Uplink transmission method and apparatuses, devices, medium and program product

Abstract

The present application relates to the field of communications. Disclosed are uplink transmission method and apparatuses, devices, a medium and a program product. The method comprises: on the basis of uplink CSI indication information and a first model, obtaining an uplink precoding matrix, the precoding matrix corresponding to a first antenna port group of at least two antenna port groups of a terminal device, and the first antenna port group being some of the at least two antenna port groups; and, by means of the first antenna port group, sending an uplink signal precoded by the precoding matrix. The method provided by the present application does not need to use an additional signaling to indicate the precoding matrix, thereby reducing signaling overheads in an uplink transmission process.

Description

Uplink transmission methods, devices, equipment, media and program products Technical Field

[0001] This application relates to the field of communications, and in particular to an uplink transmission method, apparatus, device, medium, and program product. Background Technology

[0002] When a terminal sends uplink data, it needs to perform precoding processing on the uplink data to obtain uplink precoding gain.

[0003] Related technologies include codebook-based precoding methods, which rely on a precoding instruction agreed upon by both network terminals for uplink precoding. This approach incurs significant signaling overhead because it requires separate instruction of the precoding matrix. This is particularly problematic for terminal equipment with partially coherent antennas, where the signaling overhead is even greater as both the currently used antenna port group and the precoding matrices on those ports need to be indicated simultaneously. Therefore, reducing the signaling overhead in precoding-based uplink transmissions is a pressing issue that needs to be addressed.

[0004] Summary of the Invention

[0005] This application provides an uplink transmission method, apparatus, device, medium, and program product, the technical solution of which is as follows:

[0006] According to one aspect of this application, an uplink transmission method is provided, the method being performed by a terminal device, the method comprising:

[0007] Based on the uplink CSI indication information and the first model, an uplink precoding matrix is ​​obtained. The precoding matrix corresponds to the first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is a partial antenna port group in the at least two antenna port groups.

[0008] The uplink signal, precoded by the precoding matrix, is transmitted through the first antenna port group.

[0009] According to one aspect of this application, an uplink transmission method is provided, the method being performed by a network device, the method comprising:

[0010] The receiving terminal device transmits the precoded uplink signal through the first antenna port group;

[0011] The precoded uplink signal is obtained by precoding based on a precoding matrix. The precoding matrix is ​​obtained by the terminal device based on uplink CSI indication information and a first model. The precoding matrix corresponds to the first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is a partial antenna port group in the at least two antenna port groups.

[0012] According to one aspect of this application, an uplink transmission apparatus is provided, the apparatus comprising:

[0013] The first generation module is used to obtain an uplink precoding matrix based on uplink CSI indication information and a first model. The precoding matrix corresponds to the first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is a partial antenna port group in the at least two antenna port groups.

[0014] The first transmitting module is used to transmit the uplink signal precoded by the precoding matrix through the first antenna port group.

[0015] According to one aspect of this application, an uplink transmission apparatus is provided, the apparatus comprising:

[0016] The first receiving module is used to receive the precoded uplink signal sent by the terminal device through the first antenna port group;

[0017] The precoded uplink signal is obtained by precoding based on a precoding matrix. The precoding matrix is ​​obtained by the terminal device based on uplink CSI indication information and a first model. The precoding matrix corresponds to the first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is a partial antenna port group in the at least two antenna port groups.

[0018] According to one aspect of this application, a terminal device is provided, the terminal device comprising:

[0019] A processor; a transceiver connected to the processor; a memory for storing executable instructions of the processor; wherein the processor is configured to load and execute the executable instructions to implement an uplink transmission method.

[0020] According to one aspect of this application, a network device is provided, the network device comprising:

[0021] A processor; a transceiver connected to the processor; a memory for storing executable instructions of the processor; wherein the processor is configured to load and execute the executable instructions to implement an uplink transmission method.

[0022] According to one aspect of this application, a computer-readable storage medium is provided, wherein at least one program is stored in the computer-readable storage medium, the at least one program being loaded and executed by a processor to implement an uplink transmission method.

[0023] According to one aspect of this application, a chip is provided, the chip including programmable logic circuitry and / or program instructions, which, when the chip is running on a first node, is used to implement the above-described uplink transmission method.

[0024] According to one aspect of this application, a computer program product is provided, the computer program product including computer instructions stored in a computer-readable storage medium, a processor retrieving the computer instructions from the computer-readable storage medium, and the processor executing the computer instructions to implement an uplink transmission method.

[0025] According to one aspect of this application, a computer program is provided, the computer program being stored in a computer-readable storage medium, a processor retrieving the computer program from the computer-readable storage medium, and the processor executing the computer program to implement an uplink transmission method.

[0026] The technical solutions provided in this application have at least the following beneficial effects:

[0027] The terminal device generates a precoding matrix for a portion of the antenna port group (i.e., the coherent antenna port group) based on uplink CSI indication information and a first model. Then, it precodes the uplink signal using this precoding matrix to achieve coherent precoding on the partial antenna port group. Compared to traditional precoding methods based on partially coherent codebooks, neither the terminal device nor the network device needs to use excessive indication signaling to specify the antenna port group and precoding matrix used for uplink transmission from the codebook, thus saving signaling overhead in the uplink transmission process based on precoding methods. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 shows a schematic diagram of the neuron structure in the related technology;

[0030] Figure 2 shows a schematic diagram of the structure of a neural network in the relevant technology;

[0031] Figure 3 shows a schematic diagram of a mobile communication system provided in an exemplary embodiment of this application;

[0032] Figure 4 shows a flowchart of an uplink transmission method provided in an exemplary embodiment of this application;

[0033] Figure 5 shows a flowchart of an uplink transmission method provided in an exemplary embodiment of this application;

[0034] Figure 6 shows a flowchart of an uplink transmission method provided in an exemplary embodiment of this application;

[0035] Figure 7 shows a flowchart of an uplink transmission method provided in an exemplary embodiment of this application;

[0036] Figure 8 shows a flowchart of an uplink transmission method provided in an exemplary embodiment of this application;

[0037] Figure 9 shows a flowchart of an uplink transmission method provided in an exemplary embodiment of this application;

[0038] Figure 10 shows a flowchart of an uplink transmission method provided in an exemplary embodiment of this application;

[0039] Figure 11 shows a flowchart of an uplink transmission method provided in an exemplary embodiment of this application;

[0040] Figure 12 shows a flowchart of an uplink transmission method provided in an exemplary embodiment of this application;

[0041] Figure 13 shows a structural block diagram of an uplink transmission device provided in an exemplary embodiment of this application;

[0042] Figure 14 shows a structural block diagram of an uplink transmission device provided in an exemplary embodiment of this application;

[0043] Figure 15 shows a schematic diagram of the structure of a terminal device provided in an exemplary embodiment of this application;

[0044] Figure 16 shows a schematic diagram of the structure of a network device provided in an exemplary embodiment of this application. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be further described in detail below with reference to the accompanying drawings. Exemplary embodiments will be described in detail here, examples of which are illustrated in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0046] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0047] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0048] The technical solutions described in some embodiments of this application can be applied to various communication systems, such as: Global System for 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, evolution of NR system, LTE-based access to unlicensed spectrum (LTE-U) system, NR-based access to unlicensed spectrum (NR-U) system, Non-Terrestrial Networks (NTN) system, Universal Mobile Telecommunication System (UMTS), Wireless Local Area Networks (WLAN), and Wireless Fidelity (WF). It can be used for Fidelity (WiFi), 5th-Generation (5G) systems, cellular IoT systems, cellular passive IoT systems, and can also be used for subsequent evolution systems of 5G NR systems, as well as 6G and subsequent evolution systems.

[0049] It should be understood that in some embodiments of this application, "5G" may also be referred to as "5G NR" or "NR".

[0050] It should be understood that in the description of the embodiments of this application, the term "correspondence" may indicate that there is a direct or indirect correspondence between the two, or that there is an association between the two, or that there is a relationship of instruction and being instructed, configuration and being configured, etc.

[0051] In this embodiment of the application, "predefined" can be implemented by pre-storing corresponding codes, tables, or other means that can be used to indicate relevant information in the device (e.g., including terminal devices and network devices). This application does not limit the specific implementation method. For example, predefined can refer to what is defined in the protocol.

[0052] In this application embodiment, "protocol" may refer to standard protocols in the field of communication, such as LTE protocol, NR protocol and related protocols applied to future communication systems, and this application does not limit it.

[0053] Next, we will introduce codebook-based uplink transmission.

[0054] When a terminal transmits uplink data on the Physical Uplink Shared Channel (PUSCH), it needs to perform precoding on the uplink data to obtain uplink precoding gain.

[0055] Precoding processing generally consists of two parts: analog domain processing and digital domain processing. Analog domain processing deals with the transmitted analog signal, typically using beamforming to map the RF signal onto the physical antenna. Digital domain processing deals with the digital signal, usually performed in baseband, using a precoding matrix to precode the digital signal and map the data from the transport layer onto the RF port (also called the antenna port). Due to the limited number of RF channels in a terminal, both processing methods are usually used simultaneously: precoding the digital signal and then beamforming the analog signal. PUSCH transmission is divided into codebook-based transmission and non-codebook-based transmission, depending on the precoding method.

[0056] In uplink codebook-based precoding, the network side can use DCI (Downlink Control Information) to indicate RI (Rank Indication) and TPMI (Transmit Precoding Matrix Indicator). The terminal determines the uplink precoding matrix corresponding to the TPMI from a pre-agreed codebook based on the RI and TPMI, and then uses it for uplink precoding. The codebook can be divided into fully coherent codebooks, partially coherent codebooks, and incoherent codebooks, used for terminals with different coherence capabilities. If all antenna ports of the terminal device are coherent, the network device can indicate codewords from the fully coherent codebook, and all antenna ports can be used for uplink transmission. If the antenna ports of the terminal device are partially coherent, the network device can indicate codewords from the partially coherent codebook. In this case, only one set of coherent antenna ports (called an antenna port group) can be used for uplink transmission, and the weights of other antenna ports (i.e., antenna ports in other antenna port groups) are 0. If all antenna ports of the terminal device are incoherent, the network device can only indicate codewords from the incoherent codebook. In this case, only one antenna port can be used for uplink transmission, and the weights of the other antenna ports are 0. Coherence capability is achieved by the terminal capability being pre-reported to the network device.

[0057] Neural Networks

[0058] A neural network is a computational model consisting of multiple interconnected neurons. The connections between nodes represent weighted sums from the input signal to the output signal, called weights. Each node performs a weighted summation of different input signals and outputs the result through a specific activation function. The neuron structure is shown in Figure 1.

[0059] A simple neural network, as shown in Figure 2, consists of an input layer, hidden layers, and an output layer. Through different connections, weights, and activation functions of multiple neurons, different outputs can be generated, thus fitting a mapping relationship from input to output. Each node in the previous level is connected to all its nodes in the next level. This fully connected model can also be called a DNN (Deep Neural Network).

[0060] A neural network model, also known as an AI (Artificial Intelligence) model, can be trained and obtained through the processes of dataset construction, training, validation, and testing. When a neural network is used for wireless communication between terminal devices and network devices, training can be divided into offline training and online training. Network devices can obtain a static training result through offline training on a dataset; this can be called offline training. During the use of the neural network model by network devices or terminal devices, as the terminal device further measures and / or reports, the network device can continue to collect more data and perform real-time online training to optimize the parameters of the neural network model, achieving better inference and prediction results. After obtaining the neural network model, by inputting the currently obtained information into the neural network model, the corresponding model output can be obtained through inference. When neural network models are used for wireless communication, they can be divided into single-ended models and dual-ended models. Single-ended models can be deployed only on one side of the terminal device or network device, and the model training can also be performed on only one side. Dual-ended models need to be deployed in pairs on both the terminal device and network device sides, and the models on both sides need to be trained together; that is, the models deployed on both sides are corresponding and cannot be used or updated independently.

[0061] To achieve different functions, different AI models can be introduced, with corresponding inputs and outputs defined. For example, when using AI / ML (Machine Learning) for CSI (Channel State Information) feedback, channel information (such as feature vectors, beam information, and delay information) measured based on reference signals can be used as input to the AI ​​model to infer the corresponding CSI quantized bits. On the other side, a corresponding AI model takes the CSI quantized bits as input to infer the corresponding channel information. As another example, when using AI / ML for data detection, i.e., when using an AI receiver to detect downlink signals, the terminal device can use the received signal (and additional information such as pilot sequences) as input to the AI ​​model for inference, thereby outputting the detected soft bit symbols (i.e., constellation points). Furthermore, AI models can also be used for other communication processes such as localization, channel coding, channel decoding, beam management, and channel estimation.

[0062] Figure 3 shows a schematic diagram of a mobile communication system provided in an exemplary embodiment of this application. The mobile communication system includes a network device 110 and a terminal device 120, and may or may not include a terminal device 130; this application does not limit this.

[0063] The network device 110 in this application provides wireless communication functionality. This network device 110 includes, but is not limited to: an evolved Node B (eNB), a Radio Network Controller (RNC), a Node B (NB), a Base Station Controller (BSC), a Base Transceiver Station (BTS), a Home Evolved Node B (or Home Node B, HNB), a Base Band Unit (BBU), an Access Point (AP) in a Wireless Fidelity (Wi-Fi) system, a wireless relay node, a wireless backhaul node, a Transmission Point (TP), or a Transmission and Reception Point (TRP), etc. It can also be used for next-generation Node B (Next Generation Node) systems in 5G mobile communication systems. B, gNB) or transmission point (TRP or TP), or, in a 5G system, one or a group of antenna panels (including multiple antenna panels) of a base station, or, network nodes constituting a gNB or transmission point, such as baseband unit (BBU) or distributed unit (DU), or base stations in Beyond Fifth Generation (B5G) or 6th Generation (6G) mobile communication systems, or core network (CN), fronthaul, backhaul, radio access network (RAN), network slicing, etc., or serving cell, primary cell (PCell), primary secondary cell (PSCell), special cell (SpCell), secondary cell (SCell), neighboring cell, etc. of terminal equipment.

[0064] The terminal equipment 120 in this application is also 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 equipment, user agent, or user device. This terminal includes, but is not limited to: handheld devices, wearable devices, in-vehicle devices, and IoT devices, such as: mobile phones, tablets, e-readers, laptops, desktop computers, televisions, game consoles, mobile internet devices (MID), augmented reality (AR) terminals, virtual reality (VR) terminals, mixed reality (MR) terminals, extended reality (XR) terminals, baffle reality (BR) terminals, cinematic reality (CR) terminals, deceive reality (DR) terminals, wearable devices, controllers, wireless terminals in industrial control, wireless terminals in self-driving, wireless terminals in remote medical care, wireless terminals in smart grids, wireless terminals in transportation safety, wireless terminals in smart cities, wireless terminals in smart homes, and wireless terminals in remote surgery. Wireless terminals, cellular phones, cordless phones, Session Initiation Protocol (SIP) phones, Wireless Local Loop (WLL) stations, Personal Digital Assistants (PDAs), Set-Top Boxes (STBs), and Customer Premise Equipment (CPEs) are all examples of devices used in medical surgery.

[0065] In some embodiments, network device 110 and terminal device 120 communicate with each other through some air interface technology, such as the Uu interface.

[0066] For example, there are two communication scenarios between network device 110 and terminal device 120: uplink communication scenario and downlink communication scenario. Uplink communication, or uplink transmission, refers to sending signals or data to network device 110; downlink communication, or downlink transmission, refers to sending signals or data to terminal device 120.

[0067] In some embodiments, terminal device 120 and terminal device 130 communicate with each other through some air interface technology, such as the PC5 interface.

[0068] For example, there are two communication scenarios between terminal device 120 and terminal device 130: a first side-by-side communication scenario and a second side-by-side communication scenario. The first side-by-side communication refers to terminal device 120 sending signals to terminal device 130; the second side-by-side communication refers to terminal device 130 sending signals to terminal device 120.

[0069] In some embodiments, terminal device 120 and terminal device 130 are both within network coverage and located in the same cell, or terminal device 120 and terminal device 130 are both within network coverage but located in different cells, or terminal device 120 is within network coverage but terminal device 130 is outside network coverage.

[0070] In some embodiments of this application, "NR" may also be referred to as a 5G NR system or a 5G system. The 5G mobile communication system may include non-standalone (NSA) and / or standalone (SA) networking.

[0071] The technical solutions provided in the embodiments of this application can also be applied to Machine-Type Communication (MTC), Long Term Evolution-Machine (LTE-M) technology, Device-to-Device (D2D) networks, Machine-to-Machine (M2M) networks, Internet of Things (IoT) networks, or other networks. Among them, IoT networks may include, for example, vehicle-to-everything (V2X) networks. The communication methods in V2X systems are collectively referred to as Vehicle to X (V2X), where X can represent anything. For example, V2X may include: Vehicle to Vehicle (V2V) communication, Vehicle to Infrastructure (V2I) communication, Vehicle to Pedestrian (V2P) communication, or Vehicle to Network (V2N) communication, etc.

[0072] The mobile communication system provided in this application embodiment can be applied to at least one of the following communication scenarios: uplink communication scenario, downlink communication scenario, and sidelink communication scenario.

[0073] Figure 4 illustrates a flowchart of an uplink transmission method provided in an exemplary embodiment of this application. The method is executed by a terminal device, which may be the terminal device shown in Figure 3. The method includes:

[0074] Step 220: Based on the uplink CSI indication information and the first model, obtain the uplink precoding matrix. The precoding matrix corresponds to the first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is a partial antenna port group in at least two antenna port groups.

[0075] The first model includes a neural network model. This first model is used by the terminal device to obtain the uplink precoding matrix based on the uplink CSI indication information. The uplink CSI indication information can be called uplink PMI (Precoding Matrix Indicator) indication information, first CSI indication information, or other names; this embodiment does not limit the specific name used. The uplink CSI indication information is used to indicate relevant information about the uplink channel, such as PMI information, to the terminal device.

[0076] In some embodiments, the terminal device includes at least two antenna port groups, with the first antenna port group corresponding to the precoding matrix obtained based on the first model. The first antenna port group is used to transmit the uplink signal precoded by the precoding matrix. The first antenna port group is one or a portion of the at least two antenna port groups.

[0077] In some embodiments, each antenna port group of the terminal device includes at least one antenna port, and the antenna ports included in each antenna port group are different. For example, the terminal device includes eight antenna ports a1-a8, and at least two antenna port groups of the terminal device include {a1-a4} and {a5-a8}.

[0078] In some embodiments, the precoding matrix of the terminal device corresponds to a first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is used by the terminal device to transmit precoded uplink signals. Optionally, the terminal device includes at least one transport layer, and the precoding matrix of the terminal can be divided into precoding matrices corresponding to at least one transport layer. The precoding matrix of one transport layer corresponds to one antenna port group in at least two antenna port groups. That is, the precoding matrix of the terminal device corresponds to the first antenna port group in at least two antenna port groups of the terminal device, and the first antenna port group is a portion of the antenna port groups in at least two antenna port groups. For example, the terminal device includes one transport layer b1 and two antenna port groups (including antenna port groups {a1-a4} and antenna port groups {a5-a8}), wherein the precoding matrix of the transport layer b1 corresponds to antenna port group {a1-a4}. In this case, the precoding matrix of the terminal device corresponds to the first antenna port group, and the first antenna port group contains one antenna port group (i.e., antenna port group {a1-a4}).

[0079] Step 230: Transmit the uplink signal precoded by the precoding matrix through the first antenna port group.

[0080] In some embodiments, transmitting the uplink signal precoded by the precoding matrix through the first antenna port group can be understood as "the terminal device uses the precoding matrix to precode the uplink signal to obtain the precoded uplink signal; and transmits the precoded uplink signal through the first antenna port group." It can also be described as "the terminal device uses its own precoding matrix to precode the uplink signal to obtain the precoded uplink signal; and transmits the precoded uplink signal through the first antenna port group." Alternatively, it can be described as "the terminal device uses the precoding matrices of each transmission layer to precode the uplink signal to obtain the precoded uplink signal; and transmits the precoded uplink signal through the first antenna port group of each transmission layer."

[0081] In summary, the method provided in this application involves a terminal device generating a precoding matrix on a partial antenna port group (i.e., a coherent antenna port group) based on uplink CSI indication information and a first model. The uplink signal is then precoded using this precoding matrix to achieve coherent precoding on the partial antenna port group. Compared to traditional precoding methods based on partially coherent codebooks, neither the terminal device nor the network device needs to use excessive indication signaling to specify the antenna port group and precoding matrix used for uplink transmission from the codebook, thus saving signaling overhead in the uplink transmission process based on precoding methods.

[0082] Figure 5 illustrates a flowchart of an uplink transmission method provided in an exemplary embodiment of this application. The method is performed by a network device, which may be the network device shown in Figure 3. The method includes:

[0083] Step 330: Receive the precoded uplink signal sent by the terminal device through the first antenna port group;

[0084] The precoded uplink signal is obtained by precoding based on the precoding matrix. The precoding matrix is ​​obtained by the terminal device based on the uplink CSI indication information and the first model. The precoding matrix corresponds to the first antenna port group in the at least two antenna port groups of the terminal device. The first antenna port group is a partial antenna port group in the at least two antenna port groups.

[0085] The first model includes a neural network model. This first model is used by the terminal device to obtain the uplink precoding matrix based on the uplink CSI indication information. The uplink CSI indication information may be referred to as uplink PMI indication information, first CSI indication information, or other names, and this embodiment does not limit the specific name used. The uplink CSI indication information is used to indicate relevant information about the uplink channel, such as PMI information, to the terminal device.

[0086] In some embodiments, the terminal device includes at least two antenna port groups, with the first antenna port group corresponding to the precoding matrix obtained based on the first model. The first antenna port group is used to transmit the uplink signal precoded by the precoding matrix. The first antenna port group is one or a portion of the at least two antenna port groups.

[0087] In some embodiments, the terminal device includes at least one antenna port group, which corresponds to a first antenna port group based on a precoding matrix obtained from a first model. The first antenna port group is used to transmit uplink signals precoded by the precoding matrix. The first antenna port group is one or a subset of at least two antenna port groups.

[0088] In some embodiments, each antenna port group of the terminal device includes at least one antenna port, and the antenna ports included in each antenna port group are different. For example, the terminal device includes eight antenna ports a1-a8, and at least two antenna port groups of the terminal device include {a1-a4} and {a5-a8}.

[0089] In some embodiments, the precoding matrix of the terminal device corresponds to a first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is used by the terminal device to transmit precoded uplink signals. Optionally, the terminal device includes at least one transport layer, and the precoding matrix of the terminal can be divided into precoding matrices corresponding to at least one transport layer. The precoding matrix of one transport layer corresponds to one antenna port group in at least two antenna port groups. That is, the precoding matrix of the terminal device corresponds to the first antenna port group in at least two antenna port groups of the terminal device, and the first antenna port group is a portion of the antenna port groups in at least two antenna port groups. For example, the terminal device includes one transport layer b1 and two antenna port groups (including antenna port groups {a1-a4} and antenna port groups {a5-a8}), wherein the precoding matrix of the transport layer b1 corresponds to antenna port group {a1-a4}. In this case, the precoding matrix of the terminal device corresponds to the first antenna port group, and the first antenna port group contains one antenna port group (i.e., antenna port group {a1-a4}).

[0090] In summary, the method provided in this application involves a terminal device generating a precoding matrix on a partial antenna port group (i.e., a coherent antenna port group) based on uplink CSI indication information and a first model. The uplink signal is then precoded using this precoding matrix to achieve coherent precoding on the partial antenna port group. Compared to traditional precoding methods based on partially coherent codebooks, neither the terminal device nor the network device needs to use excessive indication signaling to specify the antenna port group and precoding matrix used for uplink transmission from the codebook, thus saving signaling overhead in the uplink transmission process based on precoding methods.

[0091] Next, we will introduce the working process of the first model.

[0092] 1. Input to the first model.

[0093] In some embodiments, the input to the first model includes uplink CSI indication information. That is, the precoding matrix is ​​obtained by inputting the uplink CSI indication information into the first model. Alternatively, the terminal device obtains the uplink precoding matrix based on the uplink CSI indication information and the first model, including: the terminal device inputting the uplink CSI indication information into the first model to obtain the precoding matrix.

[0094] In some embodiments, the input to the first model further includes an antenna port group configuration of the terminal device. The antenna port group configuration indicates the result of the terminal device dividing the antenna port groups; or, the antenna port group configuration indicates information about the antenna port groups configured by the terminal device, wherein the individual antenna ports in each antenna port group are coherent. Exemplarily, the antenna port group configuration is at least one of the following: the number of antenna port groups; the number of antenna port groups and the total number of antenna ports; the number of antenna port groups and the number of antenna ports in each antenna port group; the total number of antenna ports and the number of antenna ports in each antenna port group.

[0095] In some embodiments, given any two of the three parameters—the number of antenna port groups, the total number of antenna ports, and the number of antenna ports in each antenna port group—the first model can deduce the remaining parameter. For example, the total number of antenna ports is equal to the number of antenna port groups multiplied by the number of antenna ports in each antenna port group.

[0096] In some embodiments, the input to the first model further includes the number of transport layers of the terminal device; or, the input to the first model further includes the transport layer configuration of the terminal device, which is used to indicate the number of transport layers during uplink transmission. Optionally, the number of transport layers can also be indicated by rank.

[0097] In some embodiments, the size of the precoding matrix output by the first model is related to at least one of the antenna port group configuration and the number of transport layers (or transport layer configuration). That is, the first model determines the size of the output precoding matrix based on at least one of the antenna port group configuration and the number of transport layers. Optionally, the first model determines the number of rows of the precoding matrix based on the antenna port group configuration, and / or the first model determines the number of columns of the precoding matrix based on the number of transport layers.

[0098] In some embodiments, the first model determines the number of rows of the precoding matrix based on the parameter "total number of antenna ports" in the antenna port group configuration; or, the first model determines the number of rows of the precoding matrix based on the parameter "number of antenna ports in each antenna port group" in the antenna port group configuration.

[0099] In some embodiments, the terminal device stores at least two AI models. The terminal device determines a first model from the at least two AI models based on the antenna port group configuration and / or the number of transmission layers. The first model is the AI ​​model corresponding to the antenna port group configuration and / or the number of transmission layers. Alternatively, the terminal device includes only the first model, and the size of the output precoding matrix is ​​determined based on the antenna port group configuration and / or the number of transmission layers. Alternatively, the first model includes at least one processing module, each processing module corresponding to one antenna port group configuration and / or the number of transmission layers; the first model determines the first processing module based on the antenna port group configuration and / or the number of transmission layers, and the uplink CSI indication information is input into the first processing module to obtain the precoding matrix.

[0100] In summary, the method provided in this application embodiment further includes antenna port group configuration and / or transmission layer number as input to the first model. This allows terminal devices and network devices to use different models for different antenna coherence conditions (e.g., different numbers of antenna port groups) or to use the antenna coherence condition as model input. Compared to traditional methods that require indicating the precoding matrix and the currently used antenna port group, using uplink CSI indication information and the corresponding first model to obtain the precoding matrix not only reduces the overhead of uplink CSI indication bits but also improves the accuracy of uplink CSI.

[0101] 2. Output of the first model.

[0102] In some embodiments, the first model is used to output a precoding matrix, which corresponds to a first antenna port group. That is, the precoding matrix includes at least the weights of each antenna port in the first antenna port group, so that the terminal device precodes the uplink signal based on the weights.

[0103] In some embodiments, the terminal device includes at least one transport layer; a first model outputs first antenna port group information and a precoding matrix for each transport layer; or, in addition to outputting the precoding matrix, the first model also outputs first antenna port group information for each transport layer. The first antenna port group information is used to indicate the first antenna port group in at least two antenna port groups.

[0104] For example, the terminal device includes two transport layers (including transport layer b1 and transport layer b2), eight antenna ports (including antenna ports a1-a8), and four antenna port groups (including antenna port groups {a1, a2}, {a3, a4}, {a5, a6}, and {a7, a8}, and each antenna port group and its corresponding index are shown in Table 1). If the precoding matrix is ​​[[1, 1, 0, 0, 0, 0, 0, 0]... T ,[0,0,0,0,1,-j,0,0] TThe precoding matrix is ​​defined as follows: each row corresponds to an antenna port (the first row corresponds to antenna port a1, the second row to antenna port a2, the third row to antenna port a3, the fourth row to antenna port a4, the fifth row to antenna port a5, the sixth row to antenna port a6, the seventh row to antenna port a7, and the eighth row to antenna port a8), and each column corresponds to a transport layer (the first column corresponds to transport layer b1, and the second column corresponds to transport layer b2). The first model then outputs the precoding matrix (also called the precoding submatrix, precoding vector, vector, etc.) corresponding to transport layer b1 as [1, 1, 0, 0, 0, 0, 0, 0]. T And its corresponding first antenna port group information "00", and the precoding matrix [0, 0, 0, 0, 1, -j, 0, 0] corresponding to transport layer b2. T And its corresponding first antenna port group information "10".

[0105] Table 1. Antenna Port Groups and Their Corresponding Indexes

[0106] In some embodiments, the first model outputs a precoding matrix corresponding to at least one transmission layer, and the first model outputs the first antenna port group information of each transmission layer respectively; or, the first model outputs the first antenna port group information and the precoding matrix of each transmission layer respectively.

[0107] For example, the first model outputs a precoding matrix [[1, 1, 0, 0, 0, 0, 0, 0]]. T ,[0,0,0,0,1,-j,0,0] T [00, 10] and the first antenna port group information of each transmission layer (the first output represents the first antenna port group information of the first transmission layer, and the second output represents the first antenna port group information of the second transmission layer); or, the first model outputs the first antenna port group information of each transmission layer and the precoding matrix "00, [1, 1, 0, 0, 0, 0, 0, 0]". T ,10,[0,0,0,0,1,-j,0,0] T "(The first output represents the first antenna port group information of the first transport layer, the second output represents the precoding matrix of the first transport layer, the third output represents the first antenna port group information of the second transport layer, and the fourth output represents the precoding matrix of the second transport layer.)"

[0108] It should be noted that the output format of the first model can be as shown above, first outputting the first antenna port group information of the first transmission layer, then outputting the precoding matrix of the first transmission layer, ..., outputting the first antenna port group information of the nth transmission layer, and outputting the precoding matrix of the nth transmission layer; or, the output format of the first model can also be to first output the first antenna port group information of each transmission layer, then outputting the precoding matrix of each transmission layer, such as first outputting the first antenna port group information of the first transmission layer, ..., outputting the first antenna port group information of the nth transmission layer, outputting the precoding matrix of the first transmission layer, ..., outputting the precoding matrix of the nth transmission layer; or, the output format of the first model can also be to output the index of each transmission layer, the first antenna port group information, and the precoding matrix in the form of a table, as shown in Table 2 below. In other words, the output format of the first model can also be other output formats, and this application does not limit the output format of the first model. Wherein, n is a positive integer greater than 1, and the value of n is related to the number of transmission layers.

[0109] Table 2. Output format of the first model

[0110] The precoding matrix output by the first model exists in the following two forms.

[0111] Form 1: The output precoding matrix includes the weights of each antenna port in at least two antenna port groups.

[0112] In some embodiments, the precoding matrix includes the weights of each antenna port in at least two antenna port groups of the terminal device. The weights of each antenna port in the first antenna port group are non-zero, and the weights of each antenna port in the second antenna port group are zero. The second antenna port group is the antenna port group other than the first antenna port group among the at least two antenna port groups. Because the weight of the second antenna port group is zero, it is not actually used for uplink signal transmission; therefore, the precoding matrix here corresponds only to the first antenna port group.

[0113] The antenna ports in the first antenna port group are used to transmit precoded uplink signals; the antenna ports in the second antenna port group are not used to transmit precoded uplink signals.

[0114] Optionally, a precoding matrix (or precoding vector) corresponding to a transport layer corresponds to a first antenna port group and a second antenna port group. For example, the terminal device includes one transport layer b1, eight antenna ports (including antenna ports a1-a8), and two antenna port groups (including antenna port groups {a1-a4} and {a5-a8}, with each antenna port group and its corresponding index shown in Table 3). If the terminal's precoding matrix is ​​[1, 1, 1, 1, 0, 0, 0, 0], TThe precoding matrix is ​​structured such that each row corresponds to an antenna port (the first row corresponds to antenna port a1, the second row to antenna port a2, the third row to antenna port a3, the fourth row to antenna port a4, the fifth row to antenna port a5, the sixth row to antenna port a6, the seventh row to antenna port a7, and the eighth row to antenna port a8), and each column corresponds to a transmission layer (the first column corresponds to transmission layer b1, and the second column corresponds to transmission layer b2). In this case, the first antenna port group of transmission layer b1 is {a1-a4}, and the second antenna port group is {a5-a8}; the first antenna port group of transmission layer b2 is {a5-a8}. The precoding matrix output by the first model is "[1, 1, 1, 1, 0, 0, 0, 0]". -1 ".

[0115] Table 3. Antenna Port Groups and Their Corresponding Indexes

[0116] In some embodiments, when the first model output is a precoding matrix corresponding to Form 1, the number of rows in the precoding matrix is ​​determined by the "total number of antenna ports" in the antenna port group configuration.

[0117] Format 2: Output only the weights of each antenna port in the first antenna port group and the information of the first antenna port group.

[0118] In some embodiments, the precoding matrix includes only the weights of each antenna port in the first antenna port group. In this case, the first model is also used to output first antenna port group information, which is used to indicate the first antenna port group in at least two antenna port groups. That is, the first model outputs the precoding matrix and the corresponding first antenna port group information. In some embodiments, the first antenna port group information is an index value; or, the first antenna port group information is a bitmap.

[0119] For example, the terminal device or the first model numbers at least two antenna ports to obtain an index for each antenna port group. The first model outputs the index of the first antenna port group to indicate the first antenna port group from the at least two antenna port groups. If the information of the first antenna port group is log2N bits, where N is the number of antenna port groups (i.e., at least two antenna port groups) included in the terminal device, Table 1 above is designed according to this method.

[0120] For example, the first antenna port group information is a bitmap. The number of bits in this bitmap is determined based on the number of antenna ports involved in at least two antenna port groups. For example, if at least two antenna port groups include four antenna ports (including antenna port a1, antenna port a2, antenna port a3, and antenna port a4), then the first antenna port group information is a four-bit bitmap; the first bit of the bitmap corresponds to antenna port a1, the second bit corresponds to antenna port a2, the third bit corresponds to antenna port a3, and the fourth bit corresponds to antenna port a4. A bit of 0 in the bitmap indicates that the antenna port group does not include the antenna port corresponding to that bit, and a bit of 1 in the bitmap indicates that the antenna port group includes the antenna port corresponding to that bit. For example, if the first antenna port group information is "0011", it means that the first antenna port group consists of antenna port a3 and antenna port a4. It should be noted that in the embodiments of this application, the nth bit refers to the nth number counted from the left to the right of the bitmap; however, it can also be the nth number counted from the right to the left of the bitmap. The embodiments of this application do not limit this. Furthermore, in the example given, this application embodiment illustrates that a bit of 0 in the bitmap indicates that the antenna port group does not include the antenna port corresponding to that bit, and a bit of 1 in the bitmap indicates that the antenna port group includes the antenna port corresponding to that bit. However, it can also be implemented such that a bit of 1 in the bitmap indicates that the antenna port group does not include the antenna port corresponding to that bit, and a bit of 0 in the bitmap indicates that the antenna port group includes the antenna port corresponding to that bit. This application embodiment does not limit this aspect.

[0121] In some embodiments, where the precoding matrix output by the first model may only contain the weights of each antenna port in the first antenna port group, the output of the first model also includes information about the first antenna port group. The terminal device includes two transport layers (including transport layer b1 and transport layer b2), eight antenna ports (including antenna ports a1-a8), and four antenna port groups (including antenna port groups {a1, a2}, {a3, a4}, {a5, a6}, and {a7, a8}, with each antenna port group and its corresponding index shown in Table 1). If the precoding matrix of the terminal is [[1, 1, 0, 0, 0, 0, 0, 0]] T ,[0,0,0,0,1,-j,0,0] T The precoding matrix of the terminal corresponds to an antenna port in each row (the first row corresponds to antenna port a1, the second row to antenna port a2, the third row to antenna port a3, the fourth row to antenna port a4, the fifth row to antenna port a5, the sixth row to antenna port a6, the seventh row to antenna port a7, and the eighth row to antenna port a8), and each column corresponds to a transmission layer (the first column corresponds to transmission layer b1, the second column corresponds to transmission layer b2). At this time, the output of the first model can be "00, [1, 1]".T ,10,[1,-j] T "(The first output represents the first antenna port group information of the first transport layer, the second output represents the precoding matrix of the first transport layer, the third output represents the first antenna port group information of the second transport layer, and the fourth output represents the precoding matrix of the second transport layer.)"

[0122] In some embodiments, when the first model output is a precoding matrix corresponding to form two, the number of rows in the precoding matrix is ​​determined by the "number of antenna ports in each antenna port group" in the antenna port group configuration.

[0123] In summary, the method provided in this application illustrates two indication formats for the output of the first model. Format one, where the precoding matrix includes the weights of antenna ports involved in at least two antenna port groups, shows a larger precoding matrix from the perspective of the precoding matrix itself. However, its format is consistent with that of precoding matrices in related technologies, requiring minimal protocol modification and being more easily used by current terminals. Format two, on the other hand, separately shows the first antenna port group information corresponding to the precoding matrix. When this precoding matrix is ​​a precoding matrix corresponding to a transport layer, each transport layer's precoding matrix can include only the weights of the antenna ports in the first antenna port group. For sparse precoding matrices, this makes the output of the first model more concise.

[0124] In addition, a second model corresponding to the first model also exists in this communication system. The second model is configured in the network device. The second model is used by the network device to generate uplink CSI indication information.

[0125] Figure 6 illustrates a flowchart of an uplink transmission method provided in an exemplary embodiment of this application. The method is executed by a terminal device, which may be the terminal device shown in Figure 3. The method includes:

[0126] Step 210: Receive uplink CSI indication information sent by the network device. The uplink CSI indication information is generated by the network device based on the uplink channel information and the second model corresponding to the first model.

[0127] In some embodiments, the correspondence between the first model and the second model means that the first model and the second model are trained together. The first model is used to obtain the precoding matrix based on the uplink CSI indication information; the second model is used to generate uplink CSI indication information based on the uplink channel information.

[0128] The uplink channel information is obtained by the network device based on the reference signal sent by the terminal device. That is, before step 410, the terminal device first sends a reference signal to the network device, which is used to measure the state of the uplink channel.

[0129] Optionally, the uplink channel information includes at least one of the following: feature vector; beam information; and delay information. The feature vector is the feature vector of the uplink channel, which is extracted from the received reference signal by the feature extraction module or the second model in the network device; the beam information refers to the information of the beam used when the reference signal is transmitted; and the delay information is used to indicate the delay of the reference signal during transmission.

[0130] In some embodiments, the uplink CSI indication information is compressed uplink channel information. That is, the second model is used to compress the uplink channel information to obtain the uplink CSI indication information; the first model is used to obtain the precoding matrix through the uplink CSI indication information.

[0131] In summary, the method provided in this application involves a terminal device receiving uplink CSI indication information generated by a network device based on uplink channel information and a second model. This second model corresponds one-to-one with the first model. The uplink CSI indication information obtained by compressing and reconstructing the uplink channel information based on the second model is more efficient in terms of signaling overhead compared to the network device directly indicating the channel state based on the uplink channel information.

[0132] Figure 7 illustrates a flowchart of an uplink transmission method provided in an exemplary embodiment of this application. The method is performed by a network device, which may be the network device shown in Figure 3. The method includes:

[0133] Step 310: Based on the uplink channel information and the second model corresponding to the first model, obtain the uplink CSI indication information.

[0134] In some embodiments, the correspondence between the first model and the second model means that the first model and the second model are trained together. The first model is used to obtain the precoding matrix based on the uplink CSI indication information; the second model is used to generate uplink CSI indication information based on the uplink channel information.

[0135] The uplink channel information is obtained by the network device based on the reference signal sent by the terminal device. That is, before step 410, the terminal device first sends a reference signal to the network device, which is used to measure the state of the uplink channel.

[0136] Optionally, the uplink channel information includes at least one of the following: feature vector; beam information; and delay information. The feature vector is the feature vector of the uplink channel, which is extracted from the received reference signal by the feature extraction module or the second model in the network device; the beam information refers to the information of the beam used when the reference signal is transmitted; and the delay information is used to indicate the delay of the reference signal during transmission.

[0137] In some embodiments, the uplink CSI indication information is compressed uplink channel information. That is, the second model is used to compress the uplink channel information to obtain the uplink CSI indication information; the first model is used to obtain the precoding matrix through the uplink CSI indication information.

[0138] Step 320: Send uplink CSI indication information to the terminal device.

[0139] In some embodiments, the network device indicates the uplink CSI indication information to the terminal device via the DCI. For example, the network device indicates the uplink CSI information to the terminal device via the PMI indication field in the DCI. In one embodiment, the PMI is indicated together with transport layer number indication information.

[0140] In summary, the method provided in this application obtains uplink CSI indication information based on a second model and uplink channel information, where the second model is the AI ​​model corresponding to the first model. The uplink CSI indication information obtained by compressing and reconstructing the uplink channel information based on the second model is more efficient in terms of signaling overhead compared to network devices directly indicating channel status based on uplink channel information.

[0141] Next, we will introduce the working process of the second model.

[0142] 1. Input to the second model.

[0143] In some embodiments, the input to the second model includes uplink channel information. That is, the uplink CSI indication information is obtained by inputting the uplink channel information into the second model. Alternatively, the network device obtains the uplink CSI indication information based on the uplink channel information and the second model corresponding to the first model, including: inputting the uplink channel information into the second model corresponding to the second model in the domain to obtain the uplink CSI indication information.

[0144] In some embodiments, the input to the second model further includes an antenna port group configuration of the terminal device. The antenna port group configuration indicates the result of the terminal device dividing the antenna port groups; or, the antenna port group configuration indicates information about the antenna port groups configured by the terminal device, wherein the individual antenna ports in each antenna port group are coherent. Exemplarily, the antenna port group configuration is at least one of the following: the number of antenna port groups; the number of antenna port groups and the total number of antenna ports; the number of antenna port groups and the number of antenna ports in each antenna port group; the total number of antenna ports and the number of antenna ports in each antenna port group.

[0145] In some embodiments, given any two of the three parameters—the number of antenna port groups, the total number of antenna ports, and the number of antenna ports in each antenna port group—the second model can deduce the remaining parameter. For example, the total number of antenna ports is equal to the number of antenna port groups multiplied by the number of antenna ports in each antenna port group.

[0146] In some embodiments, the second model obtains uplink CSI indication information based on at least one of the antenna port group configuration and the number of transport layers, as well as uplink channel information. The uplink CSI indication information, in addition to indicating the state of the uplink channel to the terminal device, also indicates at least one of the antenna port group configuration and the number of transport layers, or, further indicates the size of the uplink precoding matrix determined based on the antenna port group configuration and / or the number of transport layers.

[0147] In some embodiments, the network device stores at least two AI models. The network device determines a second model from the at least two AI models based on an antenna port group configuration. The second model is the AI ​​model corresponding to the antenna port group configuration. Alternatively, the network device includes only the second model, which determines the size of the uplink precoding matrix based on the antenna port group configuration, and this size will be indicated to the terminal device. Alternatively, the second model includes at least one processing module, each processing module corresponding to one antenna port group configuration; the second model determines a second processing module based on the antenna port group configuration, and inputs uplink channel information into the second processing module to obtain uplink CSI indication information.

[0148] 2. Output of the second model.

[0149] In some embodiments, the second model is used to output uplink CSI indication information. The uplink CSI indication information is used to indicate uplink channel information. The uplink CSI indication information is used by the terminal device to input the precoding matrix obtained from the first model.

[0150] In some embodiments, the second model is further used to output the number of transport layers; or, the second model is further used to output the transport layer configuration. Optionally, the second model outputs the number of transport layers based on uplink channel information. In another optional embodiment, the number of transport layers (or transport layer configuration) is determined by the network device based on the uplink channel information. In addition to sending uplink CSI indication information to the terminal device, the network device also sends transport layer indication information (or transport layer number configuration) to the terminal device, wherein the transport layer indication information (or transport layer number configuration) is used to indicate to the terminal device the number of transport layers (or transport layer configuration) determined by the network device.

[0151] The number of transport layers is used by the terminal device (or the first model) to determine the number of columns in the precoding matrix.

[0152] In some embodiments, the first model and the second model are in one-to-one correspondence, and the first model and the second model are trained together by the terminal device and / or the network device. That is, the first model of the terminal device is trained by the terminal device, or the first model of the terminal device is configured by the network device. And / or, the second model of the network device is reported by the terminal device, or the second model of the network device is trained by the network device.

[0153] That is, both the first model and the second model are trained by the terminal device, and after training, the terminal device reports the second model to the network device; or, both the first model and the second model are trained by the network device, and after training, the network device configures the first model to the terminal device; or, the first model is trained by the terminal device, and the second model is trained by the network device.

[0154] During the configuration of the first model or the reporting of the second model, instructions are required using the configuration of the first model or the second model.

[0155] In some embodiments, the first model configuration and / or the second model configuration includes at least one of the following: model parameters; model structure; model function; model identifier; input information; output information; dataset identifier.

[0156] The model parameters indicate the parameter values ​​in the trained model. The model structure indicates the network structure, such as the number of layers, the number of neurons in each layer, and the connections between neurons. The model function indicates that the model is used for uplink CSI. The model identifier indicates the specific model. Input information indicates the model's inputs, such as the number of bits, the type of input data (e.g., uplink CSI indication information), the size of the input data, and the data type (e.g., integer, string). Output information indicates the model's outputs, such as the number of bits in the output, the type of input data (e.g., uplink CSI indication information), the size of the input data, and the data type (e.g., integer, string). The dataset identifier indicates the dataset used during model training.

[0157] The training process for the first and second models is shown below.

[0158] 1. The first and second models are trained by the terminal device.

[0159] Figure 8 illustrates a flowchart of an uplink transmission method provided in an exemplary embodiment of this application. The method is executed by a terminal device, which may be the terminal device shown in Figure 3. The method includes:

[0160] Step 410: Based on the antenna port group configuration of the terminal device, or based on the antenna port group configuration and at least one transmission layer number, train to obtain a first model and a second model corresponding to the first model.

[0161] In some embodiments, the terminal device trains a first model and a corresponding second model based on the antenna port group configuration. Alternatively, the terminal device determines a training dataset based on the antenna port group configuration and trains the first model and the corresponding second model based on that dataset; in this case, the terminal device stores at least one dataset, each dataset corresponding to one antenna port group configuration. Alternatively, the terminal device trains the first model and the corresponding second model based on the antenna port group configuration and the dataset; in this case, the terminal device stores only one dataset, and the precoding matrices corresponding to the training samples in the dataset are of the same or different sizes; during training, the terminal device selects training samples from the dataset corresponding to the antenna port group configuration based on the antenna port group configuration for training the first and second models. Alternatively, the terminal device trains the first model and the corresponding second model based on the antenna port group configuration and the dataset; that is, the terminal device uses the antenna port group configuration and the training samples in the dataset as input to the first and second models to train them. Alternatively, the terminal device trains a first model and a corresponding second model based on the antenna port group configuration and at least one transmission layer number. The at least one transmission layer number includes the number of transmission layers configured for the terminal device, i.e., the at least one transmission layer number is the number of transmission layers the terminal device may be configured with during transmission. The at least one transmission layer number is determined based on the maximum number of transmission layers, or it is indicated by the network device. The terminal device uses the antenna port group configuration and at least one transmission layer number as input to the first and second models to train the first and second models, or the terminal device uses at least one transmission layer number as input to the first and second models to train the first and second models. In this case, the antenna port group configuration is used to filter training samples.

[0162] As an example and not a limitation, the training dataset includes uplink channel information and the target precoding matrix. The uplink channel information is input into the second model to obtain uplink CSI indication information; then, the uplink CSI indication information is input into the first model to obtain the prediction precoding matrix. Based on the similarity between the prediction precoding matrix and the target precoding matrix, the model parameters of the first and second models are updated using the backpropagation algorithm.

[0163] Step 420: Report the configuration of the second model corresponding to the second model to the network device.

[0164] The second model is used by network devices to generate uplink CSI indication information.

[0165] In some embodiments, after training is completed, the terminal device obtains the second model configuration based on the second model and reports the second model configuration to the network device.

[0166] In summary, the method provided in this application, by using an AI model trained by the terminal device itself, enables the AI ​​model to better match the antenna configuration of the terminal device, thereby improving the reliability of CSI compression. Furthermore, by training an AI model to achieve precoding of uplink signals on a coherent portion of the antenna port group, compared to traditional precoding methods (especially precoding based on a partially coherent codebook), transmission based on precoding can be achieved simply by the network device instructing the compressed uplink CSI indication information. This saves signaling overhead such as the indication codebook and currently used antenna port information, thus conserving air interface resources.

[0167] Figure 8 illustrates a flowchart of an uplink transmission method provided in an exemplary embodiment of this application. The method is performed by a network device, which may be the network device shown in Figure 3. The method includes:

[0168] Step 510: Receive the second model configuration sent by the terminal device, and obtain the second model based on the second model configuration.

[0169] In some embodiments, the network device receives a second model configuration sent by the terminal device and obtains a second model corresponding to the first model of the terminal device based on the second model configuration. In some embodiments, the second model is associated with an antenna port group configuration, or with an antenna port group configuration and at least one transport layer number. Alternatively, the second model can be described as being trained based on the antenna port group configuration, or based on the antenna port group configuration and at least one transport layer number.

[0170] In summary, the method provided in this application, by using an AI model trained by the terminal device itself, enables the AI ​​model to better match the antenna configuration of the terminal device, thereby improving the reliability of CSI compression. Furthermore, by training an AI model to achieve precoding of uplink signals on a coherent portion of the antenna port group, compared to traditional precoding methods (especially precoding based on a partially coherent codebook), transmission based on precoding can be achieved simply by the network device instructing the compressed uplink CSI indication information. This saves signaling overhead such as the indication codebook and currently used antenna port information, thus conserving air interface resources.

[0171] 2. The first and second models are trained by network devices.

[0172] Figure 9 illustrates a flowchart of an uplink transmission method provided in an exemplary embodiment of this application. The method is executed by a terminal device, which may be the terminal device shown in Figure 3. The method includes:

[0173] Step 610: Report the antenna port group configuration of the terminal device to the network device.

[0174] In some embodiments, the terminal device periodically reports the antenna port group configuration of the terminal device to the network device; or, after accessing the network device, the terminal device reports the antenna port group configuration of the terminal device to the network device through UE capability; or, after receiving an indication message from the network device requesting the reporting, the terminal device reports the antenna port group configuration of the terminal device to the network device.

[0175] Step 620: Receive the first model configuration and obtain the first model based on the first model configuration. The first model is associated with the antenna port group configuration, or with the antenna port group configuration and at least one transmission layer number.

[0176] In some embodiments, the first model is based on the antenna port group configuration of the terminal device, or is trained based on the antenna port group configuration of the terminal device and at least one transmission layer number.

[0177] In summary, the method provided in this application, through the AI ​​model configured in the network device, enables better matching between the models on the terminal device and the network device side, thereby improving the reliability of CSI compression. Furthermore, by training an AI model to achieve precoding of uplink signals on a coherent portion of the antenna port group, compared to traditional precoding methods (especially precoding based on a partially coherent codebook), transmission based on precoding can be achieved simply by the network device indicating the compressed uplink CSI indication information. This saves signaling overhead such as the indication codebook and currently used antenna port information, thus conserving air interface resources.

[0178] Figure 9 illustrates a flowchart of an uplink transmission method provided in an exemplary embodiment of this application. The method is performed by a network device, which may be the network device shown in Figure 3. The method includes:

[0179] Step 710: Configure the antenna port group of the receiving terminal device.

[0180] In some embodiments, the network device starts training the first model and the second model after receiving the antenna port group configuration from the terminal device; or, before starting training the first model and the second model, the network device sends an instruction message to the terminal device requesting reporting, the instruction message being used to instruct the terminal device to report the antenna port group configuration.

[0181] Step 720: Based on the antenna port group configuration of the terminal device, or based on the antenna port group configuration and at least one transmission layer number, train to obtain a second model and a first model corresponding to the second model.

[0182] In some embodiments, the network device trains a first model and a corresponding second model based on the antenna port group configuration. Alternatively, the network device determines a training dataset based on the antenna port group configuration and trains the first model and the corresponding second model based on that dataset; in this case, the network device stores at least one dataset, each dataset corresponding to one antenna port group configuration. Alternatively, the network device trains the first model and the corresponding second model based on the antenna port group configuration and the dataset; in this case, the network device stores only one dataset, and the precoding matrices corresponding to the training samples in the dataset have the same or different sizes; during training, the network device selects training samples from the dataset corresponding to the antenna port group configuration based on the antenna port group configuration for training the first and second models. Alternatively, the network device trains the first model and the corresponding second model based on the antenna port group configuration and the dataset; that is, the network device uses the antenna port group configuration and the training samples in the dataset as input to the first and second models to train them. Alternatively, the network device trains a first model and a corresponding second model based on the antenna port group configuration and at least one transmission layer number. The at least one transmission layer number includes the number of transmission layers configured for the terminal device, i.e., the at least one transmission layer number is the number of transmission layers the terminal device may be configured with during transmission. The at least one transmission layer number is determined based on the maximum number of transmission layers, or the at least one transmission layer number is indicated by the network device. The network device uses the antenna port group configuration and at least one transmission layer number as input to the first and second models to train the first and second models, or the network device uses at least one transmission layer number as input to the first and second models to train the first and second models. In this case, the antenna port group configuration is used to filter training samples.

[0183] As an example and not a limitation, the training dataset includes uplink channel information and the target precoding matrix. The uplink channel information is input into the second model to obtain uplink CSI indication information; then, the uplink CSI indication information is input into the first model to obtain the prediction precoding matrix. Based on the similarity between the prediction precoding matrix and the target precoding matrix, the model parameters of the first and second models are updated using the backpropagation algorithm.

[0184] Step 730: Send the first model configuration corresponding to the first model to the terminal device.

[0185] In some embodiments, after training is completed, the network device obtains a first model configuration based on the first model and sends the first model configuration to the terminal device.

[0186] In summary, the method provided in this application, through the AI ​​model configured in the network device, enables better matching between the models on the terminal device and the network device side, thereby improving the reliability of CSI compression. Furthermore, by training an AI model to achieve precoding of uplink signals on a coherent portion of the antenna port group, compared to traditional precoding methods (especially precoding based on a partially coherent codebook), transmission based on precoding can be achieved simply by the network device indicating the compressed uplink CSI indication information. This saves signaling overhead such as the indication codebook and currently used antenna port information, thus conserving air interface resources.

[0187] In some embodiments, the terminal device and the network device each store at least two AI models. The terminal device and the network device select a first model and a second model from the at least two AI models based on the antenna port group configuration. Alternatively, the terminal device and the network device select a first model and a second model from the at least two AI models based on the antenna port group configuration and the maximum number of transmission layers. Optionally, the at least two AI models stored in the terminal device and the network device may be the same or different.

[0188] In some embodiments, at least two AI models stored in the terminal device are trained by the terminal device; or, at least two AI models stored in the terminal device are configured by the network device; or, of the at least two AI models stored in the terminal device, some AI models are trained by the terminal device and others are configured by the network device.

[0189] In some embodiments, at least two AI models stored in the network device are trained by the network device; or, at least two AI models stored in the network device are reported by the terminal device; or, of the at least two AI models stored in the network device, some AI models are trained by the network model and others are reported by the terminal device.

[0190] Figure 10 illustrates a flowchart of an uplink transmission method provided in an exemplary embodiment of this application. The method is executed by a terminal device, which may be the terminal device shown in Figure 3. The method includes:

[0191] In some embodiments, the terminal device stores at least two AI models, each AI model corresponding to an antenna port group configuration and / or at least one transmission layer.

[0192] Step 810: Report the antenna port group configuration of the terminal device to the network device.

[0193] In some embodiments, step 810 is similar to step 610 described above, and will not be repeated here.

[0194] Step 820: Select the AI ​​model corresponding to the antenna port group configuration of the terminal device as the first model.

[0195] In another possible embodiment, the terminal device selects an AI model corresponding to its antenna port group configuration as the first model, or selects an AI model corresponding to its antenna port group configuration and a first transmission layer number as the first model. Here, the first transmission layer number is the number of transmission layers supported by the terminal device; or, the first transmission layer number is the number of transmission layers configured for the terminal device. For example, the network device indicates the first transmission layer number to the terminal device based on uplink channel information, and the terminal device selects the first model based on the antenna port group configuration and the first transmission layer number; or, the network device indicates the first transmission layer number to the terminal device based on uplink channel information, and the terminal device uses the first transmission layer number as input to the first model, where the first model is the AI ​​model selected by the terminal device corresponding to its antenna port group configuration.

[0196] In some embodiments, the first transmission layer number is a subset of the transmission layers corresponding to the first model. For example, at least one AI model contains three AI models c1-c3 corresponding to the antenna port group configuration of the terminal device. The transmission layer number corresponding to AI model c1 is 2, the transmission layer number corresponding to AI model c2 is 4, and the transmission layer number corresponding to AI model c3 is {1, 2, 4}. If the first transmission layer number is 2, either AI model c1 or AI model c3 can be selected as the first model; if the first transmission layer number is {2, 4}, only AI model c3 can be selected as the first model.

[0197] In some embodiments, each AI model stored in the terminal device corresponds to an antenna port group configuration, and the AI ​​model corresponding to the antenna port group configuration of the terminal device is selected as the first model; or, each AI model stored in the terminal device corresponds to at least one transmission layer number, and the AI ​​model corresponding to the first transmission layer number is selected as the first model; or, each AI model stored in the terminal device corresponds to an antenna port group configuration and at least one transmission layer number, and the AI ​​model corresponding to the antenna port group configuration and the first transmission layer number of the terminal device is selected as the first model.

[0198] In some embodiments, steps 810 and 820 can be executed in an alternate order or simultaneously.

[0199] In summary, the method provided in this application employs pre-training to obtain multiple AI models. During use, only the corresponding AI model needs to be selected based on the antenna port group configuration, enabling rapid configuration and application of the AI ​​models. Furthermore, by training AI models to achieve precoding of uplink signals on coherent partial antenna port groups, compared to traditional precoding methods (especially precoding based on partial coherent codebooks), transmission based on precoding can be achieved simply by the network device instructing compressed uplink CSI indication information. This saves signaling overhead such as indication codebooks and currently used antenna port information, thus conserving air interface resources.

[0200] Figure 10 illustrates a flowchart of an uplink transmission method provided in an exemplary embodiment of this application. The method is performed by a network device, which may be the network device shown in Figure 3. The method includes:

[0201] In some embodiments, the network device stores at least two AI models, each corresponding to an antenna port group configuration.

[0202] Step 910: Configure the antenna port group of the receiving terminal device.

[0203] In some embodiments, step 910 is similar to step 710 described above, and will not be repeated here.

[0204] Step 920: Select the AI ​​model corresponding to the antenna port group configuration of the terminal device as the second model. In some embodiments, each AI model stored in the network device corresponds to an antenna port group configuration, and the AI ​​model corresponding to the antenna port group configuration of the terminal device is selected as the second model.

[0205] In summary, the method provided in this application employs pre-training to obtain multiple AI models. During use, only the corresponding AI model needs to be selected based on the antenna port group configuration, enabling rapid configuration and application of the AI ​​models. Furthermore, by training AI models to achieve precoding of uplink signals on coherent partial antenna port groups, compared to traditional precoding methods (especially precoding based on partial coherent codebooks), transmission based on precoding can be achieved simply by the network device instructing compressed uplink CSI indication information. This saves signaling overhead such as indication codebooks and currently used antenna port information, thus conserving air interface resources.

[0206] In some embodiments, the terminal device reports capability information to the network device, the capability information being used to indicate that the terminal device configures some coherent antenna ports. In some embodiments, the network device receives the capability information reported by the terminal device, the capability information being used to indicate that the terminal device configures some coherent antenna ports. In some embodiments, the antenna ports within each antenna port group are coherent; the antenna ports between different antenna port groups are incoherent.

[0207] Since the first and second models can be trained by terminal devices and / or network devices, the following section introduces the complete process of training and using the first and second models.

[0208] 1. The first and second models are trained and reported based on the terminal devices.

[0209] Figure 11 shows a flowchart of an uplink transmission method provided in an exemplary embodiment of this application. The method includes:

[0210] Step 1010: The terminal device trains the first model and the second model based on the antenna port group configuration.

[0211] In some embodiments, the terminal device trains the first model and the second model based on its own antenna port group configuration; that is, the model training is performed on the terminal device. The first model and the second model are corresponding; the second model is used by the network device to obtain uplink CSI indication information based on the measured uplink channel information, and the first model is used by the terminal device to obtain the uplink precoding matrix based on the uplink CSI indication information.

[0212] Step 1020: The terminal device sends the second model configuration to the network device.

[0213] Specifically, the second model configuration includes model parameters and model structure. Optionally, the second model configuration may also include at least one of the following: model functionality; model identifier; input information; output information; dataset identifier.

[0214] In some embodiments, the terminal device may also report capability information to the network device. The capability information indicates that the terminal device configures some coherent antenna ports, that is, some antenna ports of the terminal device are coherent and some antenna ports are incoherent.

[0215] Step 1030: The network device is configured based on the second model to obtain the second model. In some embodiments, step 1030 is similar to step 510 above, and will not be described again here.

[0216] Step 1040: The terminal device sends a reference signal to the network device.

[0217] In some embodiments, the terminal device sends an SRS (Sounding Reference Signal) to the network device. The reference signal is used by the network device to measure uplink channel information.

[0218] Step 1050: The network device obtains uplink CSI indication information based on uplink channel information and the second model.

[0219] In some embodiments, the network device takes uplink channel information as input to the second model and outputs uplink CSI indication information.

[0220] In some embodiments, in addition to the uplink channel information, the input of the second model may also include the antenna port group configuration of the terminal device.

[0221] In one implementation, the antenna port group configuration includes the number of antenna port groups. For example, the number of antenna port groups can be 2, 4, 8, etc.

[0222] In another implementation, the antenna port group configuration includes the number of antenna port groups and the total number of antenna ports; or, it includes the number of antenna port groups and the number of antenna ports in each antenna port group; or, it includes the total number of antenna ports and the number of antenna ports in each antenna port group. For example, the total number of antenna ports in the terminal device can be 8, 16, 32, etc. In this case, the number of antenna port groups and the total number of antenna ports can be input as a parameter combination into the second model, such as {2, 8}, {4, 8}, {4, 16}, etc. Alternatively, the number of antenna ports in each antenna port group can be 2, 4, 8, 16, etc. In this case, the number of antenna port groups and the number of antenna ports in each antenna port group can be input as a parameter combination into the second model, such as {2, 4}, {4, 2}, {4, 4}, etc.

[0223] In some embodiments, parameters in the antenna port group configuration, such as the number of antenna port groups, the total number of antenna ports, and the number of antenna ports in each antenna port group, are pre-reported to the network device by the terminal device. For example, this can be reported to the network device via terminal capability reporting. Not all three quantities need to be reported to the network device; the terminal device can report only two, and the third quantity can be calculated from the two reported quantities. For example, the total number of antenna ports equals the number of antenna port groups multiplied by the number of antenna ports contained in each antenna port group.

[0224] Step 1060: The network device sends uplink CSI indication information to the terminal device.

[0225] Specifically, network devices use the PMI indication field in DCI to indicate uplink CSI indication information to terminal devices.

[0226] Step 1070: The terminal device obtains the precoding matrix based on the uplink CSI indication information and the first model.

[0227] The precoding matrix corresponds to the target antenna port group (also known as the first antenna port group) among the multiple antenna port groups of the terminal device. The target antenna port group is a subset of the multiple antenna port groups.

[0228] The first AI model is the AI ​​model corresponding to the second AI model, meaning that the two AI models are trained in pairs.

[0229] Specifically, the terminal device uses the uplink CSI indication information as input to the first model, thereby outputting a precoding matrix. In addition to the uplink CSI indication information, the input to the first model may also include the terminal device's antenna port group configuration and / or the number of transmission layers.

[0230] In one embodiment, the antenna port group configuration includes the number of antenna port groups. In another embodiment, the antenna port group configuration includes the number of antenna port groups and the total number of antenna ports; or, includes the number of antenna port groups and the number of antenna ports in each antenna port group; or, includes the total number of antenna ports and the number of antenna ports in each antenna port group.

[0231] In some embodiments, the output of the first model includes, in addition to the uplink precoding matrix, the target antenna port group information (also referred to as the first antenna port group information) corresponding to the precoding matrix. The target antenna port group information is used to indicate the target antenna port group from among multiple antenna port groups of the terminal device. For example, it can indicate the index of the target antenna port group using log₂N bits of information, where N is the number of antenna port groups of the terminal device. Alternatively, the weights corresponding to the target antenna port group in the precoding matrix are non-zero, while the weights corresponding to other antenna ports are 0. That is, the precoding matrix actually only contains the precoding information corresponding to the target antenna port group. The target antenna port group information and the precoding matrix can be indicated separately for each transmission layer. Specifically, the terminal device includes at least one transmission layer; the first model outputs the target antenna port group information and the precoding matrix for each transmission layer; or, the first model outputs the target antenna port group information for each transmission layer in addition to the precoding matrix; or, the first model outputs the precoding matrix for each transmission layer.

[0232] In some embodiments, the different antenna ports contained in an antenna port group are coherent, and the antenna ports of different antenna port groups are incoherent.

[0233] Step 1080: The terminal device sends a precoded uplink signal to the network device.

[0234] In some embodiments, the terminal device precodes the uplink signal using a precoding matrix. In some embodiments, the terminal device transmits the precoded uplink signal through a target antenna port group. That is, the terminal device does not transmit uplink signals on other antenna ports (the weights in the corresponding precoding matrix are 0).

[0235] In summary, the method provided in this application, for partially coherent terminal devices, can utilize AI models for CSI compression and reconstruction. Furthermore, different models are employed for different antenna coherence conditions (e.g., different numbers of antenna port groups), or the antenna coherence conditions are used as model input. This not only reduces the overhead of uplink CSI indicator bits but also improves the accuracy of uplink CSI. By using an AI model trained by the terminal device itself, the AI ​​model can better match the antenna configuration of the terminal device, thereby improving the reliability of CSI compression.

[0236] 2. The first and second models are trained and configured based on network devices.

[0237] Figure 12 shows a flowchart of an uplink transmission method provided in an exemplary embodiment of this application. The method includes:

[0238] Step 1110: The terminal device sends the antenna port group configuration to the network device. In some embodiments, step 1110 is similar to steps 610 and 810 described above, and will not be repeated here.

[0239] Step 1120: The network device trains the first model and the second model based on the antenna port group configuration. In some embodiments, step 1120 is similar to step 720 above, and will not be described again here.

[0240] Step 1130: The network device sends the first model configuration to the terminal device.

[0241] In some embodiments, before the network device sends the first model configuration, the terminal device reports capability information to the network device. The capability information indicates that the terminal device configures some coherent antenna ports, that is, some antenna ports of the terminal device are coherent and some antenna ports are incoherent.

[0242] In some implementations, the network device determines and configures the first model based on the antenna port group configuration reported by the terminal device and / or the number of transport layers determined by the network device (or the second model).

[0243] Optionally, the terminal device reports the number of antenna port groups (i.e., the antenna port group configuration includes the number of antenna port groups) to the network device, and the network device determines the first model from a pre-trained AI model based on the number of antenna port groups. Different AI models can correspond to different numbers of antenna port groups.

[0244] Optionally, the terminal device reports the number of antenna port groups and the total number of antenna ports to the network device (i.e., the antenna port group configuration includes the number of antenna port groups and the total number of antenna ports). The network device then determines a first model from a pre-trained AI model based on the number of antenna port groups and the total number of antenna ports. Different AI models can correspond to different numbers of antenna port groups and the total number of antenna ports. For example, these two numbers can form a parameter combination, such as {2, 8}, {4, 16}, etc., with different combinations corresponding to different AI models.

[0245] Optionally, the terminal device reports the number of antenna port groups and the number of antenna ports in each antenna port group to the network device (i.e., the antenna port group configuration includes the number of antenna port groups and the number of antenna ports in each antenna port group). The network device then determines a first model from a pre-trained model based on the number of antenna port groups and the number of antenna ports in each antenna port group. Different AI models can correspond to different numbers of antenna port groups and different numbers of antenna ports in each antenna port group. For example, these two numbers can form a parameter combination, such as {2, 4} or {4, 4}, with different combinations corresponding to different AI models. The antenna port group configuration can be reported to the network device via the terminal's reporting capability.

[0246] Furthermore, the second model used by the network device to obtain uplink CSI indication information is jointly trained in pairs with the first model, so the second model can also be obtained based on the antenna port group configuration. That is, the AI ​​models on both the terminal device side and the network device side are obtained based on the same antenna port group configuration, thus ensuring their consistency.

[0247] In another implementation, the network device configures multiple AI models, including a first AI model. The terminal device obtains the first model from the multiple AI models based on its own antenna port group configuration and / or the number of transport layers determined and indicated by the network device (or the second model). That is, the network device sends at least two AI model configurations to the terminal device.

[0248] Specifically, multiple AI models correspond to different antenna port group configurations and / or at least one transmission layer.

[0249] Optionally, the terminal device selects a first model from multiple AI models based on the number of antenna port groups, or based on the number of antenna port groups and the total number of antenna ports, or based on the number of antenna port groups and the number of antenna ports in each antenna port group. At the same time, the network device also determines a second model from the multiple AI models based on the antenna port group configuration reported by the terminal device for generating uplink CSI indication information. This ensures that the second model matches the first model (corresponding to the same antenna port group configuration), guaranteeing the reliability of CSI compression and recovery.

[0250] Step 1140: The terminal device is configured based on the first model to obtain the first model.

[0251] In some implementations, the terminal device receives a first model configuration sent by the network device and obtains a first model based on the first model configuration. The first model is determined based on the antenna port group configuration reported by the terminal device.

[0252] In other embodiments, the terminal device receives multiple AI model configurations sent by the network device. These multiple AI model configurations include a first model configuration. The terminal device obtains the first model configuration from the multiple AI model configurations based on its own antenna port group configuration and / or the number of transport layers determined and indicated by the network device (or a second model). Each of the multiple AI models corresponds to a different antenna port group configuration and / or a different number of transport layers.

[0253] Step 1150: The terminal device sends a reference signal to the network device. In some embodiments, step 1150 is similar to step 1040 described above, and will not be repeated here.

[0254] Step 1160: The network device obtains uplink CSI indication information based on the uplink channel information and the second model. In some embodiments, step 1160 is similar to steps 510 and 1050 described above, and will not be repeated here.

[0255] Step 1170: The network device sends uplink CSI indication information to the terminal device. In some embodiments, step 1170 is similar to steps 520 and 1060 described above, and will not be repeated here.

[0256] Step 1180: The terminal device obtains a precoding matrix based on the uplink CSI indication information and the first model. In some embodiments, step 1180 is similar to steps 210 and 1070 described above, and will not be repeated here.

[0257] Step 1190: The terminal device sends a precoded uplink signal to the network device. In some embodiments, step 1190 is similar to steps 220 and 1080 described above, and will not be repeated here.

[0258] In some embodiments, step 1150 may be implemented before step 1130, such that the network device determines the number of transport layers based on the uplink channel information corresponding to the reference signal sent by the terminal device.

[0259] In summary, the method provided in this application, for partially coherent terminal devices, can utilize AI models for CSI compression and reconstruction. Furthermore, it employs different models for different antenna coherence conditions (e.g., different numbers of antenna port groups) or uses the antenna coherence condition as model input, which not only reduces the overhead of uplink CSI indicator bits but also improves the accuracy of uplink CSI. Through the AI ​​model configured on the network device, the models on the terminal device and network device sides can be better matched, thereby improving the reliability of CSI compression.

[0260] Figure 13 shows a structural block diagram of an uplink transmission apparatus provided in an exemplary embodiment of this application. The apparatus 1200 can be implemented as part of a terminal device, which may be the terminal device shown in Figure 3. The apparatus 1200 includes:

[0261] The first generation module 1210 is used to obtain an uplink precoding matrix based on uplink CSI indication information and a first model. The precoding matrix corresponds to the first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is a partial antenna port group in at least two antenna port groups.

[0262] The first model includes a neural network model. This first model is used by the terminal device to obtain the uplink precoding matrix based on the uplink CSI indication information. The uplink CSI indication information may be referred to as uplink indication information, uplink PMI indication information, first CSI indication information, or other names, and this embodiment does not limit the specific name used. The uplink CSI indication information is used to indicate relevant information about the uplink channel, such as PMI information, to the terminal device.

[0263] In some embodiments, the terminal device includes at least two antenna port groups, with the first antenna port group corresponding to the precoding matrix obtained based on the first model. The first antenna port group is used to transmit the uplink signal precoded by the precoding matrix. The first antenna port group is one or a portion of the at least two antenna port groups.

[0264] In some embodiments, each antenna port group of the terminal device includes at least one antenna port, and the antenna ports included in each antenna port group are different. For example, the terminal device includes eight antenna ports a1-a8, and at least two antenna port groups of the terminal device include {a1-a4} and {a5-a8}. In some embodiments, the precoding matrix of the terminal device corresponds to a first antenna port group in the at least two antenna port groups of the terminal device, and the first antenna port group is used by the terminal device to transmit precoded uplink signals. Optionally, the terminal device includes at least one transport layer, and the precoding matrix of the terminal can be divided into precoding matrices corresponding to at least one transport layer. The precoding matrix of one transport layer corresponds to one antenna port group in the at least two antenna port groups; that is, the precoding matrix of the terminal device corresponds to the first antenna port group in the at least two antenna port groups of the terminal device, and the first antenna port group is a subset of the at least two antenna port groups. For example, the terminal device includes one transport layer b1 and two antenna port groups (including antenna port groups {a1-a4} and antenna port groups {a5-a8}), wherein the precoding matrix of the transport layer b1 corresponds to the antenna port group {a1-a4}. In this case, the precoding matrix of the terminal device corresponds to the first antenna port group and the first antenna port group contains one antenna port group (i.e., antenna port group {a1-a4}).

[0265] The first transmitting module 1220 is used to transmit the uplink signal precoded by the precoding matrix through the first antenna port group.

[0266] In some embodiments, transmitting the uplink signal precoded by the precoding matrix through the first antenna port group can be understood as "the terminal device uses the precoding matrix to precode the uplink signal to obtain the precoded uplink signal; and transmits the precoded uplink signal through the first antenna port group." It can also be described as "the terminal device uses its own precoding matrix to precode the uplink signal to obtain the precoded uplink signal; and transmits the precoded uplink signal through the first antenna port group." Alternatively, it can be described as "the terminal device uses the precoding matrices of each transmission layer to precode the uplink signal to obtain the precoded uplink signal; and transmits the precoded uplink signal through the first antenna port group of each transmission layer."

[0267] In summary, the apparatus provided in this application involves a terminal device generating a precoding matrix on a partial antenna port group (i.e., a coherent antenna port group) based on uplink CSI indication information and a first model. The uplink signal is then precoded using this precoding matrix to achieve coherent precoding on the partial antenna port group. Compared to traditional precoding methods based on partially coherent codebooks, neither the terminal device nor the network device needs to use excessive indication signaling to specify the antenna port group and precoding matrix used for uplink transmission from the codebook, thus saving signaling overhead in the uplink transmission process based on precoding methods.

[0268] Figure 14 shows a structural block diagram of an uplink transmission device provided in an exemplary embodiment of this application. The device 1300 can be implemented as a network device, which may be the network device shown in Figure 3. The device 1300 includes:

[0269] The first receiving module 1310 is used to receive the precoded uplink signal sent by the terminal device through the first antenna port group;

[0270] The precoded uplink signal is obtained by precoding based on a precoding matrix, which is obtained by the terminal device based on uplink CSI indication information and a first model. The precoding matrix corresponds to the first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is a subset of the at least two antenna port groups. The first model includes a neural network model. The first model is used by the terminal device to obtain the uplink precoding matrix based on the uplink CSI indication information. The uplink CSI indication information may be called uplink PMI indication information, first CSI indication information, or other names, and this application embodiment does not limit it. The uplink CSI indication information is used to indicate relevant information of the uplink channel to the terminal device, such as PMI information.

[0271] In some embodiments, the terminal device includes at least two antenna port groups, with the first antenna port group corresponding to the precoding matrix obtained based on the first model. The first antenna port group is used to transmit the uplink signal precoded by the precoding matrix. The first antenna port group is one or a portion of the at least two antenna port groups.

[0272] In some embodiments, the terminal device includes at least one antenna port group, which corresponds to a first antenna port group based on a precoding matrix obtained from a first model. The first antenna port group is used to transmit uplink signals precoded by the precoding matrix. The first antenna port group is one or a subset of at least two antenna port groups.

[0273] In some embodiments, each antenna port group of the terminal device includes at least one antenna port, and the antenna ports included in each antenna port group are different. For example, the terminal device includes eight antenna ports a1-a8, and at least two antenna port groups of the terminal device include {a1-a4} and {a5-a8}.

[0274] In some embodiments, the precoding matrix of the terminal device corresponds to a first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is used by the terminal device to transmit precoded uplink signals. Optionally, the terminal device includes at least one transport layer, and the precoding matrix of the terminal can be divided into precoding matrices corresponding to at least one transport layer. The precoding matrix of one transport layer corresponds to one antenna port group in at least two antenna port groups. That is, the precoding matrix of the terminal device corresponds to the first antenna port group in at least two antenna port groups of the terminal device, and the first antenna port group is a portion of the antenna port groups in at least two antenna port groups. For example, the terminal device includes one transport layer b1 and two antenna port groups (including antenna port groups {a1-a4} and antenna port groups {a5-a8}), wherein the precoding matrix of the transport layer b1 corresponds to antenna port group {a1-a4}. In this case, the precoding matrix of the terminal device corresponds to the first antenna port group, and the first antenna port group contains one antenna port group (i.e., antenna port group {a1-a4}).

[0275] In summary, the apparatus provided in this application involves a terminal device generating a precoding matrix on a partial antenna port group (i.e., a coherent antenna port group) based on uplink CSI indication information and a first model. The uplink signal is then precoded using this precoding matrix to achieve coherent precoding on the partial antenna port group. Compared to traditional precoding methods based on partially coherent codebooks, neither the terminal device nor the network device needs to use excessive indication signaling to specify the antenna port group and precoding matrix used for uplink transmission from the codebook, thus saving signaling overhead in the uplink transmission process based on precoding methods.

[0276] Next, we will introduce the working process of the first model.

[0277] 1. Input to the first model.

[0278] In some embodiments, the input to the first model includes uplink CSI indication information. That is, the precoding matrix is ​​obtained by inputting the uplink CSI indication information into the first model. Alternatively, the first generation module 1210 is also used by the terminal device to input the uplink CSI indication information into the first model to obtain the precoding matrix.

[0279] In some embodiments, the input to the first model further includes an antenna port group configuration of the terminal device. The antenna port group configuration indicates the result of the terminal device dividing the antenna port groups; or, the antenna port group configuration indicates information about the antenna port groups configured by the terminal device, wherein the individual antenna ports in each antenna port group are coherent. Exemplarily, the antenna port group configuration is at least one of the following: the number of antenna port groups; the number of antenna port groups and the total number of antenna ports; the number of antenna port groups and the number of antenna ports in each antenna port group; the total number of antenna ports and the number of antenna ports in each antenna port group.

[0280] In some embodiments, given any two of the three parameters—the number of antenna port groups, the total number of antenna ports, and the number of antenna ports in each antenna port group—the first model can deduce the remaining parameter. For example, the total number of antenna ports is equal to the number of antenna port groups multiplied by the number of antenna ports in each antenna port group.

[0281] In some embodiments, the input to the first model further includes the number of transport layers of the terminal device; or, the input to the first model further includes the transport layer configuration of the terminal device, which is used to indicate the number of transport layers during uplink transmission. Optionally, the number of transport layers can also be indicated by rank.

[0282] In some embodiments, the size of the precoding matrix output by the first model is related to at least one of the antenna port group configuration and the number of transport layers (or transport layer configuration). That is, the first model determines the size of the output precoding matrix based on at least one of the antenna port group configuration and the number of transport layers. Optionally, the first model determines the number of rows of the precoding matrix based on the antenna port group configuration, and / or, the first model determines the number of columns of the precoding matrix based on the number of transport layers. In some embodiments, the first model determines the number of rows of the precoding matrix based on the parameter "total number of antenna ports" in the antenna port group configuration; or, the first model determines the number of rows of the precoding matrix based on the parameter "number of antenna ports in each antenna port group" in the antenna port group configuration.

[0283] In some embodiments, the terminal device stores at least two AI models. The terminal device determines a first model from the at least two AI models based on the antenna port group configuration and / or the number of transmission layers. The first model is the AI ​​model corresponding to the antenna port group configuration and / or the number of transmission layers. Alternatively, the terminal device includes only the first model, and the size of the output precoding matrix is ​​determined based on the antenna port group configuration and / or the number of transmission layers. Alternatively, the first model includes at least one processing module, each processing module corresponding to one antenna port group configuration and / or the number of transmission layers; the first model determines the first processing module based on the antenna port group configuration and / or the number of transmission layers, and the uplink CSI indication information is input into the first processing module to obtain the precoding matrix.

[0284] In summary, the apparatus provided in this application embodiment further includes antenna port group configuration and / or transmission layer number as input to the first model. This allows terminal devices and network devices to use different models for different antenna coherence conditions (e.g., different numbers of antenna port groups) or to use the antenna coherence conditions as model input. Compared to traditional methods that require indicating the precoding matrix and the currently used antenna port group, using uplink CSI indication information and the corresponding first model to obtain the precoding matrix not only reduces the overhead of uplink CSI indication bits but also improves the accuracy of uplink CSI.

[0285] 2. Output of the first model.

[0286] In some embodiments, the first model is used to output a precoding matrix, which corresponds to a first antenna port group. That is, the precoding matrix includes at least the weights of each antenna port in the first antenna port group, so that the terminal device precodes the uplink signal based on the weights.

[0287] In some embodiments, the terminal device includes at least one transport layer; a first model outputs first antenna port group information and a precoding matrix for each transport layer; or, the first model outputs first antenna port group information for each transport layer, and the first model also outputs a precoding matrix. The first antenna port group information is used to indicate the first antenna port group in at least two antenna port groups.

[0288] For example, the terminal device includes two transport layers (including transport layer b1 and transport layer b2), eight antenna ports (including antenna ports a1-a8), and four antenna port groups (including antenna port groups {a1, a2}, {a3, a4}, {a5, a6}, and {a7, a8}, and each antenna port group and its corresponding index are shown in Table 1). If the precoding matrix is ​​[[1, 1, 0, 0, 0, 0, 0, 0]... T ,[0,0,0,0,1,-j,0,0] T The precoding matrix is ​​defined as follows: each row corresponds to an antenna port (the first row corresponds to antenna port a1, the second row to antenna port a2, the third row to antenna port a3, the fourth row to antenna port a4, the fifth row to antenna port a5, the sixth row to antenna port a6, the seventh row to antenna port a7, and the eighth row to antenna port a8), and each column corresponds to a transport layer (the first column corresponds to transport layer b1, and the second column corresponds to transport layer b2). The first model then outputs the precoding matrix (also called the precoding submatrix, precoding vector, vector, etc.) corresponding to transport layer b1 as [1, 1, 0, 0, 0, 0, 0, 0]. TAnd its corresponding first antenna port group information "00", and the precoding matrix [0, 0, 0, 0, 1, -j, 0, 0] corresponding to transport layer b2. T And its corresponding first antenna port group information "10".

[0289] In some embodiments, the first model outputs a precoding matrix corresponding to at least one transmission layer, and the first model outputs the first antenna port group information of each transmission layer respectively; or, the first model outputs the first antenna port group information and the precoding matrix of each transmission layer respectively.

[0290] For example, the first model outputs a precoding matrix [[1, 1, 0, 0, 0, 0, 0, 0]]. T ,[0,0,0,0,1,-j,0,0] T [ and the first antenna port group information of each transmission layer "00, 10" (the first output represents the first antenna port group information of the first transmission layer, and the second output represents the first antenna port group information of the second transmission layer); or, the first model outputs the first antenna port group information of each transmission layer and the precoding matrix "00, [1, 1, 0, 0, 0, 0, 0, 0]". T ,10,[0,0,0,0,1,-j,0,0] T "(The first output represents the first antenna port group information of the first transport layer, the second output represents the precoding matrix of the first transport layer, the third output represents the first antenna port group information of the second transport layer, and the fourth output represents the precoding matrix of the second transport layer.)"

[0291] It should be noted that the output format of the first model can be as shown above, first outputting the first antenna port group information of the first transmission layer, then outputting the precoding matrix of the first transmission layer, ..., outputting the first antenna port group information of the nth transmission layer, and outputting the precoding matrix of the nth transmission layer; or, the output format of the first model can also be to first output the first antenna port group information of each transmission layer, then outputting the precoding matrix of each transmission layer, such as first outputting the first antenna port group information of the first transmission layer, ..., outputting the first antenna port group information of the nth transmission layer, outputting the precoding matrix of the first transmission layer, ..., outputting the precoding matrix of the nth transmission layer; or, the output format of the first model can also be to output the index of each transmission layer, the first antenna port group information, and the precoding matrix in the form of a table, as shown in Table 2 below. In other words, the output format of the first model can also be other output formats, and this application does not limit the output format of the first model. Wherein, n is a positive integer greater than 1, and the value of n is related to the number of transmission layers.

[0292] The precoding matrix output by the first model exists in the following two forms.

[0293] Form 1: The output precoding matrix includes the weights of each antenna port in at least two antenna port groups.

[0294] In some embodiments, the precoding matrix includes the weights of each antenna port in at least two antenna port groups of the terminal device. The weights of each antenna port in the first antenna port group are non-zero, and the weights of each antenna port in the second antenna port group are zero. The second antenna port group is the antenna port group other than the first antenna port group among the at least two antenna port groups. Because the weight of the second antenna port group is zero, it is not actually used for uplink signal transmission; therefore, the precoding matrix here corresponds only to the first antenna port group.

[0295] The antenna ports in the first antenna port group are used to transmit precoded uplink signals; the antenna ports in the second antenna port group are not used to transmit precoded uplink signals.

[0296] Optionally, a precoding matrix (or precoding vector) corresponding to a transport layer corresponds to a first antenna port group and a second antenna port group. For example, the terminal device includes one transport layer b1, eight antenna ports (including antenna ports a1-a8), and two antenna port groups (including antenna port groups {a1-a4} and {a5-a8}, with each antenna port group and its corresponding index shown in Table 3). If the terminal's precoding matrix is ​​[1, 1, 1, 1, 0, 0, 0, 0], T The precoding matrix is ​​structured such that each row corresponds to an antenna port (the first row corresponds to antenna port a1, the second row to antenna port a2, the third row to antenna port a3, the fourth row to antenna port a4, the fifth row to antenna port a5, the sixth row to antenna port a6, the seventh row to antenna port a7, and the eighth row to antenna port a8), and each column corresponds to a transmission layer (the first column corresponds to transmission layer b1, and the second column corresponds to transmission layer b2). In this case, the first antenna port group of transmission layer b1 is {a1-a4}, and the second antenna port group is {a5-a8}; the first antenna port group of transmission layer b2 is {a5-a8}. The precoding matrix output by the first model is "[1, 1, 1, 1, 0, 0, 0, 0]". -1 ".

[0297] In some embodiments, when the first model output is a precoding matrix corresponding to Form 1, the number of rows in the precoding matrix is ​​determined by the "total number of antenna ports" in the antenna port group configuration.

[0298] Format 2: Output only the weights of each antenna port in the first antenna port group and the information of the first antenna port group.

[0299] In some embodiments, the precoding matrix includes only the weights of each antenna port in the first antenna port group. In this case, the first model is also used to output the first antenna port group information, which is used to indicate the first antenna port group in at least two antenna port groups. That is, the first model outputs the precoding matrix and the corresponding first antenna port group information.

[0300] In some embodiments, the first antenna port group information is an index value; or, the first antenna port group information is a bitmap.

[0301] For example, the terminal device or the first model numbers at least two antenna ports to obtain an index for each antenna port group. The first model outputs the index of the first antenna port group to indicate the first antenna port group from the at least two antenna port groups. If the information of the first antenna port group is log2N bits, where N is the number of antenna port groups (i.e., at least two antenna port groups) included in the terminal device, Table 1 above is designed according to this method.

[0302] For example, the first antenna port group information is a bitmap. The number of bits in this bitmap is determined based on the number of antenna ports involved in at least two antenna port groups. For example, if at least two antenna port groups include four antenna ports (including antenna port a1, antenna port a2, antenna port a3, and antenna port a4), then the first antenna port group information is a four-bit bitmap; the first bit of the bitmap corresponds to antenna port a1, the second bit corresponds to antenna port a2, the third bit corresponds to antenna port a3, and the fourth bit corresponds to antenna port a4. A bit of 0 in the bitmap indicates that the antenna port group does not include the antenna port corresponding to that bit, and a bit of 1 in the bitmap indicates that the antenna port group includes the antenna port corresponding to that bit. For example, if the first antenna port group information is "0011", it means that the first antenna port group consists of antenna port a3 and antenna port a4. It should be noted that in the embodiments of this application, the nth bit refers to the nth number counted from the left to the right of the bitmap; however, it can also be the nth number counted from the right to the left of the bitmap. The embodiments of this application do not limit this. Furthermore, in the example given, this application embodiment illustrates that a bit of 0 in the bitmap indicates that the antenna port group does not include the antenna port corresponding to that bit, and a bit of 1 in the bitmap indicates that the antenna port group includes the antenna port corresponding to that bit. However, it can also be implemented such that a bit of 1 in the bitmap indicates that the antenna port group does not include the antenna port corresponding to that bit, and a bit of 0 in the bitmap indicates that the antenna port group includes the antenna port corresponding to that bit. This application embodiment does not limit this aspect.

[0303] In some embodiments, where the precoding matrix output by the first model may only contain the weights of each antenna port in the first antenna port group, the output of the first model also includes information about the first antenna port group. The terminal device includes two transport layers (including transport layer b1 and transport layer b2), eight antenna ports (including antenna ports a1-a8), and four antenna port groups (including antenna port groups {a1, a2}, {a3, a4}, {a5, a6}, and {a7, a8}, with each antenna port group and its corresponding index shown in Table 1). If the precoding matrix of the terminal is [[1, 1, 0, 0, 0, 0, 0, 0]] T ,[0,0,0,0,1,-j,0,0] T The precoding matrix of the terminal corresponds to an antenna port in each row (the first row corresponds to antenna port a1, the second row to antenna port a2, the third row to antenna port a3, the fourth row to antenna port a4, the fifth row to antenna port a5, the sixth row to antenna port a6, the seventh row to antenna port a7, and the eighth row to antenna port a8), and each column corresponds to a transmission layer (the first column corresponds to transmission layer b1, the second column corresponds to transmission layer b2). At this time, the output of the first model can be "00, [1, 1]". T ,10,[1,-j] T "(The first output represents the first antenna port group information of the first transport layer, the second output represents the precoding matrix of the first transport layer, the third output represents the first antenna port group information of the second transport layer, and the fourth output represents the precoding matrix of the second transport layer.)"

[0304] In some embodiments, when the first model output is a precoding matrix corresponding to form two, the number of rows in the precoding matrix is ​​determined by the "number of antenna ports in each antenna port group" in the antenna port group configuration.

[0305] In summary, the apparatus provided in this application illustrates two indication formats for the output of the first model. Format one, where the precoding matrix includes the weights of antenna ports involved in at least two antenna port groups, shows a larger precoding matrix from the perspective of the precoding matrix itself. However, its format is consistent with that of precoding matrices in related technologies, requiring minimal protocol modification and being more easily used by current terminals. Format two, on the other hand, separately shows the first antenna port group information corresponding to the precoding matrix. When this precoding matrix is ​​a precoding matrix corresponding to a transport layer, each transport layer's precoding matrix can include only the weights of the antenna ports in the first antenna port group. For sparse precoding matrices, this makes the output of the first model more concise.

[0306] In addition, a second model corresponding to the first model also exists in this communication system. The second model is configured in the network device. The second model is used by the network device to generate uplink CSI indication information.

[0307] In an alternative embodiment based on FIG13, the device 1200 further includes a second receiving module.

[0308] The second receiving module is used to receive uplink CSI indication information sent by the network device. The uplink CSI indication information is generated by the network device based on the uplink channel information and the second model corresponding to the first model.

[0309] In some embodiments, the correspondence between the first model and the second model means that the first model and the second model are trained together. The first model is used to obtain the precoding matrix based on the uplink CSI indication information; the second model is used to generate uplink CSI indication information based on the uplink channel information.

[0310] The uplink channel information is obtained by the network device based on the reference signal sent by the terminal device. That is, before step 410, the terminal device first sends a reference signal to the network device, which is used to measure the state of the uplink channel.

[0311] Optionally, the uplink channel information includes at least one of the following: feature vector; beam information; and delay information. The feature vector is the feature vector of the uplink channel, which is extracted from the received reference signal by the feature extraction module or the second model in the network device; the beam information refers to the information of the beam used when the reference signal is transmitted; and the delay information is used to indicate the delay of the reference signal during transmission.

[0312] In some embodiments, the uplink CSI indication information is compressed uplink channel information. That is, the second model is used to compress the uplink channel information to obtain the uplink CSI indication information; the first model is used to obtain the precoding matrix through the uplink CSI indication information.

[0313] In summary, the apparatus provided in this application allows a terminal device to receive uplink CSI indication information generated by a network device based on uplink channel information and a second model. This second model corresponds one-to-one with the first model. The uplink CSI indication information obtained by compressing and reconstructing the uplink channel information based on the second model is more efficient in terms of signaling overhead compared to the network device directly indicating the channel state based on the uplink channel information.

[0314] In an alternative embodiment based on FIG14, the device 1300 further includes a second generation module and a second transmission module.

[0315] The second generation module is used to obtain uplink CSI indication information based on uplink channel information and a second model corresponding to the first model.

[0316] In some embodiments, the correspondence between the first model and the second model means that the first model and the second model are trained together. The first model is used to obtain the precoding matrix based on the uplink CSI indication information; the second model is used to generate uplink CSI indication information based on the uplink channel information.

[0317] The uplink channel information is obtained by the network device based on the reference signal sent by the terminal device. That is, before step 410, the terminal device first sends a reference signal to the network device, which is used to measure the state of the uplink channel.

[0318] Optionally, the uplink channel information includes at least one of the following: feature vector; beam information; and delay information. The feature vector is the feature vector of the uplink channel, which is extracted from the received reference signal by the feature extraction module or the second model in the network device; the beam information refers to the information of the beam used when the reference signal is transmitted; and the delay information is used to indicate the delay of the reference signal during transmission.

[0319] In some embodiments, the uplink CSI indication information is compressed uplink channel information. That is, the second model is used to compress the uplink channel information to obtain the uplink CSI indication information; the first model is used to obtain the precoding matrix through the uplink CSI indication information.

[0320] The second sending module is used to send uplink CSI indication information to the terminal device.

[0321] In some embodiments, the network device transmits the uplink CSI indication information to the terminal device via DCI.

[0322] For example, the network device indicates uplink CSI information to the terminal device via the PMI indication field in the DCI. In one implementation, the PMI is indicated together with transport layer number indication information.

[0323] In summary, the apparatus provided in this application obtains uplink CSI indication information based on a second model and uplink channel information, where the second model is the AI ​​model corresponding to the first model. The uplink CSI indication information obtained by compressing and reconstructing the uplink channel information based on the second model saves signaling overhead compared to network devices directly indicating channel status based on uplink channel information.

[0324] Next, we will introduce the working process of the second model.

[0325] 1. Input to the second model.

[0326] In some embodiments, the input to the second model includes uplink channel information. That is, the uplink CSI indication information is obtained by inputting the uplink channel information into the second model. Alternatively, the network device obtains the uplink CSI indication information based on the uplink channel information and the second model corresponding to the first model, including: inputting the uplink channel information into the second model corresponding to the second model in the domain to obtain the uplink CSI indication information.

[0327] In some embodiments, the input to the second model further includes an antenna port group configuration of the terminal device. The antenna port group configuration indicates the result of the terminal device dividing the antenna port groups; or, the antenna port group configuration indicates information about the antenna port groups configured by the terminal device, wherein the individual antenna ports in each antenna port group are coherent. Exemplarily, the antenna port group configuration is at least one of the following: the number of antenna port groups; the number of antenna port groups and the total number of antenna ports; the number of antenna port groups and the number of antenna ports in each antenna port group; the total number of antenna ports and the number of antenna ports in each antenna port group.

[0328] In some embodiments, given any two of the three parameters—the number of antenna port groups, the total number of antenna ports, and the number of antenna ports in each antenna port group—the second model can deduce the remaining parameter. For example, the total number of antenna ports is equal to the number of antenna port groups multiplied by the number of antenna ports in each antenna port group.

[0329] In some embodiments, the second model obtains uplink CSI indication information based on at least one of the antenna port group configuration and the number of transport layers, as well as uplink channel information. The uplink CSI indication information, in addition to indicating the state of the uplink channel to the terminal device, also indicates at least one of the antenna port group configuration and the number of transport layers, or, further indicates the size of the uplink precoding matrix determined based on the antenna port group configuration and / or the number of transport layers.

[0330] In some embodiments, the network device stores at least two AI models. The network device determines a second model from the at least two AI models based on an antenna port group configuration. The second model is the AI ​​model corresponding to the antenna port group configuration. Alternatively, the network device includes only the second model, which determines the size of the uplink precoding matrix based on the antenna port group configuration, and this size will be indicated to the terminal device. Alternatively, the second model includes at least one processing module, each processing module corresponding to one antenna port group configuration; the second model determines a second processing module based on the antenna port group configuration, and inputs uplink channel information into the second processing module to obtain uplink CSI indication information.

[0331] 2. Output of the second model.

[0332] In some embodiments, the second model is used to output uplink CSI indication information. The uplink CSI indication information is used to indicate uplink channel information. The uplink CSI indication information is used by the terminal device to input the precoding matrix obtained from the first model.

[0333] In some embodiments, the second model is further used to output the number of transport layers; or, the second model is further used to output the transport layer configuration. Optionally, the second model outputs the number of transport layers based on uplink channel information.

[0334] In another optional embodiment, the number of transport layers (or transport layer configuration) is determined by the network device based on uplink channel information. In addition to sending uplink CSI indication information to the terminal device, the network device also sends transport layer indication information (or transport layer number configuration) to the terminal device. This transport layer indication information (or transport layer number configuration) is used to indicate to the terminal device the number of transport layers (or transport layer configuration) determined by the network device. The number of transport layers is used by the terminal device (or the first model) to determine the number of columns in the precoding matrix.

[0335] In some embodiments, the first model and the second model are in one-to-one correspondence, and the first model and the second model are trained together by the terminal device and / or the network device. That is, the first model of the terminal device is trained by the terminal device, or the first model of the terminal device is configured by the network device. And / or, the second model of the network device is reported by the terminal device, or the second model of the network device is trained by the network device.

[0336] That is, both the first model and the second model are trained by the terminal device, and after training, the terminal device reports the second model to the network device; or, both the first model and the second model are trained by the network device, and after training, the network device configures the first model to the terminal device; or, the first model is trained by the terminal device, and the second model is trained by the network device.

[0337] During the configuration of the first model or the reporting of the second model, instructions are required using the configuration of the first model or the second model.

[0338] In some embodiments, the first model configuration and / or the second model configuration includes at least one of the following: model parameters; model structure; model function; model identifier; input information; output information; dataset identifier.

[0339] The model parameters indicate the parameter values ​​in the obtained model. The model structure indicates the network structure of the model, such as the number of layers in the neural network, the number of neurons in each layer, and the connection methods of each neuron. The model function indicates that the model is used for uplink CSI. The model identifier indicates the model itself. Input information indicates the model's input, such as the number of bits, the type of input data (e.g., uplink CSI indication information), the size of the input data, and the data type of the input data (e.g., integer, string, etc.). Output information indicates the model's output, such as the number of bits in the output, the type of input data (e.g., uplink CSI indication information), the size of the input data, and the data type of the input data (e.g., integer, string, etc.). The dataset identifier indicates the dataset used by the model during training.

[0340] The training process for the first and second models is shown below.

[0341] 1. The first and second models are trained by the terminal device.

[0342] In an alternative embodiment based on FIG13, device 1200 further includes a first training module.

[0343] The first training module is used to train a first model and a second model corresponding to the first model based on the antenna port group configuration of the terminal device, or based on the antenna port group configuration and at least one transmission layer.

[0344] In some embodiments, the terminal device trains a first model and a corresponding second model based on the antenna port group configuration. Alternatively, the terminal device determines a training dataset based on the antenna port group configuration and trains the first model and the corresponding second model based on that dataset; in this case, the terminal device stores at least one dataset, each dataset corresponding to one antenna port group configuration. Alternatively, the terminal device trains the first model and the corresponding second model based on the antenna port group configuration and the dataset; in this case, the terminal device stores only one dataset, and the precoding matrices corresponding to the training samples in the dataset are of the same or different sizes; during training, the terminal device selects training samples from the dataset corresponding to the antenna port group configuration based on the antenna port group configuration for training the first and second models. Alternatively, the terminal device trains the first model and the corresponding second model based on the antenna port group configuration and the dataset; that is, the terminal device uses the antenna port group configuration and the training samples in the dataset as input to the first and second models to train them. Alternatively, the terminal device trains a first model and a corresponding second model based on the antenna port group configuration and at least one transmission layer number. The at least one transmission layer number includes the number of transmission layers configured for the terminal device, i.e., the at least one transmission layer number is the number of transmission layers the terminal device may be configured with during transmission. The at least one transmission layer number is determined based on the maximum number of transmission layers, or it is indicated by the network device. The terminal device uses the antenna port group configuration and at least one transmission layer number as input to the first and second models to train the first and second models, or the terminal device uses at least one transmission layer number as input to the first and second models to train the first and second models. In this case, the antenna port group configuration is used to filter training samples.

[0345] As an example and not a limitation, the training dataset includes uplink channel information and the target precoding matrix. The uplink channel information is input into the second model to obtain uplink CSI indication information; then, the uplink CSI indication information is input into the first model to obtain the prediction precoding matrix. Based on the similarity between the prediction precoding matrix and the target precoding matrix, the model parameters of the first and second models are updated using the backpropagation algorithm.

[0346] The first sending module 1220 is also used to report the second model configuration corresponding to the second model to the network device. The second model is used by the network device to generate uplink CSI indication information.

[0347] In some embodiments, after training is completed, the terminal device obtains the second model configuration based on the second model and reports the second model configuration to the network device.

[0348] In summary, the apparatus provided in this application, by using an AI model trained by the terminal device itself, can better match the antenna configuration of the terminal device, thereby improving the reliability of CSI compression. Furthermore, by training an AI model to perform precoding of uplink signals on a coherent portion of the antenna port group, compared to traditional precoding methods (especially precoding based on a partially coherent codebook), transmission based on precoding can be achieved simply by the network device instructing the compressed uplink CSI indication information. This saves signaling overhead such as the indication codebook and currently used antenna port information, thus conserving air interface resources.

[0349] In an optional embodiment based on FIG14, the first receiving module 1310 is further configured to receive the second model configuration sent by the terminal device and obtain the second model based on the second model configuration.

[0350] In some embodiments, the network device receives a second model configuration sent by the terminal device, and obtains a second model corresponding to the first model of the terminal device based on the second model configuration.

[0351] In some embodiments, the second model is associated with the antenna port group configuration, or with the antenna port group configuration and the number of at least one transmission layer. Alternatively, the second model can be said to be trained based on the antenna port group configuration, or based on the antenna port group configuration and the number of at least one transmission layer.

[0352] In summary, the apparatus provided in this application, by using an AI model trained by the terminal device itself, can better match the antenna configuration of the terminal device, thereby improving the reliability of CSI compression. Furthermore, by training an AI model to perform precoding of uplink signals on a coherent portion of the antenna port group, compared to traditional precoding methods (especially precoding based on a partially coherent codebook), transmission based on precoding can be achieved simply by the network device instructing the compressed uplink CSI indication information. This saves signaling overhead such as the indication codebook and currently used antenna port information, thus conserving air interface resources.

[0353] 2. The first and second models are trained by network devices.

[0354] In an alternative embodiment based on FIG13, the first transmitting module 1220 is further configured to report the antenna port group configuration of the terminal device to the network device.

[0355] In some embodiments, the terminal device periodically reports the antenna port group configuration of the terminal device to the network device; or, after accessing the network device, the terminal device reports the antenna port group configuration of the terminal device to the network device through UE capability; or, after receiving an indication message from the network device requesting the reporting, the terminal device reports the antenna port group configuration of the terminal device to the network device.

[0356] In an optional embodiment based on Figure 13, the second receiving module is further configured to receive the first model configuration and obtain a first model based on the first model configuration. The first model is associated with the antenna port group configuration, or with the antenna port group configuration and at least one transmission layer number.

[0357] In some embodiments, the first model is based on the antenna port group configuration of the terminal device, or is trained based on the antenna port group configuration of the terminal device and at least one transmission layer number.

[0358] In summary, the apparatus provided in this application, through the AI ​​model configured on the network device, enables better matching between the models on the terminal device and the network device side, thereby improving the reliability of CSI compression. Furthermore, by training an AI model to achieve precoding of uplink signals on a coherent portion of the antenna port group, compared to traditional precoding methods (especially precoding based on a partially coherent codebook), transmission based on precoding can be achieved simply by the network device instructing the compressed uplink CSI indication information. This saves signaling overhead such as the indication codebook and currently used antenna port information, thus conserving air interface resources.

[0359] In an alternative embodiment based on FIG14, the first receiving module 1310 is further configured to receive the antenna port group configuration of the terminal device.

[0360] In some embodiments, the network device starts training the first model and the second model after receiving the antenna port group configuration from the terminal device; or, before starting training the first model and the second model, the network device sends an instruction message to the terminal device requesting reporting, the instruction message being used to instruct the terminal device to report the antenna port group configuration.

[0361] In an optional embodiment based on FIG14, the device 1300 further includes a second training module, which is used to train a second model and a first model corresponding to the second model based on the antenna port group configuration of the terminal device, or based on the antenna port group configuration and at least one transmission layer number.

[0362] In some embodiments, the network device trains a first model and a corresponding second model based on the antenna port group configuration. Alternatively, the network device determines a training dataset based on the antenna port group configuration and trains the first model and the corresponding second model based on that dataset; in this case, the network device stores at least one dataset, each dataset corresponding to one antenna port group configuration. Alternatively, the network device trains the first model and the corresponding second model based on the antenna port group configuration and the dataset; in this case, the network device stores only one dataset, and the precoding matrices corresponding to the training samples in the dataset have the same or different sizes; during training, the network device selects training samples from the dataset corresponding to the antenna port group configuration based on the antenna port group configuration for training the first and second models. Alternatively, the network device trains the first model and the corresponding second model based on the antenna port group configuration and the dataset; that is, the network device uses the antenna port group configuration and the training samples in the dataset as input to the first and second models to train them. Alternatively, the network device trains a first model and a corresponding second model based on the antenna port group configuration and at least one transmission layer number. The at least one transmission layer number includes the number of transmission layers configured for the terminal device, i.e., the at least one transmission layer number is the number of transmission layers the terminal device may be configured with during transmission. The at least one transmission layer number is determined based on the maximum number of transmission layers, or the at least one transmission layer number is indicated by the network device. The network device uses the antenna port group configuration and at least one transmission layer number as input to the first and second models to train the first and second models, or the network device uses at least one transmission layer number as input to the first and second models to train the first and second models. In this case, the antenna port group configuration is used to filter training samples.

[0363] As an example and not a limitation, the training dataset includes uplink channel information and the target precoding matrix. The uplink channel information is input into the second model to obtain uplink CSI indication information; then, the uplink CSI indication information is input into the first model to obtain the prediction precoding matrix. Based on the similarity between the prediction precoding matrix and the target precoding matrix, the model parameters of the first and second models are updated using the backpropagation algorithm.

[0364] In an optional embodiment based on Figure 14, the second sending module is further configured to send the first model configuration corresponding to the first model to the terminal device.

[0365] In some embodiments, after training is completed, the network device obtains a first model configuration based on the first model and sends the first model configuration to the terminal device.

[0366] In summary, the apparatus provided in this application, through the AI ​​model configured on the network device, enables better matching between the models on the terminal device and the network device side, thereby improving the reliability of CSI compression. Furthermore, by training an AI model to achieve precoding of uplink signals on a coherent portion of the antenna port group, compared to traditional precoding methods (especially precoding based on a partially coherent codebook), transmission based on precoding can be achieved simply by the network device instructing the compressed uplink CSI indication information. This saves signaling overhead such as the indication codebook and currently used antenna port information, thus conserving air interface resources.

[0367] In some embodiments, the terminal device and the network device each store at least two AI models. The terminal device and the network device select a first model and a second model from the at least two AI models based on the antenna port group configuration. Alternatively, the terminal device and the network device select a first model and a second model from the at least two AI models based on the antenna port group configuration and the maximum number of transmission layers. Optionally, the at least two AI models stored in the terminal device and the network device may be the same or different.

[0368] In some embodiments, at least two AI models stored in the terminal device are trained by the terminal device; or, at least two AI models stored in the terminal device are configured by the network device; or, of the at least two AI models stored in the terminal device, some AI models are trained by the terminal device and others are configured by the network device.

[0369] In some embodiments, at least two AI models stored in the network device are trained by the network device; or, at least two AI models stored in the network device are reported by the terminal device; or, of the at least two AI models stored in the network device, some AI models are trained by the network model and others are reported by the terminal device.

[0370] In some embodiments, the terminal device stores at least two AI models, each AI model corresponding to an antenna port group configuration and / or at least one transmission layer.

[0371] In an alternative embodiment based on FIG13, the first transmitting module 1220 is further configured to report the antenna port group configuration of the terminal device to the network device.

[0372] In an alternative embodiment based on FIG13, the device 1200 further includes a first selection module, which is used to select an AI model corresponding to the antenna port group configuration of the terminal device as a first model.

[0373] In another possible embodiment, the terminal device selects an AI model corresponding to its antenna port group configuration as the first model, or selects an AI model corresponding to its antenna port group configuration and a first transmission layer number as the first model. Here, the first transmission layer number is the number of transmission layers supported by the terminal device; or, the first transmission layer number is the number of transmission layers configured for the terminal device. For example, the network device indicates the first transmission layer number to the terminal device based on uplink channel information, and the terminal device selects the first model based on the antenna port group configuration and the first transmission layer number; or, the network device indicates the first transmission layer number to the terminal device based on uplink channel information, and the terminal device uses the first transmission layer number as input to the first model, where the first model is the AI ​​model selected by the terminal device corresponding to its antenna port group configuration.

[0374] In some embodiments, the first transmission layer number is a subset of the transmission layers corresponding to the first model. For example, at least one AI model contains three AI models c1-c3 corresponding to the antenna port group configuration of the terminal device. The transmission layer number corresponding to AI model c1 is 2, the transmission layer number corresponding to AI model c2 is 4, and the transmission layer number corresponding to AI model c3 is {1, 2, 4}. If the first transmission layer number is 2, either AI model c1 or AI model c3 can be selected as the first model; if the first transmission layer number is {2, 4}, only AI model c3 can be selected as the first model.

[0375] In some embodiments, each AI model stored in the terminal device corresponds to an antenna port group configuration, and the AI ​​model corresponding to the antenna port group configuration of the terminal device is selected as the first model; or, each AI model stored in the terminal device corresponds to at least one transmission layer number, and the AI ​​model corresponding to the first transmission layer number is selected as the first model; or, each AI model stored in the terminal device corresponds to an antenna port group configuration and at least one transmission layer number, and the AI ​​model corresponding to the antenna port group configuration and the first transmission layer number of the terminal device is selected as the first model.

[0376] In summary, the apparatus provided in this application uses a pre-training method to train multiple AI models. During use, only the corresponding AI model needs to be selected based on the antenna port group configuration, enabling rapid configuration and application of the AI ​​models. Furthermore, by training AI models to achieve precoding of uplink signals on a portion of the antenna port group, compared to traditional precoding methods (especially precoding based on a partially coherent codebook), transmission based on precoding can be achieved simply by the network device instructing the compressed uplink CSI indication information. This saves signaling overhead such as the indication codebook and currently used antenna port information, thus conserving air interface resources.

[0377] In some embodiments, the network device stores at least two AI models, each corresponding to an antenna port group configuration.

[0378] In an alternative embodiment based on FIG14, the first receiving module 1310 is further configured to receive the antenna port group configuration of the terminal device.

[0379] In an alternative embodiment based on FIG14, the device 1300 further includes a second selection module for selecting an AI model corresponding to the antenna port group configuration of the terminal device as a second model.

[0380] In some embodiments, each AI model stored in the network device corresponds to an antenna port group configuration, and the AI ​​model corresponding to the antenna port group configuration of the terminal device is selected as the second model.

[0381] In summary, the apparatus provided in this application uses a pre-training method to train multiple AI models. During use, only the corresponding AI model needs to be selected based on the antenna port group configuration, enabling rapid configuration and application of the AI ​​models. Furthermore, by training AI models to achieve precoding of uplink signals on a portion of the antenna port group, compared to traditional precoding methods (especially precoding based on a partially coherent codebook), transmission based on precoding can be achieved simply by the network device instructing the compressed uplink CSI indication information. This saves signaling overhead such as the indication codebook and currently used antenna port information, thus conserving air interface resources.

[0382] In an alternative embodiment based on FIG13, the first transmitting module 1220 is further configured to report capability information to the network device, the capability information being used to instruct the terminal device to configure some coherent antenna ports.

[0383] In an alternative embodiment based on FIG14, the first receiving module 1310 is further configured to receive capability information reported by the terminal device, the capability information being used to indicate the configuration of partially coherent antenna ports of the terminal device.

[0384] In some embodiments, the antenna ports within each antenna port group are coherent; the antenna ports between different antenna port groups are incoherent.

[0385] It should be noted that the device provided in the above embodiments is only an example of the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0386] Regarding the apparatus in this embodiment, the specific manner in which each module performs its operations has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0387] Figure 15 shows a schematic diagram of the structure of a terminal device provided in an exemplary embodiment of this application. The terminal device 1400 can be used to execute the method steps performed by the terminal device in the above embodiments. The terminal device 1400 may include: a processor 1401, a transceiver 1402, and a memory 1403. The processor 1401 can be used to control transmission and / or reception. The transceiver 1402 can be used to implement transmission and / or reception functions.

[0388] The processor 1401 includes one or more processing cores, and the processor 1401 executes various functional applications and information processing by running software programs and modules.

[0389] The transceiver 1402 may include a receiver and a transmitter, for example, the receiver and transmitter may be implemented as the same wireless communication component, which may include a wireless communication chip and a radio frequency antenna.

[0390] The memory 1403 can be connected to the processor 1401 and the transceiver 1402.

[0391] The memory 1403 can be used to store a computer program executed by the processor, and the processor 1401 is used to execute the computer program to implement the various steps in the above method embodiments.

[0392] Furthermore, the memory 1403 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, including but not limited to: magnetic disks or optical disks, electrically erasable programmable read-only memory, erasable programmable read-only memory, static on-demand memory, read-only memory, magnetic memory, flash memory, and programmable read-only memory.

[0393] For details not described in this embodiment, please refer to the embodiments above, which will not be repeated here.

[0394] Figure 16 shows a schematic diagram of a network device provided in an exemplary embodiment of this application. The network device 1500 can be used to execute the method steps performed by the network device in the above embodiments. The network device 1500 may include a processor 1501, a transceiver 1502, and a memory 1503. The processor 1501 can be used to control transmission and / or reception. The transceiver 1502 can be used to implement transmission and / or reception functions.

[0395] The processor 1501 includes one or more processing cores, and the processor 1501 executes various functional applications and information processing by running software programs and modules.

[0396] Transceiver 1502 may include a receiver and a transmitter. For example, transceiver 1502 may include a wired communication component, which may include a wired communication chip and a wired interface (such as a fiber optic interface). Optionally, transceiver 1502 may also include a wireless communication component, which may include a wireless communication chip and a radio frequency antenna.

[0397] The memory 1503 can be connected to the processor 1501 and the transceiver 1502.

[0398] The memory 1503 can be used to store a computer program executed by the processor, and the processor 1501 is used to execute the computer program to implement the various steps performed by the network device in the above method embodiment.

[0399] Furthermore, the memory 1503 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, including but not limited to: magnetic disks or optical disks, electrically erasable programmable read-only memory, erasable programmable read-only memory, static on-demand memory, read-only memory, magnetic memory, flash memory, and programmable read-only memory.

[0400] For details not described in this embodiment, please refer to the embodiments above, which will not be repeated here.

[0401] This application also provides a computer-readable storage medium storing a computer program for execution by a processor to implement the aforementioned random access channel transmission method. In some embodiments, the computer-readable storage medium may include ROM (Read-Only Memory), RAM (Random-Access Memory), SSD (Solid State Drives), or optical disc, etc. The random access memory may include ReRAM (Resistance Random Access Memory) and DRAM (Dynamic Random Access Memory).

[0402] This application also provides a chip, which includes programmable logic circuits and / or program instructions. When the chip is running, it is used to implement the above-described random access channel transmission method. This application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. A processor reads and executes the computer program from the computer-readable storage medium to implement the above-described random access channel transmission method. This application also provides a computer program stored in a computer-readable storage medium. A processor obtains the computer program from the computer-readable storage medium and executes the computer program to implement the above-described uplink transmission method.

[0403] It should be understood that the term "instruction" mentioned in the embodiments of this application can be a direct instruction, an indirect instruction, or an indication of an association. For example, A instructing B can mean that A directly instructs B, for example, B can be obtained through A; it can also mean that A indirectly instructs B, for example, A instructs C, B can be obtained through C; or it can mean that there is an association between A and B. In the description of the embodiments of this application, the term "corresponding" can mean that there is a direct or indirect correspondence between two things, or that there is an association between two things, or it can mean an instruction and being instructed, a configuration and being configured, etc. The term "multiple" mentioned herein refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following associated objects are in an "or" relationship. The term "greater than or equal to" mentioned herein can mean greater than or equal to, or greater than; and "less than or equal to" can mean less than or equal to, or less than. Furthermore, the step numbers described herein are merely illustrative of one possible execution order among the steps. In some other embodiments, the steps may not be executed in the numerical order, such as two steps with different numbers being executed simultaneously, or two steps with different numbers being executed in the reverse order of the illustration. This application does not limit this. Those skilled in the art should recognize that the functions described in the above one or more examples can be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any medium that facilitates the transmission of computer programs from one place to another. Storage media can be any available medium accessible to general-purpose or special-purpose computers.

[0404] The above are merely exemplary embodiments of this application and are not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application shall be included within the protection scope of this application.

Claims

1. An uplink transmission method, characterized in that, The method is executed by a terminal device, and the method includes: Based on the uplink CSI indication information and the first model, an uplink precoding matrix is ​​obtained. The precoding matrix corresponds to the first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is a partial antenna port group in the at least two antenna port groups. The uplink signal, precoded by the precoding matrix, is transmitted through the first antenna port group.

2. The method according to claim 1, characterized in that, The process of obtaining the uplink precoding matrix based on the uplink CSI indication information and the first model includes: The uplink CSI indication information is input into the first model to obtain the precoding matrix.

3. The method according to claim 1 or 2, characterized in that, The precoding matrix includes the weights of each antenna port in at least two antenna port groups of the terminal device; wherein, the weights of each antenna port in the first antenna port group are non-zero, the weights of each antenna port in the second antenna port group are zero, and the second antenna port group is the antenna port group other than the first antenna port group in the at least two antenna port groups.

4. The method according to any one of claims 1 to 3, characterized in that, The first model is also used to output first antenna port group information, which is used to indicate the first antenna port group in the at least two antenna port groups.

5. The method according to claim 4, characterized in that, The terminal device includes at least one transmission layer; The first model outputs the first antenna port group information and the precoding matrix of each of the transmission layers respectively; or, the first model outputs the first antenna port group information of each of the transmission layers respectively.

6. The method according to any one of claims 1 to 5, characterized in that, The input to the first model also includes the antenna port group configuration of the terminal device.

7. The method according to any one of claims 1 to 6, characterized in that, The input to the first model also includes the number of transmission layers of the terminal device.

8. The method according to any one of claims 1 to 6, characterized in that, The first model is trained by the terminal device; or, the first model is configured by the network device.

9. The method according to claim 8, characterized in that, The first model is trained by the terminal device, and the method further includes: Based on the antenna port group configuration of the terminal device, or based on the antenna port group configuration and at least one transmission layer number, the first model and the second model corresponding to the first model are trained. The configuration of the second model corresponding to the second model is reported to the network device; The second model is used by the network device to generate the uplink CSI indication information.

10. The method according to claim 8, characterized in that, The first model is configured by a network device, and the method further includes: Report the antenna port group configuration of the terminal device to the network device; Receive a first model configuration and obtain a first model based on the first model configuration. The first model is associated with the antenna port group configuration, or with the antenna port group configuration and at least one transmission layer number.

11. The method according to claim 9 or 10, characterized in that, The first model configuration and / or the second model configuration includes at least one of the following: model parameters; model structure; model function; model identifier; input information; output information; dataset identifier.

12. The method according to any one of claims 1 to 6, characterized in that, The terminal device stores at least two AI models, and each AI model corresponds to an antenna port group configuration. The method further includes: Report the antenna port group configuration of the terminal device to the network device; Select the AI ​​model corresponding to the antenna port group configuration of the terminal device as the first model.

13. The method according to any one of claims 7 or 9 to 12, characterized in that, The antenna port group is configured as one of the following: the number of antenna port groups; the number of antenna port groups and the total number of antenna ports; the number of antenna port groups and the number of antenna ports in each antenna port group; the total number of antenna ports and the number of antenna ports in each antenna port group.

14. The method according to any one of claims 1 to 13, characterized in that, The antenna ports within each antenna port group are coherent; the antenna ports in different antenna port groups are incoherent.

15. The method according to any one of claims 1 to 14, characterized in that, The method further includes: The terminal device reports capability information to the network device, the capability information being used to instruct the terminal device to configure some coherent antenna ports.

16. The method according to any one of claims 1 to 15, characterized in that, The method further includes: The uplink CSI indication information is received by the network device based on the uplink channel information. Information and the second model corresponding to the first model are generated.

17. An uplink transmission method, characterized in that, The method is performed by a network device, and the method includes: The receiving terminal device transmits the precoded uplink signal through the first antenna port group; The precoded uplink signal is obtained by precoding based on a precoding matrix. The precoding matrix is ​​obtained by the terminal device based on uplink CSI indication information and a first model. The precoding matrix corresponds to the first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is a partial antenna port group in the at least two antenna port groups.

18. The method according to claim 17, characterized in that, The precoding matrix is ​​obtained by inputting the uplink CSI indication information into the first model.

19. The method according to claim 17 or 18, characterized in that, The precoding matrix is ​​used to indicate the weight of each antenna port in the at least two antenna port groups; Wherein, the weight of each antenna port in the first antenna port group is non-zero, the weight of each antenna port in the second antenna port group is zero, and the second antenna port group is the antenna port group other than the first antenna port group among the at least two antenna port groups.

20. The method according to any one of claims 17 to 19, characterized in that, The first model is also used to output first antenna port group information, which is used to indicate the first antenna port group in the at least two antenna port groups.

21. The method according to claim 20, characterized in that, The terminal device includes at least one transmission layer; The first model outputs the first antenna port group information and the precoding matrix for each of the transmission layers; Alternatively, the first model may output the first antenna port group information for each of the transmission layers.

22. The method according to any one of claims 17 to 21, characterized in that, The input to the first model also includes the antenna port group configuration of the terminal device.

23. The method according to any one of claims 17 to 22, characterized in that, The method further includes: Based on the uplink channel information and the second model corresponding to the first model, the uplink CSI indication information is obtained; The uplink CSI indication information is sent to the terminal device.

24. The method according to claim 23, characterized in that, The uplink CSI indication information is obtained based on uplink channel information and a second model corresponding to the first model, including: The uplink channel information is input into the second model corresponding to the first model to obtain the uplink CSI indication information.

25. The method according to claim 23 or 24, characterized in that, The input to the second model also includes the antenna port group configuration of the terminal device.

26. The method according to any one of claims 23 to 25, characterized in that, The output of the second model also includes the number of transmission layers of the terminal device.

27. The method according to any one of claims 23 to 26, characterized in that, The second model is reported by the terminal device; or, the second model is trained by the network device.

28. The method according to claim 27, characterized in that, The second model is reported by the terminal device, and the method further includes: The system receives the second model configuration sent by the terminal device and obtains the second model based on the second model configuration.

29. The method according to claim 27, characterized in that, The method further includes: Receive the antenna port group configuration of the terminal device; Based on the antenna port group configuration of the terminal device, or based on the antenna port group configuration of the terminal device and at least one transmission layer number, the second model and the first model corresponding to the second model are trained. Send the first model configuration corresponding to the first model to the terminal device.

30. The method according to claim 28 or 29, characterized in that, The first model configuration and / or the second model configuration includes at least one of the following: model parameters; model structure; model function; model identifier; input information; output information; dataset identifier.

31. The method according to any one of claims 23 to 26, characterized in that, The network device stores at least two AI models, and each AI model corresponds to an antenna port group configuration. The method further includes: Receive the antenna port group configuration of the terminal device; Select the AI ​​model corresponding to the antenna port group configuration of the terminal device as the second model.

32. The method according to any one of claims 17 to 31, characterized in that, The antenna port group is configured as one of the following: the number of antenna port groups; the number of antenna port groups and the total number of antenna ports; the number of antenna port groups and the number of antenna ports in each antenna port group; the total number of antenna ports and the number of antenna ports in each antenna port group.

33. The method according to any one of claims 17 to 32, characterized in that, The antenna ports within each antenna port group are coherent; the antenna ports in different antenna port groups are incoherent.

34. The method according to any one of claims 17 to 33, characterized in that, The method further includes: The terminal device receives capability information reported by the terminal device, which is used to instruct the terminal device to configure some coherent antenna ports.

35. An uplink transmission device, characterized in that, The device includes: The first generation module is used to obtain an uplink precoding matrix based on uplink CSI indication information and a first model. The precoding matrix corresponds to the first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is a partial antenna port group in the at least two antenna port groups. The first transmitting module is used to transmit the uplink signal precoded by the precoding matrix through the first antenna port group.

36. An uplink transmission device, characterized in that, The device includes: The first receiving module is used to receive the precoded uplink signal sent by the terminal device through the first antenna port group; The precoded uplink signal is obtained by precoding based on a precoding matrix. The precoding matrix is ​​obtained by the terminal device based on uplink CSI indication information and a first model. The precoding matrix corresponds to the first antenna port group in at least two antenna port groups of the terminal device. The first antenna port group is a partial antenna port group in the at least two antenna port groups.

37. A terminal device, characterized in that, The terminal device includes: A processor; a transceiver connected to the processor; a memory for storing executable instructions of the processor; wherein the processor is configured to load and execute the executable instructions to implement the uplink transmission method as described in any one of claims 1 to 16.

38. A network device, characterized in that, The network device includes: A processor; a transceiver connected to the processor; a memory for storing executable instructions of the processor; wherein the processor is configured to load and execute the executable instructions to implement the uplink transmission method as described in any one of claims 17 to 34.

39. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one program, which is loaded and executed by a processor to implement the uplink transmission method as described in any one of claims 1 to 34.

40. A chip, characterized in that, The chip includes programmable logic circuits and / or program instructions, which, when running on the first node, are used to implement the uplink transmission method as described in any one of claims 1 to 34.

41. A computer program product, characterized in that, The computer program product includes computer instructions stored in a computer-readable storage medium, a processor retrieving the computer instructions from the computer-readable storage medium, and the processor executing the computer instructions to implement the uplink transmission method as described in any one of claims 1 to 34.

42. A computer program, characterized in that, The computer program is stored in a computer-readable storage medium, the processor retrieves the computer program from the computer-readable storage medium, and the processor executes the computer program to implement the uplink transmission method as described in any one of claims 1 to 34.

Images