processed, the modulus values ​​of the elements of D1 can be obtained to include {||d11||,||d12||,||d13||,||d14||}. After taking the modulus of the information to be processed and then squaring it, the square of the modulus values ​​of the elements of D1 can be obtained to include {||d11|| 2 ,||d12|| 2 ,||d13|| 2 ,||d14|| 2 The element of the information to be processed is multiplied by the conjugate dot product of the element [|d11.*d11'|,|d12.*d12'|,|d13.*d13'|,|d14.*d14'|]. Where d11.*d11' represents the conjugate dot product of d11 and its conjugate point d11'.

[0085] In one embodiment, obtaining the mean value includes at least one of the following:

[0086] The information is averaged according to the number of samples;

[0087] The information is averaged by internal elements;

[0088] After averaging the information by internal elements, the average is taken by the number of samples.

[0089] In one embodiment, averaging the information by internal elements includes:

[0090] The information is averaged by the total number of internal elements;

[0091] The information is averaged according to the number of dimensional elements of the internal elements.

[0092] In the embodiment of the present application, the method for obtaining the mean of the plurality of information to be processed may include taking the mean according to the number of samples, the number of internal elements, the number of dimensional elements, etc. For example, the number of dimensional elements may include the number of subbands, the number of transmission ports, the virtual and real dimensions of a complex number, etc.

[0093] For example, if N is the number of samples, we can take the mean of N pieces of information to be processed, based on the number of samples N, to obtain one piece of processed information. Assume that the multiple pieces of information to be processed include D1, D2, and D3. The elements of D1 include {d11, d12, d13, d14}; the elements of D2 include {d21, d22, d23, d24}; and the elements of D3 include {d31, d32, d33, d34}. If we take the mean of the elements of D1, D2, and D3 based on N = 3, we get {D11, D12, D13, D14}. Here, D11 = (d11 + d21 + d31) ÷ 3, D12 = (d12 + d22 + d32) ÷ 3, D13 = (d13 + d23 + d33) ÷ 3, and D14 = (d14 + d24 + d34) ÷ 3.

[0094] For example, if we first take the absolute values ​​of the elements in D1, D2, and D3, and then take the average of the absolute values ​​of the elements in D1, D2, and D3 based on the sample number N = 3, we can get {D21, D22, D23, D23}. Among them, D21 = (|d11| + |d21| + |d31|) ÷ 3, D22 = (|d12| + |d22| + |d32|) ÷ 3, D23 = (|d13| + |d23| + |d33|) ÷ 3, and D24 = (|d14| + |d24| + |d34|) ÷ 3.

[0095] For example, M represents the element data in a piece of information to be processed. The M elements in each of the N pieces of information to be processed can be averaged according to the number of elements M, yielding N pieces of processed information. If the elements in D1, D2, and D3 are averaged according to M = 4, we obtain {D31, D32, D33}. Here, D31 = (d11 + d12 + d13 + d14) ÷ 4, D32 = (d21 + d22 + d23 + d24) ÷ 4, and D33 = (d31 + d32 + d33 + d34) ÷ 4.

[0096] For example, M = d1*d2, where d1 and d2 are the dimensions. Each element in the N pieces of information to be processed can be first modulo-calculated and then averaged according to the dimension d1 or d2 to obtain N processed information. If the elements of D1, D2, and D3 are first modulo-calculated and then averaged according to d2 = 4, {D41, D42, D43} can be obtained. D41 = (||d11|| + ||d12|| + ||d13|| + ||d14||) ÷ 4, D42 = (||d21|| + ||d22|| + ||d23|| + ||d24||) ÷ 4, and D43 = (||d31|| + ||d32|| + ||d33|| + ||d34||) ÷ 4.

[0097] Furthermore, N processed information can be further averaged to obtain 1 processed information.

[0098] In one embodiment, the first communication device is a terminal device, the multiple information to be processed includes multiple first information to be processed in the terminal device, and the processed data includes first data obtained after processing the multiple first information using a first processing method. In an embodiment of the present application, the first information may include at least one of a first channel, a first eigenvector representing CSI, and input information of a model. The processing method of the terminal device for the first channel, the first eigenvector representing CSI, and the input information of the model can refer to the relevant descriptions and examples above. For example, the terminal device can send the first data to a network device or other terminal device.

[0099] In one embodiment, the first information includes at least one of label data, auxiliary data, and input data of the model in the terminal device. In an embodiment of the present application, at least one of the label data, auxiliary data, and input data can be input into the model for processing.

[0100] In one embodiment, the model in the terminal device is an encoder of a CSI generation model. For example, at least one of the label data, auxiliary data, and input data of the CSI generation model may include CSI that needs to be processed. The CSI that needs to be processed is input into the AI ​​encoder of the CSI generation model as shown in Figure 2A. The CSI generation model can be used in scenarios such as CSI feedback and CSI prediction. After encoding by the AI ​​encoder, compressed CSI is obtained. The terminal device can send the compressed CSI to a network device or other terminal for decoding. The terminal device can also process multiple CSIs that need to be processed, that is, multiple first information, and send them to the network device or other terminal device to assist the network device or other terminal device in performing model performance evaluation.

[0101] In one embodiment, the first data is transmitted via at least one of the following:

[0102] Radio Resource Control (RRC) messages; Uplink Control Information (UCI); Uplink messages during random access; Physical Uplink Control Channel (PUCCH); Physical Uplink Shared Channel (PUSCH); Uplink channels dedicated to AI and / or ML; Terminal device capability reporting messages, Sidelink Control Information (SCI) messages, Physical Sidelink Control Channel (PSCCH), and Physical Sidelink Shared Channel (PSSCH).

[0103] For example, the terminal device may report the first data to the network device through an RRC message, UCI, PUCCH, PUSCH, an uplink message in a random access process, an uplink channel dedicated to AI, an uplink channel dedicated to ML, or a terminal device capability reporting message. For another example, the first terminal device may send the first data to the second terminal device through SCI, PSCCH, or PSSCH. If a large amount of data needs to be transmitted, the first data may be transmitted through an RRC message, PUSCH, or PSSCH. If high timeliness is required, the first data may be transmitted through UCI, PUCCH, an uplink message in a random access process, an uplink channel dedicated to AI, an uplink channel dedicated to ML, a terminal device capability reporting message, SCI, or PSCCH, etc.

[0104] In one embodiment, the first communication device is a network device, the multiple pieces of information to be processed include multiple pieces of second information to be processed in the network device, and the processed data includes second data obtained by processing the multiple pieces of second information using a second processing method. In this embodiment of the present application, the first information may include at least one of a second channel, a second eigenvector representing CSI, and output information of a model. The manner in which the terminal device processes the second channel, the second eigenvector representing CSI, and the output information of the model can be found in the relevant descriptions and examples above. For example, the network device may send the first data to the terminal device.

[0105] In one embodiment, the second information includes at least one of evaluation data, auxiliary data, and output data of the model in the network device. In this embodiment of the present application, the model in the network device and the model in the terminal device can be the same, related, or different. The model in the network device can output at least one of the evaluation data, auxiliary data, and output data.

[0106] In one embodiment, the model in the network device is a decoder of a CSI recovery model and / or a CSI prediction model. In an embodiment of the present application, the model in the network device and the model in the terminal device may be the same, related, or different. For example, the first model in the terminal device includes an encoder of a CSI generation model, and the second model in the network device includes a decoder of a CSI recovery model and / or a CSI prediction model. For example, as shown in Figure 2A, after the terminal device sends the compressed CSI obtained after encoding processing by the AI ​​encoder to the network device, the network device can input the compressed CSI into a decoder of a CSI recovery model and / or a CSI prediction model for decoding processing to obtain decoded CSI. The network device can also compare the decoded CSI with multiple received CSIs that need to be processed to obtain a performance evaluation result.

[0107] In an embodiment of the present application, the network device may also process multiple decoded CSIs or multiple performance evaluation results and send them to the terminal device to assist the terminal device in performing performance evaluation of the model.

[0108] In one embodiment, the second data is transmitted via at least one of the following:

[0109] Broadcast message; RRC message; Medium Access Control (MAC) control element (CE); Downlink Control Information (DCI) message; Downlink message during random access; Physical Downlink Control Channel (PDCCH); Physical Downlink Shared Channel (PDSCH); Downlink channel dedicated to AI and / or ML; Capability indication on the network side; SCI; PSCCH; PSSCH.

[0110] For example, after the network device processes multiple second information to obtain the second data, it can send the second data to the terminal device through a broadcast message, RRC message, MAC CE, DCI, PDCCH, PDSCH, a downlink message in a random access process, an AI-dedicated downlink channel, and an ML-dedicated downlink channel. The capability indication of the network side is sent to the terminal device. For another example, the second terminal device can send the second data to the first terminal device through SCI, PSCCH or PSSCH. Among them, if a large amount of data needs to be transmitted, the second data can be transmitted through a broadcast message, RRC message, PDSCH or PSSCH, etc. If high timeliness is required, the second data can be transmitted through MAC CE, DCI, PDCCH, a downlink message in a random access process, an AI-dedicated downlink channel, an ML-dedicated downlink channel, a capability indication of the network side, SCI or PSCCH, etc.

[0111] FIG4 is a schematic flow chart of a data transmission method 400 according to an embodiment of the present application. The method can optionally be applied to the system shown in FIG1 , but is not limited thereto. The method includes at least part of the following contents.

[0112] S410: The second communication device receives processed data, where the processed data is obtained after the first communication device processes a plurality of information to be processed.

[0113] In one embodiment, the plurality of information to be processed includes data related to model performance monitoring, training, testing or use in the second communication device.

[0114] In one embodiment, one piece of information to be processed among the plurality of pieces of information to be processed includes at least one of the following:

[0115] channel;

[0116] The feature vector representing the CSI;

[0117] Model input information;

[0118] Output information of the model.

[0119] In one embodiment, the channel includes at least one of the following:

[0120] An initial channel composed of time domain information, frequency domain information, angle domain information, and antenna domain information;

[0121] For the initial channel, information extraction is performed in the time domain, and a channel is obtained by selecting a subset of the time domain information;

[0122] For the initial channel, information extraction is performed in the frequency domain, and a channel is obtained by selecting a subset of the frequency domain information;

[0123] For the initial channel, information extraction is performed in the angle domain, and a channel is obtained by selecting a subset of the angle domain information;

[0124] For the initial channel, information extraction is performed in the antenna domain, and a channel is obtained by selecting a subset of antenna domain information;

[0125] The channel obtained by quantizing the initial channel.

[0126] In one embodiment, the eigenvector representing the CSI includes a eigenvector representing the CSI obtained by performing SVD on the channel.

[0127] In one embodiment, the input information of the model includes the channel and / or the feature vector representing the CSI.

[0128] In one embodiment, the output information of the model includes the channel and / or the feature vector representing the CSI.

[0129] In one embodiment, the processing method of the plurality of information to be processed includes:

[0130] Obtaining the average value of the plurality of information to be processed;

[0131] The average value of the processed elements of the multiple pieces of information to be processed is obtained.

[0132] In one embodiment, element processing of the plurality of information to be processed includes at least one of the following:

[0133] Taking absolute values ​​of elements in the plurality of information to be processed;

[0134] Modulo values ​​of elements in the plurality of information to be processed;

[0135] Taking the square of the absolute value of the elements in the plurality of information to be processed;

[0136] Square the modulus value of the elements in the plurality of information to be processed;

[0137] An element in the plurality of pieces of information to be processed is multiplied by the conjugate of the element.

[0138] In one embodiment, obtaining the mean value includes at least one of the following:

[0139] The information is averaged according to the number of samples;

[0140] The information is averaged by internal elements;

[0141] After averaging the information by internal elements, the average is taken by the number of samples.

[0142] In one embodiment, averaging the information by internal elements includes:

[0143] The information is averaged by the total number of internal elements;

[0144] The information is averaged according to the number of dimensional elements of the internal elements.

[0145] In one embodiment, the second communication device is a terminal device, the multiple information to be processed include multiple first information to be processed in the terminal device, and the processed data includes first data obtained after processing the multiple first information using a first processing method.

[0146] In one embodiment, the first information includes at least one of label data, auxiliary data, and input data of the model in the terminal device.

[0147] In one embodiment, the model in the terminal device is an encoder of a CSI generation model.

[0148] In one embodiment, the first data is transmitted via at least one of the following:

[0149] RRC messages; uplink control information (UCI); uplink messages during random access; PUCCH; PUSCH; uplink channels dedicated to artificial intelligence (AI) and / or machine learning (ML); terminal device capability reporting messages, SCI, PSCCH, and PSSCH.

[0150] In one embodiment, the second communication device is a network device, the multiple information to be processed include multiple second information to be processed in the network device, and the processed data include second data obtained by processing the multiple second information using a second processing method.

[0151] In one embodiment, the second information includes at least one of evaluation data, auxiliary data, and output data of the model in the network device.

[0152] In one embodiment, the model in the network device is a decoder of a CSI recovery model and / or a CSI prediction model.

[0153] In one embodiment, the second data is transmitted via at least one of the following:

[0154] Broadcast message; RRC message; MAC CE; DCI; downlink message during random access; PDCCH; PDSCH; downlink channel dedicated to AI and / or ML; capability indication on the network side; SCI; PSCCH; PSSCH.

[0155] For a specific example of the communication device executing the data transmission method 400 in this embodiment, reference may be made to the relevant description of the communication device in the above-mentioned data transmission method 300 , which will not be repeated here for the sake of brevity.

[0156] The data transmission method of the embodiment of the present application may include the following parts:

[0157] 1. Evaluation data construction method.

[0158] The design of this construction method includes: avoiding multiple transmissions of label data and evaluation data in the performance evaluation of AI / ML solutions, and using the processed data of label data and evaluation data as label data and evaluation data.

[0159] The basic idea behind this design approach is that when AI / ML solution evaluation relies not strictly on a single evaluation result but on the combined results of multiple evaluations, the transmission of multiple evaluation data can be avoided. This approach integrates multiple evaluation data before transmitting the evaluation information, replacing the transmission of redundant air interface information with information preprocessing. Furthermore, appropriate signaling design is employed to ensure that the UE and the network have a consistent understanding of the available and selected evaluation solutions.

[0160] 1. UE transmits to the network

[0161] In AI / ML-based CSI compression, recovery, CSI prediction, and other issues, a UE transmits first data to a network. The first data is obtained by processing first information in a first manner.

[0162] The first information A1 may be one of the following:

[0163] (1) Channel H1. For example, the channel H1 obtained by the UE through channel estimation. Another example is the channel H1 composed of time domain information, frequency domain information, angle domain information, and antenna domain information. Another example is the channel H1 obtained by extracting information from the above H1 in the time domain, frequency domain, angle domain, and antenna domain, and selecting a partial time domain information subset, frequency domain information subset, angle domain information subset, and antenna domain information subset to obtain the channel H1 (which can also be marked as channel H1-1, etc., to distinguish it from the H1 before extraction). Another example is the H1 obtained by performing k-bit quantization on the above H1 (which can also be marked as channel H1-2, etc., to distinguish it from the H1 before quantization).

[0164] (2) Feature vector W1 representing CSI information. For example, the UE obtains the feature vector W1 representing CSI information by performing SVD decomposition on the channel H1.

[0165] (3) AI / ML model input information X1. For example, the AI / ML CSI compression and recovery model input information is H1 or W1.

[0166] The first processing method mentioned above can be one of the following methods:

[0167] (1) Obtaining the mean value of the first information A1, as shown in the following example:

[0168] For example, when the first information A1 consists of M1 elements, for example, the first information A1 is a matrix of dimensions [s1, s2, s3], and s1*s2*s3=M1.

[0169] The first data B1 may be at least one of the following examples:

[0170] The first data B1 is obtained by taking the average of the N first information A1 according to the number of samples (taking the average according to N).

[0171] N first information A1, each first information A1 consists of M1 elements, and the first data B1 is obtained by taking the average of the internal elements of each first information A1 (taking the average according to M1, or taking the average according to s1, or taking the average according to s2, or taking the average according to s3).

[0172] N first information A1s are formed, each of which is composed of M1 elements. The mean of each first information A1 is taken according to the internal elements (the mean is taken according to M1, or the mean is taken according to s1, or the mean is taken according to s2, or the mean is taken according to s3). The mean of the N first information A1s is taken according to the number of samples (the mean is taken according to N), and the first data B1 is obtained.

[0173] (2) Obtaining the mean of the elements of the first information A1 after processing, as shown in the following example:

[0174] For example, when the first information A1 consists of M1 elements, for example, the first information A1 is a matrix of dimensions [s1, s2, s3], and s1*s2*s3=M1.

[0175] The first data B1 may be at least one of the following examples:

[0176] N pieces of first information A1 are provided, each piece of first information A1 consisting of M1 elements. Processing is performed on the M1 elements of each piece of first information A1, such as taking the absolute value of the element in the sample, taking the modulus of the element in the sample, taking the square of the absolute value of the element in the sample, taking the square of the modulus of the element in the sample, or multiplying the element in the sample by the conjugate of the element. The N processed samples are averaged (averaged over N) to obtain first data B1.

[0177] N pieces of first information A1, each piece of first information A1 consisting of M1 elements, are processed on the M1 elements of each piece of first information A1, such as taking the absolute value, modulus, square of the absolute value, square of the modulus, square of the modulus of an element in the sample, or multiplying an element in the sample by its conjugate. For the N samples after the processing described above, the first data B1 is obtained by taking the average of the elements within the sample (taking the average according to M1, or taking the average according to s1, or taking the average according to s2, or taking the average according to s3).

[0178] N first information A1, each consisting of M1 elements, is processed on the M1 elements of each first information A1, such as taking the absolute value, modulus, square of the absolute value, square of the modulus, square of the modulus of an element in the sample, or multiplication of the conjugate of the element in the sample by the element. For the N samples after the above processing, the average of the elements within the sample is taken (average is taken according to M1, or average is taken according to s1, or average is taken according to s2, or average is taken according to s3), and the average is taken according to the number of samples (average is taken according to N), to obtain the first data B1.

[0179] 1.2 Network transmission to UE

[0180] In AI / ML-based CSI compression, recovery, CSI prediction, and other issues, the network transmits second data to the UE. The second data is obtained by processing the second information in a second manner.

[0181] The second information A2 may be one of the following:

[0182] (1) Channel H2.

[0183] For example, in the AI / ML-based channel information compression and recovery process, the network obtains channel H2 through the channel information recovery scheme. Another example is channel H2 composed of time domain information, frequency domain information, angle domain information, and antenna domain information. Another example is information extraction of the above H2 in the time domain, frequency domain, angle domain, and delay domain, and selecting a partial time domain information subset, frequency domain information subset, angle domain information subset, and antenna domain information subset to obtain channel H2 (which can also be marked as channel H2-1, etc., to distinguish it from H2 before extraction). Another example is H2 obtained by k-bit quantization of the above H2 (which can also be marked as channel H2-2, etc., to distinguish it from H2 before quantization).

[0184] For example, in the AI / ML-based channel information prediction process, the network obtains channel H2 through the channel information prediction scheme. Another example is channel H2 composed of time domain, frequency domain information, angle domain, and delay domain information. For example, information is extracted from the above H2 in the time domain, frequency domain information, angle domain, and delay domain, and a partial time domain information subset, frequency domain information subset, angle domain information subset, and antenna domain information subset is selected to obtain channel H2 (which can also be marked as channel H2-3, etc. to distinguish it from H2 before extraction). Another example is H2 obtained by performing k-bit quantization on the above H2 (which can also be marked as channel H2-4, etc. to distinguish it from H2 before quantization).

[0185] (2) The feature vector W2 representing the CSI information is as follows:

[0186] For example, in the AI / ML-based CSI compression recovery process, the network obtains the feature vector W2 representing the CSI information through the CSI information recovery scheme.

[0187] For example, in the AI / ML-based CSI prediction process, the network obtains the feature vector W2 representing the CSI information through the CSI information prediction scheme.

[0188] (3) Output information X2 of the AI / ML model.

[0189] For example, the AI / ML CSI compression and recovery model outputs H2 or W2. For example, the AI / ML CSI prediction model outputs H2 or W2.

[0190] In Examples 1 and 2, although A1 and A2 are described similarly, they can be different types of information. For example, A1 is the original information obtained by the UE, which can be reference information and / or tag information. A2 is obtained by the network through methods such as CSI feedback or CSI prediction, which can be lossy and is the information to be evaluated. Performance evaluation can compare the differences between A1 and A2.

[0191] The second processing method mentioned above can be one of the following methods:

[0192] (1) Obtaining the mean value of the second information A2, as shown in the following example:

[0193] For example, when the second information A2 consists of M2 elements, for example, the second information A2 is a matrix of dimensions [q1, q2, q3], and q1*q2*q3=M2.

[0194] The second data B2 may be at least one of the following:

[0195] The second data B2 is obtained by taking the average of the N second information A2 according to the number of samples (taking the average according to N).

[0196] N second information A2, each second information A2 consists of M2 elements. The second data B2 is obtained by taking the mean of the internal elements of each second information A2 (taking the mean by M2, or taking the mean by q1, or taking the mean by q2, or taking the mean by q3).

[0197] N pieces of second information A2 are provided, each piece of second information A2 consisting of M2 elements. For each piece of second information A2, the internal element mean is taken (the mean is taken by M2, or the mean is taken by q1, or the mean is taken by q2, or the mean is taken by q3). The N pieces of second information A2 are averaged by the number of samples (the mean is taken by N) to obtain second data B2.

[0198] (2) Obtaining the mean of the elements of the second information A2 after processing, as shown in the following example:

[0199] For example, when the second information A2 consists of M2 elements, for example, the second information A2 is a matrix of dimensions [q1, q2, q3], and q1*q2*q3=M2.

[0200] The second data B2 may be at least one of the following:

[0201] N pieces of second information A2 are present, each piece of second information A2 consisting of M2 elements. The M2 elements of each piece of second information A2 are processed. For example, the absolute value of the element in the sample is taken, or the modulus of the element in the sample is taken, or the square of the absolute value of the element in the sample is taken, or the square of the modulus of the element in the sample is taken, or the conjugate multiplication of the element in the sample by the element in the sample is performed. The average of the N samples after the above processing is taken (average is taken based on N) to obtain the second data B2.

[0202] N pieces of second information A2 are provided, each piece of second information A2 consisting of M2 elements. The M2 elements of each piece of second information A2 are processed. For example, the absolute value, modulus, square of the absolute value, square of the modulus, square of the modulus of an element in the sample, or conjugate multiplication of an element in the sample by the conjugate of the element are taken. For the N samples processed as described above, the second data B2 is obtained by taking the mean of the elements within the sample (taking the mean by M2, taking the mean by q1, taking the mean by q2, or taking the mean by q3).

[0203] N pieces of second information A2 are generated, each piece of second information A2 consists of M2 elements. The M2 elements of each piece of second information A2 are processed. For example, the absolute value, the modulus, the square of the absolute value, the square of the modulus, the square of the modulus of the element in the sample, or the conjugate multiplication of the element in the sample and the element are taken. For the N samples processed as above, the mean is taken by the elements within the sample (mean by M2, mean by q1, mean by q2, or mean by q3), and the mean is taken by the number of samples (mean by N), to obtain the second data B2.

[0204] Compared with the method of transmitting multiple samples for AI / ML performance monitoring, the solution of the embodiment of the present application can use the processing results of transmitting multiple samples for statistical processing, thereby greatly reducing the air interface overhead required for transmitting the data to be evaluated, label data, etc.

[0205] 1.3 Some examples of specific AI / ML-based CSI compression, recovery, and prediction solutions are as follows:

[0206] Example 1:

[0207] In AI / ML-based CSI feedback solutions or CSI prediction solutions, such as those where AI / ML performance evaluation is performed on the network side, the UE transmits a feature vector W1 representing the CSI information to the network. W1 may have M1 elements and dimensions [s1, s2], where s1 is the number of subbands and s2 is the number of transmit ports. W1 may also have dimensions [s1, s2, s3], where s1 is the number of subbands, s2 is the number of transmit ports, and s3 is the complex number (real and virtual).

[0208] The first information W1 consists of M1 elements. For example, W1 can be expressed as [w_1, w_2, …, w_M1]. The M1 elements of W1 are processed, for example, the absolute value of each element is [|w_1|, |w_2|, …, |w_M1|], or the modulus of each element is [||w_1||, ||w_2||, …, ||w_M1||], or the square of the absolute value of each element is [|w_1| 2 ,|w_2| 2 ,…,|w_M1| 2 ], or the square of the modulus value of each element [||w_1|| 2 ,||w_2|| 2 ,…,||w_M1|| 2 ], or the conjugate dot product of each element and the element [|w_1.*w_1'|,|w_2.*w_2'|,…,|w_M1.*w_M1'|], etc., to form the processed first information W1 (which can also be marked as W1', etc. to distinguish it from the W1 before processing, and the following examples are similar). N pieces of the above-mentioned processed first information W1. For example, N pieces of the above-mentioned processed first information W1_1,…W1_N are averaged according to the N samples to obtain the second data B1, for example

[0209] Example 2:

[0210] In AI / ML-based CSI feedback solutions or CSI prediction solutions, such as those where AI / ML performance evaluation is performed on the UE side, the network needs to transmit a feature vector W2 representing the CSI information to the UE. For example, the number of elements in W2 is M2, and the dimensions of W2 are [q1, q2], where q1 is the number of subbands and q2 is the number of transmit ports. For example, the dimensions of W2 are [q1, q2, q3], where q1 is the number of subbands, q2 is the number of transmit ports, and q3 is a complex number with both real and virtual dimensions.

[0211] The second information W2 consists of M2 elements. For example, W2 can be expressed as [w_1, w_2, ..., w_M2]. The M2 elements of W2 are processed, for example, the absolute value of each element is [|w_1|, |w_2|, ..., |w_M2|], or the modulus of each element is [||w_1||, ||w_2||, ..., ||w_M2||], or the square of the absolute value of each element is [|w_1| 2 ,|w_2| 2 ,…,|w_M2| 2 ], or the square of the modulus value of each element [||w_1|| 2 ,||w_2|| 2 ,…,||w_M2|| 2 ], or the conjugate dot product of each element and the element [|w_1.*w_1'|,|w_2.*w_2'|,…,|w_M2.*w_M2'|], etc., to form the processed first information W2. N pieces of the above processed first information W2, such as W2_1,…W2_N, are averaged according to the N samples to obtain the second data B2, such as

[0212] Example 3:

[0213] In AI / ML-based CSI feedback solutions or CSI prediction solutions, such as those where AI / ML performance evaluation is performed on the network side, the UE transmits representative channel information H1 to the network. H1 may have, for example, M1 elements and dimensions [s1, s2], where the first dimension consists of s1 time-domain information or frequency-domain information, and the second dimension consists of s2 angle-domain information or antenna-domain information. H1 may have, for example, dimensions [s1, s2, s3], where the first dimension consists of s1 time-domain information or frequency-domain information, the second dimension consists of s2 angle-domain information or antenna-domain information, and the third dimension consists of the real and imaginary parts of the complex number.

[0214] The first information H1 consists of M1 elements. For example, H1 can be expressed as [h_1, h_2, …, h_M1]. The M1 elements of H1 are processed, for example, the absolute value of each element is [|h_1|, |h_2|, …, |h_M1|], or the modulus of each element is [||h_1||, ||h_2||, …, ||h_M1||], or the square of the absolute value of each element is [|h_1| 2 ,|h_2| 2 ,…,|h_M1| 2 ], or the square of the modulus value of each element [||h_1|| 2 ,||h_2|| 2 ,…,||h_M1|| 2], or the conjugate dot product of each element and the element [|h_1.*h_1'|,|h_2.*h_2'|,…,|h_M1.*h_M1'|], etc., to form the processed first information H1. N pieces of the above processed first information H1, for example, N pieces of the above processed first information H1_1,…H1_N, are averaged according to the N samples to obtain the second data B1, for example

[0215] Example 4:

[0216] In AI / ML-based CSI feedback solutions or CSI prediction solutions, such as AI / ML performance evaluation performed on the UE side, the network transmits representative channel information H2 to the UE. For example, the number of elements in H2 is M2, and the dimensions of H2 are [s1, s2], where the first dimension is s1 time-domain or frequency-domain information, and the second dimension is s2 angle-domain or antenna-domain information. For example, the dimensions of H2 are [s1, s2, s3], where the first dimension is s1 time-domain or frequency-domain information, the second dimension is s2 angle-domain or antenna-domain information, and the third dimension is the real and imaginary parts of the complex number.

[0217] The second information H2 consists of M2 elements. For example, H2 can be expressed as [h_1, h_2, ..., h_M2]. The M2 elements of H2 are processed, for example, the absolute value of each element is [|h_1|, |h_2|, ..., |h_M2|], or the modulus of each element is [||h_1||, ||h_2||, ..., ||h_M2||], or the square of the absolute value of each element is [|h_1| 2 ,|h_2| 2 ,…,|h_M2| 2 ], or the square of the modulus value of each element [||h_1|| 2 ,||h_2|| 2 ,…,||h_M2|| 2 ], or the conjugate dot product of each element and the element [|h_1.*h_1'|,|h_2.*h_2'|,…,|h_M2.*h_M2'|], etc., to form the processed first information H2. N pieces of the above processed first information H2, for example, N pieces of the above processed first information H2_1,…H2_N, are averaged according to the N samples to obtain the second data B2, for example

[0218] 2: Evaluation data indication and / or transmission method

[0219] The first part mainly includes the evaluation data construction method, and the second part can specifically describe the indication and / or transmission scheme of the evaluation data.

[0220] 2.1 The UE transmits first data to a network, such as a base station, as shown in FIG5A .

[0221] (1) The UE indicates the provided tag data type and / or configuration, as shown in FIG5B . The UE then sends the first data to the network. Alternatively, the UE may also send the first data to the network while indicating the provided tag data type and / or configuration.

[0222] (2) The network indicates the required tag data type and / or configuration, as shown in Figure 5C. Then, the UE sends the first data to the network.

[0223] (3) The UE reports the supported tag data types and / or configurations, and the network indicates the required tag data types and / or configurations, as shown in Figure 5D. Then, the UE sends the first data to the network.

[0224] (4) The network indicates supported tag data types and / or configurations, and the UE indicates the provided tag data type and / or configuration, as shown in FIG5E . The UE then sends the first data to the network. Alternatively, the UE may also send the first data to the network while indicating the provided tag data type and / or configuration.

[0225] 2.2 The network, such as the base station, transmits the second data to the UE, as shown in FIG6A .

[0226] (1) The UE indicates the required evaluation data type and / or configuration, as shown in Figure 6B. Then, the network sends the second data to the UE.

[0227] (2) The network indicates the type of evaluation data and / or configuration provided, as shown in FIG6C . The network then sends the second data to the UE. Alternatively, the network may also send the second data to the UE while indicating the type of evaluation data and / or configuration provided.

[0228] (3) The UE reports supported evaluation data types and / or configurations, and the network indicates the provided evaluation data types and / or configurations, as shown in FIG6D . The network then sends the second data to the UE. Alternatively, the network may also send the second data to the UE while indicating the provided evaluation data types and / or configurations.

[0229] (4) The network indicates supported evaluation data types and / or configurations, and the UE indicates required evaluation data types and / or configurations, as shown in FIG6E . Then, the network sends the second data to the UE.

[0230] In the embodiments of the present application, when the UE transmits the uplink transmission to the base station, one or more of the following methods may be used to complete the transmission:

[0231] (1)RRC message, (2)UCI, (3)Uplink message in random access process, such as MsgA, Msg3, (4)PUCCH, (5)PUSCH, (6)AI / ML dedicated uplink channel, (7)UE capability reporting.

[0232] The downlink transmission involved in the embodiments of the present application can be completed in one or more of the following ways when the base station transmits to the UE:

[0233] (1) Broadcast messages, such as MIB, SIB1, SIBx, (2) RRC messages, (3) MAC CE, (4) DCI, (5) Downlink messages during random access, such as MsgB, Msg2, Msg4. (4) PDCCH, (5) PDSCH, (6) AI / ML dedicated downlink channels, (7) Capability indication on the network side.

[0234] The base station in the above example can also be replaced by other terminals in the sideline communication scenario.

[0235] 2.3 The above instructions and reports may include one or more of the following information:

[0236] (1) Label data and / or evaluation data type, for example: channel H1, feature vector W1 representing CSI information, and input information X1 of the AI / ML model.

[0237] (2) Label data and / or evaluation data configuration, for example: the number N of samples used for statistics, processing, and averaging to obtain the first data.

[0238] (3) Label data and / or evaluation data configuration, for example: a method for obtaining the first data through statistics, processing, and averaging.

[0239] For example, the method indicated is: taking the mean of the first information; taking the mean of the processed elements of the first information.

[0240] For example, the mean is the mean of N samples, the mean of each element in the sample, and the mean of N samples and different elements in the sample.

[0241] For example, the processing is taking the absolute value of the element in the sample, or the modulus of the element in the sample, or the square of the absolute value of the element in the sample, or the square of the modulus of the element in the sample, or multiplying the element in the sample by the conjugate of the element.

[0242] (4) Label data and / or evaluation data configuration, for example: the format, frequency domain configuration, time domain configuration, angle domain configuration, and antenna domain configuration of the channel information H1, a time domain information subset, a frequency domain information subset, an angle domain information subset, and an antenna domain information subset of H1, for example, channel information at a specific time delay, channel information at a specific frequency domain, channel information at a specific angle, and channel information at a specific antenna (antenna port).

[0243] (5) Tag data and / or evaluation data configuration, such as: channel frequency domain configuration, time domain configuration, angle domain configuration, antenna domain configuration, channel information W1 acquisition scheme indication, W1 format, W1 subband indication, such as subband size, subband number, etc.

[0244] (6) The configuration of the label data and / or evaluation data that can be indicated includes: quantization scheme, quantization granularity, quantization step size, and quantization accuracy information.

[0245] The data transmission method of the embodiment of the present application may include a transmission scheme for wireless AI data label transmission, mean transmission, auxiliary information, etc. In the performance evaluation of the AI / ML solution, multiple transmissions of label data and evaluation data are avoided, and the processed data of label data and evaluation data are used as label data and evaluation data. Therefore, when the evaluation of the AI / ML solution does not strictly rely on the results of a single evaluation, but relies on the comprehensive results of multiple evaluations, the transmission of multiple evaluation data can be avoided, and the method of performing multiple integration processing of the information to be evaluated before transmitting the evaluation information is adopted, and information preprocessing is used to replace the transmission of redundant air interface information.

[0246] 7 is a schematic block diagram of a first communication device 700 according to an embodiment of the present application. The first communication device 700 may include: a sending unit 701, configured to send processed data, where the processed data is obtained by processing a plurality of information to be processed.

[0247] In one embodiment, the plurality of information to be processed includes data related to model performance monitoring, training, testing or use in the first communication device.

[0248] In one embodiment, one piece of information to be processed among the plurality of pieces of information to be processed includes at least one of the following:

[0249] channel;

[0250] The feature vector representing the CSI;

[0251] Model input information;

[0252] Output information of the model.

[0253] In one embodiment, the channel includes at least one of the following:

[0254] An initial channel composed of time domain information, frequency domain information, angle domain information, and antenna domain information;

[0255] For the initial channel, information extraction is performed in the time domain, and a channel is obtained by selecting a subset of the time domain information;

[0256] For the initial channel, information extraction is performed in the frequency domain, and a channel is obtained by selecting a subset of the frequency domain information;

[0257] For the initial channel, information extraction is performed in the angle domain, and a channel is obtained by selecting a subset of the angle domain information;

[0258] For the initial channel, information extraction is performed in the antenna domain, and a channel obtained by selecting a subset of antenna domain information;

[0259] The channel obtained by quantizing the initial channel.

[0260] In one embodiment, the eigenvector representing the CSI includes a eigenvector representing the CSI obtained by performing SVD on the channel.

[0261] In one embodiment, the input information of the model includes the channel and / or the feature vector representing the CSI.

[0262] In one embodiment, the output information of the model includes the channel and / or the feature vector representing the CSI.

[0263] In one embodiment, the processing method of the plurality of information to be processed includes:

[0264] Obtaining the average value of the plurality of information to be processed;

[0265] The average value of the processed elements of the multiple pieces of information to be processed is obtained.

[0266] In one embodiment, element processing of the plurality of information to be processed includes at least one of the following:

[0267] Taking absolute values ​​of elements in the plurality of information to be processed;

[0268] Modulo values ​​of elements in the plurality of information to be processed;

[0269] Taking the square of the absolute value of the elements in the plurality of information to be processed;

[0270] Square the modulus value of the elements in the plurality of information to be processed;

[0271] An element in the plurality of pieces of information to be processed is multiplied by the conjugate of the element.

[0272] In one embodiment, obtaining the mean value includes at least one of the following:

[0273] The information is averaged according to the number of samples;

[0274] The information is averaged by internal elements;

[0275] After averaging the information by internal elements, the average is taken by the number of samples.

[0276] In one embodiment, averaging the information by internal elements includes:

[0277] The information is averaged by the total number of internal elements;

[0278] The information is averaged according to the number of dimensional elements of the internal elements.

[0279] In one embodiment, the first communication device is a terminal device, the multiple information to be processed include multiple first information to be processed in the terminal device, and the processed data includes first data obtained after processing the multiple first information using a first processing method.

[0280] In one embodiment, the first information includes at least one of label data, auxiliary data, and input data of the model in the terminal device.

[0281] In one embodiment, the model in the terminal device is an encoder of a CSI generation model.

[0282] In one embodiment, the first data is transmitted via at least one of the following:

[0283] Radio resource control RRC message; uplink control information UCI; uplink message during random access; physical uplink control channel PUCCH; physical uplink shared channel PUSCH; uplink channel dedicated to artificial intelligence AI and / or machine learning ML; terminal device capability reporting message, sidelink control information SCI, physical sidelink control channel PSCCH, physical sidelink shared channel PSSCH.

[0284] In one embodiment, the first communication device is a network device, the multiple information to be processed include multiple second information to be processed in the network device, and the processed data includes second data obtained by processing the multiple second information using a second processing method.

[0285] In one embodiment, the second information includes at least one of evaluation data, auxiliary data, and output data of a model in the network device.

[0286] In one embodiment, the model in the network device is a decoder of a CSI recovery model and / or a CSI prediction model.

[0287] In one embodiment, the second data is transmitted via at least one of the following:

[0288] Broadcast message; RRC message; Media Access Control (MAC) Control Element (CE); Downlink Control Information (DCI); Downlink message during random access; Physical Downlink Control Channel (PDCCH); Physical Downlink Shared Channel (PDSCH); Downlink channel dedicated to AI and / or ML; Capability indication on the network side; SCI; PSCCH; PSSCH.

[0289] The first communication device 700 of the embodiment of the present application can implement the corresponding functions of the first communication device in the aforementioned method embodiment. The processes, functions, implementation methods and beneficial effects corresponding to the various modules (sub-modules, units or components, etc.) in the first communication device 700 can be found in the corresponding descriptions in the above-mentioned method embodiments, and will not be repeated here. It should be noted that the functions described in the various modules (sub-modules, units or components, etc.) in the first communication device 700 of the embodiment of the application can be implemented by different modules (sub-modules, units or components, etc.) or by the same module (sub-module, unit or component, etc.).

[0290] 8 is a schematic block diagram of a second communication device 800 according to an embodiment of the present application. The second communication device 800 may include: a receiving unit 801, configured to receive processing data, where the processing data is obtained after the first communication device processes a plurality of information to be processed.

[0291] In one embodiment, the plurality of information to be processed includes data related to model performance monitoring, training, testing or use in the second communication device.

[0292] In one embodiment, one piece of information to be processed among the plurality of pieces of information to be processed includes at least one of the following:

[0293] channel;

[0294] The feature vector representing the CSI;

[0295] Model input information;

[0296] Output information of the model.

[0297] In one embodiment, the channel includes at least one of the following:

[0298] An initial channel composed of time domain information, frequency domain information, angle domain information, and antenna domain information;

[0299] For the initial channel, information extraction is performed in the time domain, and a channel is obtained by selecting a subset of the time domain information;

[0300] For the initial channel, information extraction is performed in the frequency domain, and a channel is obtained by selecting a subset of the frequency domain information;

[0301] For the initial channel, information extraction is performed in the angle domain, and a channel is obtained by selecting a subset of the angle domain information;

[0302] For the initial channel, information extraction is performed in the antenna domain, and a channel is obtained by selecting a subset of antenna domain information;

[0303] The channel obtained by quantizing the initial channel.

[0304] In one embodiment, the eigenvector representing the CSI includes a eigenvector representing the CSI obtained by performing SVD on the channel.

[0305] In one embodiment, the input information of the model includes the channel and / or the feature vector representing the CSI.

[0306] In one embodiment, the output information of the model includes the channel and / or the feature vector representing the CSI.

[0307] In one embodiment, the processing method of the plurality of information to be processed includes:

[0308] Obtaining the average value of the plurality of information to be processed;

[0309] The average value of the processed elements of the multiple pieces of information to be processed is obtained.

[0310] In one embodiment, element processing of the plurality of information to be processed includes at least one of the following:

[0311] Taking absolute values ​​of elements in the plurality of information to be processed;

[0312] Modulo values ​​of elements in the plurality of information to be processed;

[0313] Taking the square of the absolute value of the elements in the plurality of information to be processed;

[0314] Square the modulus value of the elements in the plurality of information to be processed;

[0315] An element in the plurality of pieces of information to be processed is multiplied by the conjugate of the element.

[0316] In one embodiment, obtaining the mean value includes at least one of the following:

[0317] The information is averaged according to the number of samples;

[0318] The information is averaged by internal elements;

[0319] After averaging the information by internal elements, the average is taken by the number of samples.

[0320] In one embodiment, averaging the information by internal elements includes:

[0321] The information is averaged by the total number of internal elements;

[0322] The information is averaged according to the number of dimensional elements of the internal elements.

[0323] In one embodiment, the second communication device is a terminal device, the multiple information to be processed include multiple first information to be processed in the terminal device, and the processed data includes first data obtained after processing the multiple first information using a first processing method.

[0324] In one embodiment, the first information includes at least one of label data, auxiliary data, and input data of the model in the terminal device.

[0325] In one embodiment, the model in the terminal device is an encoder of a CSI generation model.

[0326] In one embodiment, the first data is transmitted via at least one of the following:

[0327] Radio resource control (RRC) messages; uplink control information (UCI); uplink messages during random access; PUCCH; PUSCH; uplink channels dedicated to artificial intelligence (AI) and / or machine learning (ML); terminal device capability reporting messages, SCI, PSCCH, and PSSCH.

[0328] In one embodiment, the second communication device is a network device, the multiple information to be processed include multiple second information to be processed in the network device, and the processed data include second data obtained by processing the multiple second information using a second processing method.

[0329] In one embodiment, the second information includes at least one of evaluation data, auxiliary data, and output data of a model in the network device.

[0330] In one embodiment, the model in the network device is a decoder of a CSI recovery model and / or a CSI prediction model.

[0331] In one embodiment, the second data is transmitted via at least one of the following:

[0332] Broadcast message; RRC message; MACCE; DCI; downlink message during random access; PDCCH; PDSCH; downlink channel dedicated to AI and / or ML; capability indication on the network side; SCI; PSCCH; PSSCH.

[0333] The second communication device 800 of the embodiment of the present application can implement the corresponding functions of the second communication device in the aforementioned method embodiment. The processes, functions, implementation methods and beneficial effects corresponding to the various modules (sub-modules, units or components, etc.) in the second communication device 800 can be found in the corresponding descriptions in the above-mentioned method embodiments, and will not be repeated here. It should be noted that the functions described in the various modules (sub-modules, units or components, etc.) in the second communication device 800 of the embodiment of the application can be implemented by different modules (sub-modules, units or components, etc.) or by the same module (sub-module, unit or component, etc.).

[0334] Figure 9 is a schematic structural diagram of a communication device 900 according to an embodiment of the present application. The communication device 900 includes a processor 910, which can call and run a computer program from a memory to enable the communication device 900 to implement the method in the embodiment of the present application.

[0335] In one embodiment, the communication device 900 may further include a memory 920. The processor 910 may call and execute a computer program from the memory 920 to enable the communication device 900 to implement the method in the embodiment of the present application.

[0336] The memory 920 may be a separate device independent of the processor 910 , or may be integrated into the processor 910 .

[0337] In one embodiment, the communication device 900 may further include a transceiver 930 , and the processor 910 may control the transceiver 930 to communicate with other devices, specifically, to send information or data to other devices, or to receive information or data sent by other devices.

[0338] The transceiver 930 may include a transmitter and a receiver. The transceiver 930 may further include an antenna, and the number of antennas may be one or more.

[0339] In one embodiment, the communication device 900 may be the first communication device of the embodiment of the present application, and the communication device 900 may implement the corresponding processes implemented by the first communication device in each method of the embodiment of the present application. For the sake of brevity, they will not be repeated here.

[0340] In one embodiment, the communication device 900 may be the second communication device of the embodiment of the present application, and the communication device 900 may implement the corresponding processes implemented by the second communication device in each method of the embodiment of the present application. For the sake of brevity, they will not be repeated here.

[0341] 10 is a schematic structural diagram of a chip 1000 according to an embodiment of the present application. The chip 1000 includes a processor 1010, which can call and execute a computer program from a memory to implement the method according to the embodiment of the present application.

[0342] In one embodiment, the chip 1000 may further include a memory 1020. The processor 1010 may call and execute a computer program from the memory 1020 to implement the method executed by the first communication device or the second communication device in the embodiment of the present application.

[0343] The memory 1020 may be a separate device independent of the processor 1010 , or may be integrated into the processor 1010 .

[0344] In one embodiment, the chip 1000 may further include an input interface 1030. The processor 1010 may control the input interface 1030 to communicate with other devices or chips, and specifically, may obtain information or data sent by other devices or chips.

[0345] In one embodiment, the chip 1000 may further include an output interface 1040. The processor 1010 may control the output interface 1040 to communicate with other devices or chips, and specifically, may output information or data to other devices or chips.

[0346] In one embodiment, the chip can be applied to the first communication device in the embodiment of the present application, and the chip can implement the corresponding processes implemented by the first communication device in each method of the embodiment of the present application. For the sake of brevity, it will not be repeated here.

[0347] In one embodiment, the chip can be applied to the second communication device in the embodiment of the present application, and the chip can implement the corresponding processes implemented by the second communication device in each method of the embodiment of the present application. For the sake of brevity, it will not be repeated here.

[0348] The chips used in the first communication device and the second communication device may be the same chip or different chips.

[0349] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0350] The processor mentioned above may be a general-purpose processor, a digital signal processor (DSP), a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), or other programmable logic devices, transistor logic devices, discrete hardware components, etc. The general-purpose processor mentioned above may be a microprocessor or any conventional processor, etc.

[0351] The memory mentioned above may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. The non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM).

[0352] It should be understood that the above-mentioned memories are exemplary but not restrictive. For example, the memories in the embodiments of the present application may also be static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct RAM RAM (DR RAM), etc. In other words, the memories in the embodiments of the present application are intended to include, but are not limited to, these and any other suitable types of memories.

[0353] FIG11 is a schematic block diagram of a communication system 1100 according to an embodiment of the present application. The communication system 1100 includes a first communication device 1110 and a second communication device 1120 .

[0354] The first communication device 1110 is configured to send processed data, where the processed data is obtained by processing a plurality of pieces of information to be processed.

[0355] The second communication device 1120 is configured to receive processed data, where the processed data is obtained after the first communication device processes a plurality of information to be processed.

[0356] The first communication device 1110 can be used to implement the corresponding functions implemented by the first communication device in the above method, and the second communication device 1120 can be used to implement the corresponding functions implemented by the second communication device in the above method. For the sake of brevity, they are not described here in detail.

[0357] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When software is used for implementation, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function in accordance with the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can 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 can be transmitted from a website, computer, server or data center by wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.) mode to another website, computer, server or data center. The computer-readable storage medium can 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 integrations. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a DVD), or a semiconductor medium (eg, a solid state disk (SSD)).

[0358] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0359] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0360] 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 modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present application should be included within the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A data transmission method, comprising: The first communication device sends processing data, where the processing data is obtained after processing a plurality of information to be processed.

2. The method according to claim 1, wherein: The plurality of information to be processed include relevant data of model performance monitoring, training, testing or use in the first communication device.

3. The method according to claim 1, wherein: One of the plurality of pieces of information to be processed includes at least one of the following: Channel; A feature vector representing the channel state information CSI; The input information of the model; Output information of the model.

4. The method according to claim 3, wherein: The channel includes at least one of the following: An initial channel composed of time domain information, frequency domain information, angle domain information, and antenna domain information; For the initial channel, extract information in the time domain, and select a channel obtained by a partial time domain information subset; For the initial channel, extract information in the frequency domain, and select a channel obtained by a subset of the frequency domain information; For the initial channel, extract information in the angle domain, and select a channel obtained by selecting a subset of the angle domain information; For the initial channel, extract information in the antenna domain, and select a channel obtained by a subset of antenna domain information; A channel obtained by quantizing the initial channel.

5. The method according to claim 3, wherein: The eigenvector representing the CSI includes a eigenvector representing the CSI obtained by performing singular value decomposition (SVD) on the channel.

6. The method according to claim 3, wherein: The input information of the model includes the channel and / or the feature vector representing the CSI.

7. The method according to claim 3, wherein: The output information of the model includes the channel and / or the feature vector representing the CSI.

8. The method according to any one of claims 1 to 7, wherein: The processing method of the plurality of information to be processed includes: Obtaining the mean value of the plurality of information to be processed; The average value of the elements of the plurality of information to be processed after being processed is obtained.

9. The method according to claim 8, wherein: The element processing of the plurality of information to be processed includes at least one of the following: Taking absolute values ​​of elements in the plurality of information to be processed; Modulo values ​​of elements in the plurality of information to be processed; Taking the square of the absolute value of the elements in the plurality of information to be processed; The square of the modulus value of the elements in the plurality of information to be processed; The elements in the plurality of pieces of information to be processed are multiplied by the conjugate of the elements.

10. The method according to claim 8 or 9, wherein: Mean acquisition includes at least one of the following: The information is averaged according to the number of samples; The information is averaged by internal elements; After averaging the information by internal elements, the average is taken by the number of samples.

11. The method according to claim 10, wherein: The information is averaged by internal elements, including: The information is averaged by the total number of internal elements; The information is averaged according to the number of dimensional elements of the internal elements.

12. The method according to any one of claims 1 to 11, wherein: The first communication device is a terminal device, the multiple pieces of information to be processed include multiple pieces of first information to be processed in the terminal device, and the processed data include first data obtained by processing the multiple pieces of first information in a first processing manner.

13. The method according to claim 12, wherein: The first information includes at least one of label data, auxiliary data and input data of the model in the terminal device.

14. The method according to claim 13, wherein: The model in the terminal device is an encoder of the CSI generation model.

15. The method according to any one of claims 12 to 14, wherein: The first data is transmitted via at least one of the following: Radio resource control RRC message; uplink control information UCI; uplink message during random access; physical uplink control channel PUCCH; physical uplink shared channel PUSCH; uplink channel dedicated to artificial intelligence AI and / or machine learning ML; terminal device capability reporting message, sidelink control information SCI, physical sidelink control channel PSCCH, physical sidelink shared channel PSSCH.

16. The method according to any one of claims 1 to 11, wherein: The first communication device is a network device, the multiple pieces of information to be processed include multiple pieces of second information to be processed in the network device, and the processed data include second data obtained by processing the multiple pieces of second information in a second processing manner.

17. The method according to claim 16, wherein: The second information includes at least one of evaluation data, auxiliary data and output data of the model in the network device.

18. The method according to claim 17, wherein: The model in the network device is a decoder of a CSI recovery model and / or a CSI prediction model.

19. The method according to any one of claims 16 to 18, wherein: The second data is transmitted via at least one of the following: Broadcast message; RRC message; Media Access Control MAC control element CE; Downlink control information DCI; Downlink message in random access process; Physical Downlink Control Channel PDCCH; Physical Downlink Shared Channel PDSCH; Downlink channel dedicated to AI and / or ML; Capability indication on the network side; SCI; PSCCH; PSSCH.

20. A data transmission method, comprising: The second communication device receives processing data, where the processing data is obtained after the first communication device processes a plurality of information to be processed.

21. The method according to claim 20, wherein: The plurality of information to be processed include relevant data of model performance monitoring, training, testing or use in the second communication device.

22. The method according to claim 20, wherein: One of the plurality of pieces of information to be processed includes at least one of the following: Channel; The feature vector representing the CSI; The input information of the model; Output information of the model.

23. The method according to claim 22, wherein: The channel includes at least one of the following: An initial channel composed of time domain information, frequency domain information, angle domain information, and antenna domain information; For the initial channel, extract information in the time domain, and select a channel obtained by a partial time domain information subset; For the initial channel, extract information in the frequency domain, and select a channel obtained by a subset of the frequency domain information; For the initial channel, extract information in the angle domain, and select a channel obtained by selecting a subset of the angle domain information; For the initial channel, extract information in the antenna domain, and select a channel obtained by a subset of antenna domain information; A channel obtained by quantizing the initial channel.

24. The method according to claim 22, wherein: The eigenvector representing the CSI includes a eigenvector representing the CSI obtained by performing SVD on the channel.

25. The method according to claim 22, wherein: The input information of the model includes the channel and / or the feature vector representing the CSI.

26. The method of claim 22, wherein: The output information of the model includes the channel and / or the feature vector representing the CSI.

27. The method according to any one of claims 20 to 26, wherein: The processing method of the plurality of information to be processed includes: Obtaining the mean value of the plurality of information to be processed; The average value of the elements of the plurality of information to be processed after being processed is obtained.

28. The method according to claim 27, wherein: The element processing of the plurality of information to be processed includes at least one of the following: Taking absolute values ​​of elements in the plurality of information to be processed; Modulo values ​​of elements in the plurality of information to be processed; Taking the square of the absolute value of the elements in the plurality of information to be processed; The square of the modulus value of the elements in the plurality of information to be processed; The elements in the plurality of pieces of information to be processed are multiplied by the conjugate of the elements.

29. The method according to claim 27 or 28, wherein: Mean acquisition includes at least one of the following: The information is averaged according to the number of samples; The information is averaged by internal elements; After averaging the information by internal elements, the average is taken by the number of samples.

30. The method of claim 29, wherein: The information is averaged by internal elements, including: The information is averaged by the total number of internal elements; The information is averaged according to the number of dimensional elements of the internal elements.

31. The method according to any one of claims 20 to 30, wherein: The second communication device is a terminal device, the multiple pieces of information to be processed include multiple pieces of first information to be processed in the terminal device, and the processed data include first data obtained by processing the multiple pieces of first information in a first processing manner.

32. The method according to claim 31, wherein: The first information includes at least one of label data, auxiliary data and input data of the model in the terminal device.

33. The method of claim 32, wherein: The model in the terminal device is an encoder of the CSI generation model.

34. A method according to any one of claims 31 to 33, wherein: The first data is transmitted via at least one of the following: Radio resource control RRC message; uplink control information UCI; uplink message during random access; PUCCH; PUSCH; uplink channel dedicated to artificial intelligence AI and / or machine learning ML; terminal device capability reporting message, SCI, PSCCH, PSSCH.

35. The method according to any one of claims 20 to 30, wherein: The second communication device is a network device, the multiple pieces of information to be processed include multiple pieces of second information to be processed in the network device, and the processed data include second data obtained by processing the multiple pieces of second information in a second processing manner.

36. The method of claim 35, wherein: The second information includes at least one of evaluation data, auxiliary data and output data of the model in the network device.

37. The method of claim 36, wherein: The model in the network device is a decoder of a CSI recovery model and / or a CSI prediction model.

38. A method according to any one of claims 35 to 37, wherein: The second data is transmitted via at least one of the following: Broadcast message; RRC message; MAC CE; DCI; downlink message in random access process; PDCCH; PDSCH; downlink channel dedicated to AI and / or ML; capability indication on the network side; SCI; PSCCH; PSSCH.

39. A first communication device, comprising: The sending unit is used to send processing data, where the processing data is obtained after processing multiple pieces of information to be processed.

40. The first communication device according to claim 39, wherein: The plurality of information to be processed include relevant data of model performance monitoring, training, testing or use in the first communication device.

41. The first communication device according to claim 39, wherein: One of the plurality of pieces of information to be processed includes at least one of the following: Channel; The feature vector representing the CSI; The input information of the model; Output information of the model.

42. The first communication device according to claim 41, wherein: The channel includes at least one of the following: An initial channel composed of time domain information, frequency domain information, angle domain information, and antenna domain information; For the initial channel, extract information in the time domain, and select a channel obtained by a partial time domain information subset; For the initial channel, extract information in the frequency domain, and select a channel obtained by a subset of the frequency domain information; For the initial channel, extract information in the angle domain, and select a channel obtained by selecting a subset of the angle domain information; For the initial channel, extract information in the antenna domain, and select a channel obtained by a subset of antenna domain information; A channel obtained by quantizing the initial channel.

43. The first communication device according to claim 42, wherein: The eigenvector representing the CSI includes a eigenvector representing the CSI obtained by performing SVD on the channel.

44. The first communication device according to claim 42, wherein: The input information of the model includes the channel and / or the feature vector representing the CSI.

45. The first communication device according to claim 42, wherein: The output information of the model includes the channel and / or the feature vector representing the CSI.

46. ​​The first communication device according to any one of claims 39 to 45, wherein: The processing method of the plurality of information to be processed includes: Obtaining the mean value of the plurality of information to be processed; The average value of the elements of the plurality of information to be processed after being processed is obtained.

47. The first communication device according to claim 46, wherein: The element processing of the plurality of information to be processed includes at least one of the following: Taking absolute values ​​of elements in the plurality of information to be processed; Modulo values ​​of elements in the plurality of information to be processed; Taking the square of the absolute value of the elements in the plurality of information to be processed; The square of the modulus value of the elements in the plurality of information to be processed; The elements in the plurality of pieces of information to be processed are multiplied by the conjugate of the elements.

48. The first communication device according to claim 46 or 47, wherein: Mean acquisition includes at least one of the following: The information is averaged according to the number of samples; The information is averaged by internal elements; After averaging the information by internal elements, the average is taken by the number of samples.

49. The first communication device according to claim 48, wherein: The information is averaged by internal elements, including: The information is averaged by the total number of internal elements; The information is averaged according to the number of dimensional elements of the internal elements.

50. The first communication device according to any one of claims 39 to 49, wherein: The first communication device is a terminal device, the multiple pieces of information to be processed include multiple pieces of first information to be processed in the terminal device, and the processed data include first data obtained by processing the multiple pieces of first information in a first processing manner.

51. The first communication device according to claim 50, wherein: The first information includes at least one of label data, auxiliary data and input data of the model in the terminal device.

52. The first communication device according to claim 51, wherein: The model in the terminal device is an encoder of the CSI generation model.

53. The first communication device according to any one of claims 50 to 52, wherein: The first data is transmitted via at least one of the following: Radio resource control RRC message; UCI; uplink message during random access; PUCCH; PUSCH; uplink channel dedicated to AI and / or ML; terminal device capability reporting message, SCI, PSCCH, PSSCH.

54. The first communication device according to any one of claims 39 to 49, wherein: The first communication device is a network device, the multiple pieces of information to be processed include multiple pieces of second information to be processed in the network device, and the processed data include second data obtained by processing the multiple pieces of second information in a second processing manner.

55. The first communication device according to claim 54, wherein: The second information includes at least one of evaluation data, auxiliary data and output data of the model in the network device.

56. The first communication device according to claim 55, wherein: The model in the network device is a decoder of a CSI recovery model and / or a CSI prediction model.

57. The first communication device according to any one of claims 54 to 56, wherein: The second data is transmitted via at least one of the following: Broadcast message; RRC message; Media Access Control MAC control element CE; Downlink control information DCI; Downlink message in random access process; Physical Downlink Control Channel PDCCH; Physical Downlink Shared Channel PDSCH; Downlink channel dedicated to AI and / or ML; Capability indication on the network side; SCI; PSCCH; PSSCH.

58. A second communication device, comprising: The receiving unit is used to receive processing data, where the processing data is obtained after the first communication device processes a plurality of information to be processed.

59. The second communication device according to claim 58, wherein: The plurality of information to be processed include relevant data of model performance monitoring, training, testing or use in the second communication device.

60. The second communication device according to claim 58, wherein: One of the plurality of pieces of information to be processed includes at least one of the following: Channel; The feature vector representing the CSI; The input information of the model; Output information of the model.

61. The second communication device according to claim 60, wherein: The channel includes at least one of the following: An initial channel composed of time domain information, frequency domain information, angle domain information, and antenna domain information; For the initial channel, extract information in the time domain, and select a channel obtained by a partial time domain information subset; For the initial channel, extract information in the frequency domain, and select a channel obtained by a subset of the frequency domain information; For the initial channel, extract information in the angle domain, and select a channel obtained by selecting a subset of the angle domain information; For the initial channel, extract information in the antenna domain, and select a channel obtained by a subset of antenna domain information; A channel obtained by quantizing the initial channel.

62. The second communication device according to claim 60, wherein: The eigenvector representing the CSI includes a eigenvector representing the CSI obtained by performing SVD on the channel.

63. The second communication device according to claim 60, wherein: The input information of the model includes the channel and / or the feature vector representing the CSI.

64. The second communication device according to claim 60, wherein: The output information of the model includes the channel and / or the feature vector representing the CSI.

65. The second communication device according to any one of claims 58 to 64, wherein: The processing method of the plurality of information to be processed includes: Obtaining the mean value of the plurality of information to be processed; The average value of the elements of the plurality of information to be processed after being processed is obtained.

66. The second communication device according to claim 65, wherein: The element processing of the plurality of information to be processed includes at least one of the following: Taking absolute values ​​of elements in the plurality of information to be processed; Modulo values ​​of elements in the plurality of information to be processed; Taking the square of the absolute value of the elements in the plurality of information to be processed; The square of the modulus value of the elements in the plurality of information to be processed; The elements in the plurality of pieces of information to be processed are multiplied by the conjugate of the elements.

67. The second communication device according to claim 65 or 66, wherein: Mean acquisition includes at least one of the following: The information is averaged according to the number of samples; The information is averaged by internal elements; After averaging the information by internal elements, the average is taken by the number of samples.

68. The second communication device according to claim 67, wherein: The information is averaged by internal elements, including: The information is averaged by the total number of internal elements; The information is averaged according to the number of dimensional elements of the internal elements.

69. The second communication device according to any one of claims 58 to 68, wherein: The second communication device is a terminal device, the multiple pieces of information to be processed include multiple pieces of first information to be processed in the terminal device, and the processed data include first data obtained by processing the multiple pieces of first information in a first processing manner.

70. The second communication device according to claim 69, wherein: The first information includes at least one of label data, auxiliary data and input data of the model in the terminal device.

71. The second communication device according to claim 70, wherein: The model in the terminal device is an encoder of the CSI generation model.

72. The second communication device according to any one of claims 69 to 71, wherein: The first data is transmitted via at least one of the following: RRC message; UCI; uplink message during random access; PUCCH; PUSCH; uplink channel dedicated to AI and / or ML; terminal device capability reporting message, SCI, PSCCH, PSSCH.

73. The second communication device according to any one of claims 58 to 69, wherein: The second communication device is a network device, the multiple pieces of information to be processed include multiple pieces of second information to be processed in the network device, and the processed data include second data obtained by processing the multiple pieces of second information in a second processing manner.

74. The second communication device according to claim 73, wherein: The second information includes at least one of evaluation data, auxiliary data and output data of the model in the network device.

75. The second communication device according to claim 74, wherein: The model in the network device is a decoder of a CSI recovery model and / or a CSI prediction model.

76. The second communication device according to any one of claims 73 to 75, wherein: The second data is transmitted via at least one of the following: Broadcast message; RRC message; MAC CE; DCI; downlink message in random access process; PDCCH; PDSCH; downlink channel dedicated to AI and / or ML; capability indication on the network side; SCI; PSCCH; PSSCH.

77. A communication device comprising: A processor and a memory, the memory being used to store a computer program, the processor being used to call and run the computer program stored in the memory so that the communication device executes the method as claimed in any one of claims 1 to 19 or any one of claims 20 to 38.

78. A chip, comprising: A processor, configured to call and run a computer program from a memory, so that a device equipped with the chip executes a method as claimed in any one of claims 1 to 19 or any one of claims 20 to 38.

79. A computer-readable storage medium for storing a computer program, which, when executed by a device, causes the device to perform the method as claimed in any one of claims 1 to 19 or any one of claims 20 to 38.

80. A computer program product comprising computer program instructions for causing a computer to perform the method of any one of claims 1 to 19 or any one of claims 20 to 38.

81. A computer program causing a computer to perform the method of any one of claims 1 to 19 or any one of claims 20 to 38.

Images