Information transmission method and apparatus, and communication device

Abstract

The invention discloses an information transmission method and device and communication equipment, and belongs to the technical field of communication, and the information transmission method comprises the steps that the communication equipment receives or sends first information related to a first object; wherein the first object comprises at least one of the following items: a payload, at least one layer, and at least one rank; the first information is used for determining a first artificial intelligence (AI) model or configuring AI-based channel state information (CSI) reporting information, and the first AI model is used for generating the CSI reporting information or decoding the CSI reporting information.

Description

Technical Field

[0001] This application belongs to the field of communication technologies, and particularly relates to an information transmission method, apparatus, and communication device. Background Art

[0002] Currently, in a mobile communication network, tasks can be performed or services can be provided based on Artificial Intelligence (AI). For example, a terminal can perform Channel State Information (CSI) encoding based on a pre-trained AI model, and a network-side device can perform CSI decoding based on a pre-trained AI model. However, in the related art, there is no relevant solution for how to perform AI-based CSI processing or reporting, which can easily lead to the failure of AI-based CSI processing or reporting to achieve the expected effect. Summary of the Invention

[0003] Embodiments of this application provide an information transmission method, apparatus, and communication device, which can provide an AI-based CSI processing method, that is, determine an AI model for generating or decoding CSI reporting information or configure AI-based CSI reporting information based on first information related to at least one of a payload, a layer, and a rank, which is beneficial to improving the effect of AI-based CSI processing.

[0004] In a first aspect, an information transmission method is provided. The method includes:

[0005] A communication device receives or sends first information related to a first object;

[0006] wherein the first object includes at least one of the following: a payload, at least one layer, at least one rank; the first information is used to determine a first Artificial Intelligence (AI) model or to configure AI-based Channel State Information (CSI) reporting information, and the first AI model is used to generate the CSI reporting information or to decode the CSI reporting information.

[0007] In a second aspect, an information transmission apparatus is provided. The apparatus includes:

[0008] A transceiver module, configured to receive or send first information related to a first object;

[0009] wherein the first object includes at least one of the following: a payload, at least one layer, at least one rank; the first information is used to determine a first Artificial Intelligence (AI) model or to configure AI-based Channel State Information (CSI) reporting information, and the first AI model is used to generate the CSI reporting information or to decode the CSI reporting information.

[0010] In a third aspect, a communication device is provided. The terminal includes a processor and a memory. The memory stores a program or instructions that can run on the processor. When the program or instructions are executed by the processor, the steps of the method described in the first aspect are implemented.

[0011] In a fourth aspect, a communication device is provided, including a processor and a communication interface. The communication interface is used to receive or send first information related to a first object. The first object includes at least one of the following: payload, at least one layer, at least one rank. The first information is used to determine a first artificial intelligence (AI) model or to configure AI-based channel state information (CSI) reporting information. The first AI model is used to generate the CSI reporting information or to decode the CSI reporting information.

[0012] In a fifth aspect, a readable storage medium is provided. A program or instructions are stored on the readable storage medium. When the program or instructions are executed by a processor, the steps of the method described in the first aspect are implemented.

[0013] In a sixth aspect, a chip is provided. The chip includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run a program or instructions to implement the steps of the method described in the first aspect.

[0014] In a seventh aspect, a computer program / program product is provided. The computer program / program product is stored in a storage medium. The computer program / program product is executed by at least one processor to implement the steps of the method described in the first aspect.

[0015] In an embodiment of the present application, a communication device receives or sends first information related to a first object. The first object includes at least one of the following: payload, at least one layer, at least one rank. The first information is used to determine a first AI model or to configure AI-based CSI reporting information. The first AI model is used to generate the CSI reporting information or to decode the CSI reporting information. That is, in the embodiment of the present application, an AI model for CSI reporting information generation or decoding is determined or AI-based CSI reporting information is configured based on the first information related to at least one of payload, layer, and rank, which is beneficial to improving the CSI processing effect based on AI. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a block diagram of a wireless communication system to which an embodiment of the present application can be applied;

[0017] Figure 2aIt is a schematic structural diagram of a neural network provided by an embodiment of the present application;

[0018] Figure 2b It is a schematic structural diagram of a neuron provided by an embodiment of the present application;

[0019] Figure 2c It is a schematic diagram of time-frequency-spatial domain CSI encoding and decoding provided by an embodiment of the present application;

[0020] Figure 2d It is a schematic diagram of AI-based CSI encoding and decoding provided by an embodiment of the present application;

[0021] Figure 3 It is a flowchart of an information processing method provided by an embodiment of the present application;

[0022] Figures 4a to 5f It is a schematic diagram of information transmission of a layer provided by an embodiment of the present application;

[0023] Figure 6 It is a schematic diagram of multiple types of layers provided by an embodiment of the present application;

[0024] Figure 7 It is a structural diagram of an information processing device provided by an embodiment of the present application;

[0025] Figure 8 It is a structural diagram of a communication device provided by an embodiment of the present application;

[0026] Figure 9 It is a structural diagram of a terminal provided by an embodiment of the present application;

[0027] Figure 10 It is a structural diagram of a network-side device provided by an embodiment of the present application. Detailed implementation manners

[0028] Next, the technical solutions in the embodiments of the present application will be clearly described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art shall fall within the protection scope of the present application.

[0029] The terms "first", "second", etc. in this application are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that such terms can be interchanged under appropriate circumstances, so that the embodiments of this application can be implemented in an order other than those illustrated or described herein. The objects distinguished by "first" and "second" are usually of the same type, and do not limit the number of objects. For example, the first object can be one or multiple. In addition, "or" in this application means at least one of the connected objects. For example, "A or B" covers three scenarios, namely, Scenario 1: including A and not including B; Scenario 2: including B and not including A; Scenario 3: including both A and B. The character " / " generally indicates that the related objects before and after are in an "or" relationship.

[0030] The term "indication" in this application can be either a direct indication (or an explicit indication) or an indirect indication (or an implicit indication). Among them, a direct indication can be understood as that the sender clearly informs the receiver of specific information, operations to be performed, or request results, etc. in the sent indication; an indirect indication can be understood as that the receiver determines the corresponding information according to the indication sent by the sender, or makes a judgment and determines the operations to be performed or request results, etc. according to the judgment result.

[0031] It is worth noting that the technology described in the embodiments of this application is not limited to Long Term Evolution (LTE) / LTE-Advanced (LTE-A) systems, and can also be used in other wireless communication systems, such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), Single-carrier Frequency-Division Multiple Access (SC-FDMA), or other systems. The terms "system" and "network" in the embodiments of this application are often used interchangeably. The described technology can be used not only in the above-mentioned systems and radio technologies, but also in other systems and radio technologies. The following description describes the New Radio (NR) system for example purposes, and uses NR terms in most of the following descriptions, but these technologies can also be applied to systems other than the NR system, such as the 6th generation (6 thGeneration, 6G) communication system.

[0032] Figure 1The block diagram of a wireless communication system to which embodiments of the present application can be applied is shown. The wireless communication system includes a terminal 11 and a network-side device 12. Among them, the terminal 11 can be a mobile phone, a tablet personal computer, a laptop computer, a notebook computer, a personal digital assistant (PDA), a handheld computer, a netbook, an ultra-mobile personal computer (UMPC), a mobile internet device (MID), an augmented reality (AR), a virtual reality (VR) device, a robot, a wearable device, a flight vehicle, a vehicle user equipment (VUE), a shipborne device, a pedestrian user equipment (PUE), a smart home (home devices with wireless communication functions, such as refrigerators, TVs, washing machines, or furniture, etc.), a game console, a personal computer (PC), a teller machine, or a self-service machine, etc. Wearable devices include: smart watches, smart bracelets, smart earphones, smart glasses, smart jewelry (smart bracelets, smart bracelets, smart rings, smart necklaces, smart anklets, smart ankle chains, etc.), smart wristbands, smart clothing, etc. Among them, the vehicle user equipment can also be referred to as a vehicle terminal, a vehicle controller, a vehicle module, a vehicle component, a vehicle chip, or a vehicle unit, etc. It should be noted that the specific type of the terminal 11 is not limited in the embodiments of the present application. The network-side device 12 can include an access network device or a core network device. Among them, the access network device can also be referred to as a radio access network (RAN) device, a radio access network function, or a radio access network unit. The access network device can include a base station, a wireless local area network (WLAN) access point (AP), or a wireless fidelity (WiFi) node, etc.Among them, the base station may be referred to as Node B (NB), Evolved Node B (eNB), the next generation Node B (gNB), New Radio Node B (NR Node B), access point, Relay Base Station (RBS), Serving Base Station (SBS), Base Transceiver Station (BTS), radio base station, radio transceiver, Basic Service Set (BSS), Extended Service Set (ESS), home Node B (HNB), home evolved Node B, Transmission Reception Point (TRP), or some other suitable term in the art. As long as the same technical effect is achieved, the base station is not limited to specific technical terms. It should be noted that in the embodiments of this application, only the base station in the NR system is taken as an example for introduction, and the specific type of the base station is not limited.

[0033] The core network device may include, but is not limited to, at least one of the following: core network node, core network function, Mobility Management Entity (MME), Access and Mobility Management Function (AMF), Session Management Function (SMF), User Plane Function (UPF), Policy Control Function (PCF), Policy and Charging Rules Function (PCRF), Edge Application Server Discovery Function (EASDF), Unified Data Management (UDM), Unified Data Repository (UDR), Home Subscriber Server (HSS), Centralized network configuration (CNC), Network Repository Function (NRF), Network Exposure Function (NEF), Local NEF (L-NEF), Binding Support Function (BSF), Application Function (AF), etc. It should be noted that in the embodiments of this application, only the core network devices in the NR system are taken as examples for introduction, and the specific types of core network devices are not limited.

[0034] For ease of understanding, some content related to the embodiments of this application is described below:

[0035] I. Artificial Intelligence (AI)

[0036] Artificial intelligence (AI) has currently found extensive applications in various fields. Incorporating artificial intelligence into wireless communication networks to significantly improve technical indicators such as throughput, latency, and user capacity is an important task for future wireless communication networks. There are various implementation methods for the AI module, such as neural networks, decision trees, support vector machines, Bayesian classifiers, etc. This application takes neural networks as an example for illustration, but does not limit the specific type of the AI module.

[0037] Exemplarily, a neural network can be as Figure 2a shown. A neural network is composed of neurons, and each neuron can be as Figure 2b shown. Among them, a1, a2, …, aK are inputs, w is the weight (multiplicative coefficient), b is the bias (additive coefficient), and σ(.) is the activation function. Common activation functions include Sigmoid, tanh, Rectified Linear Unit (ReLU), etc.

[0038] The parameters of the neural network are optimized through gradient optimization algorithms. Gradient optimization algorithms are a class of algorithms that minimize or maximize the objective function (also known as the loss function), and the objective function is often a mathematical combination of model parameters and data. For example, given data X and its corresponding label Y, we construct a neural network model f(.). After having the model, based on the input x, we can obtain the predicted output f(x), and we can calculate the gap between the predicted value and the true value (f(x) - Y), which is the loss function. The goal is to find the appropriate W and b to minimize the value of the above loss function. Among them, the smaller the loss value, the closer the model is to the real situation.

[0039] Currently, common optimization algorithms are basically based on the error Back Propagation (BP) algorithm. The basic idea of the BP algorithm is that the learning process consists of two processes: the forward propagation of signals and the backward propagation of errors. During forward propagation, the input samples are passed from the input layer, processed layer by layer through each hidden layer, and then passed to the output layer. If the actual output of the output layer does not match the expected output, it enters the stage of backward propagation of errors. The error backpropagation is to pass the output error back to the input layer layer by layer through the hidden layer in a certain form, and distribute the error to all units of each layer, so as to obtain the error signals of each layer of units. This error signal is used as the basis for correcting the weights of each unit. This process of adjusting the weights of each layer in the forward propagation of signals and the backward propagation of errors is carried out cyclically. The process of continuously adjusting the weights is also the learning and training process of the network. This process continues until the error of the network output is reduced to an acceptable level, or until the preset number of learning times is reached.

[0040] Common optimization algorithms include Gradient Descent, Stochastic Gradient Descent (SGD), mini - batch gradient descent, Momentum, Nesterov (specifically Stochastic Gradient Descent with Momentum), Adaptive Gradient descent (Adagrad), Adadelta, Root Mean Square Prop (RMSprop), Adaptive Moment Estimation (Adam), etc.

[0041] When these optimization algorithms perform backpropagation of errors, they all calculate the derivative / partial derivative of the current neuron based on the error / loss obtained from the loss function, and then add the learning rate, previous gradients / derivatives / partial derivatives, etc. to obtain the gradient, which is then passed to the previous layer.

[0042] II. AI Unit / AI Model

[0043] The AI unit / AI model in the embodiments of this application can also be referred to as a Machine Learning (ML) model, ML unit, AI structure, AI function, AI feature, machine learning model, neural network, neural network function, neural network capability, etc. Or the above - mentioned AI unit / AI model can also refer to a processing unit that can implement specific algorithms, formulas, processing flows, capabilities, etc. related to AI. Or the AI unit / AI model can be a processing method, algorithm, function, module, or unit for a specific data set. Or the AI unit / AI model can be a processing method, algorithm, function, module, or unit running on AI / ML - related hardware such as GPUs, NPUs, TPUs, ASICs, etc. The embodiments of this application do not make specific limitations in this regard. Optionally, the specific data set includes at least one of the input and output of the AI unit / AI model.

[0044] Optionally, the identifier of the AI unit / AI model can be an AI model identifier, an AI structure identifier, an AI algorithm identifier, or the identifier of the specific data set associated with the AI unit / AI model, or the identifier of the specific scenario, environment, channel characteristics, device related to AI / ML, or the identifier of the function, feature, capability, or module related to AI / ML. The embodiments of this application do not make specific limitations in this regard.

[0045] III. Non - AI CSI Compression

[0046] Non-AI CSI compression in the current protocol, including type I, type II, and enhanced type II (e type 2 CSI), where:

[0047] 1. type I CSI

[0048] When it is not possible to report the complete channel or precoder, type I reports the Precoding Matrix Indicator (PMI) of the wideband or subband, that is, a two-dimensional Discrete Fourier Transform (DFT) vector and its phase rotation amount on the wideband or subband. Among them, type I mainly needs to report the index of the two-dimensional DFT vector and its phase rotation amount.

[0049] The reporting format of the above type I CSI is as follows:

[0050] Wideband CSI: Rank Indicator (RI)-PMI-Channel Quality Indicator (CQI), where when RI > 4, two CQIs need to be reported for two Transport Blocks (TBs), otherwise one CQI.

[0051] Subband CSI:

[0052] Part 1 CSI: RI + CQI of the first TB;

[0053] Part 2 CSI: Wideband CQI - Wideband PMI - CQI of even subbands + PMI - CQI of odd subbands + PMI; The omission principle is: omit part 2 based on priority, that is, odd subbands can be omitted first.

[0054] 2. type 2 CSI

[0055] Type 2 represents the precoding vector PMI as a linear combination of a set of basis vectors compared to a simple two-dimensional DFT vector and its phase rotation amount. Among them, type 2 needs to report the basis vector index and the projection (amplitude and phase) on the basis vector.

[0056] The reporting format of the above type 2 CSI is as follows:

[0057] Part 1: Number of non-zero wideband amplitude coefficients per layer of RI-CQI {encoded separately};

[0058] Part 2: Wideband PMI {vector} - PMI - Layer Indicator (LI) {i 1,4,l (Wideband amplitude 1)i 2,1,l (Phase)i 2，2，l (Sub - band amplitude 2)}, where the above L - vector is the basis vector.

[0059] Among them, amplitude 1: 3 bits (scalar); amplitude 2: 1 bit.

[0060] 3. e type 2 CSI

[0061] Since the overhead of type 2 is up to several hundred or even several thousand bits, e type 2 is a further compression of type 2, that is, the vector composed of weighted coefficients on different sub - bands is further compressed into a vector composed of a set of frequency - domain basis vectors.

[0062] The above e type 2 CSI reporting format is as follows:

[0063] Part 1: Number of non - zero wideband amplitude coefficients per layer {encoded separately};

[0064] Part 2: Wideband PMI {vector} - PMI: i 2,4,l Amplitude i 2,5,l Phase and i 1,7,l , {reported bitmap};

[0065] Pri(l,i,f) = 2·L·υ·π(f)+υ·i + l,

[0066] Among them, π(f) where f is the frequency - domain basis vector, π(0)=0, π(N3 - 1)=1, π(1)=2;

[0067]

[0068] Among them, l = 1,2,…,υ, i = 0,1,…,2L - 1, and f = 0,1,…,M υ -1,

[0069] The above Part 2 adopts the feedback method of unified compression for all sub - bands, where:

[0070] 0: L spatial domain basis vectors i 1,1 ,i 1,2and the strongest coefficient information {log2 2L bit}i for each layer 1,8,l (l = 1, …, υ);

[0071] 1: M frequency-domain basis vectors (i 1,5 (if reported), i 1,6,l (if reported)), reference amplitude information i 2,3,l , the v2LM - [KNZ / 2] bit i with the highest priority among the non-zero coefficient positions 1,7,l , where v is the rank, and v2LM - [KNZ / 2] coefficients with the highest priority {i 2,4,l , i 2,5,l};

[0072] 2: The coefficients with the lowest priority [KNZ / 2] among the non-zero coefficient positions.

[0073] IV. Time-Frequency-Space CSI Compression

[0074] Exemplarily, as Figure 2c shown, in traditional space-frequency domain compression, the CSI information (space-frequency domain channel) of slot (Slot) X is input to encoder (Encoder, ENC) X, generating CSI reporting information ( Figure 2c the feedback information (Feedback) in). The network (Network, NW) receives the CSI reporting information and decodes it through decoder (Decoder, DEC) X to obtain the restored CSI information (i.e., CSI’).

[0075] In multi-layer (layer) compression (where multiple layers can be layers associated with one reporting configuration ID (reportconfig ID), or can include layers associated with different reportconfig IDs), different layers can be associated with the same time or different times. For example, corresponding to different Slots, that is, compression is combined with different time domains.

[0076] In space-time-frequency domain compression, ENC X will use the intermediate information (Internal Information) of ENC X - 1, or the intermediate information of ENC X + 1, that is, use the historical information before or after to assist in encoding. Figure 2c The intermediate information shown in is sent from ENC X (i.e., the encoder of Slot X) to ENC X + 1 (the encoder of Slot X + 1). It should be noted that the intermediate information of the decoder is the same as that of the above encoder and will not be elaborated here. In addition, the intermediate state information of the encoder is generally only transmitted between encoders; the intermediate state information of the decoder is generally only transmitted between decoders.

[0077] In the 18th version (Release 18, R18), AI-based CSI / PMI compression was proposed. Among them, the CSI or codebook W expected or targeted by the UE N*B is compressed by AI, such as compressed into an AI-based PMI value, and then reported to the network device. The network device performs decompression to obtain W'. N*B , such as Figure 2d shown, where N represents the number of CSI ports and B represents the number of sub-bands (Sunband).

[0078] Next, in conjunction with the accompanying drawings, the information transmission method provided by the embodiments of the present application will be described in detail through some embodiments and their application scenarios.

[0079] Please refer to Figure 3 , Figure 3 which is a flowchart of an information transmission method provided by an embodiment of the present application. This method can be executed by a communication device, such as Figure 3 shown, and includes the following steps:

[0080] Step 301, the communication device receives or sends first information related to a first object;

[0081] Among them, the first object includes at least one of the following: payload, at least one layer, at least one rank; the first information is used to determine a first AI model or to configure AI-based CSI reporting information, and the first AI model is used to generate the CSI reporting information or to decode the CSI reporting information.

[0082] In this embodiment, the above communication device can be an encoding end or a decoding end, or the above communication device can be a terminal or a network device, etc.

[0083] The above AI-based CSI reporting information can be understood as that part or all of the CSI reporting information is encoded or compressed by an AI model. The above CSI reporting information can be understood as the information obtained by encoding or compressing part or all of the CSI information. For example, PMI, CQI, RI, etc. Among them, the above CSI information can include at least one of, but not limited to, a precoding matrix, channel information, etc. It should be noted that the above CSI reporting information can also be referred to as CSI feedback information, etc.

[0084] Exemplarily, if the above-mentioned first AI model is an encoding or compression model, the above-mentioned CSI reporting information may include information on the precoding matrix that has passed through the first AI model, or the above-mentioned CSI reporting information may include at least one of information on the precoding matrix that has passed through the first AI model, CQI information, and Rank information, etc.

[0085] Exemplarily, the above-mentioned CSI reporting information may include CSI reporting information corresponding to at least one layer or CSI reporting information corresponding to at least one rank. In some optional embodiments, the above-mentioned at least one rank may include at least one rank greater than 1.

[0086] Exemplarily, the above-mentioned payload may be the payload of the CSI reporting information, or the above-mentioned payload may be the payload of the precoding matrix that has passed through the first AI model, or the above-mentioned payload may be the payload of part of the CSI reporting information.

[0087] In some optional embodiments, different layers may respectively correspond to AI models corresponding to different payloads, or different types of layers may respectively correspond to AI models corresponding to different payloads, or different values of ranks may respectively correspond to AI models corresponding to different payloads.

[0088] The above-mentioned first information is used to determine the first AI model. Exemplarily, the above-mentioned first information is used to train an AI model to obtain the first AI model, or the above-mentioned first information is used to update or finetune an existing AI model to obtain the first model, or is used to select the first AI model from multiple AI models, etc.

[0089] In some optional embodiments, determining the first AI model according to the first information may include determining the parameters of the first AI model according to the first information. The above-mentioned determining the parameters of the first AI model according to the first information can be understood as adjusting the parameters of the first AI model according to the first information. Optionally, the above-mentioned adjusting the parameters of the first AI model may be to adjust the parameters of the first AI model to meet the requirements of the payload.

[0090] For example, the above first information may include dataset information related to the payload (e.g., at least one of the input information and output information of the AI model). Based on this dataset information, an AI model for processing the CSI reporting information corresponding to the payload, i.e., the first AI model, can be trained. Alternatively, the above first information may include the identification information of the AI model related to the payload, and the AI model identified by this identification information (i.e., the first AI model) is used to process the CSI reporting information corresponding to the payload. Or the above first information may include dataset information corresponding to at least one layer or at least one rank, and this dataset information is used to train an AI model for processing the CSI reporting information corresponding to at least one layer or at least one rank, i.e., the first AI model.

[0091] In addition, for the encoding end, the above first AI model is used to encode or compress the CSI to obtain the CSI reporting information; for the decoding end, the above first AI model is used to decode or decompress the CSI reporting information.

[0092] Optionally, the first AI model being used to generate the CSI reporting information may include the first AI model being used to generate all or part of the information of the CSI reporting information. For example, the first AI model is used to generate compressed precoding information. The first AI model being used to decode the CSI reporting information may include the first AI model being used to decode all or part of the information of the CSI reporting information. For example, the first AI model is used to decode the compressed precoding information in the CSI reporting information.

[0093] The above first information is used to configure AI-based CSI reporting information. Exemplarily, if the above first information includes the layer type, then the CSI reporting information may include the CSI reporting information corresponding to this layer type; or, if the above first information includes the layer compression method, then the CSI reporting information may be the CSI reporting information obtained by compressing using this layer compression method, or if the above first information includes the payload, then the CSI reporting information or the specified part of the CSI reporting should satisfy this payload.

[0094] The following gives examples by cases:

[0095] Case 1: When the communication device is a terminal, the terminal receives the first information related to the first object sent by the network-side device or the model management device or the data management device, determines the first AI model according to the first information, and then the terminal can compress or encode some or all of the CSI information based on the first AI model to obtain the CSI reporting information. Alternatively, the terminal can generate the CSI reporting information according to the first information and send the CSI reporting information to the network-side device;

[0096] Alternatively, the terminal sends the first information related to the first object to the network-side device, and the network-side device determines the first AI model based on the received first information. Then, the network-side device can decode or decompress the CSI reporting information received from the terminal based on the first AI model to obtain the decoded CSI information or the decompressed CSI information or the restored CSI information or the reconstructed CSI information. Alternatively, the network-side device decodes or decompresses the CSI reporting information based on the received first information to obtain the decoded CSI information or the decompressed CSI information or the restored CSI information or the reconstructed CSI information.

[0097] Case 2: When the communication device is a network-side device, the network-side device can send the first information related to the first object to the terminal, and the terminal determines the first AI model according to the first information. Then, the terminal can compress or encode some or all of the CSI based on the first AI model to obtain the CSI reporting information. Alternatively, the terminal can generate the CSI reporting information according to the first information and send the CSI reporting information to the network-side device;

[0098] Alternatively, the network-side device receives the first information related to the first object sent by the terminal or the model management device or the data management device, determines the first AI model according to the first information, and then the network-side device can decode or decompress the CSI reporting information received from the terminal based on the first AI model to obtain the decoded CSI information or the decompressed CSI information or the restored CSI information or the reconstructed CSI information. Alternatively, the network-side device decodes or decompresses the CSI reporting information based on the received first information to obtain the decoded CSI information or the decompressed CSI information or the restored CSI information or the reconstructed CSI information.

[0099] In the embodiments of the present application, a communication device receives or transmits first information related to a first object; wherein, the first object is an object related to CSI reporting information, and the first object includes at least one of the following: payload, at least one layer, at least one rank; the first information is used to determine a first AI model, and the first AI model is used to generate the CSI reporting information or decode the CSI reporting information. In the embodiments of the present application, an AI model for CSI reporting information generation or decoding is determined based on the first information related to at least one of payload, layer, and rank, or CSI reporting information based on AI is configured, which helps to improve the CSI processing effect based on AI.

[0100] Optionally, the first information related to the first object includes at least one of the following:

[0101] First dataset information related to the payload;

[0102] First AI model information related to the payload;

[0103] First quantization information related to the payload.

[0104] Exemplarily, the above first dataset information may include at least one of, but is not limited to, at least one dataset information related to the payload, payload information associated with at least one dataset information, dataset identifiers associated with at least one dataset information related to the payload, AI model information associated with at least one dataset information related to the payload, etc.

[0105] The above first AI model information related to the payload may include at least one of, but is not limited to, identification information of the AI model related to the payload, payload information associated with the AI model, type information or function information of the AI model related to the payload, parameter information of the AI model related to the payload, etc.

[0106] The above first quantization information related to the payload may include at least one of, but is not limited to, a quantization method corresponding to the output dimension of the AI model, a quantization method corresponding to the payload size, etc.

[0107] In some alternative embodiments, in order to match the first AI models of the terminal and the network side, the terminal and the network side device transmit dataset information for training the first AI model. In this case, the above first information may include the dataset information of the first AI model.

[0108] In some alternative embodiments, to match the first AI model between the terminal and the network side, the terminal and the network side device transfer the parameter information of the first AI model. In this case, the above-mentioned first information may include the parameter information of the first AI model of the first AI model.

[0109] In some alternative embodiments, to match the first AI model between the terminal and the network side, the terminal and the network side device transfer the condition information of the first AI model, for example, payload information, type information of the first AI model, etc.

[0110] Optionally, one first dataset information is associated with the CSI reporting information of one type of payload;

[0111] Or,

[0112] Different first dataset information is associated with the CSI reporting information of different payloads.

[0113] In this embodiment, one first dataset information is associated with the CSI reporting information of one type of payload, and different first dataset information is associated with the CSI reporting information of different payloads. In this way, based on the first dataset information, the AI model corresponding to the payload of the CSI reporting information can be quickly determined to process the CSI reporting information. It can be understood that the CSI reporting information obtained during the AI inference stage of the first AI model trained based on the first dataset information of the payload naturally satisfies the CSI reporting information of the payload.

[0114] Optionally, one first dataset information is associated with the CSI reporting information of multiple types of payloads.

[0115] Optionally, one first dataset information is associated with the input information or output information of multiple AI models. For example, the dimensions of the CSI information input into the AI model for compression are different, or the dimensions or payloads of the CSI reporting information output from the AI model are different.

[0116] In this embodiment, one piece of first dataset information can be associated with CSI reporting information of multiple payloads. Exemplarily, the above first dataset information may include multiple pieces of dataset information, and the multiple pieces of dataset information respectively correspond one-to-one to the CSI reporting information of multiple payloads. For example, if the first dataset information can include dataset information A, dataset information B, and dataset information C, then dataset information A corresponds to the CSI reporting information of the first payload, dataset information B corresponds to the CSI reporting information of the second payload, and dataset information C corresponds to the CSI reporting information of the third payload; or, the CSI reporting information of the above multiple payloads respectively corresponds to different data in the first dataset information. For example, data a in the above first dataset information is associated with the first payload, data b in the above first dataset information is associated with the first payload, data c in the above first dataset information is associated with the second payload, data d in the above first dataset information is associated with the first payload, data e in the above first dataset information is associated with the third payload, and so on.

[0117] It can be understood that the above first dataset information can explicitly include different data subsets, or can include different data in an unordered manner, but only belongs to different data subsets from the perspective of data characteristics. Optionally, the communication device needs to preprocess or classify the data subsets to obtain the first dataset information. Optionally, the communication device processes different data subsets into first dataset information with the same characteristics through preprocessing, or extracts different data subsets through classification. Optionally, the method may further include: the communication device receives second information, and the second information is used to assist the communication device in preprocessing or classifying the first dataset information.

[0118] In some optional embodiments, there is a corresponding relationship between the dataset information corresponding to the CSI reporting information of different payloads in the first dataset information and different AI models.

[0119] In this embodiment, one piece of first dataset information is associated with CSI reporting information of multiple payloads. In this way, based on one piece of first dataset information, the AI models corresponding to the CSI reporting information of multiple payloads can be determined to process the CSI reporting information of multiple payloads respectively.

[0120] Optionally, the first dataset information includes at least one of the following:

[0121] At least one piece of dataset information, where the at least one piece of dataset information includes at least one of the input information of the AI model and the output information of the AI model;

[0122] Payload information associated with at least one dataset information;

[0123] Dataset identifier associated with at least one dataset information;

[0124] AI model information associated with at least one dataset information.

[0125] Exemplarily, taking the above first AI model for generating CSI report information as an example, the input information of the above AI model may include, but is not limited to, at least one of the precoding matrix W N*B and channel information, where N represents the number of CSI ports and B represents the number of sub-bands (Sunband); the output information of the above AI model may include, but is not limited to, compressed CSI information (for example, compressed PMI value, compressed precoding matrix), the size of the output information, the output dimension, or the bits of the output information, etc. Among them, the output dimension of the above AI model can be understood as the number of output elements or output parameters of the above AI model.

[0126] The payload information associated with the above at least one dataset information may include, but is not limited to, at least one of the size of the payload of at least one of the input information and output information of the AI model, the type of the payload of at least one of the input information and output information of the AI model, the quantization method of the output information of the AI model, the dimension of at least one of the input information and output information of the AI model, etc. Among them, the dimension of the input information of the above AI model can be understood as the number of input elements or input parameters of the AI model, and the dimension of the output information of the above AI model can be understood as the number of output elements or output parameters of the AI model.

[0127] The dataset identifier (ID) associated with the above at least one dataset information may include, but is not limited to, the dataset ID associated with at least one of the input information and output information of the AI model. In some alternative embodiments, different dataset IDs may be associated with different payload information.

[0128] The AI model information associated with the above at least one dataset information may include, but is not limited to, at least one of the identification information of the AI model associated with the above at least one dataset information, the payload information associated with the AI model, the type information or function information of the AI model, the parameter information of the AI model, etc.

[0129] It is understandable that the at least one dataset information includes at least one of the input information of the AI model and the output information of the AI model, indicating that the at least one dataset information is used for processing such as the training and updating of the first AI model. Exemplarily, if the above AI model is an encoding or compression model, the input information of the above AI model may include uncompressed precoding information or uncompressed CSI information, and the output information of the above AI model may include compressed precoding information or compressed CSI information; if the above AI model is a decoding or decompression model, the input information of the above AI model may include compressed precoding information or compressed CSI information, and the output information of the above AI model may include uncompressed precoding information (i.e., restored precoding information) or uncompressed CSI information (i.e., restored CSI information). That is, the at least one dataset information may include at least one of uncompressed precoding information and compressed precoding information, or the at least one dataset information includes at least one of uncompressed CSI information and compressed CSI information.

[0130] Optionally, one first AI model information is associated with the CSI reporting information of one type of payload;

[0131] Or,

[0132] Different first AI model information is associated with the CSI reporting information of different payloads.

[0133] In this embodiment, one first AI model information is associated with the CSI reporting information of one type of payload, and different first AI model information is associated with the CSI reporting information of different payloads. In this way, based on the first AI model information, the AI model corresponding to the payload of the CSI reporting information can be quickly determined to process the CSI reporting information.

[0134] Optionally, the first AI model information includes:

[0135] The identification information of the AI model;

[0136] The payload information associated with the AI model;

[0137] The type information or function information of the AI model;

[0138] The parameter information of the AI model.

[0139] Exemplarily, the identification information of the above AI model may include at least one of, but not limited to, the ID of the AI model, the function ID, etc.

[0140] The payload information associated with the above AI model may include at least one of, but is not limited to, the size of the payload of at least one of the input information and output information of the AI model, the quantization method of the output information of the AI model, the type of the payload of at least one of the input information and output information of the AI model, etc.

[0141] The type information of the above AI model can be used to indicate the type of the AI model, and the function information of the above AI model can be used to indicate the function of the AI model. In some alternative embodiments, the type information or function information of the above AI model is used to indirectly indicate one or more associated payload information.

[0142] The parameter information of the above AI model may include at least one of, but is not limited to, the number of model layers, model structure, model parameters, etc. For example, the parameter information of the above AI model may include at least part of the model structure and at least part of the model parameters of the adaptation layer. Optionally, the above adaptation layer may include some layers located at the tail of the AI model (the tail of the encoder) or the head of the AI model (the head of the decoder). For example, the number of layers of the above adaptation layer is 1 layer, 2 layers, 4 layers, etc., and the structure of the above adaptation layer is a multi-layer perceptron (MLP), a convolutional neural network (CNN), etc.

[0143] Optionally, one first quantization information is associated with the CSI reporting information of one type of payload;

[0144] Or,

[0145] One first quantization information is associated with the CSI reporting information of multiple types of payloads.

[0146] In an implementation manner, one first quantization information is associated with the CSI reporting information of one type of payload. For example, in the case where the output elements of the AI model are fixed and the quantization method is fixed-length quantization, one first quantization information may be associated with the CSI reporting information of one type of payload.

[0147] In this implementation manner, one first quantization information is associated with the CSI reporting information of one type of payload. In this way, based on one first quantization information, the AI model corresponding to the CSI reporting information of one type of payload can be determined to process the CSI reporting information of this payload.

[0148] In another embodiment, a first quantization information may be associated with CSI reporting information of multiple payloads. For example, when the output elements of the AI model are fixed and the quantization method is variable-length quantization, or when the output elements of the AI model are not fixed, a first quantization information may be associated with CSI reporting information of multiple payloads.

[0149] In this embodiment, a first quantization information is associated with CSI reporting information of multiple payloads. In this way, based on a first quantization information, AI models corresponding to CSI reporting information of multiple payloads can be determined to process CSI reporting information of multiple payloads respectively.

[0150] Optionally, the first quantization information includes at least one of the following:

[0151] The quantization method corresponding to the output dimension of the AI model, where the output dimension of the AI model is used to indicate the number of output elements of the AI model;

[0152] The quantization method corresponding to the payload size, where the payload size is the payload size of the CSI reporting information.

[0153] Exemplarily, the above quantization method may include but is not limited to segmentation method, quantization bit number, quantization codebook, etc.

[0154] The above output dimension of the AI model can be understood as the number of output elements or output parameters of the AI model.

[0155] In one embodiment, the first quantization information includes the quantization method corresponding to the output dimension of the AI model. For example, if the output dimension of the above first AI model is X, the above first quantization information may include the quantization method corresponding to X.

[0156] Optionally, the quantization methods corresponding to different output dimensions of the AI model are different, or the quantization methods corresponding to the output dimensions of the AI model within different output dimension value ranges are different.

[0157] Exemplarily, if the output dimension of the AI model is X1, the corresponding quantization method is the first quantization method; if the output dimension of the AI model is X2, the corresponding quantization method is the second quantization method, where X1 and X2 have different values, and the first quantization method and the second quantization method are different. For example, the first quantization method is to quantize a floating-point number (float) or a segment of a float with 8 bits, and the second quantization method is to quantize a float or a segment of a float with 4 bits; or, the first quantization method is variable-length quantization, and the second quantization method is fixed-length quantization; or, the first quantization method is vector quantization (Vector Quantization, VQ), and the second quantization method is fixed quantization.

[0158] The above-mentioned value range of the output dimension can be understood as the value range of the output dimension of the AI model. In some alternative embodiments, multiple value ranges of the output dimension can be preset, and a corresponding relationship between different value ranges of the output dimension and different quantization methods can be established. For example, the first value range of the output dimension is [Y1, Y2), and its corresponding quantization method is the first quantization method; the second value range of the output dimension is [Y2, Y3), and its corresponding quantization method is the second quantization method; the second value range of the output dimension is [Y3, Y4), and its corresponding quantization method is the third quantization method, and so on; if the output dimension of the AI model is X1 and X1 is within [Y1, Y2), then its corresponding quantization method is the first quantization method; if the output dimension of the AI model is X2 and X2 is within [Y3, Y4), then its corresponding quantization method is the third quantization method.

[0159] In one embodiment, the above-mentioned first quantization information includes the quantization method corresponding to the payload size. For example, if the payload size of the above-mentioned CSI reporting information is Z bits, the above-mentioned first quantization information may include the quantization method corresponding to Z.

[0160] Optionally, the quantization methods corresponding to different payload sizes are different, or the quantization methods corresponding to the payload sizes within different value ranges of the payload size are different.

[0161] Exemplarily, if the size of the payload is Z1 bit, the corresponding quantization method is the first quantization method; if the size of the payload is Z2 bit, the corresponding quantization method is the second quantization method, where Z1 and Z2 are different, and the first quantization method and the second quantization method are different. For example, the first quantization method is to quantize one float or a segment of a float with 8 bits, and the second quantization method is to quantize one float or a segment of a float with 4 bits; or, the first quantization method is variable-length quantization, and the second quantization method is fixed-length quantization; or, the first quantization method is VQ quantization, and the second quantization method is fixed quantization.

[0162] The above-mentioned value range of the payload size can be understood as the value range of the payload size. In some alternative embodiments, multiple value ranges of the payload size can be preset, and the corresponding relationship between different value ranges of the payload size and different quantization methods can be established. For example, the first value range of the payload size is [K1, K2), and the corresponding quantization method is the first quantization method; the second value range of the payload size is [K2, K3), and the corresponding quantization method is the second quantization method; the second value range of the payload size is [K3, K4), and the corresponding quantization method is the third quantization method, and so on; if the size of the payload of the above CSI reporting information is Z1 and Z1 is within [K1, K2), the corresponding quantization method is the first quantization method; if the size of the payload of the above CSI reporting information is Z2 and Z2 is within [K1, K2), the corresponding quantization method is the second quantization method.

[0163] Optionally, the first information related to the first object includes at least one of the following:

[0164] The first information related to the payload of one layer in the at least one layer;

[0165] The first information related to the payload of each layer in the at least one layer;

[0166] The first information related to the payload of all layers in the at least one layer.

[0167] In one embodiment, the first information related to the first object may include the first information related to the payload of one layer in the at least one layer. For example, the first information related to the payload of the first type of layer in the at least one layer. Wherein, the first information related to the payload of one layer may include at least one of, but is not limited to, the first dataset information related to the payload of this layer, the first AI model information related to the payload of this layer, the first quantization information related to the payload of this layer, etc.

[0168] In another embodiment, the first information related to the first object may include the first information related to the payload of each layer (i.e., per layer) in the at least one layer. For example, when the CSI reporting information includes CSI reporting information corresponding to at least two layers, the first information related to the first object may include the first information related to the payload of each of the at least two layers. Wherein, the first information related to the payload of each layer may respectively include at least one of, but is not limited to, the first dataset information related to the payload of each layer, the first AI model information related to the payload of each layer, the first quantization information related to the payload of each layer, etc. Optionally, the first information related to the payload of each layer may be the same.

[0169] It can be understood that in this embodiment, different layers in the at least one layer may be associated with different payloads, or different layers may be associated with different AI models.

[0170] In yet another embodiment, the first information related to the first object may include the first information related to the payload of all layers (across layer) in the at least one layer. For example, taking all the layers related to the CSI reporting information as a whole and obtaining the first information related to the payload thereof, that is, the first information related to the payload is for all the layers related to the CSI reporting information. Wherein, the first information related to the payload of all layers in the at least one layer may include at least one of, but is not limited to, the first dataset information related to the payload of all layers in the at least one layer, the first AI model information related to the payload of all layers in the at least one layer, the first quantization information related to the payload of all layers in the at least one layer, etc.

[0171] It should be noted that the first dataset information, the first AI model information, and the first quantization information in this embodiment can refer to the relevant descriptions in the foregoing embodiments and will not be elaborated here.

[0172] Optionally, the at least one layer includes at least two types of layers;

[0173] wherein, the payloads of the CSI reporting information associated with different types of layers are different;

[0174] Or,

[0175] the input information of the AI models associated with different types of layers is different;

[0176] Or,

[0177] the AI models associated with different types of layers are different;

[0178] Or,

[0179] the CSI reporting parameters associated with different types of layers are different, and the CSI reporting parameters include at least one of the following: CSI reporting content, CSI reporting location, and CSI reporting priority.

[0180] The following illustrates this embodiment by way of example in different cases:

[0181] Case 1: The payloads of the CSI reporting information associated with different types of layers are different. For example, the layer of type 1 is associated with a first payload, or the layer of type 2 is associated with a second payload, so that the payloads are different between different types of layers. Furthermore, the correlation between the layers can be utilized to improve the effect of CSI processing based on AI. For example, for CSI information compression, the information of the layer of type 1 can be used to compress the CSI information of the layer of type 2 more.

[0182] Case 2: The input information of the AI models associated with different types of layers is different. For example, the input information of the AI model corresponding to the layer of type 1 does not include the information of other layers, and the input information of the AI model corresponding to the layer of type 2 includes the information of other layers.

[0183] Case 3: The AI models associated with different types of layers are different. For example, the layer of type 1 is associated with AI model 1, and the layer of type 2 is associated with AI model 2.

[0184] Case 4: The CSI reporting parameters associated with different types of layers are different. For example, the layer of type 1 is associated with a first CSI reporting parameter. For example, the above first CSI reporting parameter may include at least one of a first CSI reporting content, a first CSI reporting location, and a first CSI reporting priority; the layer of type 2 is associated with a second CSI reporting parameter, and the above second CSI reporting parameter may include at least one of a second CSI reporting content, a second CSI reporting location, and a second CSI reporting priority.

[0185] It should be noted that the above various situations can be reasonably combined according to actual needs, and this embodiment does not make any limitations in this regard.

[0186] Optionally, the first information related to the first object includes the first information related to the payload for each type of layer among the at least two types of layers.

[0187] For example, if the at least one layer includes a first type of layer and a second type of layer, the above first information related to the first object may include the first information related to the payload of the first type of layer and the first information related to the payload of the second type of layer. It can be understood that each of the above types of layers may include at least one layer.

[0188] It should be noted that the first information related to the payload in this embodiment can refer to the relevant descriptions in the foregoing embodiments, and will not be elaborated herein.

[0189] Optionally, the first information related to the first object includes the first information related to the layer information for each type of layer among the at least two types of layers. Exemplarily, the above layer information may include, but is not limited to, at least one of layer type, layer identifier, layer priority, etc.

[0190] Optionally, the at least two types of layers include a first type of layer and a second type of layer, wherein the AI model corresponding to the first type of layer is used to process based on the channel state information corresponding to the first type of layer, and the AI model corresponding to the second type of layer is used to process based on the channel state information corresponding to the second type of layer and the information of the associated layer of the second type of layer.

[0191] In this embodiment, the AI model corresponding to the first type of layer processes only based on the channel state information corresponding to the first type of layer, that is, the first type of layer can be understood as a layer that independently performs encoding or decoding processing.

[0192] The AI model corresponding to the second type of layer is used to process based on the channel state information corresponding to the second type of layer and the information of the associated layer of the second type of layer, wherein the associated layer of the second type of layer may include at least one layer other than the second type of layer among the at least one layer. That is, the second type of layer can be understood as a layer that needs to combine the information of other layers for encoding or decoding processing.

[0193] In some optional embodiments, the associated layer of the second type of layer may be the first type of layer, that is, the second type of layer is a layer that needs to combine the information of the first type of layer for encoding or decoding processing.

[0194] Optionally, the information of the associated layer of the second type of layer includes at least one of the following: the output information of the AI model corresponding to the associated layer of the second type of layer, the input information of the AI model corresponding to the associated layer of the second type of layer, and the output information of the intermediate layer of the AI model corresponding to the associated layer of the second type of layer.

[0195] Optionally, the associated layer of the first target layer in the second type of layer includes at least one of the following: the first layer, the second layer, and the third layer;

[0196] Wherein, both the first target layer and the first layer are layers corresponding to the first time unit, and the first layer is the lower layer of the first target layer, or the index of the first layer is the index of the first target layer minus 1;

[0197] The second layer is the layer corresponding to the second time unit, the second time unit is the time unit before the first time unit, and the index of the second layer is the same as the index of the first target layer;

[0198] The third layer is the top layer or the layer with the largest index among all the layers corresponding to the third time unit, or the third layer is the top layer or the layer with the largest index among all the layers corresponding to the third time unit where the AI model corresponding to the third time unit processes using the information of the lower layer, and the third time unit is the time unit before the first time unit.

[0199] In this embodiment, the above first target layer can be any layer in the above second type of layer.

[0200] The above first time unit can be any time unit, and the time unit can include but is not limited to time slots, sub - time slots, frames, or sub - frames, etc.

[0201] The above second time unit is the time unit before the first time unit. Exemplarily, the above second time unit can be the first K1 time units before the first time unit, K1 is a positive integer. For example, the value of K1 can be 1 or 2, etc. In some alternative embodiments, the above second time unit is the time unit when the AI model was last used for CSI processing before the first time unit.

[0202] The above third time unit and the above second time unit can be the same time unit or different time units. Exemplarily, the above second time unit is the first K2 time units before the first time unit, K2 is a positive integer. For example, the value of K2 can be 1 or 2, etc. In some alternative embodiments, the above third time unit can be the time unit when the AI model was last used for CSI processing before the first time unit.

[0203] It should be noted that in the case where the above-mentioned second type of layer includes multiple layers, the AI model corresponding to each layer in the above-mentioned multiple layers processes based on the channel state information corresponding to that layer and the information of the associated layer of that layer. In addition, the types of associated layers of different layers in the above-mentioned multiple layers can be different. For example, the associated layer of some layers is the first layer, the associated layer of some layers is the second layer, and the associated layer of some layers includes the first layer and the second layer, etc.; or the types of associated layers of different layers in the above-mentioned multiple layers can be the same. For example, the associated layer of each layer in the above-mentioned multiple layers is the first layer, or the second layer, or includes the first layer and the second layer, etc. For example, in the above-mentioned second type of layer including layer 1, layer 2, and layer 3 of the first time unit, where the associated layer of layer 1 of the first time unit can include the first layer, that is, the associated layer of layer 1 of the first time unit is layer 0 of the first time unit; the associated layer of layer 2 of the first time unit can include the first layer and the second layer, that is, the associated layer of layer 2 of the first time unit includes layer 1 of the first time unit and layer 2 of the second time unit; the associated layer of layer 3 of the first time unit can include the second layer, that is, the associated layer of layer 3 of the first time unit is layer 3 of the second time unit.

[0204] Exemplarily, the above-mentioned first layer can be understood as a layer that transmits information using the first transmission method, where the above-mentioned first transmission method is: transmitting the information of Layer x at time t to Layer (x + 1) at time t, x < K t -1, K t represents the value of the rank corresponding to time t.

[0205] The above-mentioned second layer can be understood as a layer that transmits information using the second transmission method, where the above-mentioned second transmission method is: transmitting the information of Layer x at time t to Layer x at time t + 1, x < K t -1 and x < K t+1 -1, the above-mentioned K t+1 represents the value of the rank corresponding to time t + 1.

[0206] The above-mentioned third layer can be understood as a layer that transmits information using the third transmission method, where the above-mentioned third transmission method is: transmitting the information of Layer (K t -1) or the information of the highest Layer among all the layers that transmit information using the first transmission method at time t, to Layer 0 at time t + 1, or to the lowest Layer among all the layers that transmit information using the first transmission method at time t + 1.

[0207] It should be noted that the above-mentioned transmitting information to a certain layer can be understood as transmitting the information to the AI model corresponding to that layer for processing. In addition, the above-mentioned various transmission methods can be combined arbitrarily, and the following is an example in combination with the accompanying drawings:

[0208] Example 1: Some layers transmit information in the first transmission mode, that is, transmitting the information of Layer x at time t to Layer (x + 1) at time t; some layers transmit information in the third transmission mode, that is, transmitting the information of Layer x at time t to Layer (x + 1) at time t, and transmitting the information of Layer (K t - 1) at time t to Layer 0 at time t + 1. For example, as Figure 4a shown, Layer 0 to Layer 2 at each time transmit information in the first transmission mode, and Layer 3 at each time except the last time transmits information in the third transmission mode.

[0209] Example 2: Some layers use the second transmission mode to transmit information, that is, transmitting the information of Layer x at time t to Layer x at time t + 1. For example, as Figure 4b shown, layers at each time except the last time use the second transmission mode to transmit information.

[0210] Example 3: Some low - layer Layers transmit information in the first transmission mode, some low - layer Layers transmit information in the third mode, and some high - layer Layers transmit information in the second transmission mode. For example, as Figure 4c and Figure 4d shown, among them, in Figure 4c , the above - mentioned low - layer Layers include Layer 0 and Layer 1, and the above - mentioned high - layer Layers include Layer 2 and Layer 3; in Figure 4d , the above - mentioned low - layer Layers include Layer 0, Layer 1 and Layer 2, and the above - mentioned high - layer Layers include Layer 3.

[0211] Example 4: Some low - layer Layers transmit information in the second transmission mode, some high - layer Layers transmit information in the first transmission mode, and some high - layer Layers transmit information in the third transmission mode. For example, as Figure 4e shown, among them, in Figure 4e , the above - mentioned low - layer Layers include Layer 0 and Layer 1, and the above - mentioned high - layer Layers include Layer 2 and Layer 3.

[0212] Example 5: Some Layers transmit information using both the first transmission mode and the third transmission mode at the same time, some Layers transmit information using the second transmission mode, and some Layers transmit information using the first transmission mode. For example, as Figure 4f shown.

[0213] Example 6: Some Layers transmit information in the first transmission mode, and some Layers transmit information in both the first transmission mode and the second transmission mode. For example, as Figure 4g , where, in Figure 4g , Layers 0 and 1 transmit information in both the first transmission mode and the second transmission mode, and Layer 2 transmits information in the first transmission mode.

[0214] It should be noted that the arrows in the above Figures 4a to 4g indicate the transmission of information.

[0215] Optionally, when the first target layer includes the fourth layer, the associated layer of the fourth layer includes at least one of the second layer and the third layer;

[0216] Or,

[0217] when the first target layer includes the fifth layer, the associated layer of the fifth layer includes at least one of the first layer and the second layer;

[0218] wherein, the fourth layer is the bottom layer or the layer with the smallest index among all the layers corresponding to the second time unit, or, the fourth layer is the bottom layer or the layer with the smallest index among all the layers that transmit information to its upper layer among all the layers corresponding to the second time unit;

[0219] The fifth layer is a layer different from the fourth layer among all the layers corresponding to the second time unit, or, the fifth layer is a layer different from the fourth layer among all the layers that transmit information to its upper layer among all the layers corresponding to the second time unit.

[0220] Exemplarily, as Figure 4a shown, the above fourth layer is Layer 0, and the above fifth layer includes Layers 1 to 3; or, as Figure 4e shown, the above fourth layer is Layer 2, and the above fifth layer is Layer 3.

[0221] Optionally, the associated layer of the second target layer in the second type of layer includes at least one of the following: the sixth layer, the seventh layer, and the eighth layer;

[0222] wherein, both the second target layer and the sixth layer are layers corresponding to the fourth time unit, the sixth layer is the upper layer of the second target layer, or, the index of the sixth layer is the index of the second target layer plus 1;

[0223] The seventh layer is a layer corresponding to the fifth time unit, the fifth time unit is a time unit before the fourth time unit, and the index of the seventh layer is the same as the index of the second target layer;

[0224] The eighth layer is the bottom layer or the layer with the smallest index among all the layers corresponding to the sixth time unit, or the eighth layer is the bottom layer or the layer with the smallest index among all the layers corresponding to the sixth time unit where the AI model corresponding to the sixth time unit processes information using the information of the higher layer. The sixth time unit is the time unit before the fourth time unit.

[0225] In this embodiment, the above-mentioned second target layer can be any layer in the second type of layer.

[0226] The above-mentioned fourth time unit can be any time unit, and the time unit can include but is not limited to time slots, sub-time slots, frames, or sub-frames, etc.

[0227] The above-mentioned fifth time unit is the time unit before the fourth time unit. Exemplarily, the above-mentioned fifth time unit can be the first K3 time units before the fourth time unit, where K3 is a positive integer, for example, 1 or 2, etc. In some optional embodiments, the above-mentioned fifth time unit is the time unit when the AI model was last used for CSI processing before the fourth time unit.

[0228] The above-mentioned sixth time unit and the above-mentioned fifth time unit can be the same time unit or different time units. Exemplarily, the above-mentioned fifth time unit is the first K4 time units before the fourth time unit, where K4 is a positive integer, for example, 1 or 2, etc. In some optional embodiments, the above-mentioned sixth time unit can be the time unit when the AI model was last used for CSI processing before the fourth time unit.

[0229] It should be noted that in the case where the above-mentioned second type of layer includes multiple layers, the AI model corresponding to each layer in the above-mentioned multiple layers processes based on the channel state information corresponding to the layer and the information of the associated layer of the layer. In addition, the types of the associated layers of different layers in the above-mentioned multiple layers may be different. For example, the associated layer of some layers is the sixth layer, the associated layer of some layers is the seventh layer, and the associated layer of some layers includes the sixth layer and the seventh layer, etc.; or the types of the associated layers of different layers in the above-mentioned multiple layers may be the same. For example, the associated layer of each layer in the above-mentioned multiple layers is the sixth layer, or the seventh layer, or both include the sixth layer and the seventh layer, etc. For example, the above-mentioned second type of layer includes layer 1, layer 2, and layer 3 of the fourth time unit. Among them, the associated layer of layer 1 of the fourth time unit may include the sixth layer, that is, the associated layer of layer 1 of the fourth time unit is layer 2 of the fourth time unit; the associated layer of layer 2 of the fourth time unit may include the sixth layer and the seventh layer, that is, the associated layer of layer 2 of the fourth time unit includes layer 3 of the fourth time unit and layer 2 of the fifth time unit; the associated layer of layer 3 of the fourth time unit may include the seventh layer, that is, the associated layer of layer 3 of the fourth time unit is layer 3 of the fifth time unit.

[0230] Exemplarily, the above-mentioned sixth layer can be understood as a layer that transmits information using the fourth transmission method, where the above-mentioned fourth transmission method is: transmitting the information of Layer x at time t to Layer (x - 1) at time t, where x < K t -1, K t represents the value of the rank corresponding to time t.

[0231] The above-mentioned seventh layer can be understood as a layer that transmits information using the second transmission method, where the above-mentioned second transmission method is: transmitting the information of Layer x at time t to Layer x at time t + 1, where x < K t -1 and x < K t+1 -1, the above-mentioned K t+1 represents the value of the rank corresponding to time t + 1.

[0232] The above-mentioned eighth layer can be understood as a layer that transmits information using the fifth transmission method, where the above-mentioned fifth transmission method is: transmitting the information of Layer 0 at time t or the lowest Layer of all layers that transmit information using the fourth transmission method, to Layer K at time t + 1 t+1 -1, or transmitting it to the highest Layer of all layers that transmit information using the fourth transmission method at time t + 1, where K t+1 represents the value of the rank at time t + 1.

[0233] It should be noted that the above-mentioned passing of information to a certain layer can be understood as passing the information to the AI model corresponding to that layer for processing. In addition, the above various passing methods can be combined arbitrarily, and the following is an example with reference to the accompanying drawings:

[0234] Example 1: Some layers pass information according to the fourth passing method, that is, the information of Layer x at time t is passed to Layer x - 1 at time t, and some layers pass information according to the fifth passing method, that is, the information of Layer 0 at time t is passed to Layer K at time t + 1 t+1 -1. For example, as Figure 5a shown, the information of Layer 1 to Layer 3 at each moment is passed according to the fourth passing method, and the information of Layer 0 at each moment except the last moment is passed according to the fifth passing method.

[0235] Example 2: Some low layers pass information according to the fourth passing method, some low layers pass information according to the fifth passing method, and some high layers pass information according to the second passing method. For example, as Figure 5b and Figure 5c shown, among them, in Figure 5b , the above-mentioned low layers include Layer 0 and Layer 1, and the above-mentioned high layers include Layer 2 and Layer 3; in Figure 5c , the above-mentioned low layers include Layer 0, Layer 1 and Layer 2, and the above-mentioned high layers include Layer 3.

[0236] Example 4: Some low layers pass information according to the second passing method, some high layers pass information according to the fourth passing method, and some high layers pass information according to the fifth passing method. For example, as Figure 5d shown, among them, in Figure 5d , the above-mentioned low layers include Layer 0 and Layer 1, and the above-mentioned high layers include Layer 2 and Layer 3.

[0237] Example 5: Some layers pass information according to both the second passing method and the fourth passing method at the same time, and some layers pass information according to the second passing method. For example, as Figure 5e shown.

[0238] Example 6: Some layers pass information using the fourth passing method, and some layers pass information using both the fourth passing method and the second passing method at the same time. For example, as Figure 5f , among them, in Figure 5fIn it, Layer 0 and Layer 1 simultaneously use the fourth transfer method and the second transfer method to transfer information, and Layer 2 and Layer 3 use the fourth transfer method to transfer information.

[0239] It should be noted that Figures 5a to 5f each arrow in the above indicates the information transfer path.

[0240] Optionally, when the second target layer includes the ninth layer, the associated layer of the ninth layer includes at least one of the seventh layer and the eighth layer;

[0241] Or,

[0242] when the second target layer includes the tenth layer, the associated layer of the tenth layer includes at least one of the sixth layer and the seventh layer;

[0243] Wherein, the ninth layer is the top layer or the layer with the largest index among all the layers corresponding to the sixth time unit, or, the ninth layer is the top layer or the layer with the largest index among all the layers that transfer information to its lower layer among all the layers corresponding to the sixth time unit;

[0244] The tenth layer is a layer different from the ninth layer among all the layers corresponding to the sixth time unit, or, the tenth layer is a layer different from the ninth layer among all the layers that transfer information to its upper layer among all the layers corresponding to the sixth time unit.

[0245] Exemplarily, as Figure 5a shown, the above-mentioned ninth layer is Layer 0, and the above-mentioned tenth layer includes Layer 1 to Layer 3; or, as Figure 5d shown, the above-mentioned ninth layer is Layer 2, and the above-mentioned tenth layer is Layer 3.

[0246] Optionally, the CSI reporting information corresponding to the second type of layer includes the encoded CSI information corresponding to the second type of layer and the relevant information of the associated layer of the second type of layer.

[0247] In this embodiment, the CSI reporting information corresponding to the second type of layer includes the relevant information of the associated layer of the second type of layer. For example, the identifier of the associated layer of the second type of layer, which facilitates the decoding end to quickly determine the associated layer of the second type of layer. Furthermore, the decoding end can obtain the information of the associated layer of the second type of layer to decode the encoded CSI information corresponding to the second type of layer, and obtain the decoded CSI information or decompressed CSI information or restored CSI information or reconstructed CSI information.

[0248] Optionally, the CSI reporting information corresponding to the second type of layer includes layer type information.

[0249] Optionally, the CSI reporting information includes layer type information.

[0250] Optionally, the CSI reporting information includes at least one of first data set information, layer information, payload information, and encoded CSI information.

[0251] Optionally, the relevant information of the associated layer of the second type of layer includes at least one of the index of the associated layer of the second type of layer and the time unit corresponding to the associated layer of the second type of layer.

[0252] In some alternative embodiments, the associated layer of the second type of layer above is the first type of layer, that is, the AI model corresponding to the second type of layer above needs to be compressed or decompressed based on the information of the first type of layer (i.e., the output information of the AI model corresponding to the first type of layer). Correspondingly, the CSI reporting information corresponding to the second type of layer above may include at least one of the encoded CSI information corresponding to the second type of layer and the index, time unit, etc. of its associated first type of layer. For example, as Figure 6 shown, the AI model corresponding to the P type of layer needs to be encoded or decoded based on the information of the I type of layer.

[0253] Optionally, the payload of the CSI reporting information corresponding to the second type of layer is smaller than the payload of the CSI reporting information corresponding to the first type of layer.

[0254] Optionally, the at least two types of layers further include a third type of layer, and the payload of the CSI reporting information corresponding to the third type of layer is smaller than the payload of the CSI reporting information corresponding to the second type of layer.

[0255] Exemplarily, as Figure 6 shown, the payload of the CSI reporting information corresponding to the B type of layer is smaller than the payload of the CSI reporting information corresponding to the P type of layer, and the payload of the CSI reporting information corresponding to the P type of layer is smaller than the payload of the CSI reporting information corresponding to the I type of layer.

[0256] Optionally, the AI model corresponding to the third type of layer is used to process based on the channel state information corresponding to the third type of layer and the information of the associated layer of the third type of layer, and the associated layer of the third type of layer includes the first type of layer and the second type of layer.

[0257] Exemplarily, as Figure 6As shown, the AI model corresponding to the layer of type B needs to process by combining the information of the layer of type P and the layer of type I. For example, for the encoding end, the AI model corresponding to the layer of type B needs to encode the CSI of the layer of type B by combining the information of the layer of type P and the layer of type I; for the decoding end, the AI model corresponding to the layer of type B needs to decode the CSI of the layer of type B by combining the information of the layer of type P and the layer of type I.

[0258] It can be understood that, for the encoding end, the channel state information corresponding to the above-mentioned third type of layer is the channel state information before encoding; for the decoding end, the channel state information corresponding to the above-mentioned third type of layer can be the channel state information after encoding.

[0259] Optionally, the CSI reporting information corresponding to the layer of the first type is located before the CSI reporting information corresponding to the layer of the second type, or,

[0260] the CSI reporting information corresponding to the layer of the first type is located in a specific CSI reporting group.

[0261] Exemplarily, the CSI reporting information corresponding to the layer of the first type can be located in a specific CSI reporting group among multiple CSI reporting groups (Groups) predefined by the protocol, where the above-mentioned specific CSI reporting group can be predefined by the protocol or can be configured by the network layer device.

[0262] Optionally, the Nth layer in the at least one layer is the layer of the first type, or, the layer of the Mth group in the at least one layer is the layer of the first type, and both M and N are positive integers predefined by the protocol.

[0263] Exemplarily, it can be predefined by the protocol that the Nth layer in the at least one layer related to the CSI reporting information is the layer of the first type, or, it can be predefined by the protocol that the layer of the Mth group in the at least one layer related to the CSI reporting information is the layer of the first type.

[0264] Optionally, the CSI reporting information associated with one reporting identifier only includes the CSI reporting information corresponding to one layer of the first type.

[0265] Exemplarily, it can be predefined by the protocol that the CSI reporting information associated with one reporting identifier (Report ID) only includes the CSI reporting information corresponding to one layer of the first type, which is beneficial to reducing the payload of the CSI reporting information and thus saving CSI reporting resources.

[0266] Optionally, the CSI reporting information carried on the CSI reporting resources used for one CSI reporting only includes the CSI reporting information corresponding to one layer of the first type.

[0267] In this embodiment, the CSI reporting resources used for one CSI report can carry CSI reporting information associated with one or more reporting identifiers. Among them, the above CSI reporting resources can include, but are not limited to, Physical Uplink Sharing Channel (PUSCH), Physical Uplink Control Channel (PUCCH), etc.

[0268] In this embodiment, the CSI reporting information carried on the CSI reporting resources used for one CSI report only includes the CSI reporting information corresponding to the first type of layer, which helps to reduce the payload of the CSI reporting information and thus save CSI reporting resources.

[0269] Optionally, the payload includes at least one of the following: the bit size of the CSI reporting information, the number of floating-point numbers (float) of the CSI reporting information, the number of elements of the CSI reporting information, the output dimension of the first AI model, the quantization method of each element or floating-point number in the CSI reporting information, and the quantization method of the CSI reporting information.

[0270] In this embodiment, the number of elements of the above CSI reporting information can also be referred to as the number of parameters included in the CSI reporting information. The output dimension of the first AI model can be understood as the number of output elements or output parameters of the first AI model.

[0271] Optionally, the method further includes:

[0272] The communication device sends or receives capability information related to the first object.

[0273] Exemplarily, the above capability information related to the first object can include capability information related to the payload. For example, an AI model supported for the payload, or the above capability information related to the first object can include capability information related to at least one layer or at least one rank. For example, an IA model supported for each layer.

[0274] It should be noted that if step 301 above is that the communication device receives the first information related to the first object, then this embodiment is that the communication device sends the capability information related to the first object; if step 301 above is that the communication device sends the first information related to the first object, then this embodiment is that the communication device receives the capability information related to the first object.

[0275] For example, the terminal reports to the network-side device the capability information of the terminal related to the first object. Then, the network-side device can send the first information related to the first object to the terminal based on the capability information of the terminal related to the first object. Alternatively, the network-side device can send the capability information of the network-side device related to the first object to the terminal. Then, the terminal can send the first information related to the first object to the network-side device based on the capability information of the network-side device related to the first object.

[0276] Optionally, the first object includes the payload, and the capability information related to the payload includes at least one of the following:

[0277] Candidate values of the supported payload;

[0278] Supported candidate quantization methods;

[0279] Supported AI models related to the payload.

[0280] Exemplarily, if the capability information of the terminal reported related to the payload includes the candidate values of the supported payload, the network-side device can select a payload value from the candidate values of the supported payload, determine the first information related to the payload value, and send it to the terminal.

[0281] Optionally, the first object includes the at least one layer or at least one rank; the capability information related to the at least one layer or at least one rank includes at least one of the following:

[0282] AI models supported by each layer;

[0283] Supported layer types.

[0284] Exemplarily, the above layer types may include at least one of a first type (e.g., I type), a second type (e.g., P type), and a third type (e.g., B type).

[0285] Optionally, the first information includes at least one of the following: layer type, number of layers, layer compression method, payload information.

[0286] Exemplarily, the above layer types may include at least one of the above first type, second type, and third type.

[0287] Exemplarily, CSI reporting information may be generated based on the first information. For example, if the first information includes a layer type, the communication device may generate CSI reporting information corresponding to the layer type. Or, if the first information includes a layer compression method, the communication device may use the layer compression method to compress and obtain the CSI reporting information. Or, the CSI reporting information may be decoded based on the first information. For example, if the first information includes a layer compression method, the communication device may use the decompression method corresponding to the layer compression method to decode the CSI reporting information.

[0288] Optionally, the communication device is a terminal, and the method further includes:

[0289] The communication device reports CSI reporting information related to the first information.

[0290] Exemplarily, the CSI reporting information related to the first information may be CSI reporting information determined based on the first information. It can be understood that this CSI reporting information is CSI reporting information based on AI processing.

[0291] Optionally, before the communication device receives or sends the first information related to the first object, the method further includes:

[0292] The communication device sends or receives a first request message, where the first request message is used to request the first information, and the first request message includes at least one of the following: payload, layer type, number of layers, layer compression method.

[0293] It should be noted that if step 301 above is that the communication device receives the first information related to the first object, then in this embodiment, the communication device sends the first request message; if step 301 above is that the communication device sends the first information related to the first object, then in this embodiment, the communication device receives the first request message.

[0294] Exemplarily, the terminal sends a first request message to the network device, and the first request message includes information related to the first object, such as payload, layer type, number of layers, layer compression method; the network device sends the first information related to the first object to the terminal based on the first request message; or, the network device sends a first request message to the terminal, and the first request message includes information related to the first object, such as payload, layer type, number of layers, layer compression method; the terminal sends the first information related to the first object to the network device based on the first request message. For example, the first request message sent by the network to the terminal device includes payload information and dataset request information, and the terminal sends the dataset that meets the payload information to the network device.

[0295] It should be noted that the encoding involved in the embodiments of the present application may also be referred to as compression, and the decoding involved in the embodiments of the present application may also be referred to as decompression.

[0296] It should be noted that for the information transmission method provided in the embodiments of the present application, the execution subject may be an information transmission device, or a control module in the information transmission device for executing the information transmission method. In the embodiments of the present application, taking the information transmission device executing the information transmission method as an example, the information transmission device provided in the embodiments of the present application is described.

[0297] Please refer to Figure 7 , Figure 7 which is a structural diagram of an information transmission device provided in the embodiments of the present application. As Figure 7 shown, the information transmission device 700 includes:

[0298] a transceiver module, configured to receive or send first information related to a first object;

[0299] wherein, the first object includes at least one of the following: payload, at least one layer, at least one rank; the first information is used to determine a first artificial intelligence (AI) model or to configure AI-based channel state information (CSI) reporting information, and the first AI model is used to generate the CSI reporting information or to decode the CSI reporting information.

[0300] Optionally, the first information related to the first object includes at least one of the following:

[0301] first dataset information related to the payload;

[0302] first AI model information related to the payload;

[0303] first quantization information related to the payload.

[0304] Optionally, one first dataset information is associated with one CSI reporting information of the payload;

[0305] Or,

[0306] different first dataset information is associated with different CSI reporting information of the payload.

[0307] Optionally, one first dataset information is associated with multiple CSI reporting information of the payload.

[0308] Optionally, the first dataset information includes at least one of the following:

[0309] At least one dataset information, where the at least one dataset information includes at least one of the input information of the AI model and the output information of the AI model;

[0310] Payload information associated with at least one dataset information;

[0311] Dataset identifier associated with at least one dataset information;

[0312] AI model information associated with at least one dataset information.

[0313] Optionally, one first AI model information is associated with CSI reporting information of one type of payload;

[0314] Or,

[0315] Different first AI model information is associated with CSI reporting information of different payloads.

[0316] Optionally, the first AI model information includes:

[0317] Identifier information of the AI model;

[0318] Payload information associated with the AI model;

[0319] Type information or function information of the AI model;

[0320] Parameter information of the AI model.

[0321] Optionally, one first quantization information is associated with CSI reporting information of one type of payload;

[0322] Or,

[0323] One first quantization information is associated with CSI reporting information of multiple types of payloads.

[0324] Optionally, the first quantization information includes at least one of the following:

[0325] Quantization method corresponding to the output dimension of the AI model, where the output dimension of the AI model is used to indicate the number of output elements of the AI model;

[0326] Quantization method corresponding to the payload size, where the payload size is the payload size of the CSI reporting information.

[0327] Optionally, the quantization methods corresponding to different output dimensions of the AI model are different, or the quantization methods corresponding to the output dimensions of the AI model within different output dimension value ranges are different;

[0328] Or,

[0329] The quantization methods are different for different payload sizes, or the quantization methods are different for payload sizes within different value ranges of the payload size.

[0330] Optionally, the first information related to the first object includes at least one of the following:

[0331] The first information related to the payload of one layer in the at least one layer;

[0332] The first information related to the payload of each layer in the at least one layer;

[0333] The first information related to the payload of all layers in the at least one layer.

[0334] Optionally, the at least one layer includes at least two types of layers;

[0335] Among them, the payloads of the CSI reporting information associated with different types of layers are different;

[0336] Or,

[0337] The input information of the AI models associated with different types of layers is different;

[0338] Or,

[0339] The AI models associated with different types of layers are different;

[0340] Or,

[0341] The CSI reporting parameters associated with different types of layers are different, and the CSI reporting parameters include at least one of the following: CSI reporting content, CSI reporting location, CSI reporting priority.

[0342] Optionally, the first information related to the first object includes the first information related to the payload of each type of layer in the at least two types of layers.

[0343] Optionally, the at least two types of layers include a first type of layer and a second type of layer. Among them, the AI model corresponding to the first type of layer is used to process based on the channel state information corresponding to the first type of layer, and the AI model corresponding to the second type of layer is used to process based on the channel state information corresponding to the second type of layer and the information of the associated layer of the second type of layer.

[0344] Optionally, the information of the associated layer of the second type of layer includes at least one of the following: the output information of the AI model corresponding to the associated layer of the second type of layer, the input information of the AI model corresponding to the associated layer of the second type of layer, and the output information of the intermediate layer of the AI model corresponding to the associated layer of the second type of layer.

[0345] Optionally, the associated layer of the first target layer in the second type of layer includes at least one of the following: the first layer, the second layer, and the third layer;

[0346] Wherein, both the first target layer and the first layer are layers corresponding to the first time unit, and the first layer is the layer one level lower than the first target layer, or the index of the first layer is the index of the first target layer minus 1;

[0347] The second layer is the layer corresponding to the second time unit, the second time unit is the time unit before the first time unit, and the index of the second layer is the same as the index of the first target layer;

[0348] The third layer is the top layer or the layer with the largest index among all the layers corresponding to the third time unit, or the third layer is the top layer or the layer with the largest index among all the layers corresponding to the third time unit where the AI model corresponding to these layers processes using the information of the layer one level lower. The third time unit is the time unit before the first time unit.

[0349] Optionally, when the first target layer includes the fourth layer, the associated layer of the fourth layer includes at least one of the second layer and the third layer;

[0350] Or,

[0351] When the first target layer includes the fifth layer, the associated layer of the fifth layer includes at least one of the first layer and the second layer;

[0352] Wherein, the fourth layer is the bottom layer or the layer with the smallest index among all the layers corresponding to the second time unit, or the fourth layer is the bottom layer or the layer with the smallest index among all the layers corresponding to the second time unit that transmit information to the layer one level higher;

[0353] The fifth layer is a layer different from the fourth layer among all the layers corresponding to the second time unit, or the fifth layer is a layer different from the fourth layer among all the layers corresponding to the second time unit that transmit information to the layer one level higher.

[0354] Optionally, the associated layer of the second target layer in the second type of layer includes at least one of the following: the sixth layer, the seventh layer, and the eighth layer;

[0355] Among them, both the second target layer and the sixth layer are layers corresponding to the fourth time unit, the sixth layer is the layer one level higher than the second target layer, or the index of the sixth layer is the index of the second target layer plus 1;

[0356] The seventh layer is the layer corresponding to the fifth time unit, the fifth time unit is the time unit before the fourth time unit, and the index of the seventh layer is the same as the index of the second target layer;

[0357] The eighth layer is the bottom layer or the layer with the smallest index among all layers corresponding to the sixth time unit, or the eighth layer is the bottom layer or the layer with the smallest index among all layers corresponding to the sixth time unit for which the corresponding AI model processes using information from the layer one level higher, and the sixth time unit is the time unit before the fourth time unit.

[0358] Optionally, when the second target layer includes the ninth layer, the associated layer of the ninth layer includes at least one of the seventh layer and the eighth layer;

[0359] Or,

[0360] When the second target layer includes the tenth layer, the associated layer of the tenth layer includes at least one of the sixth layer and the seventh layer;

[0361] Among them, the ninth layer is the top layer or the layer with the largest index among all layers corresponding to the sixth time unit, or the ninth layer is the top layer or the layer with the largest index among all layers corresponding to the sixth time unit that transmit information to the layer one level lower;

[0362] The tenth layer is the layer different from the ninth layer among all layers corresponding to the sixth time unit, or the tenth layer is the layer different from the ninth layer among all layers corresponding to the sixth time unit that transmit information to the layer one level higher.

[0363] Optionally, the CSI reporting information corresponding to the second type of layer includes the encoded CSI information corresponding to the second type of layer and the relevant information of the associated layer of the second type of layer.

[0364] Optionally, the relevant information of the associated layer of the second type of layer includes at least one of the index of the associated layer of the second type of layer and the time unit corresponding to the associated layer of the second type of layer.

[0365] Optionally, the payload of the CSI reporting information corresponding to the second type of layer is smaller than the payload of the CSI reporting information corresponding to the first type of layer.

[0366] Optionally, the at least two types of layers further include a third type of layer, and the payload of the CSI reporting information corresponding to the third type of layer is smaller than the payload of the CSI reporting information corresponding to the second type of layer.

[0367] Optionally, the AI model corresponding to the third type of layer is used to process based on the channel state information corresponding to the third type of layer and the information of the associated layer of the third type of layer, and the associated layer of the third type of layer includes the first type of layer and the second type of layer.

[0368] Optionally, the CSI reporting information corresponding to the first type of layer is located before the CSI reporting information corresponding to the second type of layer, or

[0369] the CSI reporting information corresponding to the first type of layer is located in a specific CSI reporting group.

[0370] Optionally, the Nth layer in the at least one layer is the first type of layer, or the layers in the Mth group in the at least one layer are the first type of layer, and both M and N are positive integers predefined by the protocol.

[0371] Optionally, the CSI reporting information associated with one reporting identifier only includes the CSI reporting information corresponding to one first type of layer.

[0372] Optionally, the CSI reporting information carried on the CSI reporting resource used for one CSI reporting only includes the CSI reporting information corresponding to one first type of layer.

[0373] Optionally, the payload includes at least one of the following: the bit size of the CSI reporting information, the number of floating-point numbers of the CSI reporting information, the number of elements of the CSI reporting information, the output dimension of the first AI model, the quantization method of each element or floating-point number in the CSI reporting information, the quantization method of the CSI reporting information.

[0374] Optionally, the transceiver module is further configured to:

[0375] send or receive capability information related to the first object.

[0376] Optionally, the first object includes the payload, and the capability information related to the payload includes at least one of the following:

[0377] candidate values of the supported payload;

[0378] supported candidate quantization methods;

[0379] Supported AI models related to the payload.

[0380] Optionally, the first object includes the at least one layer or at least one rank; the capability information related to the at least one layer or at least one rank includes at least one of the following:

[0381] AI models supported by each layer;

[0382] Types of supported layers.

[0383] Optionally, the first information includes at least one of the following: layer type, number of layers, layer compression method, payload indication information.

[0384] Optionally, the transceiver module is further configured to:

[0385] Report CSI reporting information related to the first information.

[0386] The information transmission device in the embodiments of the present application may be an electronic device, such as an electronic device with an operating system, or a component in an electronic device, such as an integrated circuit or a chip. The electronic device may be a terminal or a network-side device, or other devices other than terminals or network-side devices. Exemplarily, the terminal may include, but is not limited to, the types of the above-listed terminal 11, the network-side device may include, but is not limited to, the types of the above-listed network-side device 12, and other devices may be servers, Network Attached Storage (NAS), etc., which are not specifically limited in the embodiments of the present application.

[0387] The information transmission device provided in the embodiments of the present application can implement Figure 3 the various processes implemented by the method embodiments and achieve the same technical effects. To avoid repetition, details are not described here again.

[0388] Optionally, as Figure 8 shown, the embodiments of the present application further provide a communication device 800, including a processor 801 and a memory 802. A program or instruction that can run on the processor 801 is stored on the memory 802. For example, when the communication device 800 is a terminal, when the program or instruction is executed by the processor 801, it implements the various steps of the above information transmission method embodiments and can achieve the same technical effects. When the communication device 800 is a network-side device, when the program or instruction is executed by the processor 801, it implements the various steps of the above information transmission method embodiments and can achieve the same technical effects. To avoid repetition, details are not described here again.

[0389] An embodiment of the present application further provides a terminal, including a processor and a communication interface, where the communication interface is used to receive or send first information related to a first object; wherein, the first object includes at least one of the following: payload, at least one layer, at least one rank; the first information is used to determine a first artificial intelligence (AI) model or to configure AI-based channel state information (CSI) reporting information, and the first AI model is used to generate the CSI reporting information or to decode the CSI reporting information. Each implementation process and implementation manner of the above method embodiment can be applied to this terminal embodiment and can achieve the same technical effect. Specifically, Figure 9 FIG. is a schematic diagram of the hardware structure of a terminal according to an embodiment of the present application.

[0390] The terminal 900 includes, but is not limited to, at least some components such as a radio frequency unit 901, a network module 902, an audio output unit 903, an input unit 904, a sensor 905, a display unit 906, a user input unit 907, an interface unit 908, a memory 909, and a processor 910.

[0391] Those skilled in the art can understand that the terminal 900 may further include a power source (such as a battery) for supplying power to each component, and the power source may be logically connected to the processor 910 through a power management system, so as to implement functions such as management of charging, discharging, and power consumption management through the power management system. Figure 9 The terminal structure shown in does not limit the terminal. The terminal may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements, which will not be elaborated here.

[0392] It should be understood that in an embodiment of the present application, the input unit 904 may include a graphics processing unit (GPU) 9041 and a microphone 9042. The graphics processor 9041 processes image data of a static picture or video obtained by an image capturing device (such as a camera) in a video capture mode or an image capture mode. The display unit 906 may include a display panel 9061, and the display panel 9061 may be configured in the form of a liquid crystal display, an organic light emitting diode, etc. The user input unit 907 includes at least one of a touch panel 9071 and other input devices 9072. The touch panel 9071 is also called a touch screen. The touch panel 9071 may include two parts: a touch detection device and a touch controller. The other input devices 9072 may include, but are not limited to, a physical keyboard, function keys (such as volume control keys, power on / off keys, etc.), a trackball, a mouse, and a joystick, which will not be elaborated here.

[0393] In an embodiment of the present application, after receiving downlink data from a network-side device, the radio frequency unit 901 can transmit it to the processor 910 for processing; in addition, the radio frequency unit 901 can send uplink data to the network-side device. Generally, the radio frequency unit 901 includes, but is not limited to, an antenna, an amplifier, a transceiver, a coupler, a low-noise amplifier, a duplexer, etc.

[0394] The memory 909 can be used to store software programs or instructions and various data. The memory 909 mainly includes a first storage area for storing programs or instructions and a second storage area for storing data. Among them, the first storage area can store an operating system, applications or instructions required for at least one function (such as a sound playback function, an image playback function, etc.). In addition, the memory 909 can include volatile memory or non-volatile memory. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), a static random access memory (SRAM), a dynamic random access memory (DRAM), a synchronous dynamic random access memory (SDRAM), a double data rate synchronous dynamic random access memory (DDR SDRAM), an enhanced synchronous dynamic random access memory (ESDRAM), a synchronous link dynamic random access memory (SLDRAM), and a direct rambus random access memory (DRRAM). The memory 909 in the embodiment of the present application includes, but is not limited to, these and any other suitable types of memory.

[0395] The processor 910 can include one or more processing units; optionally, the processor 910 integrates an application processor and a modulation and demodulation processor. Among them, the application processor mainly processes operations related to the operating system, user interface, and applications, etc., and the modulation and demodulation processor mainly processes wireless communication signals, such as a baseband processor. It can be understood that the above modulation and demodulation processor may not be integrated into the processor 910.

[0396] Among them, a radio frequency unit 901 is configured to receive or transmit first information related to a first object. The first object includes at least one of the following: a payload, at least one layer, and at least one rank. The first information is used to determine a first artificial intelligence (AI) model or to configure AI-based channel state information (CSI) reporting information. The first AI model is used to generate the CSI reporting information or to decode the CSI reporting information.

[0397] It can be understood that the implementation processes of the implementation manners mentioned in this embodiment may refer to the relevant descriptions of the foregoing method embodiments and achieve the same or corresponding technical effects. To avoid repetition, they will not be elaborated herein.

[0398] An embodiment of this application further provides a network-side device, including a processor and a communication interface. The communication interface is configured to receive or transmit first information related to a first object. The first object includes at least one of the following: a payload, at least one layer, and at least one rank. The first information is used to determine a first artificial intelligence (AI) model or to configure AI-based channel state information (CSI) reporting information. The first AI model is used to generate the CSI reporting information or to decode the CSI reporting information. Each implementation process and implementation manner of the foregoing method embodiments can be applied to this network-side device embodiment and can achieve the same technical effects.

[0399] Specifically, an embodiment of this application further provides a network-side device. As Figure 10 shown, the network-side device 1000 includes: an antenna 1001, a radio frequency device 1002, a baseband device 1003, a processor 1004, and a memory 1005. The antenna 1001 is connected to the radio frequency device 1002. In the uplink direction, the radio frequency device 1002 receives information through the antenna 1001 and sends the received information to the baseband device 1003 for processing. In the downlink direction, the baseband device 1003 processes the information to be sent and sends it to the radio frequency device 1002. After processing the received information, the radio frequency device 1002 sends it out through the antenna 1001.

[0400] The method executed by the network-side device in the foregoing embodiments can be implemented in the baseband device 1003, and the baseband device 1003 includes a baseband processor.

[0401] The baseband device 1003 may include, for example, at least one baseband board, and a plurality of chips are provided on the baseband board. As Figure 10 shown, one of the chips is, for example, a baseband processor, which is connected to the memory 1005 through a bus interface to call a program in the memory 1005 and execute the operations of the network device shown in the foregoing method embodiments.

[0402] The network - side device may further include a network interface 1006, which is, for example, a Common Public Radio Interface (CPRI).

[0403] Specifically, the network - side device 1000 in the embodiments of the present application further includes: instructions or programs stored in the memory 1005 and executable on the processor 1004. The processor 1004 invokes the instructions or programs in the memory 1005 to execute Figure 7 the methods executed by the modules shown, and achieves the same technical effects. To avoid repetition, details are not described herein again.

[0404] The embodiments of the present application further provide a readable storage medium, on which a program or instructions are stored. When the program or instructions are executed by a processor, each process of the above - mentioned information - transmission method embodiment is implemented, and the same technical effects can be achieved. To avoid repetition, details are not described herein again.

[0405] Wherein, the processor is the processor in the terminal described in the above - mentioned embodiment. The readable storage medium includes a computer - readable storage medium, such as a computer read - only memory ROM, a random - access memory RAM, a magnetic disk, or an optical disc, etc. In some examples, the readable storage medium may be a non - transient readable storage medium.

[0406] The embodiments of the present application further provide a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement each process of the above - mentioned information - transmission method embodiment, and the same technical effects can be achieved. To avoid repetition, details are not described herein again.

[0407] It should be understood that the chip mentioned in the embodiments of the present application may also be referred to as a system - on - chip, system chip, chip system, or system - on - a - chip, etc.

[0408] The embodiments of the present application further provide a computer program / program product, which is stored in a storage medium. The computer program / program product is executed by at least one processor to implement each process of the above - mentioned information - transmission method embodiment, and the same technical effects can be achieved. To avoid repetition, details are not described herein again.

[0409] It should be noted that in this text, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements not only includes those elements but also other elements not explicitly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising a..." does not exclude the presence of additional identical elements in the process, method, article or device comprising that element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in the reverse order according to the functions involved. For example, the described methods may be performed in an order different from that described, and various steps may be added, omitted or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0410] From the description of the above embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of computer software products plus the necessary general hardware platforms, and of course, they can also be implemented by hardware. The computer software products are stored in storage media (such as ROM, RAM, magnetic disks, optical discs, etc.) and include several instructions for causing a terminal or a network-side device to execute the methods described in the various embodiments of the present application.

[0411] The embodiments of the present application have been described above in conjunction with the accompanying drawings. However, the present application is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present application, those of ordinary skill in the art can also make many forms of embodiments without departing from the purpose of the present application and the scope protected by the claims. These embodiments are all within the protection scope of the present application.

Claims

1. An information transmission method, characterized in that, including: A communication device receives or transmits first information related to a first object; wherein, the first object includes at least one of the following: payload, at least one layer, at least one rank; the first information is used to determine a first artificial intelligence (AI) model or to configure AI-based channel state information (CSI) reporting information, and the first AI model is used to generate the CSI reporting information or to decode the CSI reporting information.

2. The method according to claim 1, wherein The first information related to the first object includes at least one of the following: First dataset information related to the payload; First AI model information related to the payload; First quantization information related to the payload.

3. The method according to claim 2, wherein One first dataset information is associated with the CSI reporting information of one type of payload; or, Different first dataset information is associated with the CSI reporting information of different payloads.

4. The method according to claim 2, characterized in that, One first dataset information is associated with the CSI reporting information of multiple types of payloads.

5. The method according to any one of claims 2 to 4, characterized in that The first dataset information includes at least one of the following: At least one dataset information, wherein the at least one dataset information includes at least one of the input information and output information of the AI model; Payload information associated with at least one dataset information; Dataset identifier associated with at least one dataset information; AI model information associated with at least one dataset information.

6. The method according to any one of claims 2 to 5, characterized in that, One first AI model information is associated with the CSI reporting information of one type of payload; or, Different first AI model information is associated with the CSI reporting information of different payloads.

7. The method according to any one of claims 2 to 6, characterized in that, The first AI model information includes: Identifier information of the AI model; Payload information associated with the AI model; Type information or function information of the AI model; Parameter information of the AI model.

8. The method according to any one of claims 2 to 7, characterized in that One first quantization information is associated with the CSI reporting information of one type of payload; or, One first quantization information is associated with the CSI reporting information of multiple types of payloads.

9. The method according to any one of claims 2 to 8, characterized in that The first quantization information includes at least one of the following: Quantization method corresponding to the output dimension of the AI model, where the output dimension of the AI model is used to indicate the number of output elements of the AI model; Quantization method corresponding to the payload size, where the payload size is the payload size of the CSI reporting information.

10. The method according to claim 9, wherein The quantization methods corresponding to different AI model output dimensions are different, or the quantization methods corresponding to the AI model output dimensions within different output dimension value ranges are different; or, The quantization methods corresponding to different payload sizes are different, or the quantization methods corresponding to the payload sizes within different payload size value ranges are different.

11. The method according to any one of claims 1 to 10, characterized in that, The first information related to the first object includes at least one of the following: First information related to one layer in the at least one layer and related to the payload; First information related to each layer in the at least one layer and related to the payload; First information related to all layers in the at least one layer and related to the payload.

12. The method according to any one of claims 1 to 11, characterized in that, The at least one layer includes at least two types of layers; Among them, the payloads of the CSI reporting information associated with different types of layers are different; Or, the input information of the AI models associated with different types of layers is different; Or, the AI models associated with different types of layers are different; Or, the CSI reporting parameters associated with different types of layers are different, and the CSI reporting parameters include at least one of the following: CSI reporting content, CSI reporting location, CSI reporting priority.

13. The method according to claim 12, characterized in that, The first information related to the first object includes the first information related to the payload of each type of layer among the at least two types of layers.

14. The method according to claim 12 or 13, characterized in that, The at least two types of layers include a first type of layer and a second type of layer. Among them, the AI model corresponding to the first type of layer is used to process based on the channel state information corresponding to the first type of layer, and the AI model corresponding to the second type of layer is used to process based on the channel state information corresponding to the second type of layer and the information of the associated layer of the second type of layer.

15. The method according to claim 14, characterized in that, The information of the associated layer of the second type of layer includes at least one of the following: the output information of the AI model corresponding to the associated layer of the second type of layer, the input information of the AI model corresponding to the associated layer of the second type of layer, the output information of the intermediate layer of the AI model corresponding to the associated layer of the second type of layer.

16. The method according to claim 14 or 15, characterized in that, The associated layer of the first target layer in the second type of layer includes at least one of the following: the first layer, the second layer, the third layer; Among them, both the first target layer and the first layer are layers corresponding to the first time unit, and the first layer is the lower layer of the first target layer, or the index of the first layer is the index of the first target layer minus 1; The second layer is the layer corresponding to the second time unit, the second time unit is the time unit before the first time unit, and the index of the second layer is the same as the index of the first target layer; The third layer is the top layer or the layer with the largest index among all the layers corresponding to the third time unit, or the third layer is the top layer or the layer with the largest index among all the layers corresponding to the third time unit where the AI model corresponding to these layers uses the information of the lower layer for processing. The third time unit is the time unit before the first time unit.

17. The method according to claim 16, wherein In the case where the first target layer includes the fourth layer, the associated layer of the fourth layer includes at least one of the second layer and the third layer; Or, In the case where the first target layer includes the fifth layer, the associated layer of the fifth layer includes at least one of the first layer and the second layer; Among them, the fourth layer is the bottom layer or the layer with the smallest index among all the layers corresponding to the second time unit, or the fourth layer is the lowest layer or the layer with the smallest index among all the layers corresponding to the second time unit that transmit information to its upper layer; The fifth layer is the layer different from the fourth layer among all the layers corresponding to the second time unit, or the fifth layer is the layer different from the fourth layer among all the layers corresponding to the second time unit that transmit information to its upper layer.

18. The method according to any one of claims 14 to 17, characterized in that, The associated layer of the second target layer in the second type of layer includes at least one of the following: the sixth layer, the seventh layer, and the eighth layer; Wherein, both the second target layer and the sixth layer are layers corresponding to the fourth time unit, the sixth layer is the layer one higher than the second target layer, or the index of the sixth layer is the index of the second target layer plus 1; The seventh layer is the layer corresponding to the fifth time unit, the fifth time unit is the time unit before the fourth time unit, and the index of the seventh layer is the same as the index of the second target layer; The eighth layer is the bottom layer or the layer with the smallest index among all layers corresponding to the sixth time unit, or the eighth layer is the bottom layer or the layer with the smallest index among all layers corresponding to the sixth time unit where the AI model corresponding to the eighth layer processes using the information of the layer one higher, and the sixth time unit is the time unit before the fourth time unit.

19. The method according to claim 18, wherein When the second target layer includes the ninth layer, the associated layer of the ninth layer includes at least one of the seventh layer and the eighth layer; Or, When the second target layer includes the tenth layer, the associated layer of the tenth layer includes at least one of the sixth layer and the seventh layer; Wherein, the ninth layer is the top layer or the layer with the largest index among all layers corresponding to the sixth time unit, or the ninth layer is the top layer or the layer with the largest index among all layers corresponding to the sixth time unit that transmits information to the layer one lower; The tenth layer is the layer different from the ninth layer among all layers corresponding to the sixth time unit, or the tenth layer is the layer different from the ninth layer among all layers corresponding to the sixth time unit that transmits information to the layer one higher.

20. The method according to any one of claims 14 to 19, characterized in that, The CSI reporting information corresponding to the second type of layer includes the encoded CSI information corresponding to the second type of layer and the relevant information of the associated layer of the second type of layer.

21. The method according to claim 20, characterized in that, The relevant information of the associated layer of the second type of layer includes at least one of the index of the associated layer of the second type of layer and the time unit corresponding to the associated layer of the second type of layer.

22. The method according to any one of claims 14 to 21, characterized in that, The payload of the CSI reporting information corresponding to the second type of layer is less than the payload of the CSI reporting information corresponding to the first type of layer.

23. The method according to any one of claims 14 to 22, characterized in that The at least two types of layers further include a third type of layer, and the payload of the CSI reporting information corresponding to the third type of layer is less than the payload of the CSI reporting information corresponding to the second type of layer.

24. The method according to claim 23, characterized in that, The AI model corresponding to the third type of layer is used to process based on the channel state information corresponding to the third type of layer and the information of the associated layer of the third type of layer, and the associated layer of the third type of layer includes the first type of layer and the second type of layer.

25. The method according to any one of claims 14 to 24, characterized in that, The CSI reporting information corresponding to the first type of layer is located before the CSI reporting information corresponding to the second type of layer, or, The CSI reporting information corresponding to the first type of layer is located in a specific CSI reporting group.

26. The method according to any one of claims 14 to 24, characterized in that, The Nth layer in the at least one layer is a layer of the first type, or the layers in the Mth group in the at least one layer are layers of the first type, where both M and N are positive integers predefined by the protocol.

27. The method according to any one of claims 14 to 26, characterized in that, The CSI reporting information associated with one reporting identifier includes only the CSI reporting information corresponding to one layer of the first type.

28. The method according to any one of claims 14 to 26, characterized in that, The CSI reporting information carried on the CSI reporting resource used for one CSI reporting includes only the CSI reporting information corresponding to one layer of the first type.

29. The method according to any one of claims 1 to 28, characterized in that, The payload includes at least one of the following: the bit size of the CSI reporting information, the number of floating-point numbers of the CSI reporting information, the number of elements of the CSI reporting information, the output dimension of the first AI model, the quantization method for each element or floating-point number in the CSI reporting information, and the quantization method of the CSI reporting information.

30. The method according to any one of claims 1 to 29, characterized in that, The method further includes: The communication device sends or receives capability information related to the first object.

31. The method according to claim 30, characterized in that, The first object includes the payload, and the capability information related to the payload includes at least one of the following: Candidate values of the supported payload; Supported candidate quantization methods; Supported AI models related to the payload.

32. The method according to claim 30 or 31, characterized in that, The first object includes the at least one layer or at least one rank; the capability information related to the at least one layer or at least one rank includes at least one of the following: AI models supported by each layer; Supported layer types.

33. The method according to any one of claims 1 to 32, characterized in that, The first information includes at least one of the following: layer type, number of layers, layer compression method, payload indication information.

34. The method according to any one of claims 1 to 33, characterized in that, The communication device is a terminal, and the method further includes: The communication device reports CSI reporting information related to the first information.

35. An information transmission device, characterized in that, Including: A transceiver module, configured to receive or send first information related to a first object; Wherein, the first object includes at least one of the following: a payload, at least one layer, at least one rank; the first information is used to determine a first artificial intelligence (AI) model or to configure AI-based channel state information (CSI) reporting information, and the first AI model is used to generate the CSI reporting information or to decode the CSI reporting information.

36. A communication device, characterized in that, Including a processor and a memory, the memory stores a program or instruction that can run on the processor, and when the program or instruction is executed by the processor, the steps of the information transmission method according to any one of claims 1 to 34 are implemented.

37. A readable storage medium, characterized in that, A program or instruction is stored on the readable storage medium, and when the program or instruction is executed by a processor, the steps of the information transmission method according to any one of claims 1 to 34 are implemented.

Images