CSI feedback method, terminal device, and network device

Abstract

The present application relates to a CSI feedback method, a terminal device, a network device, a chip, a computer readable storage medium, a computer program product, a computer program, and a communication system. The method comprises: a terminal device sends first reporting information to a network device, wherein the first reporting information comprises CSI; and the terminal device receives first instruction information from the network device and sends second reporting information to the network device, wherein the first instruction information is used for instructing the terminal device to retransmit the CSI, and the second reporting information comprises the CSI. Embodiments of the present application enable a network device to acquire CSI of a downlink channel more efficiently and accurately, thereby improving downlink precoding performance and downlink transmission performance.

Description

CSI feedback methods, terminal equipment, and network equipment Technical Field

[0001] This application relates to the field of communications, and more specifically, to a CSI (Channel State Information) feedback method, terminal equipment, network equipment, chip, computer-readable storage medium, computer program product, computer program, and communication system. Background Technology

[0002] CSI feedback is crucial in wireless communication systems. In NR (New Radio) systems, CSI feedback schemes typically employ codebook-based feature vector feedback to enable network devices to acquire downlink CSI. Accurate CSI acquisition by network devices is fundamental to improving downlink precoding and transmission performance; therefore, enhancing the accuracy of CSI feedback has always been a hot topic in the communications field.

[0003] Summary of the Invention

[0004] This application provides a CSI feedback method, a terminal device, a network device, a chip, a computer-readable storage medium, a computer program product, a computer program, and a communication system.

[0005] This application provides a Channel State Information (CSI) feedback method, including:

[0006] The terminal device sends a first reporting information to the network device; wherein, the first reporting information includes CSI;

[0007] The terminal device receives a first indication information from the network device and sends a second reporting information to the network device; wherein, the first indication information is used to instruct the terminal device to retransmit the CSI; the second reporting information includes the CSI.

[0008] This application provides a CSI feedback method, including:

[0009] The network device receives first reported information from the terminal device; wherein, the first reported information includes CSI;

[0010] The network device sends a first indication message to the terminal device; wherein the first indication message is used to instruct the terminal device to send a second reporting message to retransmit the CSI.

[0011] This application provides a terminal device, including:

[0012] The first communication unit is configured to send a first reporting information to the network device, receive a first instruction information from the network device, and send a second reporting information to the network device; wherein the first reporting information includes CSI; the first instruction information is used to instruct the terminal device to retransmit CSI; and the second reporting information includes CSI.

[0013] This application provides a network device, including:

[0014] The second communication unit is configured to receive first reported information from the terminal device and send first instruction information to the terminal device; wherein the first reported information includes CSI; and the first instruction information is configured to instruct the terminal device to send second reported information to retransmit the CSI.

[0015] This application provides a terminal device, including: a processor, and a memory communicating with the processor. The memory is used to store instructions, which, when executed by the processor, cause the terminal device to perform the following:

[0016] Send the first report information to the network device; wherein the first report information includes CSI;

[0017] The terminal device receives a first instruction message from a network device and sends a second reporting message to the network device; wherein the first instruction message is used to instruct the terminal device to retransmit the CSI; and the second reporting message includes the CSI.

[0018] This application provides a network device, including: a processor, and a memory communicating with the processor. The memory stores instructions, which, when executed by the processor, cause the network device to perform the following:

[0019] Receive first reported information from the terminal device; wherein, the first reported information includes CSI;

[0020] Send a first instruction message to the terminal device; wherein the first instruction message is used to instruct the terminal device to send a second reporting message to retransmit CSI.

[0021] This application provides a chip for implementing the above method.

[0022] Specifically, the chip includes a processor for retrieving and running a computer program from memory, causing a device equipped with the chip to perform the methods described above.

[0023] This application provides a computer-readable storage medium for storing a computer program, which, when run by a device, causes the device to perform the above-described method.

[0024] This application provides a computer program product, including computer program instructions that cause a computer to perform the above-described method.

[0025] This application provides a computer program that, when run on a computer, causes the computer to perform the above-described method.

[0026] This application provides a communication system, including a terminal device and a network device for performing the above-described methods.

[0027] In this embodiment of the application, by instructing the terminal device to retransmit CSI through the network device, the CSI feedback performance can be enhanced, and the feedback delay and performance loss caused by the loss of CSI information under non-ideal uplink feedback conditions can be reduced. This enables the network device to obtain downlink channel CSI information more effectively and accurately, thereby improving downlink precoding and downlink transmission performance. Attached Figure Description

[0028] Figure 1 is a schematic diagram of a communication system according to an embodiment of this application.

[0029] Figure 2 is a schematic diagram of an AI-based CSI autoencoder.

[0030] Figure 3 is a schematic diagram of the CSI feedback method based on source-channel joint coding.

[0031] Figure 4 is a schematic flowchart of a CSI feedback method executed by a terminal device according to an embodiment of this application.

[0032] Figure 5 is a schematic flowchart of a CSI feedback method performed by a network device according to an embodiment of this application.

[0033] Figure 6 is a schematic diagram of the first configuration method of CSI feedback in the embodiments of this application.

[0034] Figure 7 is a schematic diagram of the second configuration method of CSI feedback in an embodiment of this application.

[0035] Figure 8 is a schematic diagram of the CSI retransmission process in an application example 1 of this application.

[0036] Figure 9 is a schematic diagram of the CSI retransmission process in application example two of the embodiments of this application.

[0037] Figure 10 is a schematic block diagram of a terminal device according to an embodiment of this application.

[0038] Figure 11 is a schematic block diagram of a terminal device according to another embodiment of this application.

[0039] Figure 12 is a schematic block diagram of a network device according to an embodiment of this application.

[0040] Figure 13 is a schematic block diagram of a network device according to another embodiment of this application.

[0041] Figure 14 is a schematic block diagram of a network device according to another embodiment of this application.

[0042] Figure 15 is a schematic block diagram of a network device according to another embodiment of this application.

[0043] Figure 16 is a schematic block diagram of a network device according to another embodiment of this application.

[0044] Figure 17 is a schematic block diagram of a communication device according to an embodiment of this application.

[0045] Figure 18 is a schematic block diagram of a chip according to an embodiment of this application.

[0046] Figure 19 is a schematic block diagram of a communication system according to an embodiment of the present application. Detailed Implementation

[0047] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings.

[0048] The technical solutions of this application embodiment can be applied to various communication systems, such as: Long Term Evolution (LTE) system, Advanced Long Term Evolution (LTE-A) system, New Radio (NR) system, evolution system 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), Wireless Fidelity (WiFi), 5th Generation (5G) system, 6th Generation (6G) system, or other communication systems, etc.

[0049] Traditional communication systems typically support a limited number of connections and are easy to implement. However, with the development of communication technology, mobile communication systems will not only support traditional communication but also, for example, device-to-device (D2D) communication, machine-to-machine (M2M) communication, machine-type communication (MTC), vehicle-to-vehicle (V2V) communication, or vehicle-to-everything (V2X) communication. The embodiments of this application can also be applied to these communication systems.

[0050] In one implementation, the communication system in this application embodiment can be applied to a carrier aggregation (CA) scenario, a dual connectivity (DC) scenario, or a standalone (SA) network deployment scenario.

[0051] In one embodiment, the communication system in this application can be applied to unlicensed spectrum, wherein the unlicensed spectrum can also be considered as shared spectrum; or, the communication system in this application can also be applied to licensed spectrum, wherein the licensed spectrum can also be considered as non-shared spectrum.

[0052] This application describes various embodiments in conjunction with network devices and terminal devices. The terminal device may also be referred to as user equipment (UE), access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication device, user agent, or user device, etc.

[0053] Terminal devices can be stations (STAION, ST) in WLANs, cellular phones, cordless phones, Session Initiation Protocol (SIP) phones, Wireless Local Loop (WLL) stations, Personal Digital Assistant (PDA) devices, handheld devices with wireless communication capabilities, computing devices or other processing devices connected to a wireless modem, in-vehicle devices, wearable devices, terminal devices in next-generation communication systems such as NR networks, or terminal devices in future evolved Public Land Mobile Network (PLMN) networks, etc.

[0054] In the embodiments of this application, the terminal device can be deployed on land, including indoor or outdoor, handheld, wearable or vehicle-mounted; it can also be deployed on water (such as ships); and it can also be deployed in the air (such as airplanes, balloons and satellites).

[0055] In the embodiments of this application, the terminal device may be a mobile phone, a tablet computer, a computer with wireless transceiver capabilities, a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal device in industrial control, a wireless terminal device in self-driving, a wireless terminal device in remote medical care, a wireless terminal device in a smart grid, a wireless terminal device in transportation safety, a wireless terminal device in a smart city, or a wireless terminal device in a smart home, etc.

[0056] By way of example and not limitation, in this embodiment, the terminal device can also be a wearable device. Wearable devices, also known as wearable smart devices, are a general term for devices that utilize wearable technology to intelligently design and develop everyday wearables, such as glasses, gloves, watches, clothing, and shoes. Wearable devices are portable devices that are worn directly on the body or integrated into the user's clothing or accessories. Wearable devices are not merely hardware devices, but also achieve powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include those that are feature-rich, large in size, and can achieve complete or partial functions without relying on a smartphone, such as smartwatches or smart glasses, as well as those that focus on a specific type of application function and require the use of other devices such as smartphones, such as various smart bracelets and smart jewelry for vital sign monitoring.

[0057] In the embodiments of this application, the network device can be a device for communicating with mobile devices, such as an access point (AP) in a WLAN, an evolved Node B (eNB or eNodeB) in LTE, a relay station or access point, or a vehicle-mounted device, a wearable device, a network device (gNB) in an NR network, or a network device in a future evolved PLMN network or an NTN network, etc.

[0058] By way of example and not limitation, in this embodiment, the network device may have mobility characteristics; for example, the network device may be a mobile device. Optionally, the network device may be a satellite or a balloon station. For example, the satellite may be a low Earth orbit (LEO) satellite, a medium Earth orbit (MEO) satellite, a geostationary earth orbit (GEO) satellite, a high elliptical orbit (HEO) satellite, etc. Optionally, the network device may also be a base station located on land, water, or other similar locations.

[0059] In this embodiment, the network device can provide services to a cell. The terminal device communicates with the network device through the transmission resources (e.g., frequency domain resources, or spectrum resources) used by the cell. The cell can be the cell corresponding to the network device (e.g., a base station). The cell can belong to a macro base station or to a base station corresponding to a small cell. The small cell can include: metro cell, micro cell, pico cell, femto cell, etc. These small cells have the characteristics of small coverage area and low transmission power, and are suitable for providing high-speed data transmission services.

[0060] Figure 1 illustrates an exemplary communication system 100. The communication system includes a network device 110 and two terminal devices 120. In one embodiment, the communication system 100 may include multiple network devices 110, and the coverage area of ​​each network device 110 may include other numbers of terminal devices 120; this embodiment does not limit the scope of the present application.

[0061] In one embodiment, the communication system 100 may also include other network entities such as a Mobility Management Entity (MME) and an Access and Mobility Management Function (AMF), which are not limited in this application.

[0062] It should be understood that devices with communication functions in the network / system of this application embodiment can be referred to as communication devices. Taking the communication system shown in Figure 1 as an example, the communication device may include network devices and terminal devices with communication functions. The network devices and terminal devices can be specific devices in this application embodiment, which will not be described in detail here. The communication device may also include other devices in the communication system, such as network controllers, mobility management entities, and other network entities. This application embodiment does not limit this.

[0063] It should be understood that the terms "system" and "network" are often used interchangeably in this document. The term "and / or" in this document merely describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Furthermore, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0064] 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 a relationship. For example, A instructing B can mean that A directly instructs B, such as B being able to obtain information through A; it can also mean that A indirectly instructs B, such as A instructing C, so B can obtain information through C; or it can mean that there is a relationship between A and B.

[0065] In the description of the embodiments of this application, the term "correspondence" may indicate that there is a direct or indirect correspondence between two things, or that there is an association between two things, or that there is a relationship of instruction and being instructed, configuration and being configured, etc.

[0066] To facilitate understanding of the technical solutions of the embodiments of this application, the relevant technologies of the embodiments of this application are described below. The following relevant technologies are optional solutions and can be combined with the technical solutions of the embodiments of this application in any way, and they all fall within the protection scope of the embodiments of this application.

[0067] (I) Artificial Intelligence and Semantic Communication

[0068] In recent years, Artificial Intelligence (AI) technology, relying on the development of different types of neural networks and deep learning algorithms, has been widely applied in various fields such as image, speech, and video processing. Typical neural network architectures include fully connected networks, Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Transformer structures with self-attention mechanisms, which can accomplish different task objectives. Drawing on the rapid development of AI technology, the combination of artificial intelligence and wireless communication technology has also attracted widespread interest from academia and industry. AI-based CSI feedback, beam management, and positioning technologies have all undergone extensive research, evaluation, and standardization efforts.

[0069] Furthermore, with a focus on the future development of wireless communication systems, semantic communication technology, combined with AI, has received widespread attention in recent years. In contrast, the traditional communication paradigm is known as "syntactic communication," which uses traditional information theory methods such as Shannon's formula, information entropy, and the Nyquist sampling theorem to compress source information and then achieve efficient and reliable transmission through channel coding. The concept of "semantic communication," however, focuses on the extraction and representation of semantic information from source information, as well as the design of semantic communication systems integrated with AI, thereby coupling various modules together to achieve the transmission of semantic information.

[0070] A key difference between traditional grammatical communication systems and modern semantic communication lies in the design of source and channel coding. In traditional grammatical communication systems, source coding and channel coding are separate. Specifically, at the transmitter, source coding first compresses source information such as images, voice, and video into the original bitstream to be transmitted. This bitstream is then transmitted via channel coding using methods such as LDPC (Low-density Parity-check) and Polar coding. At the receiver, the original bitstream is first recovered through channel decoding, and then the original source is reconstructed through source decoding. This separation of source and channel coding facilitates modular optimization and design. Source coding compresses information redundancy, saving transmission bandwidth, while channel decoding enhances redundancy, improving the information's ability to resist noise and interference, and enabling error detection and correction.

[0071] In semantic communication systems, source coding and channel coding are jointly designed. Specifically, on the transmitter side, a joint source-channel encoder is designed to directly encode the original source information into the bit stream to be transmitted; on the receiver side, a joint source-channel decoder is designed to directly recover the original source information from the received bit stream. This method of coupling source and channel can achieve joint optimization in source compression redundancy and channel redundancy. Furthermore, since both the encoder and decoder are designed using AI models, end-to-end training and deployment can be achieved, allowing the AI ​​model to be more adapted to the current wireless channel characteristics and noise conditions, thus optimizing system performance. Simultaneously, joint source-channel coding can further incorporate modulation and demodulation functions, further optimizing system performance.

[0072] (II) CSI Feedback in NR

[0073] In current NR systems, CSI feedback schemes typically employ codebook-based eigenvector feedback to enable the base station to acquire downlink CSI. Specifically, the base station sends a downlink CSI-RS (Channel-State Information Reference Signal) to the user. The user uses the CSI-RS to estimate the downlink channel's CSI and performs eigenvalue decomposition on the estimated downlink channel to obtain its corresponding eigenvector. NR provides two codebook design schemes: Type I and Type II. Type I codebooks are used for CSI feedback with standard accuracy, primarily for SU-MIMO transmissions, while Type II codebooks are mainly used to improve MU-MIMO transmission performance. Both Type I and Type II codebooks employ a two-level codebook feedback system, W = W1W2. W1 describes the channel's bandwidth and long-term characteristics, defining a set of L DFT beams; W2 describes the channel's subband and short-term characteristics. Specifically, for the Type I codebook, W2 selects one beam from L DFT beams; for the Type II codebook, W2 linearly combines the L DFT beams in W1, providing feedback in the form of amplitude and phase. Generally, the Type II codebook utilizes a higher number of feedback bits to achieve higher precision CSI feedback performance.

[0074] In NR, the CSI reporting process is carried by UCI (Uplink Control Information) on PUCCH (Physical Uplink Control Channel) or PUSCH (Physical Uplink Shared Channel). This includes periodic and semi-persistent reporting on PUCCH, and aperiodic and semi-persistent reporting on PUSCH. When the receiver fails to receive the UCI information correctly based on CRC (Cyclic Redundancy Check), causing CSI feedback failure, for periodic and semi-persistent reporting, the network side does not perform any other operations, continues to use the reporting result of the previous CSI feedback cycle, and waits for the CSI to be reported in the next cycle. For aperiodic reporting, the network side reschedules a set of CSI-RS transmission and user CSI feedback resources and re-performs CSI feedback.

[0075] (III) AI-based CSI feedback

[0076] Given the tremendous success of AI technology, especially deep learning, in computer vision and natural language processing, the communications field has begun to explore using deep learning to solve technical challenges that are difficult to address with traditional communication methods. Deep learning's commonly used neural network architectures are non-linear and data-driven, capable of extracting features from actual channel matrix data and reconstructing the compressed channel matrix information from the UE (User Equipment) at the base station side as accurately as possible. This not only ensures the reconstruction of channel information but also provides the possibility of reducing CSI (Content Support Interface) feedback overhead at the UE side. Deep learning-based CSI feedback treats channel information as an image to be compressed, using a deep learning autoencoder to compress the input channel information and then reconstructing the compressed channel image at the transmitting end, thus preserving channel information to a greater extent.

[0077] Based on the current discussion, the main implementation framework for AI-based CSI feedback is as follows:

[0078] An AI-based CSI autoencoder is employed. The entire feedback system is divided into encoder and decoder parts, deployed at the user transmitter and base station receiver, respectively. After obtaining channel information through channel estimation, the user uses it as input to the encoder. The encoder's neural network compresses and encodes the channel information matrix, and the compressed bitstream is fed back to the base station via the air interface feedback link. The base station uses the decoder to recover the channel information based on the feedback bitstream and outputs the complete feedback channel information. Figure 2 is a schematic diagram of the AI-based CSI autoencoder. As shown in Figure 2, the CSI autoencoder is a neural network containing an encoder and a decoder. It can employ a DNN composed of multiple fully connected layers, a CNN composed of multiple convolutional layers, or an RNN with structures such as LSTM and GRU. Various neural network architectures, such as residual and self-attention mechanisms, can also be used to improve the performance of the encoder and decoder.

[0079] The encoder input (also known as CSI input) and decoder output (also known as CSI output) can both be full-channel information or feature vector information (e.g., CSI input is feature vector information obtained based on full-channel information, and CSI output is feature vector information used to determine full-channel information). Therefore, current deep learning-based channel information feedback methods are mainly divided into full-channel information feedback and feature vector feedback. While the former can achieve full-channel information compression and feedback, it has high feedback bitstream overhead and is not supported in existing NR systems. Feature vector-based feedback is the feedback architecture currently supported in NR systems. AI-based feature vector feedback methods can achieve higher CSI feedback accuracy with the same feedback bit overhead, or significantly reduce feedback overhead while achieving the same CSI feedback accuracy.

[0080] (iv) CSI feedback method based on source-channel joint coding

[0081] Besides the AI-based CSI feedback mentioned above, there is another type of CSI feedback method that borrows from semantic communication ideas, which can be called the source-channel joint coding-based CSI feedback method. Figure 3 is a schematic diagram of the source-channel joint coding-based CSI feedback method. In this method, noise in the uplink feedback is considered in the CSI feedback process. The encoder deployed on the terminal side implicitly implements CSI compression, uplink channel coding, and modulation; the decoder on the network side implicitly implements demodulation, channel decoding, and CSI recovery functions. In other words, the encoder function deployed on the terminal side replaces multiple functions such as CSI compression, uplink channel coding, and modulation; and the decoder function on the network side replaces multiple functions such as demodulation, channel decoding, and CSI recovery.

[0082] This joint CSI feedback method can take into account the impact of noise and jointly implement source-channel coding. It combines the lossy source coding process of CSI compression with the redundancy-adding process of channel coding, so that with limited uplink feedback resources, the performance of CSI encoder and CSI decoder against noise and fading channels under non-ideal channels can be further optimized, and the CSI feedback performance under non-ideal uplink feedback can be significantly improved.

[0083] As can be seen, current CSI feedback processes in communication systems typically assume ideal feedback during system-level and link-level performance evaluations. This means that the CSI reporting process is error-free, and the base station can perfectly decode the bit stream carrying CSI information and recover the CSI information. However, in real-world systems, the ideal assumption of uplink CSI cannot be achieved.

[0084] For codebook-based CSI feedback in NR systems and the AI-based CSI feedback currently under discussion, the original CSI is transformed into a raw information bitstream after passing through a codebook or autoencoder. This bitstream then needs to be channel-coded along with other UCI or data information and transmitted via PUCCH / PUSCH. However, during this process, errors may occur when decoding PUCCH / PUSCH at the base station due to noise. Especially in low signal-to-noise ratio environments such as cell edges or link obstructions, the error rate of uplink feedback may be high. In such cases, for periodic and semi-persistent CSI feedback, the base station cannot obtain the current CSI feedback and can only use the previous CSI feedback result, which can affect downlink transmission performance when the channel changes rapidly. For aperiodic CSI feedback, the base station may need to schedule CSI-RS measurement resources and re-feedback once or multiple times before decoding the correct downlink CSI information for downlink transmission. This introduces significant measurement resource overhead, feedback delay, and feedback overhead.

[0085] AI-based CSI feedback can significantly improve the accuracy of CSI recovery and reduce CSI feedback overhead. However, the AI-based CSI feedback discussed here also assumes ideal uplink feedback and does not consider the impact of noise. Therefore, it faces the same problem as codebook-based CSI feedback in NR, which will not be elaborated here.

[0086] On the other hand, while CSI feedback based on joint source channel coding can significantly improve the performance of CSI feedback under non-ideal uplink feedback conditions, this scheme may not explicitly involve NR channel coding, symbol modulation, and other processes; its functionality is implemented by the joint encoder on the UE side. Correspondingly, there is no explicit channel decoding process on the network side. Therefore, the network side cannot use CRC check to detect whether the CSI information carried by the current UCI is decoded correctly, and cannot determine whether a remeasurement or other scheduling request needs to be triggered. Thus, CSI feedback based on joint source channel coding fails to have sufficient stability and practicality in real-world systems.

[0087] The technical solutions of the embodiments of this application can solve at least one of the above technical problems.

[0088] Figure 4 is a schematic flowchart of a CSI feedback method executed by a terminal device according to an embodiment of this application. The method includes at least a portion of the following.

[0089] S410, The terminal device sends a first reporting information to the network device; wherein, the first reporting information includes CSI.

[0090] S420: The terminal device receives the first instruction information from the network device;

[0091] S430. The terminal device sends a second reporting information to the network device; wherein, the first indication information is used to instruct the terminal device to retransmit the CSI; the second reporting information includes the CSI.

[0092] In this embodiment, the first reported information including CSI means that the first reported information carries information indicating CSI, or in other words, the first reported information is used to indicate CSI. For example, the first reported information may include CSI or include CSI indication information. Here, CSI indication information is information used to indicate CSI, that is, CSI indication information carries the same channel information as the original CSI but has a different data representation. It may be a feature vector corresponding to CSI, or other forms of information representing CSI. In other words, the first reported information including CSI also includes the case where the first reported information includes CSI indication information. In this embodiment, the first reported information can provide CSI feedback through the CSI itself, or it can provide CSI feedback through CSI indication information. In some descriptions, CSI indication information may also be referred to as CSI information. For example, after the terminal device estimates the downlink channel CSI based on downlink CSI-RS, it sends the CSI through the first reported information, or it determines CSI indication information based on the CSI and then sends the CSI indication information through the first reported information. It is understood that, in the embodiments of this application, any processing of CSI can be understood as processing the CSI itself (original CSI) or processing the CSI indication information.

[0093] Optionally, in addition to CSI, the first reported information may also include other indication information or data information.

[0094] Optionally, the first reported information can be carried by UCI and transmitted via PUCCH and / or PUSCH.

[0095] In this embodiment, the first indication information is used to instruct the terminal device to retransmit the CSI. That is, in this embodiment, the network device can instruct the retransmission of the CSI by issuing the first indication information. Here, retransmission means re-transmitting, that is, re-reporting the same CSI as in the first reported information to the network device. Optionally, the network device can issue the first indication information under certain conditions.

[0096] Optionally, the first indication information can be trigger indication information, i.e., used to trigger the terminal device to retransmit CSI; or, the first indication information can also be used to indicate the configuration or parameters for retransmission.

[0097] Optionally, the first indication information may be carried by DCI, MAC CE or other downlink signaling and transmitted via PDCCH.

[0098] In this embodiment, the second reporting information is used to retransmit the aforementioned CSI. Optionally, the method of transmitting CSI through the second reporting information can be the same as or different from the method of transmitting CSI through the first reporting information. For example, the CSI encoding method used in the transmission processing of the first reporting information and the CSI encoding method used in the transmission processing of the second reporting information can be the same or different. For instance, the first reporting information can be CSI feedback based on source-channel joint coding; the second reporting information can be CSI encoding feedback based on a codebook, or CSI compression based on AI followed by channel coding, or CSI feedback based on source-channel joint coding.

[0099] Similarly, "second reported information includes CSI" means that the second reported information carries information indicating CSI, or that the second reported information is used to indicate CSI. For example, the second reported information may include the original CSI or include CSI indication information. That is, "second reported information includes CSI" also includes the case where the second reported information includes CSI indication information.

[0100] Optionally, in addition to CSI, the second reporting information may also include other indication information or data information. For example, the second reporting information may include configuration selection information to instruct the terminal device whether to update the configuration according to the first indication information or select a CSI feedback method.

[0101] Optionally, the second reporting information can be carried via UCI and transmitted via PUCCH and / or PUSCH.

[0102] According to the above method, after the terminal device sends the first reporting information to report CSI, if the network device instructs the terminal device to retransmit the CSI, the terminal device retransmits the CSI by sending the second reporting information. Compared to the network device using the CSI reporting results of the previous cycle or rescheduling CSI-RS resources for measurement and then performing CSI feedback under non-ideal uplink feedback conditions, the above method can reduce the performance loss and / or feedback delay caused by CSI loss under non-ideal uplink feedback conditions, enabling the network device to obtain downlink channel CSI information more effectively and accurately, thereby improving downlink precoding and downlink transmission performance.

[0103] Figure 5 is a schematic flowchart of a CSI feedback method performed by a network device according to an embodiment of this application. The method includes at least a portion of the following.

[0104] S510, The network device receives the first reported information from the terminal device; wherein, the first reported information includes CSI.

[0105] S520, The network device sends a first instruction message to the terminal device; wherein the first instruction message is used to instruct the terminal device to send a second reporting message to retransmit the CSI.

[0106] All or part of the technical details of the CSI feedback method executed by the network device described above can be found in the corresponding content settings of the CSI feedback method executed by the terminal device in the foregoing embodiments, and have the same technical effect. They will not be repeated here.

[0107] In some embodiments, the first indication information is sent upon satisfaction of a first condition. That is, the network device sending the first indication information to the terminal device includes: the network device sending the first indication information to the terminal device upon satisfaction of the first condition. The first condition can be a pre-configured condition used by the network device to determine whether to send the first indication information, i.e., a condition used by the network device to determine whether CSI retransmission is necessary. In some descriptions, this first condition may be referred to as a retransmission condition.

[0108] In some embodiments, the CSI feedback method described above may further include: if the first condition is not met, the network device does not send the first indication information to the terminal device. Optionally, the network device may recover the CSI based on the first reported information and use the recovered information for downlink transmission. That is, if the first condition is not met, the network device obtains the CSI reporting result based on the first reported information.

[0109] In some embodiments, the first reporting information and / or the second reporting information are used to report CSI based on a codebook or AI. Specifically, the first reporting information can be based on a codebook or AI to report CSI, and the second reporting information can also be based on a codebook or AI to report CSI. The first reporting information and the second reporting information can be reported in different ways or in the same way.

[0110] In some embodiments, AI-based CSI reporting includes one of the following:

[0111] The CSI is processed into the first output information based on the first AI model; the first output information is then mapped onto the uplink channel physical resources after channel coding and modulation.

[0112] The CSI is processed into a second output information based on the second AI model; the second output information is modulated and mapped onto the uplink channel physical resources for transmission.

[0113] The CSI is processed into third output information based on the third AI model; the third output information is used to map to uplink channel physical resources for transmission.

[0114] The AI ​​model described above processes CSI into various types of output information, which may include inputting CSI or CSI indication information into the AI ​​model and having the AI ​​model output the corresponding output information.

[0115] It should be noted that in the embodiments of this application, the AI ​​model can be understood as an AI function or AI feature. Correspondingly, the first AI model can be understood as the first AI function or first AI feature, the second AI model can be understood as the second AI function or second AI feature, and so on, without listing them all here. Specifically, in different scenarios or in different descriptions, AI-based CSI processing can be implemented by any form of model, function, and feature in the terminal device. For example, the terminal device deploys an AI model that can process the CSI of the input model; or, the terminal device has an AI function that can be used to process CSI; or, the terminal device has an AI feature that enables CSI processing. In addition, in some application scenarios, ML (Machine Learning) can be used to implement the above AI processing. Therefore, in some examples or descriptions, the AI ​​model can also be referred to as an ML model, AI / ML model, ML function, AI / ML function, ML feature, AI / ML feature, etc.

[0116] In other words, AI-based CSI reporting can include one of the following:

[0117] Based on the first AI function, CSI is processed into first output information; wherein, the first output information is mapped onto the uplink channel physical resources after channel coding and modulation and then transmitted.

[0118] The CSI is processed into second output information based on the second AI function; the second output information is modulated and mapped onto the uplink channel physical resources for transmission.

[0119] The CSI is processed into third output information based on the third AI function; the third output information is used to map onto the uplink channel physical resources for transmission.

[0120] Alternatively, AI-based CSI reporting can include one of the following:

[0121] Based on the first AI characteristic, CSI is processed into first output information; wherein, the first output information is mapped onto the uplink channel physical resources after channel coding and modulation and then transmitted.

[0122] Based on the second AI characteristic, CSI is processed into second output information; wherein, the second output information is modulated and mapped onto the uplink channel physical resources for transmission;

[0123] Based on the third AI characteristic, CSI is processed into third output information; the third output information is used to map onto uplink channel physical resources for transmission.

[0124] According to the above embodiments, the terminal device can support one or more AI models to implement one or more processing methods (or feedback methods) for CSI.

[0125] Specifically, the first AI model can be used to compress and encode CSI (source coding). The first output information obtained by the first AI model is a coded bit stream, which needs to be channel-coded and modulated before being mapped onto the uplink channel physical resources for transmission.

[0126] The second AI model can implement implicit source coding and channel coding functions for CSI. The second output information obtained by the second AI model is a coded bit stream, which can be mapped onto the uplink channel physical resources for transmission after modulation.

[0127] The third AI model can implement implicit source coding, channel coding, and modulation functions for CSI. The second output information obtained by the second AI model is a complex sequence, which can be directly mapped to the uplink channel physical resources for transmission.

[0128] It is understandable that, since the second and third AI models mentioned above jointly implement source coding and channel coding, both of them can be called AI models of joint source-channel coding.

[0129] For example, when sending the first reporting information, the terminal device may process the CSI based on a codebook or one of the first, second, and third AI models. When sending the second reporting information, the terminal device may also process the CSI based on a codebook or one of the first, second, and third AI models.

[0130] In a specific example, the terminal device can process CSI based on a second or third AI model when sending the first reporting information. That is, when sending the first reporting information, an AI model using joint source-channel coding is used to process the CSI. The network device can indicate the CSI reporting configuration for the second reporting information in the first indication information, and the terminal device determines which method to use to process the CSI when sending the second reporting information based on the first indication information. This overcomes the problem in CSI feedback based on joint source-channel coding where the network device cannot detect whether the CSI has been decoded correctly through CRC check. Through CSI retransmission, CSI feedback based on joint source-channel coding can achieve better stability and practicality in real-world systems.

[0131] Corresponding to the processing of the terminal device described above, in some embodiments, on the network device side, the CSI feedback method may further include one of the following:

[0132] The network device processes the first and / or second reported information based on the codebook to recover the CSI;

[0133] The network device demodulates and decodes the received symbols corresponding to the first reported information and / or the second reported information to obtain the first input information, and uses the fifth AI model to process the first input information to recover the CSI;

[0134] The network device demodulates the received symbols corresponding to the first reported information and / or the second reported information to obtain the second input information, and uses the sixth AI model to process the second input information to recover the CSI;

[0135] The network device uses the received symbols corresponding to the first and / or second reported information as the third input information, and processes the third input information using the seventh AI model to recover the CSI.

[0136] It is understood that when a terminal device processes CSI based on an AI model, the network device should process the received information based on a corresponding AI model to recover the CSI. In this embodiment, the AI ​​model used by the network device to recover the CSI can be referred to as the fourth AI model. Optionally, the fourth AI model may include at least one of the following: a fifth AI model corresponding to the first AI model in the terminal device, a sixth AI model corresponding to the second AI model in the terminal device, and a seventh AI model corresponding to the third AI model in the terminal device. Specifically, the input information of the fifth AI model is an encoded bitstream, which needs to be obtained by the network device after demodulating and channel decoding the received symbols. Thus, the fifth AI model recovers the CSI by compressing and decoding (source decoding) the input information. The input information of the sixth AI model is an encoded bitstream, which needs to be obtained by the network device after demodulating the received symbols. Thus, the sixth AI model recovers the CSI by implicitly performing channel decoding and source decoding on the input information. The input information of the seventh AI model is the received symbol after channel estimation and equalization. The network device uses the seventh AI model to perform implicit demodulation, channel decoding and source decoding on the received symbol in order to recover the CSI.

[0137] As described above, in some embodiments, the first indication information is sent when a first condition is met. This first condition may be related to uplink channel quality and / or first reported information. For example, the first condition may be that the uplink channel quality does not meet requirements or the first reported information does not meet requirements. Thus, the network device retransmits information under non-ideal uplink feedback conditions to ensure accurate reception of the CSI information.

[0138] In one implementation, the first condition is related to the uplink channel quality. Optionally, the network device can measure the uplink feedback link quality based on an uplink reference signal (e.g., SRS), where the first condition is that the uplink feedback link quality is less than a set threshold. That is, when the measured uplink feedback link quality is less than the set threshold, the network device triggers a retransmission. The uplink feedback link quality can be characterized by physical quantities such as SNR, SINR, and RSRP, or by other possible reference indicators.

[0139] In another implementation, the first condition is related to the first reported information, that is, the first reported information is used by the network device to determine whether to retransmit the CSI.

[0140] In some embodiments, the first reporting information is used by the network device to determine the input information of the fourth AI model; the input information of the fourth AI model is used by the network device to recover CSI based on the fourth AI model and to determine whether to retransmit CSI based on the distribution characteristics of the input information of the fourth AI model.

[0141] In other words, in some embodiments, the CSI feedback method performed by the network device may further include:

[0142] The network device determines the input information of the fourth AI model based on the received symbols corresponding to the first reported information; wherein, the fourth AI model is used to recover CSI;

[0143] Network devices determine whether the first condition is met based on the distribution characteristics of the input information.

[0144] In practical applications, the distribution characteristics of the input information of the fourth AI model can characterize the quality of demodulation performance. Therefore, based on the distribution characteristics of the input information of the fourth AI model, the network device can determine whether the first condition is met, and thus determine whether to send the first indication information.

[0145] In some embodiments, the first reporting information is used by the network device to obtain a log-likelihood ratio sequence through demodulation; the log-likelihood ratio sequence is used by the network device to determine whether to retransmit CSI based on the statistical value of the distribution characteristics of the log-likelihood ratio sequence.

[0146] Accordingly, the input information for the fourth AI model is a log-likelihood ratio sequence. That is, in some embodiments, in the CSI feedback method executed by the network device described above, the network device determines whether the first condition is met based on the distribution characteristics of the input information, including: the network device determines whether the first condition is met based on the statistical value of the distribution characteristics of the log-likelihood ratio sequence.

[0147] Specifically, when the terminal device uses the second AI model to process CSI, the network device obtains a log-likelihood ratio sequence after demodulating the received symbols. When the uplink feedback link quality is poor, the absolute value of the log-likelihood ratio is small, meaning the demodulation algorithm is weak in judging whether the bits carried at the constellation point are 0 or 1, resulting in poor demodulation performance. When the uplink feedback link quality is good, the absolute value of the log-likelihood ratio is large, meaning the demodulation algorithm is strong in judging whether the bits carried at that constellation point are 0 or 1, resulting in good demodulation performance. Therefore, the network device detects the distribution characteristics of the input information of the fourth AI model (which may include a sixth AI model corresponding to the second AI model), such as the absolute value mean, variance, standard deviation, and other statistical values ​​of the input information. When the statistical characteristics meet the set threshold requirements (e.g., the absolute value mean is greater than a preset first threshold, or the variance, standard deviation, etc. are less than a preset second threshold), a retransmission condition is triggered. When the set threshold requirements of the statistical characteristics are not met, the retransmission condition is not triggered. In this way, retransmission conditions can be triggered when the uplink feedback link quality is poor, and retransmission conditions can be not triggered when the uplink feedback link quality is good.

[0148] In some embodiments, the first reporting information is used by the network device to obtain received symbols through channel equalization; the received symbols are used by the network device to compare the distribution of the received symbols with the distribution of the unit circle or the third output information to determine whether to retransmit CSI; wherein, the third output information is the output information of the third AI model corresponding to the fourth AI model in the terminal device.

[0149] Accordingly, the input information of the fourth AI model includes the received symbol. That is, in some embodiments, in the CSI feedback method executed by the network device described above, the network device determines whether the first condition is met based on the distribution characteristics of the input information, including: the network device compares the distribution of the received symbol with the distribution of the unit circle or the third output information to determine whether the first condition is met; wherein, the third output information is the output information of the third AI model corresponding to the fourth AI model in the terminal device.

[0150] Specifically, when the terminal device processes CSI using the third AI model, the network device obtains the received symbol through channel equalization. Since the third AI model implicitly implements source coding, channel coding, and modulation functions, the network device-side model corresponding to the third AI model can implicitly implement demodulation, channel decoding, and source decoding. In other words, the network device can directly use this received symbol as input information for the fourth AI model. At this point, the following two situations exist:

[0151] Scenario 1: The output information of the third AI model (i.e., the third output information) satisfies the constant modulus constraint (i.e., all complex numbers are distributed on the unit circle). In this case, the network device can compare the distribution of received symbols with the distribution of the unit circle to determine whether the first condition is satisfied. Specifically, when the uplink feedback link quality is better, the distribution of received symbols is closer to the unit circle, and the first condition is not satisfied; when the uplink feedback link quality is poor, the distribution of received symbols differs significantly from the distribution of the unit circle, and the first condition is satisfied.

[0152] Scenario 2: The output information of the third AI model (i.e., the third output information) only satisfies the total power constraint but not the constant modulus constraint. In this case, the network device can compare the distribution of received symbols with the distribution of the third output information. Specifically, when the uplink feedback link quality is better, the distribution of received symbols is closer to the distribution of the third output information, and the first condition is not met, i.e., CSI retransmission is not triggered. When the uplink feedback link quality is poor, the distribution of received symbols differs significantly from the distribution of the unit circle, satisfying the first condition, i.e., CSI retransmission is triggered.

[0153] In scenario 2 above, the characteristics of the third output information distribution depend on the implementation method and parameters of the third AI model on the terminal device side. When the third AI model on the terminal device side is trained and distributed by the network device side, the distribution characteristics of the third output information are known to the network device. When the third AI model on the terminal device side is trained and deployed by the user side, the terminal device can report a sample set according to the network device's instructions for the network device side to compare the distribution of received symbols with the distribution of the third output information.

[0154] Specifically, in some embodiments, the CSI feedback method executed by the terminal device further includes:

[0155] The terminal device sends a sample set of output information from the third AI model to the network device; the sample set of output information is used by the network device to determine the distribution of the third output information.

[0156] Accordingly, the CSI feedback methods performed by network devices also include:

[0157] The network device receives a sample set of output information from a third AI model from a terminal device; the sample set of output information from the third AI model is used by the network device to determine the distribution of the third output information.

[0158] In some embodiments, the first indication information is used to indicate at least one of the following information A to D:

[0159] A. Whether to send the second reporting information based on the CSI reporting configuration of the first reporting information.

[0160] Optionally, the first indication information may include a retransmission configuration reuse indication, which indicates whether to reuse the CSI reporting configuration of the first reported information, i.e., whether to send the second reported information based on the CSI reporting configuration of the first reported information. Here, the CSI reporting configuration may refer to the CSI processing method (also known as the CSI feedback method), such as based on a codebook or based on an AI model. Furthermore, the CSI reporting configuration may also include the type of AI model.

[0161] Optionally, when the retransmission configuration reuse indication is a first value, the current CSI feedback method is used by default. For example, the terminal device is instructed to still use the current AI model for CSI compression reporting, and the network device is instructed to still use the corresponding fifth AI model for CSI recovery. When the retransmission configuration reuse indication is a second value, the network device can further instruct the terminal device to use the CSI reporting configuration for the next retransmission, or the terminal device can use other pre-configured CSI feedback methods, such as pre-configuring to use the codebook for retransmission when the first indication information is received and the reuse configuration is a second value. For example, the first value can be 0 and the second value can be 1.

[0162] B. CSI reporting configuration for the second reporting information.

[0163] Optionally, the CSI reporting configuration for the second reporting information may include at least one of the following:

[0164] The handling method for CSI when sending the second reporting information;

[0165] Configuration information related to this processing method.

[0166] Specifically, the processing method for CSI when sending the second reporting information may include at least: processing CSI based on the codebook, or processing CSI based on AI-based non-joint source coding and channel coding (i.e., processing CSI based on the first AI model), or processing CSI based on AI-based joint source coding and channel coding (i.e., processing CSI based on the second AI model or the third AI model).

[0167] Optionally, the configuration information related to this processing method may include relevant parameters of this processing method.

[0168] For example, the configuration information related to this processing method may include at least one of the following:

[0169] The length of the output information of the AI ​​model used to implement the processing method;

[0170] Constraint information for the AI ​​model used to implement the processing method; wherein, the constraint information is used to indicate whether the output information of the AI ​​model needs to satisfy the constant modulus constraint;

[0171] The modulation methods included in the processing method.

[0172] For example, when the instruction to send the second reporting information is to use AI-based joint source coding and channel coding (i.e., to process the CSI based on the second AI model or the third AI model), the relevant configuration information may include the length of the encoded bit stream output by the second AI model or the third AI model, i.e., the number of bits in the encoded bit stream.

[0173] For example, when the processing method is to process CSI based on a third AI model, the relevant configuration information may also include the aforementioned constraint information, so that the terminal device can determine whether it needs to send the output information sample set to the network device.

[0174] For example, when the processing method is to process CSI based on the second AI model, the relevant configuration information can also be the modulation method of the second output information after the second AI model outputs the second output information.

[0175] C. Uplink channel resources used for transmitting second reporting information.

[0176] Optionally, uplink channel resources include PUSCH resources or PUCCH resources.

[0177] Optionally, when the first reported information is used for periodic CSI feedback or semi-persistent CSI feedback, the uplink channel resources include PUSCH resources or PUCCH resources. Since periodic or semi-persistent CSI feedback is performed via PUCCH, retransmissions in periodic or semi-persistent CSI feedback can be performed via PUCCH or PUSCH. If retransmission is performed via PUCCH, only the reporting configuration for CSI retransmission is supported, and the same configuration is used for the current periodic CSI feedback. If retransmission is performed via PUSCH, a portion of the uplink PUSCH resources needs to be allocated for CSI retransmission, and this needs to be indicated through the first indication information.

[0178] Optionally, when the first reported information is used for aperiodic CSI feedback, the uplink channel resources include PUSCH resources. Since aperiodic CSI feedback only occurs on the PUSCH, aperiodic CSI retransmission is also configured only on the PUSCH. The network device needs to allocate a portion of the uplink PUSCH resources for CSI retransmission and indicate this through the first indication information.

[0179] D. Time-domain information used to transmit the second reporting information.

[0180] Optionally, the time-domain information includes the time slot offset between the second reported information and the first reported information.

[0181] For periodic or semi-continuous CSI feedback, the time slot offset is specifically the time slot offset δ of the uplink time slot occupied by the retransmitted CSI (second reporting information) in the periodic CSI feedback relative to the current periodic CSI report (first reporting information). For shorter CSI feedback cycles, only δ=1 can be supported, that is, CSI feedback is performed on the adjacent uplink time slot; for longer CSI feedback cycles, different time slot offset configurations can be supported, such as δ=1, 3, 5, etc.

[0182] For retransmission of aperiodic CSI, the time slot offset is specifically the time slot offset δ of the retransmitted CSI (second reported information) relative to the current aperiodic CSI report (first reported information), and can have various configurations such as δ = 1, 3, 5, etc.

[0183] It is understood that in practical applications, the first instruction information can be used to indicate one or more of the above information. The specific information can be determined according to the agreement, system agreement, or application scenario requirements. This application does not limit this, and for the sake of brevity, it does not list the combination methods of the above information one by one.

[0184] In some embodiments, the second reporting information may include CSI and configuration selection information. The configuration selection information is used to instruct the terminal device whether to update or select a CSI feedback method according to the configuration of the first indication information. If the terminal device does not report the configuration selection information or the configuration selection information indicates a third value, the network device may assume that the terminal device continues to use the current CSI feedback method for CSI retransmission reporting, and the network device will use the corresponding method, such as the corresponding AI model, to perform CSI recovery. If the configuration selection information indicates a fourth value, the network device may consider that the terminal device has selected the CSI feedback method indicated by the first indication information for CSI retransmission reporting, and the network device will also select the corresponding method, such as the corresponding AI model, for CSI recovery.

[0185] In some embodiments, when the first reporting information is used for non-periodic CSI feedback, the terminal device can receive the first indication information multiple times. Accordingly, in the CSI feedback method performed by the terminal device described above, when the terminal device receives the first indication information from the network device, sending the second reporting information to the network device may include: the terminal device sending the second reporting information to the network device each time it receives the first indication information.

[0186] To avoid increasing transmission overhead, the number of retransmissions can be limited. Optionally, the network device can send a maximum of N first indication messages; where N is determined based on the uplink channel quality. For example, the network device can determine the uplink channel quality and then automatically determine the maximum number of retransmissions N based on that quality. For instance, when the detected uplink signal-to-noise ratio is high, the maximum number of retransmissions is 1 to reduce retransmission overhead; when the detected uplink signal-to-noise ratio is low, the maximum number of retransmissions is 4 to ensure higher CSI feedback accuracy.

[0187] In some embodiments, when the first reported information is used for non-periodic CSI feedback, the CSI feedback method performed by the network device may further include: if it is determined that the first condition is not met or the number of retransmissions reaches N, the network device stops sending the first indication information.

[0188] In some embodiments, when the first reported information is used for periodic CSI feedback, the network device needs to consider whether a periodic CSI retransmission mechanism is configured, i.e., whether it is configured to support CSI retransmission in the case of periodic CSI feedback. Specifically, the network device sends a first indication message to the terminal device, including: the network device determining whether the terminal device has a periodic CSI retransmission mechanism configured; if the terminal device has a periodic CSI retransmission mechanism configured, the network device sends the first indication message.

[0189] Optionally, the above method may further include: if the terminal device is not configured with a periodic CSI retransmission mechanism, the network device does not send the first indication information.

[0190] In the above embodiments, configuring a periodic CSI retransmission mechanism for the terminal device can be done in several ways:

[0191] (1) It can be pre-configured by network devices through RRC signaling. If the RRC signaling is configured with a periodic CSI retransmission mechanism, the terminal device will support the CSI retransmission mechanism in each periodic CSI feedback; otherwise, it will not support it.

[0192] (2) It can be pre-configured by network devices via RRC and activated and deactivated by MAC CE or DCI. That is, when MAC CE or DCI is activated, the periodic CSI feedback supports the CSI retransmission mechanism, and when MAC CE or DCI is deactivated, the periodic CSI feedback does not support the CSI retransmission mechanism.

[0193] (3) It can be pre-configured by the network device through MAC CE. If the MAC CE signaling is configured with a periodic CSI retransmission mechanism, the terminal device will support the CSI retransmission mechanism in each periodic CSI feedback cycle; otherwise, it will not support it.

[0194] (4) It can be pre-configured by network devices through MAC CE and activated and deactivated by DCI. That is, when DCI is activated, the periodic CSI feedback supports the CSI retransmission mechanism, and when DCI is deactivated, the periodic CSI feedback does not support the CSI retransmission mechanism.

[0195] (5) It can be dynamically configured by network devices through DCI, and DCI indicates whether or not the CSI retransmission mechanism in periodic CSI feedback is supported.

[0196] (6) Other possible higher-level or physical-level signaling, or combinations of other signaling and the above signaling, are not listed one by one, but are all within the scope of this application.

[0197] In some embodiments, the second reporting information is used by the network device to merge the CSI recovered based on the second reporting information with the CSI recovered based on the first reporting information, or to replace the CSI recovered based on the first reporting information with the CSI recovered based on the second reporting information.

[0198] In other words, in some embodiments, the CSI feedback method performed by the network device described above may further include:

[0199] The network device will merge the CSI recovered based on the second reported information with the CSI recovered based on the first reported information;

[0200] or,

[0201] The network device uses the CSI recovered based on the second reported information instead of the CSI recovered based on the first reported information.

[0202] The method of merging the CSI recovered based on the second reported information with the CSI recovered based on the first reported information can be called merged reception. In merged reception, the same CSI, after passing through two different uplink channels and noise interference conditions, achieves diversity reception gain after merging at the network side, which can improve the CSI recovery performance of the network device. Optionally, the merged reception method can be to add or average the input information of the network device's AI model during the two transmissions and use this as the merged input information. Alternatively, feature extraction can be performed on the output information of the network device's AI model during the two transmissions to obtain the feature vectors of each output information, and then the two output information feature vectors can be merged. The above are only examples of different implementation methods of the merging method. Since the merging method only depends on the network side implementation and does not affect the air interface protocol, other possible merging methods are also applicable and included within the scope of this application.

[0203] The above-mentioned replacement of the CSI recovered from the first reporting information with the CSI recovered from the second reporting information can be called non-combined reception. Non-combined reception can discard the first reporting information transmitted under non-ideal uplink feedback conditions, and can guarantee CSI recovery performance to a certain extent when the channel state changes significantly.

[0204] In some embodiments, upon receiving first indication information from a network device, the terminal device may send second reporting information to the network device, which may include:

[0205] After sending the first reporting information, the terminal device stores the CSI in the first storage unit;

[0206] Upon receiving the first instruction information from the network device, the terminal device reads the CSI from the first storage unit and sends the second reporting information to the network device based on the CSI.

[0207] In some embodiments, the above method may further include:

[0208] When the terminal device determines that it will no longer receive the first instruction information for the first reported information, it clears the first storage unit.

[0209] In other words, a first storage unit is set up in the terminal device to cache CSI data to support CSI retransmission. This mechanism can be called a buffer mechanism for CSI retransmission. The buffer can be a storage unit that stores the current CSI measurement and feedback results, or a cache, or other functional partition that implements similar functionality. For the first storage unit, after each periodic, non-periodic, or semi-continuous CSI feedback, the CSI data contained in the first reported information can be automatically saved into the first storage unit. When the first indication information is received from the network side, the content of the first storage unit is not updated, and the content of the first storage unit is used for reporting the CSI data contained in the second reported information. When the first indication information is no longer received, the content of the first storage unit is cleared.

[0210] Optionally, if the terminal device does not receive the first instruction information within a predetermined time period after sending the first reporting information or the second reporting information, it can be determined that it will no longer receive the first instruction information for the first reporting information.

[0211] Optionally, the capacity of the first storage unit is one unit in size, meaning it can only store one CSI (CSI measurement result).

[0212] Similar to the embodiments described above, a buffer mechanism for CSI retransmission can also be configured on the network device side. Specifically, in some embodiments, the CSI feedback method executed by the network device may further include:

[0213] After receiving the first report information, the network device stores the CSI recovered based on the first report information in the second storage unit;

[0214] Upon receiving the second reported information, the network device stores the second reported information or the CSI recovered based on the second reported information in the second storage unit.

[0215] In some embodiments, the above method may further include:

[0216] When the network device determines that the first condition is not met or the number of retransmissions reaches the preset number, it clears the first storage unit.

[0217] In other words, a second storage unit is set up in the network device to cache the received second reported information or the recovered CSI, in order to support the merging of CSI after retransmission. Specifically, each time a periodic, non-periodic, or semi-persistent CSI feedback is received, the network device automatically saves the CSI contained in the recovered first reported information into the second storage unit. When a second reported information is received from a terminal device, the CSI contained in the second reported information is processed, and the second reported information or the CSI recovered based on the second reported information is saved into the second storage unit. This process is repeated for multiple retransmissions until the network-side retransmission condition detection passes and no further retransmission is required, or the maximum number of retransmissions is reached, at which point the second storage unit is cleared.

[0218] Optionally, the capacity of the second storage unit, for retransmission mechanisms with periodic CSI feedback and semi-persistent CSI feedback, is also only one unit size, meaning it can only store the second input information or the second output information result once. For retransmission mechanisms with non-periodic CSI feedback, it can be one or more unit sizes, depending on the maximum allowed number of retransmissions.

[0219] The aforementioned buffer mechanism allows both network and terminal devices to store current CSI measurements and feedback results, which are then soft-merged with retransmitted CSI information. This further improves CSI recovery accuracy and ensures downlink precoding and transmission performance.

[0220] In some embodiments, the CSI feedback method performed by the terminal device further includes:

[0221] The terminal device sends a second indication message to the network device; wherein the second indication message is used by the network device to determine that the terminal device possesses a first capability; the first capability includes at least one of the following:

[0222] Terminal devices support CSI feedback and retransmission based on AI;

[0223] The terminal device supports the deployment of AI models for CSI feedback;

[0224] The terminal device supports receiving the first instruction information;

[0225] Terminal devices support caching CSI for retransmission.

[0226] Optionally, the step of sending the second instruction information is performed before the terminal device sends the first reporting information.

[0227] Accordingly, the CSI feedback methods performed by network devices also include:

[0228] The network device receives second indication information from the terminal device; wherein the second indication information is used by the network device to determine that the terminal device possesses a first capability; the first capability includes at least one of the following:

[0229] Terminal devices support CSI feedback and retransmission based on AI;

[0230] The terminal device supports the deployment of AI models for CSI feedback;

[0231] The terminal device supports receiving the first instruction information;

[0232] Terminal devices support caching CSI for retransmission.

[0233] The above embodiments provide a terminal capability reporting method that can support a retransmission mechanism based on CSI feedback.

[0234] In practical applications, the first capability reported by the terminal device may include one or more of the above capabilities. Optionally, if the first capability includes some of the above four capabilities, then the terminal device also supports the other capabilities among the above four capabilities. That is, when the terminal device reports that it possesses some of the above four capabilities, it can be assumed that the terminal device also possesses the other capabilities among the above four capabilities.

[0235] For example, the following reporting methods are possible:

[0236] Method 1: Terminal devices can report CSI feedback and retransmission based on AI. Terminals with this first capability can also support the deployment of AI models for CSI feedback, receive first indication information, and cache CSI for retransmission.

[0237] Method 2: Terminal devices can report and deploy AI models for CSI feedback. Terminal devices with the first capability can also receive the first indication information, perform CSI feedback and retransmission based on AI, and cache CSI for retransmission.

[0238] Method 3: The terminal device reports the ability to receive the first indication information. The terminal device with the first capability can also support the deployment of an AI model for CSI feedback, support AI-based CSI feedback and retransmission, and cache CSI for retransmission.

[0239] Method 4: The terminal device reports CSI with the ability to cache for retransmission. The terminal device with the first capability can also support the deployment of AI models for CSI feedback, support AI-based CSI feedback and retransmission, and support receiving the first indication information.

[0240] Through the above terminal capability reporting mechanism, network devices can obtain more accurate CSI feedback results with lower measurement feedback overhead and latency, thus ensuring downlink precoding and downlink transmission performance.

[0241] To facilitate understanding of the above technical solutions, several specific application examples are provided below for different CSI feedback scenarios.

[0242] Application Example 1

[0243] This application example provides a CSI retransmission method for joint source channel coding.

[0244] In this application example, CSI feedback using joint source-channel coding includes at least the following two configuration methods.

[0245] Referring to Figure 6, the first configuration method involves the terminal device-side AI / ML model / function / feature implementing implicit source coding and channel coding. The terminal device-side AI / ML model / function / feature can refer to the second AI model / function / feature in the above embodiments. The input information for the terminal device-side AI / ML model / function / feature is CSI information (which can be a channel feature vector, the original channel, or other input forms representing CSI information; here, a channel feature vector is used as an example, but other input forms are within the scope of this application). The output information for the terminal device-side AI / ML model / function / feature is the jointly coded bitstream. After passing through the modulation module, the output information of the terminal device-side AI / ML model / function / feature forms constellation points, which are mapped to the corresponding uplink channel physical resources and transmitted. After the network side receives the symbol, after channel estimation and equalization, the log-likelihood ratio of the bits carried by each constellation point is obtained through the demodulation module as the input information of the AI / ML model / function / characteristic on the network device side. The AI / ML model / function / characteristic on the network device side implements implicit channel decoding and source decoding functions. The AI / ML model / function / characteristic on the network device side can refer to the sixth AI model / function / characteristic in the above embodiment. The output information of the AI / ML model / function / characteristic on the network device side is the recovered CSI information.

[0246] Referring to Figure 7, the second configuration method involves the terminal device-side AI / ML model / function / feature implementing implicit source coding, channel coding, and modulation. The terminal device-side AI / ML model / function / feature can refer to the third AI model / function / feature in the above embodiments. The input information and output information of both the terminal device-side AI / ML model / function / feature and the network device-side AI / ML model / function / feature are the same as in the first configuration method. The difference is that in the second configuration method, the output information of the terminal device-side AI / ML model / function / feature is a complex sequence, which can be directly mapped to the corresponding uplink channel physical resources and transmitted. The network device-side AI / ML model / function / feature implements implicit demodulation, channel decoding, and source decoding functions, specifically referring to the seventh AI model / function / feature in the above embodiments, with the input information being the received symbols after channel estimation and equalization.

[0247] The CSI retransmission process is shown in Figure 8:

[0248] The specific steps include:

[0249] S1, the terminal device sends a first reporting message to the network. The first reporting message includes at least one CSI or CSI indication message from the terminal device; the first reporting message is carried by UCI and reported via PUCCH and / or PUSCH.

[0250] S2, the network side performs retransmission condition detection. If the retransmission condition is met, a new CSI report is required; if the condition is not met, a new CSI report is not required. Retransmission condition detection can take various forms, which can be determined by the network side, for example:

[0251] (1) A retransmission detection method based on uplink channel quality: The network side measures the uplink feedback link quality (which can be physical quantities that characterize the uplink channel link quality, such as SNR, SINR, RSRP, etc., or other possible reference indicators) based on the uplink reference signal (e.g., SRS, etc.). When the measured uplink feedback link quality is less than a set threshold, the retransmission condition is triggered.

[0252] (2) A retransmission detection method based on the distribution characteristics of input information of AI / ML models / functions / characteristics on the network device side. Specifically: a) For the first configuration method, the input information of AI / ML models / functions / characteristics on the network device side is the demodulated log-likelihood ratio sequence. When the uplink feedback link quality is poor, the absolute value of the log-likelihood ratio is small, that is, the demodulation algorithm is weak in judging whether the bit carried at this constellation point is 0 or 1, and the final demodulation performance is poor; when the uplink feedback link quality is good, the absolute value of the log-likelihood ratio is large, that is, the demodulation algorithm is strong in judging whether the bit carried at this constellation point is 0 or 1, and the final demodulation performance is good. Therefore, the network side can detect the distribution characteristics of the input information of the AI / ML model / function / feature on the network device side, such as the absolute mean, variance, standard deviation, and other statistical values ​​of the input information. When the statistical characteristics meet the set threshold requirements, a retransmission condition is triggered; when the set threshold requirements are not met, a retransmission condition is not triggered. b) For the second configuration, the input information of the AI / ML model / function / feature on the network device side is the received symbol after channel equalization. In this case, the network side can detect the distribution characteristics of the input information of the AI / ML model / function / feature on the network device side. For example, when the complex sequence of the output information of the AI / ML model / function / feature on the terminal device side satisfies the constant modulus constraint (i.e., all complex numbers are distributed on the unit circle), the difference between the distribution of the input information of the AI / ML model / function / feature on the network device side and the unit circle can be compared. When the uplink feedback link quality is better, the distribution of received symbols approaches the unit circle more closely. When the complex sequence of output information of the AI / ML model / function / feature on the terminal device side only satisfies the total power constraint but not the constant modulus constraint, the difference between the input information distribution of the AI / ML model / function / feature on the network device side and the output information distribution of the AI / ML model / function / feature on the terminal device side can be compared. The characteristics of the output information distribution of the AI / ML model / function / feature on the terminal device side depend on the implementation method and parameters of the AI / ML model / function / feature on the terminal device side. When the AI / ML model / function / feature on the terminal device side is trained by the network side and sent to the terminal device, the output information distribution characteristics of the AI / ML model / function / feature on the terminal device side are known on the network side. When the AI / ML model / function / feature on the terminal device side is trained and deployed by the terminal device side, the terminal device can report the output information sample set of the AI / ML model / function / feature on the terminal device side according to network instructions for the network side to realize distribution feature detection.

[0253] S3: The network sends a first indication message to the terminal device. This first indication message can be carried by DCI, MAC CE, or other downlink signaling and transmitted via PDCCH. The first indication message is used to instruct the terminal device on retransmission configuration information. When the network does not trigger the S3 procedure (i.e., does not send the first indication message), the terminal device does not retransmit CSI. When the network configures the first indication message to be sent, the first indication message includes at least the following:

[0254] (1) Retransmission configuration reuse indication: When the retransmission configuration reuse indication is 0, the current CSI feedback method is used by default, that is, the terminal device is still used to compress and report CSI using the current AI / ML model / function / feature, and the network is still used to restore CSI using the corresponding AI / ML model / function / feature; when the retransmission configuration reuse indication is 1, the network side needs to further indicate the configuration of CSI reporting used by the terminal device for the next retransmission, such as the CSI reporting configuration in (2) below.

[0255] (2) The CSI reporting configuration used for retransmission by the terminal device shall include at least the following: the terminal device adopts a traditional codebook-based CSI reporting method, or an AI-based non-joint source coding and channel coding method, or a first configuration method or a second configuration method based on joint source and channel coding. If the first configuration method is indicated, it may further include the output length and modulation scheme of the AI / ML model / function / feature on the terminal device side; if the second configuration method is indicated, it may further include the output length of the AI / ML model / function / feature on the terminal device side, and configuration information such as whether the output needs to meet constant modulus constraints.

[0256] S4, the terminal device performs CSI compression and retransmission based on the configuration information indicated by the first indication information and using the terminal device-side AI / ML model / function / feature. If the terminal device can support the CSI reporting configuration used for retransmission in the first indication information, then the configuration in the first indication information is used; if the terminal device cannot support the configuration in the first indication information, then the current terminal device-side AI / ML model / function / feature configuration is used for CSI reporting by default.

[0257] S5, the terminal device sends second reporting information to the network. This second reporting information is carried via UCI and reported through PUCCH and / or PUSCH. The second reporting information includes at least the terminal device's CSI retransmission information and configuration selection information. The configuration selection information indicates whether the terminal device updates and selects its AI / ML model / function / feature according to the configuration specified in the first indication information. If the terminal device does not report this configuration selection information or the configuration selection information is 0, the network side defaults to the terminal device continuing to use the current terminal device-side AI / ML model / function / feature for CSI retransmission reporting, and the network side uses the current network device-side AI / ML model / function / feature for CSI recovery. If the configuration selection information is 1, the network side considers that the terminal device has selected the terminal device-side AI / ML model / function / feature configuration indicated by the first indication information for CSI retransmission reporting, and the network side will also select the corresponding network device-side AI / ML model / function / feature for CSI recovery.

[0258] S6, the network side receives the second reported information and performs CSI recovery based on the corresponding network device's AI / ML model / function / feature. Retransmitted CSI information can be utilized in two ways: non-merged reception and merged reception. For non-merged reception, the network side only uses the retransmitted CSI reported information contained in the second reported information and selects the matching network device's AI / ML model / function / feature for CSI recovery. For merged reception, there are several different methods:

[0259] (1) Soft combining of input information of AI / ML model / function / feature on the network device side under the first configuration mode: Under the first configuration mode, if the selected terminal device side AI / ML model / function / feature and modulation mode are the same in the two transmissions, the log-likelihood ratio sequence of the input information of the network device side AI / ML model / function / feature can be combined using a soft combining method. That is, the sum of the second information inputs of the two transmissions can be used as the combined input information of the network device side AI / ML model / function / feature. The advantage of this combining method is that the same CSI information passes through two different uplink channels and noise interference conditions, and after combining on the network side, diversity reception gain is achieved, improving the CSI recovery performance of the network device side AI / ML model / function / feature.

[0260] (2) Soft combining of input information of AI / ML model / function / characteristic on the network device side under the second configuration: Under the second configuration, if the selected terminal device side AI / ML model / function / characteristic is the same in the two transmissions, the received symbols of the input information of the AI / ML model / function / characteristic on the network device side can be combined using a soft combining method. That is, the average of the second information inputs of the two transmissions can be used as the input information of the combined network device side AI / ML model / function / characteristic. The advantages of this combining method are similar to those of (1), that is, the same CSI information passes through two different uplink channels and noise interference conditions, and after combining on the network side, diversity reception gain is achieved, which improves the CSI recovery performance of the AI / ML model / function / characteristic on the network device side.

[0261] (3) Soft merging based on feature extraction of output information of AI / ML model / function / characteristics on network device side: When the configuration information used in the two transmissions is different and soft merging in (1) or (2) is not possible, a soft merging method based on feature extraction of output information of AI / ML model / function / characteristics on network device side can be adopted. Specifically, the output information of AI / ML model / function / characteristics on network device side includes the recovered CSI information. As an example, a merging method based on eigenvalue decomposition can be adopted, and the feature vectors on the kth subband of the two transmissions are respectively denoted as... and Calculate the autocorrelation matrix of the merged eigenvectors for:

[0262] right The eigenvalue decomposition yields the eigenvector corresponding to the largest eigenvalue, which is the CSI eigenvector of the kth subband after merging.

[0263] The above examples are merely illustrations of different implementation methods for merging. Since the merging method depends only on the network side implementation and does not affect the air interface protocol, other possible merging methods can be applied to the above signaling and configuration methods and are included within the scope of this application.

[0264] The CSI retransmission signaling procedure only shows the process of one retransmission. Multiple retransmissions can be performed by repeating steps S2-S6 until the retransmission condition check in S2 passes and no further retransmission is needed, or until the maximum number of retransmissions is reached. The maximum number of retransmissions is not visible to the terminal device; the terminal device only triggers retransmission based on whether it receives the first indication message. Therefore, the maximum number of retransmissions can be specifically implemented by the network side based on different CSI feedback situations (e.g., application examples two / three / four below).

[0265] The beneficial effects of the CSI retransmission method for joint source channel coding presented in this application example are: it can enhance the CSI feedback performance of joint source channel coding, reduce the feedback delay and performance loss caused by the loss of CSI information under non-ideal uplink feedback conditions, ensure that the network side can obtain downlink channel CSI information more effectively and accurately, and thus improve downlink precoding and transmission performance.

[0266] Application Example 2

[0267] This application example provides a retransmission mechanism based on a periodic CSI feedback method.

[0268] In the periodic feedback configuration of an NR system, when a CSI feedback is not received correctly, the network side will continue to use the CSI feedback result from the previous cycle and wait for the next CSI feedback, without rescheduling the CSI feedback. When the channel changes rapidly, the CSI feedback from the previous cycle may not match the downlink channel of the current cycle. This mechanism can lead to inaccurate downlink precoding, resulting in poor downlink transmission performance.

[0269] This application example provides a retransmission mechanism under periodic CSI feedback using joint source channel coding. For joint source channel coding CSI feedback, the network side always receives the current CSI feedback result, therefore it is not necessary to use the CSI result from the previous period. When the current CSI feedback satisfies the retransmission condition detection in step S2 of Application Example 1, the process shown in Figure 9 needs to be executed:

[0270] First, the network side needs to determine whether the current cell and the current terminal device are configured with a periodic CSI retransmission mechanism. There are several ways to configure whether a periodic CSI retransmission mechanism is supported:

[0271] (1) It can be pre-configured by the network side through RRC. If the RRC signaling is configured with a periodic CSI retransmission mechanism, the terminal device will support the CSI retransmission mechanism in each periodic CSI feedback; otherwise, it will not support it.

[0272] (2) It can be pre-configured by the network side through RRC and activated and deactivated by MAC CE or DCI. That is, when MAC CE or DCI is activated, the periodic CSI feedback supports the CSI retransmission mechanism, and when MAC CE or DCI is deactivated, the periodic CSI feedback does not support the CSI retransmission mechanism.

[0273] (3) It can be pre-configured by the network side through MAC CE. If the MAC CE signaling is configured with a periodic CSI retransmission mechanism, the terminal device will support the CSI retransmission mechanism in each periodic CSI feedback cycle; otherwise, it will not support it.

[0274] (4) It can be pre-configured by the network side through MAC CE and activated and deactivated by DCI. That is, when DCI is activated, the periodic CSI feedback supports the CSI retransmission mechanism, and when DCI is deactivated, the periodic CSI feedback does not support the CSI retransmission mechanism.

[0275] (5) It can be dynamically configured by the network side through DCI. DCI indicates whether the CSI retransmission mechanism in the periodic CSI feedback is supported or not.

[0276] (6) Other possible higher-level or physical-level signaling, or combinations of other signaling and the above signaling, are not listed one by one, but are all within the scope of this application.

[0277] If both the cell base station and the terminal device support and activate the periodic CSI retransmission mechanism, then step S3 in Application Example 1 is triggered, sending the first indication information to the terminal device and configuring uplink retransmission resources. Otherwise, no operation is performed.

[0278] Configure uplink retransmission resources, including at least:

[0279] (1) Configuring uplink channel resources for retransmission: Since periodic CSI feedback is performed via PUCCH, retransmissions in periodic CSI feedback can be performed via PUCCH or PUSCH. If retransmission is performed via PUCCH, only the reporting configuration for CSI retransmission is supported to be the same as the current periodic CSI feedback, i.e., the retransmission configuration reuse indication in the first indication information in S3 is 0. If retransmission is performed via PUSCH, a portion of uplink PUSCH resources needs to be allocated for CSI retransmission, and this needs to be indicated through the first indication information.

[0280] (2) Configure retransmission time slot offset: The uplink time slot occupied by the retransmission CSI in the periodic CSI feedback is offset by δ relative to the time slot of the current periodic CSI report. For shorter CSI feedback periods, it is recommended to only support δ=1, that is, CSI feedback is performed on the adjacent uplink time slot; for longer CSI feedback periods, different time slot offset configurations can be supported, such as δ=1, 3, 5, etc.

[0281] For retransmissions of periodic CSI feedback, the maximum number of retransmissions is 1, meaning only one retransmission is supported. After receiving the first indication information and uplink retransmission resource configuration, the terminal device retransmits and reports the CSI according to the indication information. The beneficial effect of this application example is that it can support CSI retransmission under periodic CSI reporting. Compared with the NR system, which can only use the CSI reporting results of the previous period, this method allows the network side to obtain more accurate CSI reports, improving downlink precoding and downlink transmission performance.

[0282] Application Example 3

[0283] This application example provides a retransmission mechanism based on an aperiodic CSI feedback method.

[0284] In NR systems, aperiodic CSI reporting is supported. This reporting is configured and triggered using MAC CE combined with DCI triggering and is reported via PUSCH. For NR systems, when a CSI report cannot be received correctly, the network side schedules new CSI-RS measurements and aperiodic CSI feedback resources.

[0285] This application example provides a retransmission mechanism under aperiodic CSI feedback based on joint source channel coding. Unlike the retransmission mechanism under periodic CSI feedback, this mechanism is enabled by default and does not require configuration or dynamic activation by higher-layer or physical-layer signaling. Specifically, when the network receives aperiodic CSI feedback based on joint source channel coding, the network directly sends a first indication message to the terminal device and configures uplink retransmission resources.

[0286] Configure uplink retransmission resources, including at least:

[0287] (1) Configuring uplink channel resources for retransmission: Since aperiodic CSI feedback only occurs on the PUSCH, aperiodic CSI retransmission is also configured only on the PUSCH. The network needs to allocate a portion of uplink PUSCH resources for CSI retransmission and indicate this through the first indication information;

[0288] (2) Configure retransmission time slot offset: For the retransmission of non-periodic CSI, the time slot offset δ relative to the current non-periodic CSI report can be configured in various ways, such as δ = 1, 3, 5, etc.

[0289] For non-periodic CSI feedback and retransmission mechanisms, the maximum number of retransmissions can be specifically implemented by the network side. For example, when the detected uplink signal-to-noise ratio is high, the maximum number of retransmissions is 1 to reduce retransmission overhead; when the detected uplink signal-to-noise ratio is low, the maximum number of retransmissions is 4 to ensure higher CSI feedback accuracy.

[0290] After receiving the first instruction information and uplink retransmission resource configuration, the terminal device retransmits and reports CSI according to the instruction information. The beneficial effects of this application example are: it can realize CSI retransmission under non-periodic CSI reporting. Compared with the NR system, which needs to reschedule CSI-RS resources for measurement and then perform CSI feedback, this scheme can save CSI-RS measurement overhead, and enable the network side to obtain more accurate CSI reports faster, reduce feedback latency, and improve downlink precoding and downlink transmission performance.

[0291] Application Example 4

[0292] This application example provides a retransmission mechanism based on a semi-persistent CSI feedback method.

[0293] In NR systems, there is also a semi-persistent CSI reporting method, which falls between periodic and aperiodic CSI reporting. CSI reporting occurs at regular intervals, between activation and deactivation. Semi-persistent CSI can be reported via PUSCH or PUCCH.

[0294] This application example provides a retransmission mechanism based on semi-persistent CSI feedback. After semi-persistent CSI reporting is activated, the terminal device performs CSI feedback at certain intervals. If the system supports a periodic CSI feedback retransmission mechanism, it automatically supports the semi-persistent CSI feedback retransmission mechanism. The configuration of this retransmission mechanism is similar to that of the periodic CSI feedback retransmission mechanism until CSI reporting is deactivated.

[0295] The specific process and beneficial effects of this application example will not be elaborated here.

[0296] Figure 10 is a schematic diagram of the composition structure of a terminal device 1000 according to an embodiment of this application, including:

[0297] The first communication unit 1001 is configured to send a first reporting information to the network device, and receive a first indication information from the network device, and send a second reporting information to the network device; wherein, the first reporting information includes CSI; the first indication information is used to instruct the terminal device to retransmit CSI; and the second reporting information includes CSI.

[0298] In some embodiments, the first reporting information and / or the second reporting information are used to report CSI based on the codebook or AI.

[0299] In some embodiments, as shown in FIG10, the terminal device 1000 further includes a first processing unit 1002, configured to perform one of the following:

[0300] The CSI is processed into the first output information based on the first AI model; the first output information is then mapped onto the uplink channel physical resources after channel coding and modulation.

[0301] The CSI is processed into a second output information based on the second AI model; the second output information is modulated and mapped onto the uplink channel physical resources for transmission.

[0302] The CSI is processed into third output information based on the third AI model; the third output information is used to map to uplink channel physical resources for transmission.

[0303] In some embodiments, the first reporting information is used by the network device to determine whether to retransmit the CSI.

[0304] In some embodiments, the first reporting information is used by the network device to determine the input information of the fourth AI model; the input information of the fourth AI model is used by the network device to recover CSI based on the fourth AI model and to determine whether to retransmit CSI based on the distribution characteristics of the input information of the fourth AI model.

[0305] In some embodiments, the first reporting information is used by the network device to obtain the log-likelihood ratio sequence through demodulation; the log-likelihood ratio sequence is used by the network device to determine whether to retransmit CSI based on the statistical value of the distribution characteristics of the log-likelihood ratio sequence.

[0306] In some embodiments, the first reporting information is used by the network device to obtain received symbols through channel equalization; the received symbols are used by the network device to compare the distribution of the received symbols with the distribution of the unit circle or the third output information to determine whether to retransmit CSI; wherein, the third output information is the output information of the third AI model corresponding to the fourth AI model in the terminal device.

[0307] In some embodiments, the first communication unit 1001 is further configured to send a sample set of output information of the third AI model to the network device; wherein the sample set of output information is used by the network device to determine the distribution of the third output information.

[0308] In some embodiments, the first indication information is used to indicate at least one of the following: whether to send the second reporting information based on the CSI reporting configuration of the first reporting information; the CSI reporting configuration of the second reporting information; uplink channel resources for transmitting the second reporting information; and time domain information for transmitting the second reporting information.

[0309] In some embodiments, the CSI reporting configuration for the second reporting information includes at least one of the following: a processing method for CSI when sending the second reporting information; and configuration information related to the processing method.

[0310] In some embodiments, the configuration information related to the processing method includes at least one of the following:

[0311] The length of the output information of the AI ​​model used to implement the processing method;

[0312] Constraint information for the AI ​​model used to implement the processing method; wherein, the constraint information is used to indicate whether the output information of the AI ​​model needs to satisfy the constant modulus constraint;

[0313] The modulation methods included in the processing method.

[0314] In some embodiments, uplink channel resources include PUSCH resources or PUCCH resources.

[0315] In some embodiments, when the first reported information is used for periodic CSI feedback or semi-persistent CSI feedback, the uplink channel resources include PUSCH resources or PUCCH resources.

[0316] In some embodiments, when the first reported information is used for aperiodic CSI feedback, the uplink channel resources include PUSCH resources.

[0317] In some embodiments, the time-domain information includes the time slot offset between the second reported information and the first reported information.

[0318] In some embodiments, when the first reporting information is used for non-periodic CSI feedback, the first communication unit 1001 is used to receive multiple first indication information, and each time the first indication information is received, it sends a second reporting information to the network device based on the first indication information.

[0319] In some embodiments, as shown in FIG11, the terminal device 1000 further includes a second processing unit 1003, configured to store the CSI in a first storage unit after sending the first reporting information; and to read the CSI in the first storage unit and send the second reporting information to the network device based on the CSI through the first communication unit 1001 upon receiving the first indication information from the network device.

[0320] In some embodiments, the second processing unit 1003 is configured to clear the first storage unit when it is determined that no more first instruction information for the first reporting information is received.

[0321] In some embodiments, the first communication unit 1001 is configured to send second indication information to a network device; wherein the second indication information is used by the network device to determine that the terminal device has a first capability; the first capability includes at least one of the following: the terminal device supports CSI feedback and retransmission based on AI; the terminal device supports deploying an AI model for CSI feedback; the terminal device supports receiving the first indication information; the terminal device supports caching CSI for retransmission.

[0322] In some embodiments, the second reporting information is used by the network device to merge the CSI recovered based on the second reporting information with the CSI recovered based on the first reporting information, or to replace the CSI recovered based on the first reporting information with the CSI recovered based on the second reporting information.

[0323] Figure 12 is a schematic diagram of the composition structure of a network device 1200 according to an embodiment of the present application, including:

[0324] The second communication unit 1201 is used to receive first reporting information from the terminal device and send first instruction information to the terminal device; wherein, the first reporting information includes CSI; the first instruction information is used to instruct the terminal device to send second reporting information to retransmit CSI.

[0325] In some embodiments, the first reporting information and / or the second reporting information are used to report CSI based on the codebook or AI.

[0326] In some embodiments, as shown in FIG12, the network device 1200 further includes a third processing unit 1202 for performing one of the following:

[0327] The first and / or second reported information is processed based on the codebook to recover CSI;

[0328] The received symbols corresponding to the first reported information and / or the second reported information are demodulated and the channel is decoded to obtain the first input information. The first input information is then processed using the fifth AI model to recover the CSI.

[0329] The received symbols corresponding to the first reported information and / or the second reported information are demodulated to obtain the second input information, and the second input information is processed by the sixth AI model to recover the CSI;

[0330] The received symbols corresponding to the first and / or second reported information are used as the third input information, and the seventh AI model is used to process the third input information to recover CSI.

[0331] In some embodiments, the first condition is related to at least one of the following: uplink channel quality; first reporting information.

[0332] In some embodiments, as shown in FIG13, the network device 1200 further includes a fourth processing unit 1203, configured to determine the input information of the fourth AI model based on the received symbol corresponding to the first reported information, and to determine whether the first condition is met based on the distribution characteristics of the input information; wherein, the fourth AI model is used to recover CSI.

[0333] In some embodiments, the input information includes a log-likelihood ratio sequence; the fourth processing unit 1203 is used to determine whether the first condition is met based on the statistical value of the distribution characteristics of the log-likelihood ratio sequence.

[0334] In some embodiments, the input information includes received symbols; the fourth processing unit 1203 is used to compare the distribution of the received symbols with the distribution of the unit circle or the third output information to determine whether the first condition is met; wherein, the third output information is the output information of the third AI model corresponding to the fourth AI model in the terminal device.

[0335] In some embodiments, the second communication unit 1201 is further configured to receive a sample set of output information from a third AI model from a terminal device; wherein the sample set of output information from the third AI model is used by the network device to determine the distribution of the third output information.

[0336] In some embodiments, the first indication information is used to indicate at least one of the following: whether to send the second reporting information based on the CSI reporting configuration of the first reporting information; the CSI reporting configuration of the second reporting information; uplink channel resources for transmitting the second reporting information; and time domain information for transmitting the second reporting information.

[0337] In some embodiments, the CSI reporting configuration for the second reporting information includes at least one of the following: a processing method for CSI when sending the second reporting information; and configuration information related to the processing method.

[0338] In some embodiments, configuration information related to the processing method includes at least one of the following: the length of the output information of the AI ​​model used to implement the processing method; constraint information of the AI ​​model; wherein the constraint information is used to indicate whether the output information of the AI ​​model needs to satisfy constant modulus constraints; and the modulation method included in the processing method.

[0339] In some embodiments, uplink channel resources include PUSCH resources or PUCCH resources.

[0340] In some embodiments, when the first reported information is used for periodic CSI feedback or semi-persistent CSI feedback, the uplink channel resources include PUSCH resources or PUCCH resources.

[0341] In some embodiments, when the first reported information is used for aperiodic CSI feedback, the uplink channel resources include PUSCH resources.

[0342] In some embodiments, the time-domain information includes the time slot offset between the second reported information and the first reported information.

[0343] In some embodiments, as shown in FIG14, the network device 1200 further includes a fifth processing unit 1204, which is used to determine whether the terminal device is configured with a periodic CSI retransmission mechanism; if the terminal device is configured with a periodic CSI retransmission mechanism, the first indication information is sent through the second communication unit 1201.

[0344] In some embodiments, the second communication unit 1201 is configured to send a maximum of N first indication messages; wherein N is determined based on the uplink channel quality.

[0345] In some embodiments, when the first reporting information is used for non-periodic CSI feedback, the second communication unit 1201 is used to stop sending the first indication information if it is determined that the first condition is not met or the number of retransmissions reaches N.

[0346] In some embodiments, as shown in FIG15, the network device 1200 further includes a sixth processing unit 1205, configured to, after receiving the first reporting information, store the CSI recovered based on the first reporting information in a second storage unit; and, upon receiving the second reporting information, store the second reporting information or the CSI recovered based on the second reporting information in the second storage unit.

[0347] In some embodiments, the sixth processing unit 1205 is configured to clear the first storage unit when it is determined that the first condition is not met or the number of retransmissions reaches a preset number.

[0348] In some embodiments, the second communication unit 1201 is further configured to receive second indication information from the terminal device; wherein the second indication information is used by the network device to determine that the terminal device has a first capability; the first capability includes at least one of the following: the terminal device supports CSI feedback and retransmission based on AI; the terminal device supports deploying an AI model for CSI feedback; the terminal device supports receiving the first indication information; the terminal device supports caching CSI for retransmission.

[0349] In some embodiments, as shown in FIG16, the network device 1200 further includes a seventh processing unit 1206, configured to merge the CSI recovered based on the second reported information with the CSI recovered based on the first reported information; or, to replace the CSI recovered based on the first reported information with the CSI recovered based on the second reported information.

[0350] The device in this application embodiment can realize the corresponding functions of each device in the aforementioned CSI feedback method embodiments. The processes, functions, implementation methods, and beneficial effects of each module (sub-module, unit, or component, etc.) in this device can be found in the corresponding descriptions in the above method embodiments, and will not be repeated here. It should be noted that the functions described for each module (sub-module, unit, or component, etc.) in the device of this application embodiment can be implemented by different modules (sub-modules, units, or components, etc.) or by the same module (sub-module, unit, or component, etc.).

[0351] Figure 17 is a schematic structural diagram of a communication device 1700 according to an embodiment of this application. The communication device 1700 includes a processor 1710, which can call and run computer programs from memory to enable the communication device 1700 to implement the methods in the embodiments of this application.

[0352] In one possible implementation, the communication device 1700 may further include a memory 1720. The processor 1710 can retrieve and run computer programs from the memory 1720 to enable the communication device 1700 to implement the methods described in this embodiment. The memory 1720 may be a separate device independent of the processor 1710, or it may be integrated into the processor 1710. In one possible implementation, the communication device 1700 may further include a transceiver 1730. The processor 1710 can control the transceiver 1730 to communicate with other devices; specifically, it can send information or data to other devices, or receive information or data sent by other devices. The transceiver 1730 may include a transmitter and a receiver. The transceiver 1730 may further include antennas, and the number of antennas may be one or more.

[0353] In one embodiment, the communication device 1700 may be a network device in the embodiments of this application, and the communication device 1700 may implement the corresponding processes implemented by the network device in the various methods of the embodiments of this application. For the sake of brevity, it will not be described in detail here.

[0354] In one embodiment, the communication device 1700 may be a terminal device in the embodiments of this application, and the communication device 1700 may implement the corresponding processes implemented by the terminal device in the various methods of the embodiments of this application. For the sake of brevity, it will not be described in detail here.

[0355] Figure 18 is a schematic structural diagram of a chip 1800 according to an embodiment of this application. The chip 1800 includes a processor 1810, which can call and run computer programs from memory to implement the methods in the embodiments of this application. In one possible implementation, the chip 1800 may further include a memory 1820. The processor 1810 can call and run computer programs from the memory 1820 to implement the methods executed by various devices in the embodiments of this application. The memory 1820 may be a separate device independent of the processor 1810, or it may be integrated into the processor 1810. In one possible implementation, the chip 1800 may further include an input interface 1830. The processor 1810 can control the input interface 1830 to communicate with other devices or chips; specifically, it can acquire information or data sent by other devices or chips. In one possible implementation, the chip 1800 may further include an output interface 1840. The processor 1810 can control the output interface 1840 to communicate with other devices or chips; specifically, it can output information or data to other devices or chips. In one possible implementation, the chip can be applied to various devices in the embodiments of this application, and the chip can implement the corresponding processes implemented by each device in the various methods of the embodiments of this application. For simplicity, further details are omitted here. It should be understood that the chip mentioned in the embodiments of this application can also be called a system-on-a-chip (SoC), system-on-a-chip (SoC), chip system, or system-on-a-chip, etc.

[0356] The processors mentioned above can be general-purpose processors, digital signal processors, off-the-shelf programmable gate arrays, application-specific integrated circuits (ASICs), or other programmable logic devices, transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processors mentioned above can be microprocessors or any conventional processor. The memory mentioned above can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory.

[0357] It should be understood that the above-described memory is exemplary but not a limiting description. For example, the memory in the embodiments of this application may also be static random access memory, dynamic random access memory, etc.

[0358] Figure 19 is a schematic block diagram of a communication system 1900 according to an embodiment of this application. The communication system 1900 includes a terminal device 1910 and a network device 1920. In the above embodiments, it can be implemented entirely or partially by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (such as a hard disk) or a semiconductor medium (such as a solid-state drive).

[0359] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0360] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0361] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A channel state information (CSI) feedback method, comprising: The terminal device sends a first reporting information to the network device; wherein, the first reporting information includes CSI; The terminal device receives a first indication information from the network device and sends a second reporting information to the network device; wherein, the first indication information is used to instruct the terminal device to retransmit the CSI; and the second reporting information includes the CSI.

2. The method according to claim 1, wherein, The first reported information and / or the second reported information are used to report the CSI based on the codebook or artificial intelligence (AI).

3. The method according to claim 2, wherein, The AI-based reporting of the CSI includes one of the following: The CSI is processed into first output information based on the first AI model; wherein the first output information is mapped onto uplink channel physical resources and transmitted after channel coding and modulation. The CSI is processed into second output information based on the second AI model; wherein, the second output information is modulated and mapped onto uplink channel physical resources for transmission; The CSI is processed into third output information based on the third AI model; wherein the third output information is used to map onto uplink channel physical resources for transmission.

4. The method according to any one of claims 1-3, wherein, The first reported information is used by the network device to determine whether to retransmit the CSI.

5. The method according to claim 4, wherein, The first reported information is used by the network device to determine the input information for the fourth AI model; The input information of the fourth AI model is used by the network device to recover the CSI based on the fourth AI model and to determine whether to retransmit the CSI based on the distribution characteristics of the input information of the fourth AI model.

6. The method according to claim 5, wherein, The first reported information is used by the network device to obtain the log-likelihood ratio sequence through demodulation; The log-likelihood ratio sequence is used by the network device to determine whether to retransmit the CSI based on the statistical values ​​of the distribution characteristics of the log-likelihood ratio sequence.

7. The method according to claim 5, wherein, The first reported information is used by the network device to obtain received symbols through channel equalization; The received symbols are used by the network device to compare the distribution of the received symbols with the distribution of the unit circle or the third output information to determine whether to retransmit the CSI; wherein, the third output information is the output information of the third AI model corresponding to the fourth AI model in the terminal device.

8. The method according to claim 7, wherein, The method further includes: The terminal device sends the output information sample set of the third AI model to the network device; wherein, the output information sample set is used by the network device to determine the distribution of the third output information.

9. The method according to any one of claims 1-8, wherein, The first indication information is used to indicate at least one of the following: Whether to send the second reporting information based on the CSI reporting configuration of the first reporting information; The CSI reporting configuration for the second reported information; Uplink channel resources used for transmitting the second reported information; Time-domain information used to transmit the second reported information.

10. The method according to claim 9, wherein, The CSI reporting configuration for the second reported information includes at least one of the following: The processing method for the CSI when sending the second reporting information; Configuration information related to the processing method.

11. The method according to claim 10, wherein, Configuration information related to the processing method includes at least one of the following: The length of the output information of the AI ​​model used to implement the processing method; Constraint information for the AI ​​model used to implement the processing method; wherein, the constraint information is used to indicate whether the output information of the AI ​​model needs to satisfy the constant modulus constraint; The modulation method included in the processing method.

12. The method according to any one of claims 9-11, wherein, The uplink channel resources include Physical Uplink Shared Channel (PUSCH) resources or Physical Uplink Control Channel (PUCCH) resources.

13. The method according to any one of claims 9-12, wherein, When the first reported information is used for periodic CSI feedback or semi-persistent CSI feedback, the uplink channel resources include PUSCH resources or PUCCH resources.

14. The method according to any one of claims 9-13, wherein, When the first reported information is used for aperiodic CSI feedback, the uplink channel resources include PUSCH resources.

15. The method according to any one of claims 9-14, wherein, The time-domain information includes the time slot offset between the second reported information and the first reported information.

16. The method according to any one of claims 1-15, wherein, When the first reported information is used for non-periodic CSI feedback, the terminal device can receive the first indication information multiple times; The terminal device receives first indication information from the network device and sends second reporting information to the network device, including: Each time the terminal device receives the first indication information, it sends the second reporting information to the network device based on the first indication information.

17. The method according to any one of claims 1-16, wherein, The terminal device receives first indication information from the network device and sends second reporting information to the network device, including: After sending the first reporting information, the terminal device stores the CSI in the first storage unit; Upon receiving a first indication from the network device, the terminal device reads the CSI from the first storage unit and sends a second reporting information to the network device based on the CSI.

18. The method according to claim 17, wherein, The method further includes: When the terminal device determines that it will no longer receive the first instruction information for the first reported information, it clears the first storage unit.

19. The method according to any one of claims 1-18, wherein, The method further includes: The terminal device sends a second indication message to the network device; wherein the second indication message is used by the network device to determine that the terminal device possesses a first capability; the first capability includes at least one of the following: The terminal device supports CSI feedback and retransmission based on AI; The terminal device supports the deployment of AI models for CSI feedback; The terminal device supports receiving the first indication information; The terminal device supports caching CSI for retransmission.

20. The method according to any one of claims 1-19, wherein, The second reporting information is used by the network device to merge the CSI recovered based on the second reporting information with the CSI recovered based on the first reporting information, or to replace the CSI recovered based on the first reporting information with the CSI recovered based on the second reporting information.

21. A CSI feedback method, comprising: The network device receives first reported information from the terminal device; wherein, the first reported information includes CSI; The network device sends a first indication message to the terminal device; wherein the first indication message is used to instruct the terminal device to send a second reporting message to retransmit the CSI.

22. The method according to claim 21, wherein, The first reported information and / or the second reported information are used to report the CSI based on the codebook or artificial intelligence (AI).

23. The method according to claim 21 or 22, wherein, The method also includes one of the following: The network device processes the first reported information and / or the second reported information based on the codebook to recover the CSI; The network device demodulates and decodes the received symbols corresponding to the first reported information and / or the second reported information to obtain the first input information, and uses the fifth AI model to process the first input information to recover the CSI; The network device demodulates the received symbols corresponding to the first reported information and / or the second reported information to obtain the second input information, and uses the sixth AI model to process the second input information to recover the CSI; The network device uses the received symbols corresponding to the first reported information and / or the second reported information as the third input information, and processes the third input information using the seventh AI model to recover the CSI.

24. The method according to any one of claims 21-23, wherein, The first indication information is sent if a first condition is met, and the first condition is related to at least one of the following: Uplink channel quality; The first reported information.

25. The method according to claim 24, wherein, The method further includes: The network device determines the input information of the fourth AI model based on the received symbol corresponding to the first reported information; wherein, the fourth AI model is used to recover the CSI; The network device determines whether the first condition is met based on the distribution characteristics of the input information.

26. The method according to claim 25, wherein, The input information includes a log-likelihood ratio sequence; The network device determines whether the first condition is met based on the distribution characteristics of the input information, including: The network device determines whether the first condition is met based on the statistical value of the distribution characteristics of the log-likelihood ratio sequence.

27. The method according to claim 25, wherein, The input information includes the received symbol; The network device determines whether the first condition is met based on the distribution characteristics of the input information, including: The network device compares the distribution of the received symbols with the distribution of the unit circle or the third output information to determine whether the first condition is met; wherein, the third output information is the output information of the third AI model corresponding to the fourth AI model in the terminal device.

28. The method according to claim 27, wherein, The method further includes: The network device receives a sample set of output information from the third AI model from the terminal device; wherein the sample set of output information from the third AI model is used by the network device to determine the distribution of the third output information.

29. The method according to any one of claims 21-28, wherein, The first indication information is used to indicate at least one of the following: Whether to send the second reporting information based on the CSI reporting configuration of the first reporting information; The CSI reporting configuration for the second reported information; Uplink channel resources used for transmitting the second reported information; Time-domain information used to transmit the second reported information.

30. The method according to claim 29, wherein, The CSI reporting configuration for the second reported information includes at least one of the following: The processing method for CSI when sending the second reporting information; Configuration information related to the processing method.

31. The method according to claim 30, wherein, Configuration information related to the processing method includes at least one of the following: The length of the output information of the AI ​​model used to implement the processing method; The constraint information of the AI ​​model; wherein, the constraint information is used to indicate whether the output information of the AI ​​model needs to satisfy the constant modulus constraint; The modulation method included in the processing method.

32. The method according to any one of claims 29-31, wherein, The uplink channel resources include PUSCH resources or PUCCH resources.

33. The method according to any one of claims 29-32, wherein, When the first reported information is used for periodic CSI feedback or semi-persistent CSI feedback, the uplink channel resources include PUSCH resources or PUCCH resources.

34. The method according to any one of claims 29-33, wherein, When the first reported information is used for aperiodic CSI feedback, the uplink channel resources include PUSCH resources.

35. The method according to any one of claims 29-34, wherein, The time-domain information includes the time slot offset between the second reported information and the first reported information.

36. The method according to any one of claims 21-35, wherein, The network device sends a first indication message to the terminal device, including: The network device determines whether the terminal device is configured with a periodic CSI retransmission mechanism. If the terminal device is configured with a periodic CSI retransmission mechanism, the network device sends the first indication information.

37. The method according to any one of claims 21-35, wherein, The network device can send a maximum of N first indication messages; where N is determined based on the uplink channel quality.

38. The method according to claim 37, wherein, When the first reported information is used for non-periodic CSI feedback, the method further includes: If the first condition is not met or the number of retransmissions reaches N, the network device stops sending the first indication information.

39. The method according to any one of claims 21-38, wherein, The method further includes: After receiving the first reported information, the network device stores the CSI recovered based on the first reported information in the second storage unit; Upon receiving the second reported information, the network device stores the second reported information or the CSI recovered based on the second reported information in the second storage unit.

40. The method according to claim 39, wherein, The method further includes: When the network device determines that the first condition is not met or the number of retransmissions reaches a preset number, it clears the second storage unit.

41. The method according to any one of claims 21-40, wherein, The method further includes: The network device receives second indication information from the terminal device; wherein the second indication information is used by the network device to determine that the terminal device possesses a first capability; the first capability includes at least one of the following: The terminal device supports CSI feedback and retransmission based on AI; The terminal device supports the deployment of AI models for CSI feedback; The terminal device supports receiving the first indication information; The terminal device supports caching CSI for retransmission.

42. The method according to any one of claims 21-41, wherein, The method further includes: The network device merges the CSI recovered based on the second reported information with the CSI recovered based on the first reported information; or, The network device uses the CSI recovered based on the second reported information instead of the CSI recovered based on the first reported information.

43. A terminal device, comprising: A first communication unit is configured to send a first reporting information to a network device, and receive a first indication information from the network device, and send a second reporting information to the network device; wherein the first reporting information includes CSI; the first indication information is used to instruct the terminal device to retransmit the CSI; and the second reporting information includes the CSI.

44. The terminal device according to claim 43, wherein, The first and / or the second reporting information are used to report the CSI based on the codebook or AI.

45. The terminal device according to claim 44, wherein, The terminal device further includes a first processing unit, configured to perform one of the following: The CSI is processed into first output information based on the first AI model; wherein the first output information is mapped onto uplink channel physical resources and transmitted after channel coding and modulation. The CSI is processed into second output information based on the second AI model; wherein, the second output information is modulated and mapped onto uplink channel physical resources for transmission; The CSI is processed into third output information based on the third AI model; wherein the third output information is used to map onto uplink channel physical resources for transmission.

46. ​​The terminal device according to any one of claims 43-45, wherein, The first reported information is used by the network device to determine whether to retransmit the CSI.

47. The terminal device according to claim 46, wherein, The first reported information is used by the network device to determine the input information of the fourth AI model; the input information of the fourth AI model is used by the network device to recover the CSI based on the fourth AI model and to determine whether to retransmit the CSI based on the distribution characteristics of the input information of the fourth AI model.

48. The terminal device according to claim 47, wherein, The first reported information is used by the network device to obtain a log-likelihood ratio sequence through demodulation; the log-likelihood ratio sequence is used by the network device to determine whether to retransmit the CSI based on the statistical value of the distribution characteristics of the log-likelihood ratio sequence.

49. The terminal device according to claim 47, wherein, The first reported information is used by the network device to obtain received symbols through channel equalization; the received symbols are used by the network device to compare the distribution of the received symbols with the distribution of the unit circle or the third output information to determine whether to retransmit the CSI; wherein, the third output information is the output information of the third AI model corresponding to the fourth AI model in the terminal device.

50. The terminal device according to claim 49, wherein, The first communication unit is further configured to send the output information sample set of the third AI model to the network device; wherein the output information sample set is used by the network device to determine the distribution of the third output information.

51. The terminal device according to any one of claims 43-50, wherein, The first indication information is used to indicate at least one of the following: whether to send the second reporting information based on the CSI reporting configuration of the first reporting information; the CSI reporting configuration of the second reporting information; and uplink channel resources for transmitting the second reporting information. Time-domain information used to transmit the second reported information.

52. The terminal device according to claim 51, wherein, The CSI reporting configuration for the second reported information includes at least one of the following: the processing method for CSI when sending the second reported information; and configuration information related to the processing method.

53. The terminal device according to claim 52, wherein, Configuration information related to the processing method includes at least one of the following: The length of the output information of the AI ​​model used to implement the processing method; Constraint information for the AI ​​model used to implement the processing method; wherein, the constraint information is used to indicate whether the output information of the AI ​​model needs to satisfy the constant modulus constraint; The modulation method included in the processing method.

54. The terminal device according to any one of claims 51-53, wherein, The uplink channel resources include PUSCH resources or PUCCH resources.

55. The terminal device according to any one of claims 51-54, wherein, When the first reported information is used for periodic CSI feedback or semi-persistent CSI feedback, the uplink channel resources include PUSCH resources or PUCCH resources.

56. The terminal device according to any one of claims 51-55, wherein, When the first reported information is used for aperiodic CSI feedback, the uplink channel resources include PUSCH resources.

57. The terminal device according to any one of claims 51-56, wherein, The time-domain information includes the time slot offset between the second reported information and the first reported information.

58. The terminal device according to any one of claims 43-57, wherein, When the first reporting information is used for non-periodic CSI feedback, the first communication unit is used to receive multiple first indication information, and each time the first indication information is received, it sends the second reporting information to the network device based on the first indication information.

59. The terminal device according to any one of claims 43-58, wherein, The terminal device further includes a second processing unit, used to store the CSI in a first storage unit after sending the first reporting information; Upon receiving a first indication from the network device, the CSI is read from the first storage unit, and a second reporting information is sent to the network device based on the CSI via the first communication unit.

60. The terminal device according to claim 59, wherein, The second processing unit is configured to clear the first storage unit when it determines that it will no longer receive a first instruction message for the first reported information.

61. The terminal device according to any one of claims 43-60, wherein, The first communication unit is configured to send second indication information to the network device; wherein the second indication information is used by the network device to determine that the terminal device possesses a first capability; the first capability includes at least one of the following: the terminal device supports CSI feedback and retransmission based on AI; the terminal device supports deploying an AI model for CSI feedback; the terminal device supports receiving the first indication information; the terminal device supports caching CSI for retransmission.

62. The terminal device according to any one of claims 43-61, wherein, The second reporting information is used by the network device to merge the CSI recovered based on the second reporting information with the CSI recovered based on the first reporting information, or to replace the CSI recovered based on the first reporting information with the CSI recovered based on the second reporting information.

63. A network device, comprising: The second communication unit is configured to receive first reported information from the terminal device and send first indication information to the terminal device; wherein the first reported information includes CSI; and the first indication information is configured to instruct the terminal device to send second reported information to retransmit the CSI.

64. The network device according to claim 63, wherein, The first and / or the second reporting information are used to report the CSI based on the codebook or AI.

65. The network device according to claim 63 or 64, wherein, The network device further includes a third processing unit for performing one of the following: The first reported information and / or the second reported information are processed based on the codebook to recover the CSI; The received symbols corresponding to the first reported information and / or the second reported information are demodulated and the channel is decoded to obtain the first input information. The first input information is then processed using the fifth AI model to recover the CSI. The received symbols corresponding to the first reported information and / or the second reported information are demodulated to obtain the second input information, and the second input information is processed by the sixth AI model to recover the CSI; The received symbols corresponding to the first reported information and / or the second reported information are used as the third input information, and the seventh AI model is used to process the third input information to recover the CSI.

66. The network device according to any one of claims 63-65, wherein, The first indication information is sent if a first condition is met, the first condition being related to at least one of the following: uplink channel quality; the first reporting information.

67. The network device according to claim 66, wherein, The network device further includes a fourth processing unit, configured to determine the input information of a fourth AI model based on the received symbol corresponding to the first reported information, and to determine whether the first condition is met based on the distribution characteristics of the input information; wherein, the fourth AI model is used to recover the CSI.

68. The network device according to claim 67, wherein, The input information includes a log-likelihood ratio sequence; the fourth processing unit is used to determine whether the first condition is met based on the statistical value of the distribution characteristics of the log-likelihood ratio sequence.

69. The network device according to claim 67, wherein, The input information includes the received symbols; the fourth processing unit is used to compare the distribution of the received symbols with the distribution of the unit circle or the third output information to determine whether the first condition is met; wherein, the third output information is the output information of the third AI model corresponding to the fourth AI model in the terminal device.

70. The network device according to claim 69, wherein, The second communication unit is further configured to receive a sample set of output information from the third AI model of the terminal device; wherein the sample set of output information from the third AI model is used by the network device to determine the distribution of the third output information.

71. The network device according to any one of claims 63-70, wherein, The first indication information is used to indicate at least one of the following: whether to send the second reporting information based on the CSI reporting configuration of the first reporting information; the CSI reporting configuration of the second reporting information; and uplink channel resources for transmitting the second reporting information. Time-domain information used to transmit the second reported information.

72. The network device according to claim 71, wherein, The CSI reporting configuration for the second reported information includes at least one of the following: the processing method for CSI when sending the second reported information; and configuration information related to the processing method.

73. The network device according to claim 72, wherein, Configuration information related to the processing method includes at least one of the following: the length of the output information of the AI ​​model used to implement the processing method; constraint information of the AI ​​model; wherein the constraint information is used to indicate whether the output information of the AI ​​model needs to satisfy constant modulus constraints; and the modulation method included in the processing method.

74. The network device according to any one of claims 71-73, wherein, The uplink channel resources include PUSCH resources or PUCCH resources.

75. The network device according to any one of claims 71-74, wherein, When the first reported information is used for periodic CSI feedback or semi-persistent CSI feedback, the uplink channel resources include PUSCH resources or PUCCH resources.

76. The network device according to any one of claims 71-75, wherein, When the first reported information is used for aperiodic CSI feedback, the uplink channel resources include PUSCH resources.

77. The network device according to any one of claims 71-76, wherein, The time-domain information includes the time slot offset between the second reported information and the first reported information.

78. The network device according to any one of claims 63-77, wherein, The network device further includes a fifth processing unit, used to determine whether the terminal device is configured with a periodic CSI retransmission mechanism; if the terminal device is configured with a periodic CSI retransmission mechanism, the first indication information is sent through the second communication unit.

79. The network device according to any one of claims 63-77, wherein, The second communication unit is used to send a maximum of N first indication messages; where N is determined based on the uplink channel quality.

80. The network device according to claim 79, wherein, When the first reported information is used for non-periodic CSI feedback, the second communication unit is used to stop sending the first indication information if it is determined that the first condition is not met or the number of retransmissions reaches N.

81. The network device according to any one of claims 63-80, wherein, The network device further includes a sixth processing unit, which, after receiving the first reported information, stores the CSI recovered based on the first reported information in a second storage unit; Upon receiving the second reported information, the second reported information or the CSI recovered based on the second reported information is stored in the second storage unit.

82. The network device according to claim 81, wherein, The sixth processing unit is used to clear the second storage unit when it is determined that the first condition is not met or the number of retransmissions reaches a preset number.

83. The network device according to any one of claims 63-82, wherein, The second communication unit is further configured to receive second indication information from the terminal device; wherein the second indication information is used by the network device to determine that the terminal device possesses a first capability; the first capability includes at least one of the following: the terminal device supports CSI feedback and retransmission based on AI; the terminal device supports deploying an AI model for CSI feedback; the terminal device supports receiving the first indication information; the terminal device supports caching CSI for retransmission.

84. The network device according to any one of claims 63-83, wherein, The network device further includes a seventh processing unit, configured to merge the CSI recovered based on the second reported information with the CSI recovered based on the first reported information; or, to replace the CSI recovered based on the first reported information with the CSI recovered based on the second reported information.

85. A terminal device, comprising: A processor, and a memory communicating with the processor, the memory storing instructions that, when executed by the processor, cause the terminal device to perform the following: Send a first reporting information to the network device; wherein the first reporting information includes CSI; The terminal device receives a first indication message from the network device and sends a second reporting message to the network device; wherein the first indication message is used to instruct the terminal device to retransmit the CSI; and the second reporting message includes the CSI.

86. The terminal device according to claim 85, wherein, The first reported information and / or the second reported information are used to report the CSI based on the codebook or artificial intelligence (AI).

87. The terminal device according to claim 86, wherein, The instruction also causes the terminal device to perform one of the following: process the CSI into first output information based on a first AI model; wherein the first output information is mapped onto uplink channel physical resources and transmitted after channel coding and modulation; process the CSI into second output information based on a second AI model; wherein the second output information is mapped onto uplink channel physical resources and transmitted after modulation; process the CSI into third output information based on a third AI model; wherein the third output information is used to map onto uplink channel physical resources and transmit.

88. The terminal device according to any one of claims 85-87, wherein, The first reported information is used by the network device to determine whether to retransmit the CSI.

89. The terminal device according to claim 88, wherein, The first reported information is used by the network device to determine the input information of the fourth AI model; the input information of the fourth AI model is used by the network device to recover the CSI based on the fourth AI model and to determine whether to retransmit the CSI based on the distribution characteristics of the input information of the fourth AI model.

90. The terminal device according to claim 89, wherein, The first reported information is used by the network device to obtain a log-likelihood ratio sequence through demodulation; the log-likelihood ratio sequence is used by the network device to determine whether to retransmit the CSI based on the statistical value of the distribution characteristics of the log-likelihood ratio sequence.

91. The terminal device according to claim 89, wherein, The first reported information is used by the network device to obtain received symbols through channel equalization; the received symbols are used by the network device to compare the distribution of the received symbols with the distribution of the unit circle or the third output information to determine whether to retransmit the CSI; wherein, the third output information is the output information of the third AI model corresponding to the fourth AI model in the terminal device.

92. The terminal device according to claim 91, wherein, The instruction also causes the terminal device to: send a sample set of output information from the third AI model to the network device; wherein the sample set of output information is used by the network device to determine the distribution of the third output information.

93. The terminal device according to any one of claims 85-92, wherein, The first indication information is used to indicate at least one of the following: whether to send the second reporting information based on the CSI reporting configuration of the first reporting information; the CSI reporting configuration of the second reporting information; and uplink channel resources for transmitting the second reporting information. Time-domain information used to transmit the second reported information.

94. The terminal device according to claim 93, wherein, The CSI reporting configuration for the second reported information includes at least one of the following: the processing method for CSI when sending the second reported information; and configuration information related to the processing method.

95. The terminal device according to claim 94, wherein, Configuration information related to the processing method includes at least one of the following: the length of the output information of the AI ​​model used to implement the processing method; constraint information of the AI ​​model used to implement the processing method; wherein the constraint information is used to indicate whether the output information of the AI ​​model needs to satisfy constant modulus constraints; and the modulation method included in the processing method.

96. The terminal device according to any one of claims 93-95, wherein, The uplink channel resources include Physical Uplink Shared Channel (PUSCH) resources or Physical Uplink Control Channel (PUCCH) resources.

97. The terminal device according to any one of claims 93-96, wherein, When the first reported information is used for periodic CSI feedback or semi-persistent CSI feedback, the uplink channel resources include PUSCH resources or PUCCH resources.

98. The terminal device according to any one of claims 93-97, wherein, When the first reported information is used for aperiodic CSI feedback, the uplink channel resources include PUSCH resources.

99. The terminal device according to any one of claims 93-98, wherein, The time-domain information includes the time slot offset between the second reported information and the first reported information.

100. The terminal device according to any one of claims 85-99, wherein, When the first reporting information is used for non-periodic CSI feedback, the instruction also causes the terminal device to perform: receiving multiple first indication messages; and each time the first indication message is received, sending the second reporting information to the network device based on the first indication message.

101. The terminal device according to any one of claims 85-100, wherein, The instruction also causes the terminal device to: after sending the first reporting information, store the CSI in a first storage unit; and upon receiving a first indication information from the network device, read the CSI from the first storage unit and send a second reporting information to the network device based on the CSI.

102. The terminal device according to claim 101, wherein, The instruction also causes the terminal device to: clear the first storage unit when it determines that it will no longer receive the first instruction information for the first reported information.

103. The terminal device according to any one of claims 85-102, wherein, The instruction further causes the terminal device to perform: sending second indication information to the network device; wherein the second indication information is used by the network device to determine that the terminal device possesses a first capability; the first capability includes at least one of the following: the terminal device supports CSI feedback and retransmission based on AI; the terminal device supports deploying an AI model for CSI feedback; the terminal device supports receiving the first indication information; the terminal device supports caching CSI for retransmission.

104. The terminal device according to any one of claims 85-103, wherein, The second reporting information is used by the network device to merge the CSI recovered based on the second reporting information with the CSI recovered based on the first reporting information, or to replace the CSI recovered based on the first reporting information with the CSI recovered based on the second reporting information.

105. A network device, comprising: A processor, and a memory communicating with the processor, the memory storing instructions that, when executed by the processor, cause the network device to perform the following: Receive first reported information from the terminal device; wherein, the first reported information includes CSI; Send a first indication message to the terminal device; wherein the first indication message is used to instruct the terminal device to send a second reporting message to retransmit the CSI.

106. The network device according to claim 105, wherein, The first reported information and / or the second reported information are used to report the CSI based on the codebook or artificial intelligence (AI).

107. The network device according to claim 105 or 106, wherein, The instructions also cause the network device to perform one of the following: process the first reported information and / or the second reported information based on the codebook to recover the CSI; demodulate and channel decode the received symbols corresponding to the first reported information and / or the second reported information to obtain first input information, and process the first input information using a fifth AI model to recover the CSI; demodulate the received symbols corresponding to the first reported information and / or the second reported information to obtain second input information, and process the second input information using a sixth AI model to recover the CSI; use the received symbols corresponding to the first reported information and / or the second reported information as third input information, and process the third input information using a seventh AI model to recover the CSI.

108. The network device according to any one of claims 105-107, wherein, The first indication information is sent when a first condition is met, the first condition being related to at least one of the following: uplink channel quality; the first reporting information.

109. The network device according to claim 108, wherein, The instruction also causes the network device to perform: determining the input information of the fourth AI model based on the received symbol corresponding to the first reported information; wherein the fourth AI model is used to recover the CSI; and determining whether the first condition is met based on the distribution characteristics of the input information.

110. The network device according to claim 109, wherein, The input information includes a log-likelihood ratio sequence; the instruction further causes the network device to perform: determining whether the first condition is met based on the statistical values ​​of the distribution characteristics of the log-likelihood ratio sequence.

111. The network device according to claim 109, wherein, The input information includes the received symbols; the instruction further causes the network device to perform: comparing the distribution of the received symbols with the distribution of the unit circle or the third output information to determine whether the first condition is met; wherein, the third output information is the output information of the third AI model corresponding to the fourth AI model in the terminal device.

112. The network device according to claim 111, wherein, The instruction also causes the network device to perform: receiving a sample set of output information from the third AI model of the terminal device; wherein the sample set of output information from the third AI model is used by the network device to determine the distribution of the third output information.

113. The network device according to any one of claims 105-112, wherein, The first indication information is used to indicate at least one of the following: whether to send the second reporting information based on the CSI reporting configuration of the first reporting information; the CSI reporting configuration of the second reporting information; and uplink channel resources for transmitting the second reporting information. Time-domain information used to transmit the second reported information.

114. The network device according to claim 113, wherein, The CSI reporting configuration for the second reported information includes at least one of the following: the processing method for CSI when sending the second reported information; and configuration information related to the processing method.

115. The network device according to claim 114, wherein, Configuration information related to the processing method includes at least one of the following: the length of the output information of the AI ​​model used to implement the processing method; constraint information of the AI ​​model; wherein the constraint information is used to indicate whether the output information of the AI ​​model needs to satisfy constant modulus constraints; and the modulation method included in the processing method.

116. The network device according to any one of claims 113-115, wherein, The uplink channel resources include PUSCH resources or PUCCH resources.

117. The network device according to any one of claims 113-116, wherein, When the first reported information is used for periodic CSI feedback or semi-persistent CSI feedback, the uplink channel resources include PUSCH resources or PUCCH resources.

118. The network device according to any one of claims 113-117, wherein, When the first reported information is used for aperiodic CSI feedback, the uplink channel resources include PUSCH resources.

119. The network device according to any one of claims 113-118, wherein, The time-domain information includes the time slot offset between the second reported information and the first reported information.

120. The network device according to any one of claims 105-119, wherein, The instruction also causes the network device to: determine whether the terminal device is configured with a periodic CSI retransmission mechanism; if the terminal device is configured with a periodic CSI retransmission mechanism, then send the first indication information.

121. The network device according to any one of claims 105-119, wherein, The instruction also causes the network device to: send a maximum of N first indication messages; where N is determined based on the uplink channel quality.

122. The network device according to claim 121, wherein, In the case where the first reported information is used for non-periodic CSI feedback, the instruction also causes the network device to: stop sending the first indication information if it is determined that the first condition is not met or the number of retransmissions reaches N.

123. The network device according to any one of claims 105-122, wherein, The instruction also causes the network device to: upon receiving the first reporting information, store the CSI recovered based on the first reporting information in the second storage unit; and upon receiving the second reporting information, store the second reporting information or the CSI recovered based on the second reporting information in the second storage unit.

124. The network device according to claim 123, wherein, The instruction also causes the network device to: clear the second storage unit when it is determined that the first condition is not met or the number of retransmissions reaches a preset number.

125. The network device according to any one of claims 105-124, wherein, The instruction also causes the network device to perform: receiving second indication information from the terminal device; wherein the second indication information is used by the network device to determine that the terminal device has a first capability; the first capability includes at least one of the following: the terminal device supports CSI feedback and retransmission based on AI; the terminal device supports deploying an AI model for CSI feedback; the terminal device supports receiving the first indication information; the terminal device supports caching CSI for retransmission.

126. The network device according to any one of claims 105-125, wherein, The instruction also causes the network device to: merge the CSI recovered based on the second reported information with the CSI recovered based on the first reported information; or, replace the CSI recovered based on the first reported information with the CSI recovered based on the second reported information.

127. A chip, comprising: A processor for retrieving and running a computer program from memory, causing a device having the chip mounted to perform the method as claimed in any one of claims 1 to 20 or 21 to 42.

128. A computer-readable storage medium for storing a computer program that, when run by a device, causes the device to perform the method as claimed in any one of claims 1 to 20 or 21 to 42.

129. A computer program product comprising computer program instructions that cause a computer to perform the method as claimed in any one of claims 1 to 20 or 21 to 42.

130. A computer program that causes a computer to perform the method as claimed in any one of claims 1 to 20 or 21 to 42.

131. A communication system, comprising: A terminal device for performing the method as described in any one of claims 1 to 20; A network device for performing the method as described in any one of claims 21 to 42.

Images