WO2025146034A1 - Ai-based csi compression method and apparatus, and terminal and network-side device - Google Patents

Abstract

The present application belongs to the technical field of communications. Disclosed are an AI-based CSI compression method and apparatus, and a terminal and a network-side device. The AI-based CSI compression method of the embodiments of the present application comprises: a terminal sending to a network-side device capability information of the terminal regarding an artificial intelligence (AI) unit, wherein the AI unit is configured to compress channel state information (CSI); the terminal receiving CSI configuration information sent by the network-side device on the basis of the capability information; and on the basis of the CSI configuration information, the terminal compressing the CSI by means of the AI unit.

Description

A CSI compression method, device, terminal and network side equipment based on AI

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims the priority of the Chinese patent application filed with the China Patent Office on January 4, 2024, with application number 202410015933.1 and titled “A CSI compression method, device, terminal and network side equipment based on Al”, the entire contents of which are incorporated by reference in this application.

Technical Field

The present application belongs to the field of communication technology, and specifically relates to an AI-based CSI compression method, device, terminal and network-side equipment.

Background Art

From information theory, we know that accurate channel state information (CSI) is crucial to channel capacity. Especially for multi-antenna systems, the transmitter can optimize the signal transmission based on CSI to make it more compatible with the channel state. For example, the channel quality indicator (CQI) can be used to select the appropriate modulation and coding scheme (MCS) to achieve link adaptation; the precoding matrix indicator (PMI) can be used to achieve eigen beamforming to maximize the strength of the received signal, or to suppress interference (such as interference between cells, interference between multiple users, etc.). Therefore, since the multi-input multi-output (MIMO) technology was proposed, CSI acquisition has always been a research hotspot.

Usually, the base station sends a Channel State Information-Reference Signal (CSI-RS) on certain time-frequency resources in a certain time slot. The terminal performs channel estimation based on the CSI-RS, calculates the channel information on this slot, and feeds back the PMI to the base station through the codebook. The base station combines the channel information based on the codebook information fed back by the terminal, and uses this to perform data precoding and multi-user scheduling before the next CSI report.

In order to further reduce the CSI feedback overhead, an evolved codebook solution is provided, that is, the terminal can change the PMI reported for each subband to reporting the PMI according to the delay. Since the channels in the delay domain are more concentrated, the PMIs with fewer delays can approximately represent the PMIs of all subbands, that is, the delay domain information is compressed before reporting.

Furthermore, in order to better compress the channel information, a neural network or machine learning method can be used to compress the CSI, that is, the AI unit is used to compress the CSI. Among them, the terminal and the network side device need to exchange necessary information to realize the compression of CSI based on the AI unit. However, the information exchange method when the terminal and the network side device compress the CSI based on the AI unit is not given at present.

Summary of the invention

The embodiments of the present application provide a CSI compression method, apparatus, terminal and network-side equipment based on AI, and provide an information interaction method when the terminal and the network-side equipment compress CSI based on the AI unit.

In a first aspect, a CSI compression method based on AI is provided, the method comprising:

The terminal sends capability information of the terminal about an artificial intelligence AI unit to the network side device, where the AI unit is used to compress the channel state information CSI;

The terminal receives CSI configuration information sent by the network side device according to the capability information;

The terminal compresses the CSI through the AI unit according to the CSI configuration information.

In a second aspect, a CSI compression method based on AI is provided, the method comprising:

The network side device receives capability information of the terminal about an artificial intelligence AI unit sent by the terminal, where the AI unit is used to compress channel state information CSI;

The network side device determines, according to the capability information, CSI configuration information for instructing the terminal to compress the CSI;

The network side device sends the CSI configuration information to the terminal.

In a third aspect, an AI-based CSI compression device is provided, which is applied to a terminal, and the device includes:

A first sending module is used to send capability information of the terminal about an artificial intelligence AI unit to a network side device, where the AI unit is used to compress channel state information CSI;

A first receiving module, configured to receive CSI configuration information sent by the network side device according to the capability information;

The CSI compression module is used to compress the CSI through the AI unit according to the CSI configuration information.

In a fourth aspect, an AI-based CSI compression device is provided, which is applied to a network side device, and the device includes:

A second receiving module is used to receive capability information of an artificial intelligence AI unit of the terminal sent by the terminal, where the AI unit is used to compress channel state information CSI;

an information determination module, configured to determine, according to the capability information, CSI configuration information for instructing the terminal to compress the CSI;

The second sending module is used to send the CSI configuration information to the terminal.

In a fifth aspect, a terminal is provided, comprising a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the method described in the first aspect are implemented.

In a sixth aspect, a terminal is provided, including a processor and a communication interface;

Wherein, the communication interface is used for:

Sending capability information of the terminal about an artificial intelligence AI unit to a network side device, where the AI unit is used to compress channel state information CSI;

Receiving CSI configuration information sent by the network side device according to the capability information;

The processor is used to: compress the CSI through the AI unit according to the CSI configuration information.

In the seventh aspect, a network side device is provided, which includes a processor and a memory, wherein the memory stores programs or instructions that can be run on the processor, and when the program or instructions are executed by the processor, the steps of the method described in the second aspect are implemented.

In an eighth aspect, a network side device is provided, including a processor and a communication interface;

The communication interface is used to: receive capability information of the terminal about an artificial intelligence AI unit sent by the terminal, and the AI unit is used to compress channel state information CSI;

The processor is used to: determine, according to the capability information, CSI configuration information for instructing the terminal to compress the CSI;

The communication interface is further used to: send the CSI configuration information to the terminal.

In a ninth aspect, a readable storage medium is provided, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method described in the first aspect are implemented, or the steps of the method described in the second aspect are implemented.

In the tenth aspect, an AI-based CSI compression system is provided, comprising: a terminal and a network side device, wherein the terminal can be used to execute the steps of the method described in the first aspect, and the network side device can be used to execute the steps of the method described in the second aspect.

In the eleventh aspect, a chip is provided, comprising a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instructions to implement the steps of the method described in the first aspect, or to implement the steps of the method described in the second aspect.

In the twelfth aspect, a computer program/program product is provided, wherein the computer program/program product is stored in a storage medium, and the program/program product is executed by at least one processor to implement the steps of the method described in the first aspect, or to implement the steps of the method described in the second aspect.

In the thirteenth aspect, an embodiment of the present application provides an AI-based CSI compression device, which is used to execute the steps of the AI-based CSI compression method as described in the first aspect or the second aspect.

In an embodiment of the present application, the terminal can send the capability information of the terminal about the AI unit to the network side device, thereby receiving the CSI configuration information sent by the network side device according to the capability information, and then compressing the CSI through the AI unit according to the received CSI configuration information. It can be seen that in an embodiment of the present application, the terminal can interact with the network side device about its capability information about the AI unit used to compress the CSI, so that the network side device can configure how the terminal performs CSI compression based on the capability information of the terminal about the AI unit. Therefore, an embodiment of the present application provides an information interaction method when the terminal and the network side device compress CSI based on the AI unit.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a block diagram of a wireless communication system to which an embodiment of the present application can be applied;

FIG2 is a schematic diagram of a neural network in an embodiment of the present application;

FIG3 is a schematic diagram of a neuron of a neural network in an embodiment of the present application;

FIG4 is a flow chart of an AI-based CSI compression method in an embodiment of the present application;

FIG5 is a schematic diagram of a packaged time-frequency-spatial domain CSI compression solution in an embodiment of the present application;

FIG6 is a schematic diagram of a progressive time-frequency-spatial domain CSI compression scheme for a shared model on multiple slots in an embodiment of the present application;

FIG7 is a schematic diagram of a progressive time-frequency-spatial domain CSI compression scheme for a dedicated model on multiple slots in an embodiment of the present application;

FIG8 is a schematic diagram of a slot interval pattern in the reasoning phase in an embodiment of the present application;

FIG9 is a flowchart of another AI-based CSI compression method in an embodiment of the present application;

FIG10 is a structural block diagram of an AI-based CSI compression device in an embodiment of the present application;

FIG11 is a structural block diagram of another AI-based CSI compression device in an embodiment of the present application;

FIG12 is a structural block diagram of a communication device in an embodiment of the present application;

FIG13 is a block diagram of a terminal in an embodiment of the present application;

FIG14 is a structural block diagram of a network side device in an embodiment of the present application.

Specific embodiments

The following will be combined with the drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of this application.

The terms "first", "second", etc. in this application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the terms used in this way are interchangeable where appropriate, so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of one type, and the number of objects is not limited, for example, the first object can be one or more. In addition, "or" in this application represents at least one of the connected objects. For example, "A or B" covers three schemes, namely, Scheme 1: including A but not including B; Scheme 2: including B but not including A; Scheme 3: including both A and B. The character "/" generally indicates that the objects associated with each other are in an "or" relationship.

The term "indication" in this application can be a direct indication (or explicit indication) or an indirect indication (or implicit indication). A direct indication can be understood as the sender explicitly informing the receiver of specific information, operations to be performed, or request results in the sent indication; an indirect indication can be understood as the receiver determining the corresponding information according to the indication sent by the sender, or making a judgment and determining the operation to be performed or the request result according to the judgment result.

It is worth noting that the technology described in the embodiments of the present application is not limited to the Long Term Evolution (LTE)/LTE-Advanced (LTE-A) system, but 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 the present application are often used interchangeably, and the described technology can be used for the above-mentioned systems and radio technologies as well as other systems and radio technologies. The following description describes a New Radio (NR) system for example purposes, and NR terms are used in most of the following descriptions, but these technologies can also be applied to systems other than NR systems, such as ^6th Generation (6G) communication systems.

FIG1 shows a block diagram of a wireless communication system applicable to the embodiment of the present application. 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 computer (Tablet Personal Computer), a laptop computer (Laptop Computer), a notebook computer, a personal digital assistant (PDA), a handheld computer, a netbook, an ultra-mobile personal computer (Ultra-mobile Personal Computer, UMPC), a mobile Internet device (Mobile Internet Device, MID), an augmented reality (Augmented Reality, AR), a virtual reality (Virtual Reality, VR) device, a robot, a wearable device (Wearable Device), a flight vehicle (flight vehicle), a vehicle user equipment (VUE), a shipborne equipment, a pedestrian terminal (Pedestrian User Equipment, PUE), a smart home (home appliances with wireless communication functions, such as refrigerators, televisions, washing machines or furniture, etc.), a game console, a personal computer (Personal Computer, PC), a teller machine or a self-service machine and other terminal side devices. Wearable devices include: smart watches, smart bracelets, smart headphones, smart glasses, smart jewelry (smart bracelets, smart bracelets, smart rings, smart necklaces, smart anklets, smart anklets, etc.), smart wristbands, smart clothing, etc. Among them, the vehicle-mounted device can also be called a vehicle-mounted terminal, a vehicle-mounted controller, a vehicle-mounted module, a vehicle-mounted component, a vehicle-mounted chip or a vehicle-mounted unit, etc. It should be noted that the specific type of the terminal 11 is not limited in the embodiment of the present application.

The network side equipment 12 may include access network equipment or core network equipment, wherein the access network equipment may also be referred to as radio access network (RAN) equipment, radio access network function or radio access network unit. The access network equipment may 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 a node B (Node B, NB), an evolved node B (Evolved Node B, eNB), the next generation Node B (the next generation Node B, gNB), a new radio node B (New Radio Node B, NR Node B), an access point, a relay station (Relay Base Station, RBS), a serving base station (Serving Base Station, SBS), a base transceiver station (Base Transceiver Station, BTS), a radio base station, a radio transceiver, a base Basic Service Set (BSS), Extended Service Set (ESS), home Node B (HNB), home evolved Node B (home evolved Node B), Transmission Reception Point (TRP) or other appropriate term in the field, as long as the same technical effect is achieved, the base station is not limited to specific technical vocabulary. It should be noted that, in the embodiments of the present 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.

In order to facilitate understanding of the AI-based CSI compression method of the embodiment of the present application, the following related technologies are first introduced:

1. Introduction to CSI compression based on artificial intelligence (AI)/machine learning (ML)

In order to reduce overhead, the base station can pre-code the CSI-RS in advance and send the encoded CSI-RS to each terminal. The terminal sees the channel corresponding to the encoded CSI-RS. The terminal only needs to select several ports with higher strength from the ports indicated by the network side and report the coefficients corresponding to these ports.

Furthermore, in order to better compress channel information, neural network or machine learning methods can be used.

Among them, artificial intelligence has been widely used in various fields. There are many ways to implement AI units, such as neural networks, decision trees, support vector machines, Bayesian classifiers, etc. This application uses neural networks as an example for illustration, but does not limit the specific type of AI modules. A schematic diagram of the structure of a simple neural network is shown in Figure 2.

In addition, the neural network is composed of neurons, and the schematic diagram of neurons is shown in Figure 3. In Figure 3, a ₁ , a ₂ , ... a _K represent inputs, w represents weights (i.e., multiplicative coefficients), b represents biases (i.e., additive coefficients), and σ(.) represents activation functions. Common activation functions include Sigmoid (mapping variables between 0 and 1), tanh (translation and contraction of Sigmoid), linear rectification function/rectified linear unit (Rectified Linear Unit, ReLU), etc.

Among them, the parameters of the neural network can be optimized through the gradient optimization algorithm. The gradient optimization algorithm is a type of algorithm that minimizes or maximizes the objective function (sometimes also called the loss function), and the objective function is often a mathematical combination of model parameters and data. For example, given the data X and its corresponding label Y, a neural network model f(.) can be constructed, then the predicted output f(x) can be obtained based on the input x, and the difference between the predicted value and the true value (f(x)-Y) can be calculated, which is the loss function. Among them, the optimization goal of the gradient optimization algorithm is to find the appropriate w (i.e. weight) and b (i.e. bias) to minimize the value of the above loss function, and the smaller the loss value, the closer the model is to the actual situation.

At present, the 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 the signal and the back propagation of the error. During the forward propagation, the input sample is transmitted from the input layer, processed by each hidden layer layer by layer, and then transmitted to the output layer. If the actual output of the output layer does not match the expected output, it will enter the back propagation stage of the error. Error back propagation is to propagate the output error layer by layer through the hidden layer to the input layer in some form, and distribute the error to all units in each layer, so as to obtain the error signal of each layer unit, and this error signal is used as the basis for correcting the weights of each unit. This process of adjusting the weights of each layer of the signal forward propagation and error back propagation is repeated. Among them, the process of continuous adjustment of weights is 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 pre-set number of learning times is reached.

In addition, common optimization algorithms include Gradient Descent, Stochastic Gradient Descent (SGD), mini-batch gradient descent, Momentum, Nesterov (the name of the inventor, specifically stochastic gradient descent with momentum), Adaptive Gradient Descent (Adagrad), Adagrad's extended algorithm (Adadelta), root mean square prop (RMSprop), Adaptive Moment Estimation (Adam), etc.

When the error is back-propagated, these optimization algorithms calculate the derivative/partial derivative of the current neuron based on the error/loss obtained by the loss function, add the influence of the learning rate, the previous gradient/derivative/partial derivative, etc., get the gradient, and pass the gradient to the previous layer.

Specifically, the terminal compresses and encodes the channel information, and the base station decodes the compressed content to restore the channel information. At this time, the decoding network of the base station and the encoding network of the terminal need to be jointly trained to achieve a reasonable match. The neural network is composed of a joint neural network through the encoder of the terminal and the decoder of the base station. The network side conducts joint training. After the training is completed, the base station sends the encoder network to the terminal. During inference, the terminal estimates the CSI-RS, calculates the channel information, and obtains the encoding result through the encoding network of the calculated channel information or the original estimated channel information. The encoding result is sent to the base station. The base station receives the encoded result and inputs it into the decoding network to restore the channel information.

Scalability of AI/ML CSI Compression Model

The scalability of a model refers to the ability of a single model to adapt to multiple input/output configurations at the same time. In CSI compression, the scalability of the model mainly considers the following indicators: the number of subbands of the (encoder) input, the number of ports of the (encoder) input, and the length of the (encoder) output payload. That is, it is hoped that a CSI compression model can support as many of the above configuration combinations as possible (for example, a model can support both 64-bit and 116-bit payloads. When the network needs to change the payload from 64 bits to 116 bits, it does not need to retrain the model but rely on the existing model implementation), thereby reducing the cost of replacing the model and maintaining at least one of the multiple models at the same time. Generally speaking, in order to achieve the scalability of the CSI compression model to a certain extent, it is necessary to consider multiple situations in the model training stage and combine special model structures (such as adaptation layers, etc.) to achieve this.

Theoretically, all CSI compression model scalability possibilities can be considered during the model training phase, so that one model can handle all configurations. However, considering the difficulty of model training and the fact that some network parameter configurations cannot be known in advance during the training phase, it is difficult to fully consider all scalability possibilities that all models need to meet in advance. In this case, the model still needs to be retrained to solve this problem.

3. Training collaboration types for AI/ML CSI compression models

The CSI compression use case is a typical two-end model use case, that is, the complete CSI compression model needs to be deployed on different network nodes. Currently, most of the cases considered are to deploy the encoder on the UE side and the decoder on the network side (NW). The (sub) models deployed on multiple nodes need to be paired with each other to work properly. Considering the above characteristics of the two-end model, 3GPP has identified several basic types of training collaboration:

(1) Joint training at single entity (also called type 1):

The training framework aims to train a complete encoder-decoder model on a network node (UE or NW or a third-party server node, etc.), and then deploy the corresponding model module to the target node through methods such as model transfer (for example, the encoder part is transferred to the UE and the decoder part is transferred to the NW).

(2) Joint training at multiple entities (also called type 2)

The training framework refers to the joint participation of multiple nodes in the training process, and each node independently calculates the forward/backward propagation information required for local model training and updates the model parameters of its own node. Since the training process requires forward/backward propagation of the entire model (including encoder and decoder), the corresponding forward/backward propagation information needs to be transmitted between the nodes participating in the training. After the training is completed, the model no longer needs to be transmitted between the nodes.

(3) Separate (or step-by-step) training on multiple nodes (also called type 3)

The training framework means that a model used as a reference is first trained on a certain node, and then the relevant information of the reference model is sent to the target node. Finally, the target node trains the model required by the node based on the information, thereby ensuring that each node (sub) model can be paired with each other. For example, the NW side first trains a set of complete models of encoders and decoders, and determines that the decoder obtained is the decoder actually used in the future, and then sends the relevant information of the encoder corresponding to (the decoder) (generally the input and output data of the encoder) to the UE side, and the UE side trains the encoder used by itself based on the information.

This training framework can be further divided into two cases: UE-first training and NW-first training. UE-first training means training the complete model on the UE side first, and then sending the information required for NW training to match the model (generally the input and output data of the model to be trained on the NW side) to the NW side. In contrast, NW-first training means training the complete model on the NW side first, and then sending the information required for UE training to match the model (generally the input and output data of the model to be trained on the UE side) to the UE side.

The following, in combination with the accompanying drawings, describes in detail the AI-based CSI compression method provided in the embodiments of the present application through some embodiments and their application scenarios.

As shown in FIG. 4 , an embodiment of the present application provides an AI-based CSI compression method, which may include the following steps 401 to 403:

Step 401: The terminal sends capability information of the terminal about the artificial intelligence AI unit to the network side device.

The AI unit is used to compress the channel state information CSI.

It should be noted that the AI unit can be used to perform CSI compression in at least one of the time domain, frequency domain, and spatial domain on the CSI, that is, the AI unit can perform at least one of CSI time domain compression, CSI frequency domain compression, and CSI spatial domain compression.

In addition, the AI unit may also be referred to as an AI model, a machine learning (ML) model, an ML unit, an AI structure, an AI function, an AI characteristic, a neural network, a neural network function, a neural network function, etc.; or the AI unit may also refer to a processing unit that can implement specific algorithms, formulas, processing procedures, capabilities, etc. related to AI, or the AI unit may be a processing method, algorithm, function, module or unit for a specific data set, or the AI unit may be a processing method, algorithm, function, module or unit running on AI/ML related hardware such as a graphics processing unit (GPU), a neural network processor (NPU), a tensor processor (TPU), an application specific integrated circuit (ASIC), etc., and this application does not make specific limitations on this. Optionally, the specific data set includes at least one of the input and output of the AI unit/AI model.

Optionally, the identifier of the AI unit may be an AI model identifier, an AI structure identifier, an AI algorithm identifier, or an identifier of a specific data set associated with the AI unit, or an identifier of a specific scenario, environment, channel feature, or device related to the AI/ML, or an identifier of a function, feature, capability, or module related to the AI/ML, and this application does not make any specific limitations on this.

In addition, the capability information is used to indicate the terminal's support for the AI unit, that is, the terminal's capability in performing CSI compression.

Step 402: The terminal receives CSI configuration information sent by the network side device according to the capability information.

The CSI configuration information includes configuration information for performing CSI compression, that is, the CSI configuration information is used to instruct the terminal how to perform CSI compression.

In addition, the CSI configuration information may also be referred to as CSI association information, CSI trigger information, or CSI scheduling information, and is used to indicate how to perform CSI compression.

It can be seen from step 402 that after receiving the capability information reported by the terminal, the network side device determines CSI configuration information according to the capability information to instruct the terminal how to perform CSI compression.

Step 403: The terminal compresses the CSI through the AI unit according to the CSI configuration information.

As mentioned above, the AI unit can be used to perform CSI compression in at least one of the time domain, frequency domain, and spatial domain on the CSI, that is, the AI unit can perform at least one of CSI time domain compression, CSI frequency domain compression, and CSI spatial domain compression. In step 403, the terminal specifically performs one or more CSI compressions in the time domain, frequency domain, and spatial domain, depending on the function of the AI unit and the CSI configuration information.

It should be noted that time domain compression is time domain joint compression, that is, CSI on multiple time units (such as slots) are combined for compression, so as to further reduce the overhead of CSI reporting or improve the accuracy of CSI reporting.

In addition, time domain joint compression, that is, the reporting method of CSI on multiple time units, is divided into two types: packaged reporting and progressive reporting. Packaged reporting is to report CSI on multiple time units (such as time slots) at one time (as shown in Figure 5 below), while progressive reporting is to report CSI on each time unit (such as time slot) in turn in an autoregressive manner (as shown in Figure 6 above).

In addition, the traditional CSI reporting based on the Type II codebook supports two modes: aperiodic (AP) and semi-persistent. For the AP mode, from the system perspective, since the network side device (such as the base station) only needs the latest CSI for one scheduling, it only needs to trigger the CSI report of the terminal once to indicate the most recent measurement result, and there is no need to continuously report the CSI at multiple times. Even if this progressive CSI reporting is adopted in the AP mode, there will be a problem of no way to control the time of the last CSI transmission. If the network side device may be the terminal that was scheduled a long time ago, the time domain correlation between the two CSIs that need to be reported is very weak, and there is no room for the time domain compression scheme to play. Therefore, continuous scheduling (such as the SP mode) can lead to a scenario where the CSI is compressed in the time domain. That is, the progressive time domain CSI compression is mainly carried out by the semi-persistent (SP) CSI reporting mechanism.

It can be seen from the above steps 401 to 403 that in the embodiment of the present application, the terminal can send the capability information of the terminal about the AI unit to the network side device, thereby receiving the CSI configuration information sent by the network side device according to the capability information, and then compressing the CSI through the AI unit according to the received CSI configuration information. It can be seen that in the embodiment of the present application, the terminal can interact with the network side device about its capability information about the AI unit used to compress the CSI, so that the network side device can configure how the terminal performs CSI compression based on the terminal's capability information about the AI unit. Therefore, the embodiment of the present application provides a method for information interaction between the terminal and the network side device when compressing CSI based on the AI unit.

Optionally, the capability information includes at least one of the following items A-1 to A-4:

Item A-1: whether the terminal supports AI-based CSI compression;

Item A-2: the type of the AI unit supported by the terminal;

The type of the AI unit may include at least one of the following types:

The first type (also called a dedicated model): different encoders and decoders are used to compress CSI in different time units; as shown in FIG7 , ENC0 and DEC0 are specifically used for reporting CSI on slot0, and so on, where ENC represents an encoder and DEC represents a decoder;

The second type (also called the shared model): the same encoder and decoder are used to compress CSI in different time units; as shown in FIG6 , a set of ENC0 and DEC0 is applied to CSI reporting on all slots;

The third type: in some time units, different encoders and decoders are used to compress CSI in different time units; in another part of time units, the same encoder and decoder are used to compress CSI in different time units; for example, the AI unit can compress CSI of 4 time units, among which the same encoder and decoder can be used to compress CSI in the first two time units, and different encoders and decoders can be used to compress CSI in the last two time units.

Item A-3: First indication information, the first indication information is used to indicate the length of a target time window supported by the terminal, the target time window including at least one time unit (e.g., time slot) for the AI unit to perform CSI compression; that is, it can be understood that: the target time window is used to indicate the maximum number of time units processed when the AI unit performs an inference (i.e., performs CSI compression).

It should be noted that an important feature of the dedicated model is that it has a clear concept of time window, that is, it processes CSI at a maximum of several different moments (generally speaking, the time window needs to be determined in the model training phase, and the time window length in the inference phase is consistent with the training phase). Once the number of CSIs to be reported exceeds the maximum number of CSIs that the model can process in the time domain, it is necessary to restart the inference from slot0 (that is, it is difficult for this solution to process CSI reports on slots that exceed the window length in one inference process, as shown in Figure 7. The possible reason is that no set of ENC and DEC models can process the intermediate information flow (internal information) output by ENC3 and DEC3); in addition, during inference, the position of the current slot in the time window must be aligned with the model used (that is, ENC1 and DEC1 can only process CSI on the second slot in the window, and cannot be used to process CSI on the third slot), otherwise the model performance is likely to degrade due to mismatching.

Although the dedicated model has the above limitations, its performance ceiling is relatively high, especially when processing CSI in slots later in the time window. The shared model has weaker requirements on the time window because it shares the same set of ENC0 and DEC0 relationships, and its reasoning process can continue in the time domain without obvious performance loss (for example, a model is trained using CSI in 4 consecutive slots, but the model can be used to compress CSI in 6 or more consecutive slots because ENC0 and DEC0 can always identify the intermediate information flow). Therefore, this solution has less demand for information interaction. Although the shared model solution is easy to implement, it is limited by the parameter restrictions brought by the contribution model, and its performance is generally weaker than the dedicated model solution, especially for slots later in time.

It is understandable that although the shared model has weaker requirements on the time window, a corresponding time window may exist when the shared model is used for CSI compression; similarly, when the third type of AI unit mentioned above performs CSI compression, the corresponding time window may also be stored.

Optionally, the first indication information includes at least one of the following items B-1 to B-3:

Item B-1: The maximum length of the target time window supported by the terminal; that is, the available target time window length indicated in the capability information reported by the terminal is the maximum length, on this basis, the network side device can schedule any number of CSI reports that are less than, equal to or less than the maximum length.

Item B-2: a set of target time window lengths supported by the terminal; that is, the capability information reported by the terminal may indicate a set of target time window lengths supported by the terminal, and the network-side device may select the number of CSIs to be reported from the set when subsequently scheduling CSI reporting.

Item B-3: identification information of the target time window length configuration supported by the terminal; that is, during the terminal capability reporting stage, the terminal can not only directly report the supported target time window length, but also report the supported typical target time window length configuration. For example, some typical target time window configurations are agreed upon in the protocol, and the identification (such as index or serial number) of these typical configurations can be directly reported during reporting.

Item A-4: the interval mode of CSI reporting supported by the terminal, such as equal interval mode or non-equal interval mode; here, the interval refers to the time unit between reporting CSI;

Among them, even if the data used by the AI unit in the model training phase is CSI at continuous equal intervals, the AI unit can still have a certain generalization ability and expand to non-equally spaced CSI work, as shown in Figure 8. That is, the CSI data on continuous slots used in the training phase in Figure 8, and the non-equally spaced CSI data (CSI on slot1 is missing) are used in the inference phase. There may be a certain performance loss in the above-mentioned slot pattern expansion. For example, as the time interval between adjacent slots in the pattern increases, the recovery accuracy of CSI will decrease (it can be understood that CSI with a longer time interval can provide less information to each other), but the experimental results show that as long as the interval is not too large (for example, one or two slots), the degree of decline in recovery accuracy is still within an acceptable range.

It can be seen from item A-4 that a CSI report can be performed once in a time unit within a target time window, where the time unit within a target time window may include time units with equal intervals in the time domain (for example, time slots 0, 1, 2, and 3), and may also include time units with unequal intervals (for example, time slots 0, 2, 3, and 4).

Optionally, the CSI configuration information includes at least one of the following C-1 to C-6:

Item C-1: Number of CSI reports (i.e., the number of CSI reports performed by terminals scheduled by the network-side device);

Item C-2: the type of the AI unit used; wherein the type of the AI unit includes at least one of the first type, the second type, and the third type described above, which will not be described in detail here;

Item C-3: CSI grouping information, where the CSI grouping information is used to indicate grouping of CSI to be reported according to the length of a target time window, where the target time window includes at least one time unit for CSI compression by the AI unit;

If the number of CSI reports that the terminal needs to perform exceeds the length of the target time window, the CSI reported by the terminal needs to be grouped, so that the CSI reports in one group are within one target time window (ie, one group corresponds to one target time window).

That is, it can be understood as: CSI grouping information is used to indicate which CSIs are reported within the same target time window. When the number of CSIs to be scheduled belongs to one of the target time window lengths supported by the terminal capability, or the number of CSIs to be scheduled is less than the maximum target time window length supported by the terminal capability, the network side device can directly schedule the CSI report of this number. When the number of CSIs to be scheduled does not belong to the target time window length supported by the terminal capability (or is greater than the length of the maximum target time window supported), the network side device can further indicate the grouping information of the scheduled CSI.

For example, the network side device schedules 10 CSI reports, and the terminal capability supports time domain compression of up to 4 CSIs as a group. At this time, the network side device can indicate that the first 4 CSIs are a group, the next 4 CSIs are another group, and the last 2 CSIs are a group; or the first 2 CSIs are a group, the next 4 are a group, and the last 4 are a group.

Item C-4: second indication information, the second indication information is used to indicate whether, when reporting CSI, it is necessary to carry the position of the reported CSI in the group to which it belongs; wherein, one group corresponds to one target time window, therefore, the second indication information can also be understood as: used to indicate whether, when reporting CSI, it is necessary to carry the position of the reported CSI in the corresponding target time window.

Item C-5: The load length of at least part of the CSI reports within the target time window, where the load length is the length of the uplink control information UCI resources occupied by the CSI reports; that is, the network side device can indicate to the terminal the load length of part of the CSI reports within a target time window, or can indicate the load length of all CSI reports within a target time window; wherein, when the network side indicates the load length of part of the CSI reports within a target time window, the load length of the remaining CSI reports within the target time window can be calculated according to the arrangement rules agreed in advance by the protocol.

Item C-6: Target resources used to measure CSI within the target time window.

It should be noted that the terminal measures the CSI in a time unit within the target time window, and thus reports the CSI. Accordingly, the network-side device may also indicate to the terminal the target resource for measuring the CSI within the target time window.

Optionally, the CSI grouping information in the above item C-3 includes at least one of the following items D-1 to D-3:

Item D-1: identification information of the group to which each CSI to be reported belongs;

For example, the network side device schedules 10 CSI reports, and the terminal capability supports time domain compression of up to 4 CSIs as a group. If the network side device indicates that the 1st to 4th CSIs are a group, the 5th to 8th CSIs are a group, and the 9th to 10th CSIs are a group, then specifically, the network side device can respectively indicate the identification information of the group to which each CSI belongs, for example, the first CSI belongs to the first group, the second CSI belongs to the first group, the third CSI belongs to the first group, the fourth CSI belongs to the first group….

Item D-2: location information of the first CSI report in each group in the CSI to be reported, and the number of CSI reports included in each group; in this way, the terminal can determine the group to which each CSI belongs;

For example, the network side device schedules 10 CSI reports, and the terminal capability supports time domain compression of up to 4 CSIs as a group. If the network side device indicates that the 1st to 4th CSIs are a group, the 5th to 8th CSIs are a group, and the 9th to 10th CSIs are a group, then specifically, the network side device can respectively indicate the position of the first CSI in each group among the 10 CSIs, and the number of CSIs included in each group, that is, the first CSI in the first group is the first of the 10 CSIs, the first CSI in the second group is the fifth of the 10 CSIs, and the first CSI in the third group is the ninth of the 10 CSIs. The first group includes 4 CSIs, the second group includes 4 CSIs, and the third group includes 2 CSIs.

It should be noted that if a group includes a CSI (ie, the length of a target time window is 1), the CSI (ie, the CSI within the target time window) is not compressed in the time domain (eg, only compressed in the space-frequency domain).

Item D-3: target CSI that needs to be jointly compressed in the time domain, and identification information of the group to which the target CSI belongs.

As can be seen from item D-3, the network side device can only indicate the CSI and grouping that need to be compressed in the time domain, and the remaining CSIs are not compressed in the time domain by default. When no grouping indication is given, all are not compressed in the time domain by default.

Optionally, in the above item C-5, the payload length of at least part of the CSI reported within the target time window satisfies at least one of the following items E-1 to E-5:

Item E-1: The payload length of each CSI report within the target time window is the same (i.e., the payload length of each CSI report within the same target time window is the same); in this case, the network side device only needs to indicate one payload length for one target time window; that is, if the network side device indicates that one target time window corresponds to one payload length X, it means that the payload length of each CSI report within the target time window is X;

Item E-2: the length of the payload of the first CSI report within the target time window is greater than the length of the payload of the CSI report after the first CSI report within the target time window (i.e., the length of the payload of the first CSI report within the same target time window is greater than the length of the payload of the CSI report after the first CSI report);

Item E-3: the load length of the CSI report after the first CSI report in the target time window is arranged in the form of an arithmetic progression (that is, the load length of the CSI report after the first report in the same target time window is arranged in the form of an arithmetic progression); in this case, for the CSI report after the first CSI report in a target time window, the network side device can indicate a load length and a load difference between adjacent CSI reports, wherein the load length is the load length of the second CSI report in the target time window, so that according to the load difference between adjacent CSI reports, the load length of the third and subsequent CSI reports in the target time window can be determined;

In item E-3, exemplarily, there may be a certain correlation between the load lengths of CSI reports within a target time window, for example, the feedback load length of the first CSI in the target time window is higher, followed by the second, until the last CSI load in the window is the lowest value reported in the window.

Item E-4: The payload length of the CSI reports after the first CSI report within the target time window is the same (i.e., the payload length of each CSI report after the first CSI report within the same target time window is the same); in this case, the network side device only needs to indicate one payload length for the CSI report after the first CSI report.

Item E-5: When the load lengths of each CSI report within the target time window are different, the load lengths of each CSI report within the target time window implicitly indicate the position of each CSI report within the target time window within the target time window; for example, a target time window length is 4 time slots, and the loads of the CSI reports in each time slot are 100, 80, 64, and 50, then the CSI report with a load of 100 is located in the first time slot within the time window, the CSI report with a load of 80 is located in the second time slot within the time window, the CSI report with a load of 64 is located in the third time slot within the time window, and the CSI report with a load of 50 is located in the fourth time slot within the time window. In this case, the network-side device can configure the terminal to report CSI without carrying the position of the reported CSI in the group to which it belongs (i.e., the position of the reported CSI in the corresponding target time window).

Optionally, the target resource in the above item C-6 meets at least one of the following items F-1 to F-4:

Item F-1: the resources for measuring CSI each time within the target time window are the same (that is, the resources for measuring CSI in different time units within the same target time window are the same);

Item F-2: the number of resources used to measure CSI in the first time unit in the target time window is the largest (i.e., the number of resources used to measure CSI in the first time unit in the same target time window is greater than the number of resources used to measure CSI in the time units after the first time unit);

Item F-3: the number of resources used for measuring CSI in each time unit after the first time unit in the target time window is the same (that is, the number of resources used for measuring CSI in each time unit after the first time unit in the same target time window is the same);

Item F-4: The number of resources used to measure CSI in each time unit after the first time unit in the target time window decreases in sequence in the time domain (i.e., the number of resources used to measure CSI in the i-th time unit in the same target time window is greater than the number of resources used to measure CSI in the i+1-th time unit, where i is an integer greater than 2).

It can be seen from the above items F-1 to F-4 that the resources for measuring CSI in different time units within the same target time window may be the same or different.

As shown in FIG. 9 , an embodiment of the present application further provides an AI-based CSI compression method, which may include the following steps 901 to 903:

Step 901: The network side device receives capability information about the artificial intelligence AI unit of the terminal sent by the terminal.

The AI unit is used to compress the channel state information CSI.

It should be noted that the AI unit can be used to perform CSI compression in at least one of the time domain, frequency domain, and spatial domain on the CSI, that is, the AI unit can perform at least one of CSI time domain compression, CSI frequency domain compression, and CSI spatial domain compression.

In addition, the capability information is used to indicate the terminal's support for the AI unit, that is, the terminal's capability in performing CSI compression.

Step 902: The network-side device determines, according to the capability information, CSI configuration information for instructing the terminal to compress the CSI.

The CSI configuration information includes configuration information for performing CSI compression, that is, the CSI configuration information is used to instruct the terminal how to perform CSI compression.

In addition, the CSI configuration information may also be referred to as CSI association information, CSI trigger information, or CSI scheduling information, and is used to indicate how to perform CSI compression.

It can be seen from step 902 that after receiving the capability information reported by the terminal, the network side device determines CSI configuration information according to the capability information to instruct the terminal how to perform CSI compression.

Step 903: The network side device sends the CSI configuration information to the terminal.

Among them, after the terminal receives the CSI configuration information, it compresses the CSI through the AI unit according to the CSI configuration information.

In addition, the AI unit can be used to perform CSI compression in at least one of the time domain, frequency domain, and spatial domain on the CSI, that is, the AI unit can perform at least one of CSI time domain compression, CSI frequency domain compression, and CSI spatial domain compression. Among them, the terminal specifically performs which one or more CSI compressions in the time domain, frequency domain, and spatial domain depends on the function of the AI unit and the CSI configuration information.

It can be seen from the above steps 901 to 903 that in the embodiment of the present application, the network side device can receive the capability information of the terminal about the AI unit sent by the terminal, thereby determining the CSI configuration information based on the capability information, and then sending the CSI configuration information to the terminal, so that the terminal compresses the CSI through the AI unit according to the received CSI configuration information. It can be seen that in the embodiment of the present application, the terminal can interact with the network side device about its capability information about the AI unit used to compress the CSI, so that the network side device can configure how the terminal performs CSI compression based on the terminal's capability information about the AI unit. Therefore, the embodiment of the present application provides a method for information interaction between the terminal and the network side device when compressing CSI based on the AI unit.

Optionally, the capability information includes at least one of the following items A-1 to A-4:

Item A-1: whether the terminal supports AI-based CSI compression;

Item A-2: the type of the AI unit supported by the terminal;

Item A-3: First indication information, where the first indication information is used to indicate the length of a target time window supported by the terminal, where the target time window includes at least one time unit (eg, a time slot) for performing CSI compression by the AI unit.

Item A-4: The CSI reporting interval mode supported by the terminal.

It can be understood that the relevant explanations of items A-1 to A-4 here can be found in the previous text and will not be repeated here.

Optionally, the first indication information includes at least one of the following items B-1 to B-3:

Item B-1: the maximum length of the target time window supported by the terminal;

Item B-2: a set of lengths of the target time window supported by the terminal;

Item B-3: identification information of the length configuration of the target time window supported by the terminal.

It can be understood that the relevant explanations of items B-1 to B-3 here can be found in the previous text and will not be repeated here.

Optionally, the CSI configuration information includes at least one of the following C-1 to C-6:

Item C-1: CSI reporting quantity;

Item C-2: Type of AI unit used;

Item C-3: CSI grouping information, where the CSI grouping information is used to indicate grouping of CSI to be reported according to the length of a target time window, where the target time window includes at least one time unit for CSI compression by the AI unit;

Item C-4: second indication information, where the second indication information is used to indicate whether the position of the reported CSI in the group to which it belongs needs to be carried when the CSI is reported;

Item C-5: the payload length of at least part of the CSI report within the target time window, where the payload length is the length of the uplink control information UCI resources occupied by the CSI report;

Item C-6: Target resources used to measure CSI within the target time window.

It can be understood that the relevant explanations of items C-1 to C-6 here can be found in the previous text and will not be repeated here.

Optionally, the CSI grouping information in the above item C-3 includes at least one of the following items D-1 to D-3:

Item D-1: identification information of the group to which each CSI to be reported belongs;

Item D-2: location information of the first CSI report in each group in the CSI to be reported, and the number of CSI reports included in each group;

Item D-3: target CSI that needs to be jointly compressed in the time domain, and identification information of the group to which the target CSI belongs.

It can be understood that the relevant explanations of items D-1 to D-3 here can be found in the previous text and will not be repeated here.

Optionally, in the above item C-5, the payload length of at least part of the CSI reported within the target time window satisfies at least one of the following items E-1 to E-5:

Item E-1: The payload length of each CSI report within the target time window is the same;

Item E-2: the length of the payload of the first CSI report within the target time window is greater than the length of the payload of the CSI report after the first CSI report within the target time window;

Item E-3: the payload length of the CSI report after the first CSI report within the target time window, arranged in the form of an arithmetic progression;

Item E-4: the payload lengths of CSI reports after the first CSI report within the target time window are the same;

Item E-5: When the payload lengths of each CSI report within the target time window are different, the payload lengths of each CSI report within the target time window implicitly indicate the position of each CSI report within the target time window.

It is to be understood that the relevant explanations of items E-1 to E-5 here can be found in the previous text and will not be repeated here.

Optionally, the target resource in the above item C-6 meets at least one of the following items F-1 to F-4:

Item F-1: the resources for each CSI measurement within the target time window are the same;

F-2: the resources used to measure CSI in the first time unit in the target time window are the largest;

Item F-3: the number of resources used to measure CSI in each time unit after the first time unit in the target time window is the same;

Item F-4: The number of resources used to measure CSI in each time unit after the first time unit in the target time window decreases in sequence according to the order of time domain.

It can be understood that the relevant explanations of items F-1 to F-4 here can be found in the previous text and will not be repeated here.

In summary, the current CSI compression is mainly limited to CSI compression in the space-frequency domain. In fact, there is also an obvious correlation (referred to as time domain correlation) between different (but close in time) slots. Increasing the use of time domain CSI correlation on the basis of the space-frequency domain can significantly improve the performance of CSI compression. In an embodiment of the present application, progressive time-frequency space-domain CSI compression can be used: under this compression framework, CSI reporting still occurs once per slot, but the CSI report on a certain slot will refer to the CSI content that has been reported in the previous slot, thereby improving the reporting accuracy. In addition, in an embodiment of the present application, additional information (compared to traditional space-frequency domain CSI compression) that needs to be interacted or determined between the network side device and the terminal when performing CSI compression in the above manner is also provided.

The AI-based CSI compression method provided in the embodiment of the present application may be executed by an AI-based CSI compression device. In the embodiment of the present application, an AI-based CSI compression device executing the AI-based CSI compression method is taken as an example to illustrate the AI-based CSI compression method device provided in the embodiment of the present application.

The embodiment of the present application further provides an AI-based CSI compression device, which is applied to a terminal. As shown in FIG10 , the AI-based CSI compression device 100 includes the following modules:

The first sending module 1001 is used to send the capability information of the terminal about the artificial intelligence AI unit to the network side device, where the AI unit is used to compress the channel state information CSI;

A first receiving module 1002 is configured to receive CSI configuration information sent by the network side device according to the capability information;

The CSI compression module 1002 is used to compress the CSI through the AI unit according to the CSI configuration information.

Optionally, the capability information includes at least one of the following:

Whether the terminal supports AI-based CSI compression;

The type of the AI unit supported by the terminal;

first indication information, where the first indication information is used to indicate a length of a target time window supported by the terminal, where the target time window includes at least one time unit for the AI unit to perform CSI compression;

The CSI reporting interval mode supported by the terminal.

Optionally, the first indication information includes at least one of the following:

The maximum length of the target time window supported by the terminal;

a set of lengths of the target time window supported by the terminal;

Identification information of the length configuration of the target time window supported by the terminal.

Optionally, the CSI configuration information includes at least one of the following:

Number of CSI reports;

the type of said AI unit used;

CSI grouping information, where the CSI grouping information is used to indicate grouping of CSI to be reported according to the length of a target time window, where the target time window includes at least one time unit for CSI compression by the AI unit;

second indication information, where the second indication information is used to indicate whether, when reporting the CSI, it is necessary to carry a position of the reported CSI in the group to which it belongs;

The payload length of at least part of the CSI report within the target time window, where the payload length is the uplink control information UCI resource length occupied by the CSI report;

The target resource for measuring CSI in the target time window.

Optionally, the CSI grouping information includes at least one of the following:

Identification information of the group to which each CSI to be reported belongs;

The location information of the first CSI report in each group in the CSI reports that need to be reported, and the number of CSI reports included in each group;

The target CSI that needs to be jointly compressed in the time domain, and the identification information of the group to which the target CSI belongs.

Optionally, a payload length of at least part of the CSI reported within the target time window satisfies at least one of the following:

The payload length of each CSI report within the target time window is the same;

The length of the payload of the first CSI report within the target time window is greater than the length of the payload of the CSI report after the first CSI report within the target time window;

The payload lengths of the CSI reports after the first CSI report within the target time window are arranged in an arithmetic progression;

The payload lengths of CSI reports after the first CSI report within the target time window are the same;

In the case where the payload lengths of each CSI report within the target time window are different, the payload lengths of each CSI report within the target time window implicitly indicate the position of each CSI report within the target time window.

Optionally, the target resource meets at least one of the following conditions:

The resources for measuring the CSI each time within the target time window are the same;

The resources used for measuring CSI in the first time unit in the target time window are the largest;

The number of resources used for measuring CSI in each time unit after the first time unit in the target time window is the same;

The number of resources used to measure CSI in each time unit after the first time unit in the target time window decreases in sequence according to the time domain.

The AI-based CSI compression device in the embodiment 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; illustratively, the terminal may include but is not limited to the types of terminals 11 listed above, and the embodiment of the present application does not specifically limit this.

The AI-based CSI compression device provided in the embodiment of the present application can implement the various processes implemented in the method embodiment of Figure 4 and achieve the same technical effect. To avoid repetition, it will not be repeated here.

The embodiment of the present application further provides an AI-based CSI compression device, which is applied to a network-side device. As shown in FIG11 , the AI-based CSI compression device 110 includes the following modules:

The second receiving module 1101 is used to receive capability information of an artificial intelligence AI unit of the terminal sent by the terminal, where the AI unit is used to compress channel state information CSI;

An information determining module 1102 is used to determine, according to the capability information, CSI configuration information for instructing the terminal to compress the CSI;

The second sending module 1103 is configured to send the CSI configuration information to the terminal.

Optionally, the capability information includes at least one of the following:

Whether the terminal supports AI-based CSI compression;

The type of the AI unit supported by the terminal;

first indication information, where the first indication information is used to indicate a length of a target time window supported by the terminal, where the target time window includes at least one time unit for the AI unit to perform CSI compression;

The CSI reporting interval mode supported by the terminal.

Optionally, the first indication information includes at least one of the following:

The maximum length of the target time window supported by the terminal;

a set of lengths of the target time window supported by the terminal;

Identification information of the length configuration of the target time window supported by the terminal.

Optionally, the CSI configuration information includes at least one of the following:

Number of CSI reports;

the type of said AI unit used;

CSI grouping information, where the CSI grouping information is used to indicate grouping of CSI to be reported according to the length of a target time window, where the target time window includes at least one time unit for CSI compression by the AI unit;

second indication information, where the second indication information is used to indicate whether, when reporting the CSI, it is necessary to carry a position of the reported CSI in the group to which it belongs;

The payload length of at least part of the CSI report within the target time window, where the payload length is the uplink control information UCI resource length occupied by the CSI report;

The target resource for measuring CSI in the target time window.

Optionally, the CSI grouping information includes at least one of the following:

Identification information of the group to which each CSI to be reported belongs;

The location information of the first CSI report in each group in the CSI reports that need to be reported, and the number of CSI reports included in each group;

The target CSI that needs to be jointly compressed in the time domain, and the identification information of the group to which the target CSI belongs.

Optionally, a payload length of at least part of the CSI reported within the target time window satisfies at least one of the following:

The payload length of each CSI report within the target time window is the same;

The length of the payload of the first CSI report within the target time window is greater than the length of the payload of the CSI report after the first CSI report within the target time window;

The payload lengths of the CSI reports after the first CSI report within the target time window are arranged in an arithmetic progression;

The payload lengths of CSI reports after the first CSI report within the target time window are the same;

In the case where the payload lengths of each CSI report within the target time window are different, the payload lengths of each CSI report within the target time window implicitly indicate the position of each CSI report within the target time window.

Optionally, the target resource meets at least one of the following conditions:

The resources for measuring the CSI each time within the target time window are the same;

The resources used for measuring CSI in the first time unit in the target time window are the largest;

The number of resources used for measuring CSI in each time unit after the first time unit in the target time window is the same;

The number of resources used to measure CSI in each time unit after the first time unit in the target time window decreases in sequence according to the time domain.

The AI-based CSI compression device in the embodiment of the present application can 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 can be a network-side device; illustratively, the network-side device can include but is not limited to the types of network-side devices 12 listed above, and the embodiment of the present application does not specifically limit this.

The AI-based CSI compression device provided in the embodiment of the present application can implement the various processes implemented in the method embodiment of Figure 9 and achieve the same technical effect. To avoid repetition, it will not be repeated here.

As shown in FIG12 , the embodiment of the present application further provides a communication device 1200, including a processor 1201 and a memory 1202, wherein the memory 1202 stores a program or instruction that can be run on the processor 1201. For example, when the communication device 1200 is a terminal, the program or instruction is executed by the processor 1201 to implement the various steps of the above-mentioned AI-based CSI compression method embodiment applied to the terminal, and can achieve the same technical effect. When the communication device 1200 is a network side device, the program or instruction is executed by the processor 1201 to implement the various steps of the above-mentioned AI-based CSI compression method embodiment applied to the network side device, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

The embodiment of the present application also provides a terminal, including a processor and a communication interface, the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the steps in the method embodiment shown in Figure 4. This terminal embodiment corresponds to the above-mentioned terminal side method embodiment, and each implementation process and implementation method of the above-mentioned method embodiment can be applied to the terminal embodiment and can achieve the same technical effect. Specifically, Figure 13 is a schematic diagram of the hardware structure of a terminal implementing an embodiment of the present application.

The terminal 1300 includes but is not limited to: a radio frequency unit 1301, a network module 1302, an audio output unit 1303, an input unit 1304, a sensor 1305, a display unit 1306, a user input unit 1307, an interface unit 1308, a memory 1309 and at least some of the components of a processor 1310.

Those skilled in the art will appreciate that the terminal 1300 may also 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 1310 through a power management system, so as to implement functions such as managing charging, discharging, and power consumption management through the power management system. The terminal structure shown in FIG13 does not constitute a limitation on the terminal, and the terminal may include more or fewer components than shown in the figure, or combine certain components, or arrange components differently, which will not be described in detail here.

It should be understood that in the embodiment of the present application, the input unit 1304 may include a graphics processing unit (GPU) 13041 and a microphone 13042, and the graphics processor 13041 processes the image data of the static picture or video obtained by the image capture device (such as a camera) in the video capture mode or the image capture mode. The display unit 1306 may include a display panel 13061, and the display panel 13061 may be configured in the form of a liquid crystal display, an organic light emitting diode, etc. The user input unit 1307 includes a touch panel 13071 and at least one of other input devices 13072. The touch panel 13071 is also called a touch screen. The touch panel 13071 may include two parts: a touch detection device and a touch controller. Other input devices 13072 may include, but are not limited to, a physical keyboard, function keys (such as a volume control key, a switch key, etc.), a trackball, a mouse, and a joystick, which will not be repeated here.

In the embodiment of the present application, after receiving downlink data from the network side device, the RF unit 1301 can transmit the data to the processor 1310 for processing; in addition, the RF unit 1301 can send uplink data to the network side device. Generally, the RF unit 1301 includes but is not limited to an antenna, an amplifier, a transceiver, a coupler, a low noise amplifier, a duplexer, etc.

The memory 1309 can be used to store software programs or instructions and various data. The memory 1309 may mainly include a first storage area for storing programs or instructions and a second storage area for storing data, wherein the first storage area may store an operating system, an application program or instruction required for at least one function (such as a sound playback function, an image playback function, etc.), etc. In addition, the memory 1309 may include a volatile memory or a non-volatile memory. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), 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 (DDRSDRAM), an enhanced synchronous dynamic random access memory (ESDRAM), a synchronous link dynamic random access memory (SLDRAM) and a direct memory bus random access memory (DRRAM). The memory 1309 in the embodiment of the present application includes but is not limited to these and any other suitable types of memory.

The processor 1310 may include one or more processing units; optionally, the processor 1310 integrates an application processor and a modem processor, wherein the application processor mainly processes operations related to an operating system, a user interface, and application programs, and the modem processor mainly processes wireless communication signals, such as a baseband processor. It is understandable that the modem processor may not be integrated into the processor 1310.

The radio frequency unit 1301 is used for:

Sending capability information of the terminal about an artificial intelligence AI unit to a network side device, where the AI unit is used to compress channel state information CSI;

Receiving CSI configuration information sent by the network side device according to the capability information;

The processor 1310 is used to: compress the CSI through the AI unit according to the CSI configuration information.

Optionally, the capability information includes at least one of the following:

Whether the terminal supports AI-based CSI compression;

The type of the AI unit supported by the terminal;

first indication information, where the first indication information is used to indicate a length of a target time window supported by the terminal, where the target time window includes at least one time unit for the AI unit to perform CSI compression;

The CSI reporting interval mode supported by the terminal.

Optionally, the first indication information includes at least one of the following:

The maximum length of the target time window supported by the terminal;

a set of lengths of the target time window supported by the terminal;

Identification information of the length configuration of the target time window supported by the terminal.

Optionally, the CSI configuration information includes at least one of the following:

Number of CSI reports;

the type of said AI unit used;

CSI grouping information, where the CSI grouping information is used to indicate grouping of CSI to be reported according to the length of a target time window, where the target time window includes at least one time unit for CSI compression by the AI unit;

second indication information, where the second indication information is used to indicate whether, when reporting the CSI, it is necessary to carry a position of the reported CSI in the group to which it belongs;

The payload length of at least part of the CSI report within the target time window, where the payload length is the uplink control information UCI resource length occupied by the CSI report;

The target resource for measuring CSI in the target time window.

Optionally, the CSI grouping information includes at least one of the following:

Identification information of the group to which each CSI to be reported belongs;

The location information of the first CSI report in each group in the CSI reports that need to be reported, and the number of CSI reports included in each group;

The target CSI that needs to be jointly compressed in the time domain, and the identification information of the group to which the target CSI belongs.

Optionally, a payload length of at least part of the CSI reported within the target time window satisfies at least one of the following:

The payload length of each CSI report within the target time window is the same;

The length of the payload of the first CSI report within the target time window is greater than the length of the payload of the CSI report after the first CSI report within the target time window;

The payload lengths of the CSI reports after the first CSI report within the target time window are arranged in an arithmetic progression;

The payload lengths of CSI reports after the first CSI report within the target time window are the same;

In the case where the payload lengths of each CSI report within the target time window are different, the payload lengths of each CSI report within the target time window implicitly indicate the position of each CSI report within the target time window.

Optionally, the target resource meets at least one of the following conditions:

The resources for measuring the CSI each time within the target time window are the same;

The resources used for measuring CSI in the first time unit in the target time window are the largest;

The number of resources used for measuring CSI in each time unit after the first time unit in the target time window is the same;

The number of resources used to measure CSI in each time unit after the first time unit in the target time window decreases in sequence according to the time domain.

It can be understood that the implementation process of each implementation method mentioned in this embodiment can refer to the relevant description of the method embodiment and achieve the same or corresponding technical effect. To avoid repetition, it will not be repeated here.

The embodiment of the present application also provides a network side device, including a processor and a communication interface, the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the steps of the method embodiment shown in Figure 9. The network side device embodiment corresponds to the above network side device method embodiment, and each implementation process and implementation method of the above method embodiment can be applied to the network side device embodiment, and can achieve the same technical effect.

Specifically, the embodiment of the present application also provides a network side device. As shown in Figure 14, the network side device 1400 includes: an antenna 141, a radio frequency device 142, a baseband device 143, a processor 144 and a memory 145. The antenna 141 is connected to the radio frequency device 142. In the uplink direction, the radio frequency device 142 receives information through the antenna 141 and sends the received information to the baseband device 143 for processing. In the downlink direction, the baseband device 143 processes the information to be sent and sends it to the radio frequency device 142. The radio frequency device 142 processes the received information and sends it out through the antenna 141.

The method executed by the network-side device in the above embodiment may be implemented in the baseband device 143, which includes a baseband processor.

The baseband device 143 may include, for example, at least one baseband board, on which multiple chips are arranged, as shown in Figure 14, one of which is, for example, a baseband processor, which is connected to the memory 145 through a bus interface to call the program in the memory 145 to execute the network device operations shown in the above method embodiment.

The network side device may also include a network interface 146, which is, for example, a Common Public Radio Interface (CPRI).

Specifically, the network side device 1400 of the embodiment of the present invention also includes: instructions or programs stored in the memory 145 and executable on the processor 144. The processor 144 calls the instructions or programs in the memory 145 to execute the methods executed by the modules shown in Figure 11 and achieve the same technical effect. To avoid repetition, it will not be described here.

An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the various processes of the above-mentioned AI-based CSI compression method embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

The processor is the processor in the terminal described in the above 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 disk. In some examples, the readable storage medium may be a non-transient readable storage medium.

An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned AI-based CSI compression method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

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

The embodiments of the present application further provide a computer program/program product, which is stored in a storage medium, and is executed by at least one processor to implement the various processes of the above-mentioned AI-based CSI compression method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

An embodiment of the present application also provides an AI-based CSI compression system, including: a terminal and a network side device, wherein the terminal can be used to execute the steps of the AI-based CSI compression method applied to the terminal as above, and the network side device can be used to execute the steps of the method applied to the network side device as above.

It should be noted that, in this article, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises one..." does not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the method and device in the embodiment 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 reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted or combined. In addition, the features described with reference to certain examples may be combined in other examples.

Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of a computer software product plus a necessary general hardware platform, and of course, can also be implemented by hardware. The computer software product is stored in a storage medium (such as ROM, RAM, disk, CD, etc.), including several instructions to enable a terminal or a network-side device to execute the methods described in each embodiment of the present application.

The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms of implementation methods without departing from the purpose of the present application and the scope of protection of the claims, and these implementation methods are all within the protection of the present application.

Claims (23)

A CSI compression method based on AI, wherein the method comprises: The terminal sends capability information of the terminal about an artificial intelligence AI unit to the network side device, where the AI unit is used to compress the channel state information CSI; The terminal receives CSI configuration information sent by the network side device according to the capability information; The terminal compresses the CSI through the AI unit according to the CSI configuration information. The method according to claim 1, wherein the capability information includes at least one of the following: Whether the terminal supports AI-based CSI compression; The type of the AI unit supported by the terminal; first indication information, where the first indication information is used to indicate a length of a target time window supported by the terminal, where the target time window includes at least one time unit for the AI unit to perform CSI compression; The CSI reporting interval mode supported by the terminal. The method according to claim 2, wherein the first indication information includes at least one of the following: The maximum length of the target time window supported by the terminal; a set of lengths of the target time window supported by the terminal; Identification information of the length configuration of the target time window supported by the terminal. The method according to any one of claims 1 to 3, wherein the CSI configuration information includes at least one of the following: Number of CSI reports; the type of said AI unit used; CSI grouping information, where the CSI grouping information is used to indicate grouping of CSI to be reported according to the length of a target time window, where the target time window includes at least one time unit for CSI compression by the AI unit; second indication information, where the second indication information is used to indicate whether, when reporting the CSI, it is necessary to carry a position of the reported CSI in the group to which it belongs; The payload length of at least part of the CSI report within the target time window, where the payload length is the uplink control information UCI resource length occupied by the CSI report; The target resource for measuring CSI in the target time window. The method according to claim 4, wherein the CSI grouping information includes at least one of the following: Identification information of the group to which each CSI to be reported belongs; The location information of the first CSI report in each group in the CSI reports that need to be reported, and the number of CSI reports included in each group; The target CSI that needs to be jointly compressed in the time domain, and the identification information of the group to which the target CSI belongs. The method according to claim 4 or 5, wherein the payload length of at least part of the CSI reported within the target time window satisfies at least one of the following: The payload length of each CSI report within the target time window is the same; The length of the payload of the first CSI report within the target time window is greater than the length of the payload of the CSI report after the first CSI report within the target time window; The payload lengths of the CSI reports after the first CSI report within the target time window are arranged in an arithmetic progression; The payload lengths of CSI reports after the first CSI report within the target time window are the same; In the case where the payload lengths of each CSI report within the target time window are different, the payload lengths of each CSI report within the target time window implicitly indicate the position of each CSI report within the target time window. The method according to any one of claims 4 to 6, wherein the target resource meets at least one of the following conditions: The resources for measuring the CSI each time within the target time window are the same; The resources used for measuring CSI in the first time unit in the target time window are the largest; The number of resources used for measuring CSI in each time unit after the first time unit in the target time window is the same; The number of resources used to measure CSI in each time unit after the first time unit in the target time window decreases in sequence according to the time domain. A CSI compression method based on AI, wherein the method comprises: The network side device receives capability information of the terminal about an artificial intelligence AI unit sent by the terminal, where the AI unit is used to compress channel state information CSI; The network side device determines, according to the capability information, CSI configuration information for instructing the terminal to compress the CSI; The network side device sends the CSI configuration information to the terminal. The method according to claim 8, wherein the capability information includes at least one of the following: Whether the terminal supports AI-based CSI compression; The type of the AI unit supported by the terminal; first indication information, where the first indication information is used to indicate a length of a target time window supported by the terminal, where the target time window includes at least one time unit for the AI unit to perform CSI compression; The CSI reporting interval mode supported by the terminal. The method according to claim 9, wherein the first indication information includes at least one of the following: The maximum length of the target time window supported by the terminal; a set of lengths of the target time window supported by the terminal; Identification information of the length configuration of the target time window supported by the terminal. The method according to any one of claims 8 to 10, wherein the CSI configuration information includes at least one of the following: Number of CSI reports; the type of said AI unit used; CSI grouping information, where the CSI grouping information is used to indicate grouping of CSI to be reported according to a length of a target time window, where the target time window includes at least one time unit for CSI compression by the AI unit; second indication information, where the second indication information is used to indicate whether, when reporting the CSI, it is necessary to carry a position of the reported CSI in the group to which it belongs; The payload length of at least part of the CSI report within the target time window, where the payload length is the uplink control information UCI resource length occupied by the CSI report; The target resource for measuring CSI in the target time window. The method according to claim 11, wherein the CSI grouping information includes at least one of the following: Identification information of the group to which each CSI to be reported belongs; The location information of the first CSI report in each group in the CSI reports that need to be reported, and the number of CSI reports included in each group; The target CSI that needs to be jointly compressed in the time domain, and the identification information of the group to which the target CSI belongs. The method according to claim 11 or 12, wherein the payload length of at least part of the CSI reported within the target time window satisfies at least one of the following: The payload length of each CSI report within the target time window is the same; The length of the payload of the first CSI report within the target time window is greater than the length of the payload of the CSI report after the first CSI report within the target time window; The payload lengths of the CSI reports after the first CSI report within the target time window are arranged in an arithmetic progression; The payload lengths of CSI reports after the first CSI report within the target time window are the same; In the case where the payload lengths of each CSI report within the target time window are different, the payload lengths of each CSI report within the target time window implicitly indicate the position of each CSI report within the target time window. The method according to any one of claims 11 to 13, wherein the target resource meets at least one of the following conditions: The resources for measuring the CSI each time within the target time window are the same; The resources used for measuring CSI in the first time unit in the target time window are the largest; The number of resources used for measuring CSI in each time unit after the first time unit in the target time window is the same; The number of resources used to measure CSI in each time unit after the first time unit in the target time window decreases in sequence according to the time domain. An AI-based CSI compression device, wherein the device is applied to a terminal, and the device includes: A first sending module is used to send capability information of the terminal about an artificial intelligence AI unit to a network side device, where the AI unit is used to compress channel state information CSI; A first receiving module, configured to receive CSI configuration information sent by the network side device according to the capability information; The CSI compression module is used to compress the CSI through the AI unit according to the CSI configuration information. The apparatus according to claim 15, wherein the capability information comprises at least one of the following: Whether the terminal supports AI-based CSI compression; The type of the AI unit supported by the terminal; first indication information, where the first indication information is used to indicate a length of a target time window supported by the terminal, where the target time window includes at least one time unit for the AI unit to perform CSI compression; The CSI reporting interval mode supported by the terminal. The apparatus according to claim 15 or 16, wherein the CSI configuration information includes at least one of the following: Number of CSI reports; the type of said AI unit used; CSI grouping information, where the CSI grouping information is used to indicate grouping of CSI to be reported according to a length of a target time window, where the target time window includes at least one time unit for CSI compression by the AI unit; second indication information, where the second indication information is used to indicate whether, when reporting the CSI, it is necessary to carry a position of the reported CSI in the group to which it belongs; The payload length of at least part of the CSI report within the target time window, where the payload length is the uplink control information UCI resource length occupied by the CSI report; The target resource for measuring CSI in the target time window. An AI-based CSI compression device, wherein the device is applied to a network side device, and the device includes: A second receiving module is used to receive capability information of an artificial intelligence AI unit of the terminal sent by the terminal, where the AI unit is used to compress channel state information CSI; an information determination module, configured to determine, according to the capability information, CSI configuration information for instructing the terminal to compress the CSI; The second sending module is used to send the CSI configuration information to the terminal. The apparatus according to claim 18, wherein the capability information comprises at least one of the following: Whether the terminal supports AI-based CSI compression; The type of the AI unit supported by the terminal; first indication information, where the first indication information is used to indicate a length of a target time window supported by the terminal, where the target time window includes at least one time unit for the AI unit to perform CSI compression; The CSI reporting interval mode supported by the terminal. The apparatus according to claim 18 or 19, wherein the CSI configuration information includes at least one of the following: Number of CSI reports; the type of said AI unit used; CSI grouping information, where the CSI grouping information is used to indicate grouping of CSI to be reported according to a length of a target time window, where the target time window includes at least one time unit for CSI compression by the AI unit; second indication information, where the second indication information is used to indicate whether, when reporting the CSI, it is necessary to carry a position of the reported CSI in the group to which it belongs; The payload length of at least part of the CSI report within the target time window, where the payload length is the uplink control information UCI resource length occupied by the CSI report; The target resource for measuring CSI in the target time window. A terminal, comprising a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the AI-based CSI compression method as described in any one of claims 1 to 7 are implemented. A network side device, comprising a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the AI-based CSI compression method as described in any one of claims 8 to 14 are implemented. A readable storage medium, wherein the readable storage medium stores a program or instruction, and when the program or instruction is executed by a processor, the method for CSI compression based on AI as described in any one of claims 1 to 7 is implemented, or the steps of the method for CSI compression based on AI as described in any one of claims 8 to 14 are implemented.