Information compression method and apparatus, and terminal device and network device

Abstract

Provided in the embodiments of the present application are an information compression method and apparatus, and a terminal device and a network device. The method comprises: according to the priority of first channel state information (CSI) in CSI that is born on an uplink channel, a terminal device determining that it is necessary to perform compression on the first CSI, wherein the first CSI is AI / ML-based CSI, and the uplink channel is used for bearing the first CSI; and the terminal device performing compression on the first CSI by means of adjusting a first model, wherein the first model is used for generating the first CSI.

Description

Information compression method and device, terminal equipment, and network equipment Technical Field

[0001] The embodiments of the present application relate to the field of mobile communication technology, and specifically to an information compression method and apparatus, terminal equipment, and network equipment. Background Art

[0002] In order to enable the network device to perform reasonable scheduling, the terminal device can send channel state information (CSI) to the network device, so that the network device can determine the scheduling information of the terminal device such as the number of transmission layers, precoding matrix, transmission beam, modulation and coding method based on the CSI.

[0003] When a CSI carries a large number of bits, the terminal device can selectively omit or discard the CSI. This means that the terminal device will omit or discard some or all of the CSI information. Omitting or discarding some or all of the CSI information will result in a loss of channel information, affecting communication performance.

[0004] Summary of the Invention

[0005] The embodiments of the present application provide an information compression method and apparatus, a terminal device, and a network device.

[0006] The information compression method provided in the embodiment of the present application includes:

[0007] The terminal device determines, based on a priority of the first channel state information CSI in the CSI carried by the uplink channel, that the first CSI needs to be compressed, wherein the first CSI is AI / ML-based CSI and the uplink channel is used to carry the first CSI;

[0008] The terminal device compresses the first CSI by adjusting a first model, wherein the first model is used to generate the first CSI.

[0009] The information compression method provided in the embodiment of the present application includes:

[0010] The network device determines, based on a priority of the first channel state information CSI in the CSI carried by the uplink channel, that the terminal device needs to compress the first CSI, wherein the first CSI is AI / ML-based CSI and the uplink channel is used to carry the first CSI;

[0011] The network device receives the compressed first CSI; the compressed first CSI is obtained by the terminal device compressing the first CSI by adjusting the first model.

[0012] The information compression device provided in the embodiment of the present application is applied to a terminal device, including:

[0013] A first determining unit is configured to determine, based on a priority of first channel state information CSI in CSI carried by an uplink channel, that the first CSI needs to be compressed, wherein the first CSI is AI / ML-based CSI and the uplink channel is used to carry the first CSI;

[0014] A compression unit is configured to compress the first CSI by adjusting a first model, wherein the first model is used to generate the first CSI.

[0015] The information compression device provided in the embodiment of the present application is applied to a network device, including:

[0016] A second determining unit is configured to determine, based on a priority of the first channel state information CSI in the CSI carried by the uplink channel, that the terminal device needs to compress the first CSI, wherein the first CSI is AI / ML-based CSI and the uplink channel is used to carry the first CSI;

[0017] The receiving unit is configured to receive a compressed first CSI; the compressed first CSI is obtained by the terminal device compressing the first CSI by adjusting the first model.

[0018] The communication device provided in an embodiment of the present application may be a terminal device or a network device in the above-mentioned solution, and includes a processor and a memory. The memory is used to store a computer program, and the processor is used to call and execute the computer program stored in the memory to perform the above-mentioned information compression method.

[0019] The chip provided in the embodiment of the present application is used to implement the above-mentioned information compression method.

[0020] Specifically, the chip includes: a processor, which is used to call and run a computer program from a memory, so that a device equipped with the chip executes the above-mentioned information compression method.

[0021] The computer-readable storage medium provided in an embodiment of the present application is used to store a computer program, which enables a computer to execute the above-mentioned information compression method.

[0022] The computer program product provided in the embodiment of the present application includes computer program instructions, which enable a computer to execute the above-mentioned information compression method.

[0023] The computer program provided in the embodiment of the present application, when executed on a computer, enables the computer to execute the above-mentioned information compression method.

[0024] An embodiment of the present application provides an information compression method, wherein, when the number of bits carried in the first CSI based on AI / ML is large, the terminal device can adjust the first model for generating the first CSI, and use the adjusted first model to process the channel information to generate a compressed first CSI to reduce the number of bits of the first CSI. At the same time, by adjusting the first model to compress the first CSI, the original channel information is not omitted / discarded during the compression process. The compressed first CSI can also be used by the network device to restore the complete channel information, and the compressed first CSI can represent the complete channel information. The information compression method provided in the embodiment of the present application can reduce the number of CSI bits while ensuring the integrity of the channel information, thereby further ensuring the communication performance of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0026] FIG1 is a schematic diagram of a communication architecture provided in an embodiment of the present application;

[0027] FIG2 is a schematic diagram of a neuron structure provided in an embodiment of the present application;

[0028] FIG3 is a schematic diagram of a neural network structure provided in an embodiment of the present application;

[0029] FIG4 is a schematic diagram of a CSI reporting method provided in an embodiment of the present application;

[0030] FIG5 is a flowchart of an information compression method according to an embodiment of the present application;

[0031] FIG6 is a schematic diagram of a neural network for CSI feedback provided in an embodiment of the present application;

[0032] FIG7 is a second flow chart of an information compression method provided in an embodiment of the present application;

[0033] FIG8 is a third flow chart of an information compression method provided in an embodiment of the present application;

[0034] FIG9 is a schematic diagram of the structure of a communication processing device 900 provided in an embodiment of the present application;

[0035] FIG10 is a schematic diagram of the structure of a communication processing device 1000 provided in an embodiment of the present application;

[0036] FIG11 is a schematic structural diagram of a communication device provided in an embodiment of the present application;

[0037] FIG12 is a schematic structural diagram of a chip according to an embodiment of the present application;

[0038] FIG13 is a schematic block diagram of a communication system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0039] The following will describe the technical solutions in the embodiments of this application in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0040] FIG1 is a schematic diagram of an application scenario of an embodiment of the present application.

[0041] As shown in Figure 1, a communication system 100 may include a terminal device 110 and a network device 120. The network device 120 may communicate with the terminal device 110 via an air interface. The terminal device 110 and the network device 120 support multi-service transmission.

[0042] It should be understood that the embodiments of the present application are only illustrative of the communication system 100, but the embodiments of the present application are not limited thereto. That is, the technical solutions of the embodiments of the present application can be applied to various communication systems, such as: Long Term Evolution (LTE) system, LTE Time Division Duplex (TDD), Universal Mobile Telecommunication System (UMTS), Internet of Things (IoT) system, Narrow Band Internet of Things (NB-IoT) system, enhanced Machine-Type Communications (eMTC) system, 5G communication system (also known as New Radio (NR) communication system), or future communication systems.

[0043] In the communication system 100 shown in Figure 1, the network device 120 may be an access network device that communicates with the terminal device 110. The access network device may provide communication coverage for a specific geographical area and may communicate with the terminal device 110 (eg, UE) located within the coverage area.

[0044] The network device 120 can be an evolved base station (eNB or eNodeB) in a Long Term Evolution (LTE) system, or a Next Generation Radio Access Network (NG RAN) device, or a base station (gNB) in an NR system, or a wireless controller in a Cloud Radio Access Network (CRAN), or the network device 120 can be a relay station, an access point, a vehicle-mounted device, a wearable device, a hub, a switch, a bridge, a router, or a network device in a future evolved Public Land Mobile Network (PLMN), etc.

[0045] The terminal device 110 may be any terminal device, including but not limited to a terminal device connected to the network device 120 or other terminal devices by wire or wireless connection.

[0046] For example, the terminal device 110 may refer to an access terminal, user equipment (UE), a subscriber unit, a subscriber station, a mobile station, a mobile station, a remote station, a remote terminal, a mobile device, a user terminal, a terminal, a wireless communication device, a user agent, or a user apparatus. An access terminal may be a cellular phone, a cordless phone, a Session Initiation Protocol (SIP) phone, an IoT device, a satellite handheld terminal, a Wireless Local Loop (WLL) station, a Personal Digital Assistant (PDA), a handheld device with wireless communication capabilities, a computing device or other processing device connected to a wireless modem, an in-vehicle device, a wearable device, a terminal device in a 5G network, or a terminal device in a future evolution network, etc.

[0047] The terminal device 110 can be used for device-to-device (D2D) communication.

[0048] The wireless communication system 100 may further include a core network device 130 for communicating with the base station. The core network device 130 may be a 5G core network (5G Core, 5GC) device, such as an Access and Mobility Management Function (AMF), an Authentication Server Function (AUSF), a User Plane Function (UPF), or a Session Management Function (SMF). Optionally, the core network device 130 may also be an Evolved Packet Core (EPC) device of an LTE network, such as a Session Management Function + Core Packet Gateway (SMF+PGW-C) device. It should be understood that SMF+PGW-C can simultaneously implement the functions that can be implemented by SMF and PGW-C. During the network evolution process, the above-mentioned core network device may also be called other names, or a new network entity may be formed by dividing the functions of the core network, which is not limited in the embodiments of the present application.

[0049] The functional units in the communication system 100 may also establish connections and implement communication via next generation (NG) network interfaces.

[0050] For example, the terminal device establishes an air interface connection with the access network device through the NR interface for transmitting user plane data and control plane signaling; the terminal device can establish a control plane signaling connection with the AMF through the NG interface 1 (referred to as N1); the access network device, such as the next generation wireless access base station (gNB), can establish a user plane data connection with the UPF through the NG interface 3 (referred to as N3); the access network device can establish a control plane signaling connection with the AMF through the NG interface 2 (referred to as N2); the UPF can establish a control plane signaling connection with the SMF through the NG interface 4 (referred to as N4); the UPF can exchange user plane data with the data network through the NG interface 6 (referred to as N6); the AMF can establish a control plane signaling connection with the SMF through the NG interface 11 (referred to as N11); the SMF can establish a control plane signaling connection with the PCF through the NG interface 7 (referred to as N7).

[0051] Figure 1 exemplarily shows a base station, a core network device and two terminal devices. Optionally, the wireless communication system 100 may include multiple base station devices and each base station may include other numbers of terminal devices within its coverage area, which is not limited in this embodiment of the present application.

[0052] It should be noted that Figure 1 is merely an example of a system applicable to this application. Of course, the methods described in the embodiments of this application can also be applied to other systems. Furthermore, the terms "system" and "network" are often used interchangeably herein. The term "and / or" herein simply describes an association relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " generally indicates that the associated objects are in an "or" relationship. It should also be understood that the "indication" mentioned in the embodiments of this application can be a direct indication, an indirect indication, or an indication of an association relationship. For example, "A indicates B" can mean that A directly indicates B, for example, B can obtain information through A; it can also mean that A indirectly indicates B, for example, A indicates C, and B can obtain information through C; or it can mean that A and B have an association relationship. It should also be understood that the "correspondence" mentioned in the embodiments of this application can mean that there is a direct or indirect correspondence between two objects, or that there is an association relationship between the two objects, or a relationship between an indicator and the indicated, a configuration and the configured, and so on. It should also be understood that the “predefined” or “predefined rules” mentioned in the embodiments of the present application can be implemented by pre-saving corresponding codes, tables or other methods that can be used to indicate relevant information in a device (for example, including a terminal device and a network device), and the present application does not limit its specific implementation method. For example, predefined can refer to a definition in a protocol. It should also be understood that in the embodiments of the present application, the “protocol” can refer to a standard protocol in the field of communications, such as an LTE protocol, an NR protocol, and related protocols used in future communication systems, and the present application does not limit this.

[0053] To facilitate understanding of the technical solutions of the embodiments of the present application, the relevant technologies of the embodiments of the present application are described below. The following relevant technologies can be arbitrarily combined with the technical solutions of the embodiments of the present application as optional solutions, and they all fall within the protection scope of the embodiments of the present application.

[0054] Given the tremendous success of artificial intelligence (AI) / machine learning (ML) technologies in computer vision, natural language processing, and other fields, the communications field has begun to try to use AI / ML technologies to seek new technical ideas to solve technical problems that are limited by traditional methods.

[0055] The following is a detailed introduction to the neural network model.

[0056] A neural network is a computational model consisting of multiple interconnected neuron nodes. The connections between nodes represent weighted values ​​from input signals to output signals, called weights. Each node performs a weighted summation of different input signals and outputs the result through a specific activation function. Figure 2 shows a schematic diagram of a neuron structure, where a1, a2, ..., an, and 1 represent the neuron's inputs; w1, w2, ..., wn, and b represent weights; Sum represents the summation function; f represents the activation function; and t represents the output.

[0057] A simple neural network, shown in Figure 3, consists of an input layer, hidden layers, and an output layer. Different connections, weights, and activation functions between multiple neurons produce different outputs, thereby fitting the mapping from input to output. Each node in the previous level is connected to all nodes in the next level. This fully connected model is also called a DNN, or deep neural network.

[0058] A neural network model can be trained and obtained through the process of dataset construction, training, validation, and testing. Training can be divided into offline and online training. Offline training with a dataset can produce static training results, which is referred to as offline training. As network devices or terminal devices use the neural network model, as the terminal devices further measure and / or report, the network devices can continue to collect more data and perform real-time online training to optimize the neural network model parameters and achieve better inference and prediction results. After the neural network model is obtained, the current information is input into the model to infer the corresponding model output.

[0059] The following details the CSI feedback.

[0060] To enable reasonable scheduling by network devices, terminal devices can send downlink CSI to the network devices, so that the network devices can determine scheduling information for the terminal devices, such as the number of transmission layers, precoding matrix, transmit beam, modulation and coding scheme, etc., based on the downlink CSI. The CSI sent by the terminal devices to the network devices is based on the CSI reporting configuration indicated by the network devices. The uplink resources used by the terminal devices to report CSI and the downlink reference signals used for CSI measurements are both indicated by the CSI reporting configuration.

[0061] It should be noted that the CSI in the embodiment of the present application may also be referred to as CSI reporting. The CSI and CSI reporting in the embodiment of the present application are equivalent and interchangeable.

[0062] In practical applications, each CSI reporting configuration may correspond to one CSI report, and each CSI report may include different information such as the channel state information reference signal resource indicator (CSI-RS Resource Indicator, CRI), rank indicator (Rank Indicator, RI), precoding matrix indicator (Precoding Matrix Indicator, PMI), and channel quality indicator (Channel Quality Indicator, CQI).

[0063] The content / information included in the CSI can be determined by the reporting quantity information (reportQuantity) in the CSI reporting configuration. The reporting quantity information can indicate one of the following reporting quantities:

[0064] The CRI is used to determine a CSI-RS resource currently used for channel measurement and an Interference Measurement Resource (IMR) currently used for interference measurement from multiple CSI-RS resources.

[0065] RI is used to report the recommended number of transmission layers.

[0066] The PMI is used to determine the recommended precoding matrix from a predefined codebook.

[0067] The CQI is used to report the current channel quality and can be determined based on the SINR estimated by the terminal device. The channel portion of the SINR is determined based on the non-zero power CSI-RS configured by the network device for channel measurement, and the interference portion is determined based on the CSI-IM or non-zero power CSI-RS configured by the network device for interference measurement. The CQI is calculated based on the reported RI and PMI.

[0068] The Reference Signal Received Power (RSRP) is used to report the RSRP of the SSB or CSI-RS corresponding to the fed-back index, so that the network side can determine the beam used for downlink transmission.

[0069] The layer indicator (LI) is used to report the index of the transmission layer associated with the phase tracking reference signal (PTRS).

[0070] In some embodiments, CSI reporting can be performed in three modes: periodic CSI reporting, quasi-continuous CSI reporting, and aperiodic CSI reporting. As shown in Figure 4 , periodic CSI reporting is transmitted on the Physical Uplink Control Channel (PUCCH). The CSI reporting configuration is configured by the Radio Resource Control (RRC). After receiving the corresponding RRC configuration, the terminal device periodically reports CSI to the network device.

[0071] Quasi-persistent CSI reporting can be transmitted on the PUCCH or the Physical Uplink Shared Channel (PUSCH). The CSI reporting configuration corresponding to the CSI transmitted on the PUCCH is pre-configured by RRC signaling and activated or deactivated by the Media Access Control (MAC) layer signaling; the CSI reporting configuration corresponding to the CSI transmitted on the PUSCH is dynamically indicated (activated or deactivated) by the Downlink Control Information (DCI). After receiving the activation or indication signaling configured by the network device, the terminal device periodically transmits the CSI on the PUCCH or PUSCH until it stops reporting after receiving the deactivation signaling.

[0072] The CSI reporting configuration corresponding to aperiodic CSI reporting is also pre-configured via RRC signaling, with some of the configurations activated via MAC layer signaling. The CSI reporting configuration for CSI reporting is then indicated via CSI trigger signaling in the DCI. Upon receiving the CSI trigger signaling, the terminal device reports the corresponding CSI on the scheduled PUSCH in one go according to the indicated CSI reporting configuration.

[0073] When a CSI carries a large number of bits, in order to prioritize the transmission of important CSI information, a CSI can be divided into two parts: CSI part 1 and part 2. As shown in Table 1, for different types of codebooks, CSI part 1 and part 2 contain different information.

[0074] Table 1

[0075] It should be noted that the number of bits in CSI part 1 is fixed and is used to carry a small amount of important information such as RI and CQI; CSI part 2

[0076] The number of bits is determined by CSI Part 1 and is used to carry information with a large number of bits, such as the PMI. When the code rate of the PUSCH / PUCCH carrying CSI exceeds a certain value, the terminal device needs to perform CSI omission on CSI Part 2. In other words, the terminal device needs to omit / discard some or all of the information in CSI Part 2 to ensure PUSCH / PUCCH transmission performance, so that the code rate does not exceed the maximum transmission code rate configured by the network equipment.

[0077] Among them, the information of CSI part 1 will not be omitted / discarded, and the terminal device can first omit / discard part or all of the information of CSI part 2 in the CSI with lower priority according to the priority of CSI reporting.

[0078] In some embodiments, the priority of CSI reporting is determined based on the periodicity of CSI (periodic, semi-continuous or non-periodic), the content of CSI reporting (whether RSRP reporting), the carrier corresponding to the CSI reporting (i.e., carrier index), and the ID of the CSI reporting configuration. If the priority of a CSI is low and the information of CSI part 2 in the CSI needs to be discarded, the terminal device can first omit / discard the CSI corresponding to the odd subband in CSI part 2, and then omit / discard the CSI corresponding to the even subband, and finally omit / discard the broadband CSI, and even omit / discard all the information of the entire CSI part 2 until the code rate meets the requirements. It can be seen that performing CSI omission on CSI means discarding some information in the CSI and not sending it. Omitting / discarding part or all of the information in CSI part 2 will cause the loss of channel information and affect communication performance.

[0079] Based on this, an embodiment of the present application provides an information compression method, wherein, when the number of bits carried in the first CSI based on AI / ML is large and the uplink channel carrying the first CSI does not meet the maximum transmission code rate limit, the terminal device can adjust the first model for generating the first CSI, and use the adjusted first model to process the channel information to generate a compressed first CSI to reduce the number of bits of the first CSI. At the same time, the compressed first CSI can also be used by the network device to restore the complete channel information, and the compressed first CSI can represent the complete channel information. The information compression method provided in the embodiment of the present application can reduce the number of CSI bits while ensuring the integrity of the channel information, thereby further ensuring the communication performance of the system.

[0080] To facilitate understanding of the technical solutions of the embodiments of the present application, the technical solutions of the present application are described in detail below through specific embodiments. The above related technologies can be combined arbitrarily with the technical solutions of the embodiments of the present application as optional solutions, and all of them fall within the scope of protection of the embodiments of the present application. The embodiments of the present application include at least part of the following contents.

[0081] FIG5 shows an information compression method provided by an embodiment of the present application, which may include step S110 and step S120.

[0082] S110. The terminal device determines that the first CSI needs to be compressed according to the priority of the first CSI in the CSI carried by the uplink channel, where the first CSI is AI / ML-based CSI and the uplink channel is used to carry the first CSI.

[0083] S120. The terminal device compresses the first CSI by adjusting the first model, where the first model is used to generate the first CSI.

[0084] It should be noted that the embodiments of the present application can be applied to a scenario where a terminal device reports CSI to a network device. The terminal device sends CSI to the network device via an uplink channel. The uplink channel here includes PUCCH and / or PUSCH.

[0085] It is understandable that the uplink channel can carry one or more CSIs. The one or more CSIs here may include CSI obtained based on AI / ML methods, CSI obtained based on non-AI / ML methods, and may also additionally include RSRP obtained based on AI / ML methods, RSRP obtained based on non-AI / ML methods, beam information obtained based on AI / ML methods, feedback information for model monitoring, etc.

[0086] Among them, the CSI obtained based on the AI / ML method can be CSI obtained based on neural network model reasoning, or CSI obtained based on bilateral model compression (compressed CSI based on two-sides model), or CSI based on AI training. The embodiments of the present application are not limited to this.

[0087] CSI derived through AI / ML can be obtained by the terminal device using the obtained channel information (such as eigenvectors, beam information, delay information, etc.) as the input of the model, and the model can infer the corresponding CSI quantization bits. Correspondingly, the network device can set up a corresponding model, use the CSI quantization bits as input, and infer the corresponding channel information through the model.

[0088] For example, refer to the two-sided network model for CSI feedback shown in Figure 6. A model for encoding (encoder) is deployed on the terminal device side, and a corresponding model for decoding (decoder) is deployed on the network device side. The terminal device outputs the CSI obtained based on the AI / ML method based on the encoder model, typically such as the PMI bit, and then feeds it back to the network device through the uplink channel (PUSCH / PUCCH). The network device uses the CSI obtained based on the AI / ML method fed back by the terminal device as the input of the decoder model, thereby outputting channel information corresponding to the input on the terminal device side, such as the eigenvectors of each subband, etc., for downlink precoding. At the same time, the terminal device also needs to monitor the model performance and report it to the network device when the model performance is not good, so that the network device can update the model. Since the neural networks on both sides are matched, if the model is updated, the terminal device needs to be notified through signaling.

[0089] It can be understood that the CSI obtained based on the AI / ML method is different from the CSI obtained based on the non-AI / ML method. Among them, in combination with Figures 2, 3 and 6, it can be seen that the encoder model deployed on the terminal device side can use the connection mode, weights and activation functions of multiple neurons to fit the mapping relationship between the input of the encoder model (i.e., channel information) and the output of the encoder model (i.e., CSI obtained based on the AI / ML method). The decoder model deployed on the network device side corresponds to the encoder model deployed on the terminal device side, and can use the connection mode, weights and activation functions of multiple neurons to inversely fit the input of the decoder model (i.e., CSI obtained based on the AI / ML method) to the output of the decoder model (i.e., channel information).

[0090] As can be seen, the mapping between channel information and CSI quantization bits is achieved through a model. As long as the models of the terminal device and the network device match, CSI obtained using AI / ML methods can be transmitted between the terminal device and the network device to complete CSI reporting. Conversely, if the network device does not deploy a model corresponding to the terminal device, or if the model deployed on the network device for the decoder does not correspond to the model deployed on the terminal device for the encoder, the network device cannot recover the channel information corresponding to the CSI obtained using AI / ML methods.

[0091] In other words, the CSI obtained using AI / ML is quantized using a model, and the different bit fields do not carry specific information. In fact, the CSI quantized bits obtained using AI / ML are treated as a whole. Only when network equipment processes the entire CSI quantized bit using a specific model can the corresponding channel information be recovered.

[0092] CSI obtained using non-AI / ML methods is bit information obtained by mapping channel information using preset physical rules, with different bit fields carrying specific channel information. For example, the first half of the bits in the non-AI / ML CSI carry the PMI, and the second half carry the CQI. Alternatively, bits 0 to N-1 in the non-AI / ML CSI can carry the PMI for odd subbands, and bits N to 2N-1 can carry the PMI corresponding to even subbands, and so on.

[0093] As can be seen, performing CSI omission on CSI obtained using non-AI / ML methods essentially omits or discards specific channel information carried in the CSI bits. This results in a loss of channel information and affects communication performance.

[0094] In the embodiment of the present application, the first CSI may be any one of the AI / ML-based CSIs carried by the uplink channel. The first CSI is generated based on the first model.

[0095] Exemplarily, the first model may be the model for the encoder deployed on the terminal device side in the above embodiment.

[0096] Exemplarily, the first model may be obtained by training the terminal device through processes such as data set construction, training, verification, and testing.

[0097] Exemplarily, the first model may be obtained by training the terminal device in advance through offline training and / or online training.

[0098] In an embodiment of the present application, when the terminal device determines that the first CSI needs to be compressed, the first model can be adjusted to compress the first CSI.

[0099] It should be noted that compressing the first CSI refers to reducing the number of bits of the first CSI so that the uplink channel carrying the compressed first CSI can meet the maximum transmission code rate limit.

[0100] In an embodiment of the present application, the terminal device can compress the first CSI by adjusting the first model.

[0101] It can be understood that the terminal device can adjust the first model for generating the first CSI, and use the adjusted first model to process the channel information, and output the compressed first CSI, so that the number of bits of the compressed first CSI is less than the number of bits of the original first CSI, thereby making the uplink information carrying the compressed first CSI meet the maximum transmission code rate limit.

[0102] The purpose of adjusting the first model is to use the adjusted first model to process the channel information, so that the number of bits of the generated compressed first CSI is less than the number of bits of the original first CSI, thereby making the uplink information carrying the compressed first CSI meet the maximum transmission code rate limit.

[0103] As described above, the first model can use the connection method, weights, and activation functions of multiple neurons to fit the mapping relationship between channel information and the first CSI. Adjusting the first model can adjust this mapping relationship so that the input channel information can be mapped to CSI with fewer bits. Accordingly, the model corresponding to the adjusted first model deployed on the network device side can recover the complete channel information based on the CSI with fewer bits.

[0104] That is to say, in an embodiment of the present application, the CSI with fewer bits output by the adjusted first model (i.e., the compressed first CSI) can restore the complete channel information on the network device side, or in other words, the compressed first CSI can still represent the complete channel information.

[0105] In current technologies, terminal devices typically rely on CSI obtained using non-AI / ML methods. As mentioned above, different bit fields in CSI obtained using non-AI / ML methods can carry specific channel information. In related technologies, performing CSI omission on CSI essentially omits or discards the information carrying specific meanings in the CSI bits, resulting in a loss of channel information and impacting communication performance.

[0106] In the embodiment of the present application, the first CSI is compressed, which is different from the CSI omission in the related art. In the embodiment of the present application, the number of bits of the first CSI is reduced by adjusting the first model. Although the number of bits output by the adjusted first model is reduced, the original channel information is not omitted / discarded during the compression process. Accordingly, the compressed first CSI can still restore the original channel information on the network device side. That is to say, compared with the CSI omission operation, in the embodiment of the present application, the terminal device sends relatively complete information to the network device, avoiding the loss of channel information.

[0107] It can be seen that the information compression method provided in the embodiment of the present application can be used as an alternative to CSI omission. In the information compression method provided in the embodiment of the present application, when the number of bits carried in the first CSI based on AI / ML is large and the uplink channel carrying the first CSI does not meet the maximum transmission code rate limit, the terminal device can adjust the first model for generating the first CSI, and use the adjusted first model to process the channel information to generate a compressed first CSI to reduce the number of bits of the first CSI. At the same time, the compressed first CSI can also be used by the network device to restore the complete channel information, and the compressed first CSI can represent the complete channel information. The information compression method provided in the embodiment of the present application can reduce the number of CSI bits while ensuring the integrity of the channel information, thereby further ensuring the communication performance of the system.

[0108] It should be noted that in the embodiments of the present application, the CSI obtained based on the AI / ML method can also be referred to as AI / ML-based CSI, and the two are equivalent or interchangeable.

[0109] In some embodiments, the AI / ML-based CSI may include CSI obtained by AI / ML-based CSI compression and / or CSI obtained by AI / ML-based CSI prediction.

[0110] Among them, the CSI obtained by AI / ML-based CSI compression is essentially different from the compression of the first CSI in the embodiment of the present application. The CSI obtained by AI / ML-based CSI compression refers to compressing the measured channel information (such as eigenvectors, beam information, delay information, etc.) into CSI quantization bits, while the compression of the first CSI in the embodiment of the present application refers to the compression of the CSI quantization bit size, that is, reducing / lowering the size of the CSI quantization bits. In addition, the CSI obtained by AI / ML-based CSI prediction refers to using the channel information obtained by historical measurements to predict future channel information based on AI / ML.

[0111] In an embodiment of the present application, the first CSI may include at least an AI / ML-based PMI, i.e., indication information for obtaining a precoding matrix, and may also include an AI / ML-based CQI and / or an AI / ML-based RI. This embodiment of the present application does not impose any restrictions on this.

[0112] It should be noted that the CSI obtained based on non-AI / ML may include CSI obtained based on a codebook and / or CSI obtained based on a non-codebook method. Among them, the CSI obtained based on a non-codebook method may include CSI obtained based on channel reciprocity.

[0113] In some embodiments, AI / ML-based CSI may also include three reporting modes: periodic CSI reporting, quasi-persistent CSI reporting, and aperiodic CSI reporting.

[0114] In some embodiments, the first CSI may include CSI part 1 and CSI part 2. In the embodiments of the present application, compressing the first CSI actually compresses CSI part 2 of the first CSI. That is, the terminal device may input CSI part 2 of the first CSI into the adjusted first model for processing, and use the output of the adjusted first model and CSI part 1 as the compressed first CSI.

[0115] In some embodiments, adjusting the first model may effectively be selecting a suitable number of CSI output bits from a plurality of candidate CSI output bit numbers, so that the number of bits of the compressed first CSI is less than the number of bits of the original first CSI, and the uplink information carrying the compressed first CSI meets the maximum transmission code rate limit.

[0116] Exemplarily, adjusting the first model may be switching the first model to another model with fewer output bits, or setting the current output bits of the first model to a smaller number of output bits, or compressing the input information of the first model to make the output bits smaller.

[0117] It should be noted that there are many ways to adjust the first model. By adjusting the first model so that the number of bits of the compressed first CSI is less than the number of bits of the original first CSI, and the uplink information carrying the compressed first CSI meets the maximum transmission code rate limit, all of these fall within the scope of protection of this application.

[0118] In some embodiments, when the code rate of the uplink channel is higher than the maximum transmission code rate configured by the network device, the terminal device can determine whether the first CSI needs to be compressed based on the priority order of the first CSI in the CSI carried by the uplink channel.

[0119] In some embodiments, when the code rate of the uplink channel is higher than the maximum transmission code rate configured by the network device, and the first CSI is the CSI with the lowest priority among the CSI carried by the uplink channel (or called non-discarded CSI), the terminal device can determine to compress the first CSI.

[0120] That is, when the code rate of the uplink channel is higher than the maximum transmission code rate configured by the network device, the terminal device may choose to compress the CSI with the lowest priority among one or more CSIs currently carried by the uplink channel.

[0121] It should be noted that, in some embodiments, the AI / ML-based CSI priority determination method can be the same as the CSI priority determination method in the related art. For example, the CSI priority is determined based on the periodicity of the CSI (periodic, semi-continuous, or aperiodic), the carrier corresponding to the CSI report (i.e., the carrier index), and the ID of the CSI reporting configuration.

[0122] In some other embodiments, a new priority determination method may be introduced to determine the priority of the AI / ML-based CSI. For example, the priority of the first CSI in the CSI carried by the uplink channel is determined based on one or more of the following:

[0123] The priority of AI / ML-based CSI is higher than that of CSI obtained by non-AI / ML methods;

[0124] The priority of the AI / ML-based CSI is equal to the priority of the CSI obtained based on the non-AI / ML method. The priority of the AI / ML-based CSI and the priority of the CSI obtained based on the non-AI / ML method are determined according to the corresponding serving cell identification information and / or CSI reporting configuration identification information;

[0125] The priority of CSI based on AI / ML is lower than that of CSI obtained by non-AI / ML methods;

[0126] The priority of CSI based on AI / ML is lower than the priority of RSRP obtained based on non-AI / ML methods;

[0127] The priority of AI / ML-based CSI is lower than the priority of AI / ML-based RSRP reporting;

[0128] AI / ML-based CSI has lower priority than AI / ML-based beam information;

[0129] AI / ML-based CSI has a lower priority than feedback information used for model monitoring;

[0130] Model information of the first model.

[0131] In one possible implementation, when AI / ML-based CSI is introduced into a communication system, a new priority order may be defined based on the AI / ML-based CSI.

[0132] Among them, it can be defined that the priority of AI / ML-based CSI is higher than the priority of CSI obtained based on non-AI / ML methods, or that the priority of AI / ML-based CSI is equal to the priority of CSI obtained based on non-AI / ML methods, or that the priority of AI / ML-based CSI is equal to the priority of CSI obtained based on non-AI / ML methods.

[0133] In addition, it can also be defined that the priority of AI / ML-based CSI is lower than the priority of RSRP obtained based on non-AI / ML methods, the priority of AI / ML-based CSI is lower than the priority of AI / ML-based RSRP reporting, the priority of AI / ML-based CSI is lower than the priority of AI / ML-based beam information, and the priority of AI / ML-based CSI is lower than the priority of feedback information used for model monitoring.

[0134] It should be noted that the terminal device can monitor the model corresponding to the current AI / ML-based CSI and report feedback information of the current model monitoring to the network device. Exemplary feedback information for model monitoring may include, but is not limited to, block error rate (BLER) estimation, throughput estimation, vector correlation estimation, model failure reporting, model update request, etc. of the current model, and the embodiments of the present application do not limit this.

[0135] It should also be noted that when the priority order of two CSIs is the same in the predefined or network device configured priority order, the priority of the two can be further determined based on the service cell identification information and / or CSI reporting configuration identification information corresponding to each CSI.

[0136] For example, when the priority of the AI / ML-based CSI is equal to the priority of the CSI obtained based on the non-AI / ML method, the priority order of the AI / ML-based CSI and the CSI obtained based on the non-AI / ML method can be further determined based on the service cell ID and / or CSI reporting configuration identification information corresponding to the AI / ML-based CSI, and the service cell ID and / or CSI reporting configuration identification information corresponding to the CSI obtained based on the non-AI / ML method.

[0137] Each serving cell identification information and / or each CSI reporting configuration identification information may correspond to one CSI priority, and different serving cell identification information and / or different CSI reporting configuration identification information may correspond to different CSI priorities.

[0138] In some embodiments, each serving cell identification information and / or each CSI reporting configuration identification information corresponding to the CSI priority may be predefined or configured by a network device, and the embodiments of the present application do not impose any restrictions on this.

[0139] It should be noted that the new priority order introduced above can be predefined or configured to the terminal device through the network device, and the embodiments of the present application do not limit this.

[0140] In another possible implementation, the priority of the AI / ML-based CSI may also be related to the model information used to generate the CSI. The model information may include one or more of the following:

[0141] Functionality of the model, type of model, and identification information of the model.

[0142] In other words, for two CSIs based on AI / ML, but generated using different models, different types of models, or different functions, they can have different priorities. In other words, different models, different types of models, or different functions correspond to different CSI priorities.

[0143] For example, the model can be divided into a low-precision model and a high-precision model. The priority of the AI / ML-based CSI generated using the high-precision model is higher than the priority of the CSI obtained based on the codebook; the priority of the AI / ML-based CSI generated using the low-precision model is lower than the priority of the CSI obtained based on the codebook. In this case, the priority of the AI / ML-based CSI generated using the high-precision model is higher than the priority of the AI / ML-based CSI generated using the low-precision model.

[0144] For example, priority can be distinguished based on model ID. The priority of generating AI / ML-based CSI using the model corresponding to the first model ID is higher than the priority of generating AI / ML-based CSI using the model corresponding to the second model ID. It is assumed here that each model has a model ID, that is, there is a one-to-one correspondence between models and model IDs.

[0145] For example, the network device may pre-configure a first model ID set and a second model ID set, wherein the priority of the AI / ML-based CSI generated by the model corresponding to the first model ID set is higher than the priority of the AI / ML-based CSI generated by the model corresponding to the second model ID set. Alternatively, the priority of the AI / ML-based CSI generated by the model corresponding to the first model ID set is higher than the priority of the CSI obtained based on the codebook; and the priority of the CSI obtained based on the codebook is higher than the priority of the AI / ML-based CSI generated by the model corresponding to the second model ID set.

[0146] It should be noted that the CSI priorities corresponding to different models / different types of models / models with different functions can be predefined or configured by the network device, and the embodiments of the present application do not limit this.

[0147] It is understandable that in the embodiment of the present application, the first CSI is AI / ML-based CSI. When determining the first CSI, the terminal device may determine the priority of the first CSI based on the above priority order and / or the model information of the first model.

[0148] In some embodiments, the terminal device may combine the method for determining CSI priority in related technologies with the above-mentioned method for determining CSI priority to obtain the priority of the first CSI.

[0149] Exemplarily, the terminal device may first determine the priority of the first CSI based on the periodicity of the CSI. If the priority of the first CSI is determined to be the same as the priority of other CSIs based on the periodicity, the priority of the first CSI generated based on AI / ML may be further determined based on the above-mentioned predefined or network configured priority order, and / or the model information of the first model. If the priority of the first CSI and other CSIs still cannot be determined, the terminal device may further determine the priority order of the two based on the carriers, service cell IDs, and CSI reporting configuration IDs corresponding to the first CSI and other CSIs.

[0150] It should be noted that the AI / ML-based CSI priority determination method can also be combined with the CSI priority determination method in related technologies through other methods, and the embodiments of the present application do not provide examples one by one for this.

[0151] It can be seen that in the method provided in the embodiment of the present application, the terminal device can determine the priority of the AI / ML-based CSI according to the priority order predefined or configured by the network device, and / or the model information of the first model, clarifying the method for determining the priority of the AI / ML-based CSI and ensuring the normal operation of the communication system.

[0152] In an embodiment of the present application, the terminal device S120 compresses the first CSI by adjusting the first model, which may be achieved by one or more of the following:

[0153] Switching the first model to the second model, processing the channel information using the second model to obtain compressed first CSI; the output bit number of the second model is smaller than the output bit number of the first model;

[0154] Adjusting an initial output bit number of the first model to a target output bit number, processing the channel information using the first model after adjusting the output bit number to obtain a compressed first CSI; wherein the target output bit number is less than the initial output bit number;

[0155] By compressing the channel information input by the first model, the number of output bits of the first model is reduced, and a compressed first CSI is obtained.

[0156] In one possible implementation, the terminal device may support multiple candidate models, and the model structures and / or model parameters corresponding to different candidate models may be different. Accordingly, the output bit numbers of different candidate models may be different.

[0157] It is understood that a terminal device can use multiple candidate models to generate CSI. The input information of each candidate model can be channel information (such as eigenvectors, beam information, delay information, etc.) measured by the terminal device. Each candidate model can infer CSI quantization bits based on the input channel information.

[0158] It should be noted that the multiple candidate models supported by the terminal device can be configured / indicated by the network device or predefined, and the embodiments of the present application do not impose any restrictions on this.

[0159] It should also be noted that in the embodiment of the present application, if the parameters and structures of two models are exactly the same, and only the quantization granularity or quantization method of the CSI bits is different (such as vector quantization or scalar quantization), they are still regarded as the same model.

[0160] In this implementation, if it is determined that the first CSI needs to be compressed, the terminal device can select a second model from multiple supported candidate models that has fewer output bits than the first model, switch the currently used first model to the second model, and process the channel information input by the terminal device through the second model, thereby reducing the model's output bits and the corresponding uplink channel bit rate. This reduces the bit rate of the uplink channel carrying the compressed first CSI to less than or equal to the maximum transmission bit rate configured by the network device. This avoids discarding the entire CSI report and improves the performance of the entire communication system.

[0161] In one embodiment, the second model is a candidate model among multiple candidate models, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0162] It is understandable that the terminal device can select any one of the candidate models from multiple candidate models that makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device. In other words, the second model is any one of the multiple candidate models that makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0163] Exemplarily, the network device can pre-configure 4 models, with output bits of 60, 80, 100, and 120 bits respectively. Assuming that the first model currently used by the terminal device is a model with an output bit number of 120, and the first CSI generated by the first model needs to be compressed, the terminal device can select any one of the other three models that enables the uplink channel (PUSCH or PUCCH) to meet the maximum transmission code rate limit (that is, the code rate of the uplink channel is less than or equal to the maximum transmission code rate configured by the network device). Assuming that the output bit number is less than 90 to meet the maximum transmission code rate limit, the terminal device can select any one of the models with an output bit number of 60 or 80 as the second model. At this time, the terminal device can use the second model instead of the first model. Although reducing the output bit number will affect the accuracy of the CSI report, it avoids discarding the entire CSI report and improves the performance of the entire system.

[0164] In another embodiment, the second model is a candidate model among multiple candidate models that makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device and has the largest output bit number.

[0165] It can be understood that in order to ensure the accuracy of CSI reporting as much as possible, the terminal device can select from multiple candidate models a candidate model that makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device and has the largest output bit number as the second model.

[0166] Exemplarily, the network device can pre-configure 4 models, with output bits of 60, 80, 100, and 120 bits respectively. Assuming that the first model currently used by the terminal device is a model with an output bit number of 120, and the first CSI generated by the first model needs to be compressed, the terminal device can select the model with the most output bits that enables the uplink channel (PUSCH or PUCCH) to meet the maximum code rate limit from the other three supported models. Assuming that the output bit number is less than 90 to meet the maximum code rate limit, the model with an output bit number of 80 can be selected as the second model. At this time, the terminal device can use the second model instead of the first model. Although reducing the output bit number will affect the accuracy of the CSI report, it avoids discarding the entire CSI report and improves the performance of the entire system.

[0167] In another possible implementation, the first model may support multiple output bit numbers. The output bit number of the first model may be switched according to demand.

[0168] It should be noted that the multiple output bit numbers supported by the first model may be multiple different PMI output bit numbers and / or CQI output bit numbers supported by the first model.

[0169] It should also be noted that in the embodiments of the present application, the first model supports multiple output bits. This means that the number of output bits of the first model can be controlled by configuration during model training, and the model is generalizable to different output bit numbers. The multiple output bit numbers supported by the first model here refer to the model itself supporting different output sizes, rather than adjusting the output bit size by adjusting the quantization granularity or quantization method of the CSI bits.

[0170] It should also be noted that the multiple output bit numbers supported by the first model are predefined or configured by the network device. The multiple output bit numbers supported by the first model can also be reported by the terminal device to the network device, which is not limited in this embodiment of the application.

[0171] In this implementation, if it is determined that the first CSI needs to be compressed, the terminal device can adjust the initial output bit number currently used by the first model to a target output bit number with a smaller number of output bits, thereby using the adjusted first model to process the channel information and obtain the compressed first CSI. In other words, the terminal device can select a smaller number of output bits from the multiple output bit numbers supported by the first model, thereby reducing the model's output bit number and the corresponding uplink channel bit rate, thereby making the uplink channel bit rate carrying the compressed first CSI less than or equal to the maximum transmission bit rate configured by the network device. In this way, the entire CSI report can be avoided from being discarded, thereby improving the performance of the entire communication system.

[0172] It should be noted that the multiple output bit numbers supported by the first model can be multiple different PMI bit numbers and / or CQI bit numbers supported by the first model. Accordingly, the terminal device can only adjust the PMI output bit number to achieve CSI omission, the terminal device can also only adjust the CQI output bit number to achieve CSI compression, and the terminal device can also adjust the PMI output bit and the CQI output bit number at the same time to achieve CSI compression. The embodiments of the present application do not impose any restrictions on this.

[0173] In one embodiment, the target target output bit number is an output bit number among multiple output bit numbers supported by the first model, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0174] It is understandable that the terminal device can select, from among the multiple output bit numbers supported by the first model, any output bit number that makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device, as the target output bit number. In other words, the target output bit number is any output bit number among the multiple output bit numbers supported by the first model that makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0175] Exemplarily, the first model can support four different output bit numbers (such as the output bit number of PMI / the output bit number of CQI), which are 60, 70, 100, and 120 bits respectively. Assuming that the output bit number of the first model currently used by the terminal device is 100, and the first CSI generated by the first model needs to be CSI compressed, the terminal device can select any output bit number from the other three output bit numbers so that the uplink channel (PUSCH or PUCCH) can meet the maximum code rate limit. Assuming that the output bit number is less than 80 to meet the maximum code rate limit, the terminal device can select 60 bits or 70 bits as the output bit number of the first model. Although reducing the output bit number will affect the accuracy of the CSI report, it avoids discarding the entire CSI report and improves the performance of the entire system.

[0176] In another embodiment, the target output bit number is the maximum output bit number among the multiple output bit numbers supported by the first model, so that the bit rate of the uplink channel carrying the compressed first CSI is less than or equal to the maximum transmission bit rate configured by the network device, and the output bit number is the largest output bit number.

[0177] It can be understood that in order to ensure the accuracy of the CSI uplink as much as possible, the terminal device can select from the multiple output bit numbers supported by the first model, so that the code rate of the uplink channel carrying the compressed first CSI is less than or equal to the maximum transmission code rate configured by the network device, and the output bit number is the maximum output bit number.

[0178] Exemplarily, the first model can support four different output bit numbers (such as the output bit number of PMI / the output bit number of CQI), which are 60, 70, 100, and 120 bits respectively. Assuming that the output bit number of the first model currently used by the terminal device is 100, and the first CSI report generated by the first model requires CSI compression, the terminal can select the maximum output bit number from the other three output bit numbers so that the uplink channel (PUSCH or PUCCH) can meet the maximum code rate limit. Assuming that the output bit number is less than 80 to meet the maximum code rate limit, the terminal device can select 70 bits as the output bit number of the first model. Although reducing the output bit number will affect the accuracy of the CSI report, it avoids discarding the entire CSI report and improves the performance of the entire system.

[0179] In another possible implementation, the terminal device may compress the channel information input to the first model to reduce the number of output bits of the first model, thereby obtaining compressed first CSI. In other words, the terminal device may compress the first CSI by constraining the input of the first model to adjust the number of output bits of the first model.

[0180] It should be noted that there is a corresponding relationship between the input information of the model and the number of output bits of the model. The more input information the model has, the more output bits it has.

[0181] It can be understood that in an embodiment of the present application, the input information of the first model can be channel information (such as eigenvectors, beam information, delay information, etc.) measured by the terminal device. When the channel information is less, the required PMI quantization bits are also fewer, so the number of output bits can be reduced. Therefore, the number of output bits of the first model can be reduced by reducing or constraining the input information of the first model, thereby reducing the code rate of the uplink channel, and then making the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device. In this way, the entire CSI report can be avoided from being discarded, and the performance of the entire communication system can be improved.

[0182] In some embodiments, compressing the channel information output by the first model to reduce the number of output bits of the first model to obtain the compressed first CSI may include: using part of the channel information as input to the first model to obtain the compressed first CSI;

[0183] The partial channel information includes one or more of the following:

[0184] Channel information of some sub-bands;

[0185] Broadband channel information;

[0186] Partial beam information in the beam information;

[0187] Partial delay information in the delay information.

[0188] In one embodiment, the terminal device may use only the channel information of some subbands as input to the first model. For example, the terminal device may use only the channel information of even subbands as input to the first model (i.e., discarding the information of odd subbands), or only the channel information of odd subbands as input to the first model (i.e., discarding the information of even subbands). The subband channel information here can be a subband eigenvector, a subband beam, a subband phase, etc., and this embodiment of the application is not limited to this.

[0189] In one embodiment, the terminal device may use only wideband channel information as input to the first model. That is, the terminal device may not use subband channel information as model input. The wideband channel information may include wideband eigenvectors, beam information, or delay information, and is not limited in this embodiment of the present application.

[0190] In one embodiment, the terminal device may use part of the beam information as input to the first model. That is, the terminal device may not use all of the beam information as model input, but may select part of the beam information as output of the first model, or may sample the beam information and use the sampled beam information as input to the first model.

[0191] In one embodiment, the terminal device may use only part of the delay information as input to the first model. That is, the terminal device does not use all of the channel delay information as input to the model, but only selects part of it as input to the first model.

[0192] It should be noted that the content of the aforementioned partial channel information may be agreed upon between the terminal device and the network device. For example, the terminal device and the network device may agree to use one or more of partial subband channel information, wideband channel information, partial beam information, and partial delay information as inputs to the first model. Furthermore, the content of the aforementioned partial channel information may also be selected by the terminal device based on implementation. For example, the terminal device may choose to use partial subband channel information, wideband channel information, partial beam information, or partial delay information as inputs to the first model. This embodiment of the present application does not impose any restrictions on this.

[0193] It should also be noted that the above three compression implementation methods can also be implemented in combination.

[0194] In one example, if it is determined that the first CSI needs to be compressed, the terminal device can select a second model with fewer output bits than the first model from multiple supported candidate models, and simultaneously select an appropriate output bit number from the multiple output bit numbers supported by the second model to process the channel information to generate the compressed first CSI, thereby reducing the output bit number of the model and the corresponding uplink channel bit rate, thereby making the bit rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission bit rate configured by the network device. In this way, the number of bits of the first CSI can be further compressed, avoiding discarding the entire CSI report, and improving the performance of the entire communication system.

[0195] In one example, if it is determined that the first CSI needs to be compressed, the terminal device can select a second model from multiple supported candidate models that has fewer output bits than the first model, and simultaneously compress the channel information input by the second model to obtain the compressed first CSI, thereby reducing the model's output bits and the corresponding uplink channel bit rate, thereby making the uplink channel bit rate carrying the compressed first CSI less than or equal to the maximum transmission bit rate configured by the network device. In this way, the number of bits of the first CSI can be further compressed, avoiding discarding the entire CSI report, and improving the performance of the entire communication system.

[0196] In one example, if it is determined that the first CSI needs to be compressed, the terminal device can select a target output bit number with fewer output bits from the multiple output bit numbers supported by the first model, and simultaneously compress the channel information input by the first model to obtain the compressed first CSI, thereby reducing the output bit number of the model and the corresponding uplink channel bit rate, and further making the bit rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission bit rate configured by the network device. In this way, the number of bits of the first CSI can be further compressed, avoiding discarding the entire CSI report, and improving the performance of the entire communication system.

[0197] In one example, if it is determined that the first CSI needs to be compressed, the terminal device can select a second model with fewer output bits than the first model from multiple supported candidate models, and at the same time select an appropriate output bit number from the multiple output bit numbers supported by the second model, and compress the channel information input by the second model to obtain the compressed first CSI, thereby reducing the output bit number of the model and the corresponding uplink channel bit rate, and further making the bit rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission bit rate configured by the network device. In this way, the number of bits of the first CSI can be further compressed, avoiding discarding the entire CSI report, and improving the performance of the entire communication system.

[0198] In an embodiment of the present application, referring to FIG7 , the information compression method provided in the embodiment of the present application may further include the following steps:

[0199] S130. The terminal device sends the compressed first CSI through an uplink channel, and the code rate of the uplink channel is less than or equal to the maximum transmission code rate configured by the network device.

[0200] It is understandable that after the terminal device compresses the first CSI by adjusting the first model, it can send the compressed first CSI to the network device via an uplink channel, wherein the code rate of the uplink channel carrying the compressed first CSI is less than or equal to the maximum transmission code rate configured by the network device.

[0201] Optionally, in some embodiments, referring to FIG7 , the information compression method provided in the embodiment of the present application may further include the following steps:

[0202] S140: The terminal device sends first information. The first information may indicate one or more of the following:

[0203] model information of the second model;

[0204] Target number of output bits;

[0205] Partial channel information;

[0206] The number of output bits after partial channel information is processed by the first model.

[0207] It should be noted that after the terminal device performs CSI omission on the first CSI by adjusting the first model, it can report the relevant information of the adjustment of the first model to the network device, so that the network device can determine the bit size of the compressed first CSI carried in the uplink information based on the reported relevant information to realize the reception of CSI.

[0208] It can be understood that if the terminal device chooses to switch the first model to the second model, and processes the channel information through the second model to obtain the compressed first CSI, the terminal device can report the model information of the second model to the network device. The model information of the second model can be the index information of the second model in the multiple candidate models supported by the terminal device, or it can be the identification information of the second model. The embodiment of the present application does not limit this. In this way, the network device can determine the number of bits of the compressed first CSI carried in the uplink information based on the model information of the second model reported by the terminal device, and realize the reception of CSI.

[0209] If the terminal device chooses to adjust the initial output bit number of the first model to the target output bit number, and processes the channel information through the first model after adjusting the output bit number to obtain the compressed first CSI, the terminal device can report the selected target output bit number to the network device. The first information can indicate the specific value of the target output bit number, or it can indicate the index information of the target output bit number among the multiple output bit numbers supported by the first model, and the embodiment of the present application does not limit this. In this way, the network device can determine the bit size of the compressed first CSI carried in the uplink information based on the target output bit number reported by the terminal device, and realize the reception of CSI.

[0210] If the terminal device compresses the channel information input by the first model, reduces the output bit number of the first model, and obtains the compressed first CSI, the terminal device can report the selected partial channel information to the network device, and / or the output bit number after the partial channel information is processed by the first model.

[0211] In one embodiment, if the content included in the partial channel information is agreed upon by the terminal device and the network device, the terminal device also needs to report to the network device the number of output bits after the first model processes the partial channel information. In this way, the network device can determine the size of the compressed first CSI based on the size of the output bit number reported by the terminal device, and thus perform corresponding CSI reception.

[0212] In one embodiment, if the content of the partial channel information is selected by the terminal device based on implementation, the terminal device may report the specific content of the partial channel information and / or the number of output bits after processing the partial channel information using the first model to the network device. In this way, the network device may determine the size of the compressed first CSI based on the specific content of the partial channel information reported by the terminal device and / or the number of output bits after processing the partial channel information using the first model, thereby performing corresponding CSI reception.

[0213] In an embodiment of the present application, the compressed first CSI may include CSI part 1 and CSI part 2. The number of bits of CSI part 1 is fixed, and the number of bits of CSI part 2 is determined according to the content indicated by CSI part 1;

[0214] The above-mentioned first information can be carried by CSI part 1.

[0215] It should be noted that the compression of the first CSI in the embodiment of the present application is essentially the compression of CSI part 2 in the first CSI. The CSI part 1 in the compressed first CSI is still the CSI part 1 of the original first CSI. The CSI part 2 in the compressed first CSI refers to the information after the CSI part 2 in the original first CSI is compressed. In other words, the CSI part 1 in the compressed first CSI is the same information as the CSI part 1 in the original first CSI. The number of bits of the CSI part 1 of the compressed first CSI is fixed and is used to carry important information such as RI and CQI in the original first CSI.

[0216] It can be understood that the terminal device can report the relevant information of the above-mentioned adjustment of the first model to the network device through CSI part 1.

[0217] In one embodiment of the present application, the information compression method provided in the embodiment of the present application may further include the following steps:

[0218] When one or more of the following conditions are met, the terminal device discards the first CSI;

[0219] The output bit number of the second model is less than the first threshold;

[0220] The target number of output bits is less than a first threshold;

[0221] Any candidate model among multiple candidate models supported by the terminal device makes the code rate of the uplink channel carrying the compressed first CSI greater than the maximum transmission code rate configured by the network device;

[0222] Any output bit number among multiple output bit numbers supported by the first model makes the code rate of the uplink channel carrying the compressed first CSI greater than the maximum transmission code rate configured by the network device;

[0223] The terminal device compresses the channel information input by the first model so that the code rate of the uplink channel carrying the compressed first CSI is greater than the maximum transmission code rate configured by the network device.

[0224] It is understandable that in some specific cases, the terminal device may directly discard the first CSI and not send the first CSI to the network device.

[0225] In one embodiment, if the output bit number of the second model selected by the terminal device is less than a first threshold, the terminal device may discard the first CSI. The first threshold may be the minimum requirement for the number of CSI bits (or CSI feedback accuracy). In other words, if the output bit number of the second model is too low to meet the CSI feedback accuracy requirement, the corresponding first CSI may be directly discarded, thereby saving overhead.

[0226] It should be noted that when it is determined that the output bit number of the second model is less than the first threshold, the terminal device may no longer perform inference based on the second model and output the corresponding compressed first CSI, thereby reducing the operational complexity of the terminal device and improving the processing efficiency of the terminal device.

[0227] Likewise, if the target number of output bits selected by the terminal device is less than the first threshold (that is, the target number of output bits is too low to meet the CSI feedback accuracy requirement), the terminal device may discard the first CSI.

[0228] It should be noted that when it is determined that the selected target output bit number is less than the first threshold, the terminal device does not need to perform inference based on the model and output the corresponding compressed first CSI, thereby reducing the operational complexity of the terminal device and improving the processing efficiency of the terminal device.

[0229] It should also be noted that the above-mentioned first threshold can be predefined or configured by the network device, and the embodiment of the present application does not limit this.

[0230] In one embodiment, if any candidate model among multiple candidate models supported by the terminal device causes the code rate of the uplink channel carrying the compressed first CSI to be greater than the maximum transmission code rate configured by the network device, the terminal device may discard the first CSI. In other words, if the output bit number of all candidate models supported by the terminal device cannot make the uplink channel (PUCCH and / or PUSCH) carrying the compressed first CSI meet the maximum code rate limit, the terminal device may discard the first CSI.

[0231] For example, the network device can pre-configure four models with output bits of 60, 80, 100, and 120 bits, respectively. If the lowest output bit number among the four candidate models cannot meet the maximum transmission rate limit of the uplink channel (for example, the output bit number must be less than 60 to meet the maximum rate limit), the first CSI can be directly discarded. In this case, the terminal device does not need to perform model-based reasoning and output the corresponding compressed first CSI.

[0232] In one embodiment, if any of the multiple output bit numbers supported by the first model causes the code rate of the uplink channel carrying the compressed first CSI to exceed the maximum transmission code rate configured by the network device, the terminal device may discard the first CSI. In other words, if all output bit numbers supported by the first model cannot cause the uplink channel carrying the compressed first CSI to meet the maximum code rate limit, the terminal device discards the first CSI.

[0233] Exemplarily, the first model can support four different output bit numbers (such as the output bit number of PMI / the output bit number of CQI), which are 60, 70, 100, and 120 bits respectively. If the lowest output bit number among the above four different output bit numbers cannot meet the maximum transmission code rate limit of the uplink channel (for example, the output bit number must be less than 60 to meet the maximum code rate limit), the corresponding CSI report can be directly discarded. At this time, the terminal device does not need to perform inference based on the model and output the corresponding compressed first CSI.

[0234] In one embodiment, when a terminal device compresses the channel information input by the first model so that the code rate of the uplink channel carrying the compressed first CSI exceeds the maximum transmission code rate configured by the network device, the terminal device may discard the first CSI. In other words, if the terminal device cannot make the uplink channel carrying the compressed first CSI meet the maximum code rate limit regardless of the compression method used to compress the channel information input by the first model, the terminal device discards the first CSI.

[0235] In summary, in the information compression method provided in the embodiment of the present application, when the priority of the first AI / ML-based CSI is low and CSI omission is required, the number of CSI bits can be reduced by adjusting the first model, thereby avoiding the performance loss caused by directly discarding the entire CSI. When the compression reaches a certain threshold (less than the first threshold) causing CSI reporting to be unavailable, the first CSI can be directly discarded to avoid unnecessary CSI feedback.

[0236] The above text combines Figure 6 to describe in detail the communication processing method of an embodiment of the present invention from the perspective of the terminal. The following text combines Figure 8 to describe in detail the communication processing method of an embodiment of the present invention from the perspective of the network device. It should be understood that some of the steps executed by the network device may correspond to some of the steps executed by the terminal. For the sake of brevity, repeated descriptions are appropriately omitted below.

[0237] FIG8 shows an information compression method provided by an embodiment of the present application, which may include:

[0238] S210. The network device determines, based on a priority of the first CSI in the CSI carried by the uplink channel, that the terminal device needs to compress the first CSI, where the first CSI is AI / ML-based CSI and the uplink channel is used to carry the first CSI.

[0239] S220. The network device receives the compressed first CSI; the compressed first CSI is obtained by the terminal device compressing the first CSI by adjusting the first model.

[0240] It is understandable that the network device needs to perform similar operations as the terminal device to determine whether the terminal device has compressed the first CSI and the output bit size of the compressed first CSI, so as to correctly receive the CSI feedback information.

[0241] In the embodiment of the present application, the compressed first CSI is sent through an uplink channel, and the code rate of the uplink channel is less than or equal to the maximum transmission code rate configured by the network device.

[0242] It should be noted that the first CSI may include CSI obtained by AI / ML-based CSI compression and / or CSI obtained by AI / ML-based CSI prediction.

[0243] It should also be noted that after the network device receives the compressed first CSI sent by the terminal device, it can use the compressed first CSI as the input of the decoder model to restore the actual channel information, such as the eigenvector of the subband.

[0244] In some embodiments, the priority of the first CSI in the CSI carried by the uplink channel is determined based on one or more of the following:

[0245] The priority of AI / ML-based CSI is higher than that of CSI obtained by non-AI / ML methods;

[0246] The priority of the AI / ML-based CSI is equal to the priority of the CSI obtained based on the non-AI / ML method, and the priority of the AI / ML-based CSI and the priority of the CSI obtained based on the non-AI / ML method are determined according to the corresponding serving cell identification information and / or CSI reporting configuration identification information;

[0247] The priority of CSI based on AI / ML is lower than that of CSI obtained by non-AI / ML methods;

[0248] The priority of CSI based on AI / ML is lower than the priority of RSRP obtained based on non-AI / ML methods;

[0249] The priority of AI / ML-based CSI is lower than the priority of AI / ML-based RSRP;

[0250] AI / ML-based CSI has lower priority than AI / ML-based beam information;

[0251] AI / ML-based CSI has a lower priority than feedback information used for model monitoring;

[0252] Model information of the first model.

[0253] In some embodiments, the model information includes one or more of the following:

[0254] The function of the first model, the type of the first model, and identification information of the first model.

[0255] In some embodiments, the CSI obtained based on a non-AI / ML method includes:

[0256] CSI obtained based on a codebook, and / or CSI obtained based on a non-codebook approach.

[0257] In some embodiments, when the code rate of the uplink channel is higher than the maximum transmission code rate configured by the network device and the first CSI is the CSI with the lowest priority among the CSI carried by the uplink channel, the network device can determine that the terminal device has compressed the first CSI.

[0258] In some embodiments, the compressed first CSI is obtained by the terminal device switching the first model to the second model and processing the channel information using the second model; the output bit number of the second model is less than the output bit number of the first model;

[0259] and / or,

[0260] The compressed first CSI is obtained by the terminal device adjusting the initial output bit number of the first model to the target output bit number, and processing the channel information using the first model after the output bit number is adjusted; the target output bit number is less than the initial output bit number;

[0261] and / or,

[0262] The compressed first CSI is obtained by the terminal device by compressing the channel information input by the first model and reducing the number of output bits of the first model.

[0263] In some embodiments, the terminal device supports multiple candidate models, and the multiple candidate models include at least the first model and the second model;

[0264] The second model is a candidate model among multiple candidate models, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0265] In some embodiments, the second model is a candidate model among the multiple candidate models, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device and has the largest output bit number.

[0266] It should be noted that the multiple candidate models supported by the terminal device are predefined or configured by the network device.

[0267] In some embodiments, the first model supports multiple output bit numbers, and the multiple output bit numbers include at least the initial output bit number and the target output bit number;

[0268] The target output bit number is an output bit number among the multiple output bit numbers that makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0269] In some embodiments, the target output bit number is the maximum output bit number among the multiple output bit numbers, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0270] It should be noted that the multiple output bit numbers supported by the first model are predefined or configured by the network device.

[0271] In some embodiments, compressing the channel information input by the first model to reduce the number of output bits of the first model to obtain the compressed first CSI includes:

[0272] Inputting partial channel information in the channel information into the first model to obtain the compressed first CSI, wherein the partial channel information includes one or more of the following:

[0273] Channel information of some sub-bands;

[0274] Broadband channel information;

[0275] Partial beam information in the beam information;

[0276] Partial delay information in the delay information.

[0277] In some embodiments, the information compression method provided in the embodiments of the present application may further include the following steps:

[0278] The network device receives first information sent by the terminal device, where the first information indicates one or more of the following:

[0279] model information of the second model;

[0280] Target number of output bits;

[0281] Partial channel information;

[0282] The number of output bits after partial channel information is processed by the first model.

[0283] In one embodiment, if the terminal device reports the model information of the selected second model through the first information, the network device can determine the size of the corresponding output bit number based on the reported model information, and then obtain the size of the compressed first CSI and perform corresponding CSI reception.

[0284] In one embodiment, if the terminal device reports the selected target output bit number through the first information, the network device can obtain the size of the compressed first CSI based on the size of the reported output bit number, and thus perform corresponding CSI reception.

[0285] In one embodiment, if the terminal device reports the output bit number information and partial channel information of the model through the first information, then the network device can determine the size of the output bit number of the first model based on the reported partial channel information, thereby obtaining the size of the compressed first CSI and performing corresponding CSI reception.

[0286] In one embodiment, if the terminal device reports the number of output bits after processing part of the channel information through the first model through the first information, then the network device can directly determine the size of the compressed first CSI based on the output bit number and perform corresponding CSI reception.

[0287] In some embodiments, the compressed first CSI includes CSI part 1 and CSI part 2, the number of bits of the CSI part 1 is fixed, and the number of bits of the CSI part 2 is determined according to the content indicated by the CSI part 1;

[0288] The first information is carried by the CSI part 1.

[0289] In some embodiments, the network device determines that the terminal device discards the first CSI when one or more of the following conditions are met:

[0290] The output bit number of the second model is less than the first threshold;

[0291] The target number of output bits is less than the first threshold;

[0292] Any candidate model among multiple candidate models supported by the terminal device, such that the code rate of the uplink channel carrying the compressed first CSI is greater than the maximum transmission code rate configured by the network device;

[0293] The first model supports any one of a plurality of output bit numbers, so that the code rate of the uplink channel carrying the compressed first CSI is greater than the maximum transmission code rate configured for the network device;

[0294] The terminal device compresses the channel information input by the first model so that the code rate of the uplink channel carrying the compressed first CSI is greater than the maximum transmission code rate configured by the network device.

[0295] It should be noted that the first threshold is predefined or configured by the network device.

[0296] The following describes in detail the information compression method provided in the embodiments of the present application in combination with specific application scenarios.

[0297] In this application scenario, the terminal device can send a PUCCH or PUSCH to the network device. The PUCCH or PUSCH can carry the first CSI and the second CSI. The first CSI is CSI obtained based on AI / ML, and the second CSI is CSI obtained based on non-AI / ML or AI / ML.

[0298] S1. When the code rate of PUCCH or PUSCH is higher than the maximum transmission code rate configured by the network device, the terminal device compares the priorities of the first CSI and the second CSI.

[0299] For example, the first CSI is CSI obtained based on an AI / ML method, and the second CSI is CSI obtained based on a non-AI / ML method. The first CSI and the second CSI are of different types. The terminal device may determine the priority order of the first CSI and the second CSI based on a predefined priority order. If the predefined priority order specifies that the priority of the AI / ML-based CSI is higher than the priority of the CSI obtained based on a non-AI / ML method, the priority of the first CSI is higher than the priority of the second CSI. If the predefined priority order specifies that the priority of the AI / ML-based CSI is lower than the priority of the CSI obtained based on a non-AI / ML method, the priority of the first CSI is lower than the priority of the second CSI. If the predefined priority order specifies that the priority of the AI / ML-based CSI is equal to the priority of the CSI obtained based on a non-AI / ML method, the terminal device may determine the priority order of the first CSI and the second CSI based on the serving cell ID corresponding to the first CSI and the serving cell ID corresponding to the second CSI, or determine the priority order of the first CSI and the second CSI based on the CSI reporting configuration ID corresponding to the first CSI and the CSI reporting configuration ID corresponding to the second CSI.

[0300] For example, the first CSI and the second CSI are both CSI obtained based on AI / ML methods, the first CSI is generated by a first model, and the second CSI is generated by a third model. The terminal device can determine the priority of the two CSIs based on the model information of the first model and the model information of the third model. The determination of the priority of the two CSIs based on the model information can be referred to the description in the above embodiment and will not be repeated here for the sake of brevity.

[0301] S2. When the priority of the first CSI is lower than that of the second CSI, the terminal device compresses the first CSI.

[0302] It should be noted that the first CSI may include CSI part 1 and CSI part 2.

[0303] It is understandable that if discarding CSI part 2 of the first CSI is necessary to enable the PUSCH or PUCCH to meet the maximum transmission code rate limit, the terminal device directly discards CSI part 2 of the first CSI. Otherwise, the terminal device can reduce the number of bits of the first CSI by adjusting the first model for generating the first CSI to avoid discarding the entire CSI part 2 of the first CSI.

[0304] S2.1. In one embodiment, the terminal device selects a second model from the candidate models that has fewer output bits than the first model, and compresses the CSI part 2 of the first CSI, thereby reducing the number of output bits.

[0305] Exemplarily, the network device can pre-configure 4 candidate models, and the output bit numbers of the 4 candidate models are 60, 80, 100, and 120 bits respectively. If the output bit number of the current first model is 120, the terminal can select the model with the most output bits from the other three candidate models so that the above-mentioned PUSCH or PUCCH can meet the maximum transmission code rate limit. Assuming that the output bit number is less than 90 to meet the maximum transmission code rate limit, the candidate model with an output bit number of 80 can be selected to replace the first model. At this time, the terminal device can use the second model to replace the first model. Although reducing the output bit number will affect the accuracy of the first CSI report, it avoids discarding the entire CSI part 2 of the first CSI, thereby improving the performance of the entire system.

[0306] If the number of output bits of the selected second model is lower than the first threshold, the terminal device may discard CSI Part 2 of the first CSI. In other words, if the number of output bits of the second model is too low to meet the CSI feedback accuracy requirements, the corresponding CSI Part 2 may be directly discarded, thereby saving overhead. In this case, the terminal device does not need to perform model-based inference and output the corresponding CSI.

[0307] If the output bit numbers of all candidate models cannot make the above-mentioned PUCCH or PUSCH meet the maximum transmission code rate limit, the terminal device discards the entire CSI part 2 of the first CSI. For example, if the lowest output bit number among the above four candidate models cannot meet the maximum code rate limit of PUSCH or PUCCH (for example, the output bit number must be less than 60 to meet the maximum code rate limit), the entire CSI part 2 of the first CSI can be directly discarded.

[0308] Furthermore, the terminal device may report the model information of the selected second model to the network device through the CSI part 1 of the first CSI.

[0309] S2.2. In one embodiment, the terminal device selects a suitable output bit number from a plurality of different output bit numbers supported by the first model, and compresses the CSI part 2 of the first CSI, thereby reducing the output bit number.

[0310] Exemplarily, the first model can support four different output bit numbers (such as the number of PMI / CQI bits), which are 60, 70, 100, and 120 bits respectively. Assuming that the output bit number of the first model currently used by the terminal device is 100, the terminal can select the maximum output bit number from the other three output bit numbers so that PUSCH or PUCCH can meet the maximum transmission code rate limit. Assuming that the output bit number is less than 80 to meet the maximum code rate limit, the terminal device can select 70 bits as the output bit number of the first model. Although reducing the output bit number will affect the accuracy of the first CSI report, it avoids discarding the entire CSI part 2 of the first CSI, thereby improving the performance of the entire system.

[0311] If the number of output bits of the selected first model is lower than the first threshold, the terminal device may discard the entire CSI Part 2 of the first CSI. In other words, if the selected number of output bits is too low to meet the CSI feedback accuracy requirements, the entire CSI Part 2 of the first CSI may be directly discarded, thereby saving overhead. In this case, the terminal device does not need to perform model-based inference and output the corresponding CSI.

[0312] If all output bit numbers supported by the first model cannot make the above-mentioned PUCCH or PUSCH meet the maximum transmission code rate limit, the terminal device discards the entire CSI part 2 of the first CSI. For example, if the lowest output bit number among the four different output bit numbers supported by the first model cannot meet the maximum code rate limit of PUSCH or PUCCH (for example, the output bit number must be less than 60 to meet the maximum code rate limit), the entire CSI part 2 of the first CSI can be directly discarded.

[0313] Furthermore, the terminal device may report the selected number of output bits to the network device via the CSI part 1 of the first CSI. The information indicated by the terminal device may be an index of the selected number of output bits among multiple numbers of output bits supported by the current model.

[0314] S2.3. In one embodiment, the terminal device adjusts the corresponding number of output bits by compressing the input of the first model, and compresses the CSI part 2 of the first CSI, thereby reducing the number of output bits.

[0315] In some embodiments, the terminal device uses only the channel information of some subbands as input to the first model. For example, only the channel information of even subbands is used as input to the first model, while the channel information of odd subbands is not used as input to the first model. The subband channel information here can include the eigenvector of the subband, the phase of the subband, etc.

[0316] In some embodiments, the terminal device uses only wideband channel information as input to the first model. For example, the terminal device does not use subband channel information as input to the model. The wideband channel information here can be a wideband eigenvector, beam, or delay.

[0317] In some embodiments, the terminal device uses only part of the beam information as input to the first model. For example, the terminal device does not use all of the beam information as input to the first model, but only selects a portion of the sampled beams as input.

[0318] In some embodiments, the terminal device uses only part of the delay information as input to the first model. For example, the terminal device does not use all of the channel delay information as input to the first model, but only selects part of it as input.

[0319] Furthermore, the terminal device may report the output bit number of the model after the constraint input to the network device via the CSI part 1 of the first CSI. For example, the PMI quantization bit number after the constraint input is reported to the network device via the CSI part 1.

[0320] It should be noted that if the terminal device does not discard the entire CSI part 2 of the first CSI, the terminal device can send the CSI part 1 of the first CSI and the compressed CSI part 2 to the network device through PUSCH or PUCCH.

[0321] S3. The network device receives the PUSCH or PUCCH sent by the terminal device.

[0322] It should be noted that the network device also needs to perform similar operations as the terminal device to determine whether the terminal device has compressed the first CSI and the size of the compressed output bits, so as to correctly receive the CSI feedback information.

[0323] If the terminal device reports the relevant information for adjusting the first model through the CSI part 1 of the first CSI (for example, the model information of the second model selected by the terminal device in S2.1, the output bit number selected by the terminal device in S2.2, or the output bit number of the model after constrained input in S2.3), the network device can obtain the bit size of the compressed CSI part 2 in the first CSI based on the information reported by the CSI part 1 of the first CSI, and thus perform corresponding CSI reception.

[0324] After receiving the compressed CSI part 2 of the first CSI, the network device can use the compressed CSI part 2 as the input of the decoder model to restore the actual channel information, such as the eigenvector of the subband, or the beam and delay information.

[0325] The information compression method provided in the embodiment of the present application can be used as an alternative to CSI omission. In the information compression method provided in the embodiment of the present application, when the number of bits carried in the first CSI based on AI / ML is large and the uplink channel carrying the first CSI does not meet the maximum transmission code rate limit, the terminal device can adjust the first model for generating the first CSI, and use the adjusted first model to process the channel information to generate a compressed first CSI to reduce the number of bits of the first CSI. At the same time, the compressed first CSI can also be used by the network device to restore the complete channel information, and the compressed first CSI can represent the complete channel information. The information compression method provided in the embodiment of the present application can reduce the number of CSI bits while ensuring the integrity of the channel information, thereby further ensuring the communication performance of the system.

[0326] The preferred embodiments of the present application are described in detail above in conjunction with the accompanying drawings. However, the present application is not limited to the specific details in the above embodiments. Within the technical concept of the present application, the technical solution of the present application can be subjected to a variety of simple modifications, and these simple modifications all fall within the scope of protection of the present application. For example, the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present application will not further explain the various possible combinations. For another example, the various different embodiments of the present application can also be arbitrarily combined, as long as they do not violate the idea of ​​the present application, they should also be regarded as the contents disclosed in the present application. For another example, under the premise of no conflict, the various embodiments and / or the technical features in each embodiment described in the present application can be arbitrarily combined with the prior art, and the technical solution obtained after the combination should also fall within the scope of protection of the present application.

[0327] It should also be understood that in the various method embodiments of the present application, the sequence numbers of the above-mentioned processes do not imply a precedence in the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application. In addition, in the embodiments of the present application, the terms "downlink," "uplink," and "sidelink" are used to indicate the transmission direction of signals or data, where "downlink" is used to indicate the first direction of transmission of signals or data from a site to a user equipment in a cell, "uplink" is used to indicate the second direction of transmission of signals or data from a user equipment in a cell to a site, and "sidelink" is used to indicate the third direction of transmission of signals or data from user equipment 1 to user equipment 2. For example, "downlink signal" indicates that the transmission direction of the signal is the first direction. In addition, in the embodiments of the present application, the term "and / or" is merely a description of the association relationship between associated objects, indicating that three relationships can exist. Specifically, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this document generally indicates that the associated objects are in an "or" relationship.

[0328] FIG9 is a schematic diagram of the structure of an information compression device 900 provided in an embodiment of the present application, which is applied to a terminal device. As shown in FIG9 , the information compression device 900 includes:

[0329] A first determining unit 901 is configured to determine, based on a priority of first channel state information CSI in CSI carried by an uplink channel, that compression of the first CSI is required, wherein the first CSI is AI / ML-based CSI and the uplink channel is used to carry the first CSI;

[0330] The compression unit 902 is configured to compress the first CSI by adjusting a first model, wherein the first model is used to generate the first CSI.

[0331] In some embodiments, the priority of the first CSI in the CSI carried by the uplink channel is determined based on one or more of the following:

[0332] The priority of AI / ML-based CSI is higher than that of CSI obtained by non-AI / ML methods;

[0333] The priority of the AI / ML-based CSI is equal to the priority of the CSI obtained based on the non-AI / ML method, and the priority of the AI / ML-based CSI and the priority of the CSI obtained based on the non-AI / ML method are determined according to the corresponding serving cell identification information and / or CSI reporting configuration identification information;

[0334] The priority of CSI based on AI / ML is lower than that of CSI obtained by non-AI / ML methods;

[0335] The priority of CSI based on AI / ML is lower than the priority of RSRP obtained based on non-AI / ML methods;

[0336] The priority of AI / ML-based CSI is lower than the priority of AI / ML-based RSRP reporting;

[0337] AI / ML-based CSI has lower priority than AI / ML-based beam information;

[0338] AI / ML-based CSI has a lower priority than feedback information used for model monitoring;

[0339] Model information of the first model.

[0340] In some embodiments, the model information includes one or more of the following:

[0341] The function of the first model, the type of the first model, and identification information of the first model.

[0342] In some embodiments, the CSI obtained based on a non-AI / ML method includes:

[0343] CSI obtained based on a codebook, and / or CSI obtained based on a non-codebook approach.

[0344] In some embodiments, the first determination unit 901 is configured to, when the code rate of the uplink channel is higher than the maximum transmission code rate configured by the network device, and the first CSI is the CSI with the lowest priority among the CSI carried by the uplink channel, the terminal device determines that the first CSI needs to be compressed.

[0345] In some embodiments, the compression unit 902 is configured to do one or more of the following:

[0346] Switching the first model to a second model, processing the channel information using the second model to obtain compressed first CSI; the number of output bits of the second model is smaller than the number of output bits of the first model;

[0347] Adjusting an initial output bit number of the first model to a target output bit number, and processing the channel information using the first model after adjusting the output bit number to obtain a compressed first CSI; wherein the target output bit number is less than the initial output bit number;

[0348] By compressing the channel information input by the first model, the number of output bits of the first model is reduced, and a compressed first CSI is obtained.

[0349] In some embodiments, the terminal device supports multiple candidate models, and the multiple candidate models include at least the first model and the second model;

[0350] The second model is a candidate model among the multiple candidate models, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0351] In some embodiments, the second model is a candidate model among the multiple candidate models that makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device and has the largest output bit number.

[0352] In some embodiments, the plurality of candidate models are predefined or configured by the network device.

[0353] In some embodiments, the first model supports multiple output bit numbers, and the multiple output bit numbers include at least the initial output bit number and the target output bit number;

[0354] The target output bit number is the output bit number among the multiple output bit numbers that makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0355] In some embodiments, the target output bit number is the maximum output bit number among the multiple output bit numbers, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0356] In some embodiments, the multiple output bit numbers supported by the first model are predefined or configured by the network device.

[0357] In some embodiments, the compression unit 902 is further configured to use partial channel information in the channel information as input to the first model to obtain the compressed first CSI; the partial channel information includes one or more of the following:

[0358] Channel information of some sub-bands;

[0359] Broadband channel information;

[0360] Partial beam information in the beam information;

[0361] Partial delay information in the delay information.

[0362] In some embodiments, the information compression apparatus 900 further includes a sending unit, wherein the sending unit is configured to send the compressed first CSI via the uplink channel, wherein the code rate of the uplink channel is less than or equal to the maximum transmission code rate configured by the network device.

[0363] In some embodiments, the sending unit is further configured to send first information, where the first information indicates one or more of the following:

[0364] model information of the second model;

[0365] Target number of output bits;

[0366] Partial channel information;

[0367] The number of output bits after partial channel information is processed by the first model.

[0368] In some embodiments, the compressed first CSI includes CSI part 1 and CSI part 2, the number of bits of the CSI part 1 is fixed, and the number of bits of the CSI part 2 is determined according to the content indicated by the CSI part 1; the first information is carried by the CSI part 1.

[0369] In some embodiments, the first determining unit 901 is further configured to discard the first CSI if one or more of the following conditions are met:

[0370] The output bit number of the second model is less than the first threshold;

[0371] The target number of output bits is less than the first threshold;

[0372] Any candidate model among multiple candidate models supported by the terminal device makes the code rate of the uplink channel carrying the compressed first CSI greater than the maximum transmission code rate configured by the network device;

[0373] Any output bit number among multiple output bit numbers supported by the first model makes the code rate of the uplink channel carrying the compressed first CSI greater than the maximum transmission code rate configured by the network device;

[0374] The terminal device compresses the channel information input by the first model so that the code rate of the uplink channel carrying the compressed first CSI is greater than the maximum transmission code rate configured by the network device.

[0375] In some embodiments, the first threshold is predefined or configured by a network device.

[0376] In some embodiments, the first CSI includes CSI obtained by AI / ML-based CSI compression and / or CSI obtained by AI / ML-based CSI prediction.

[0377] FIG10 is a schematic diagram of the structure of an information compression device 1000 provided in an embodiment of the present application, which is applied to a network device. As shown in FIG10 , the information compression device 1000 includes:

[0378] The second determining unit 1001 is configured to determine, based on a priority of the first channel state information CSI in the CSI carried by the uplink channel, that the terminal device needs to compress the first CSI, wherein the first CSI is AI / ML-based CSI and the uplink channel is used to carry the first CSI;

[0379] The receiving unit 1002 is configured to receive a compressed first CSI; the compressed first CSI is obtained by the terminal device compressing the first CSI by adjusting the first model.

[0380] In some embodiments, the priority of the first CSI in the CSI carried by the uplink channel is determined based on one or more of the following:

[0381] The priority of AI / ML-based CSI is higher than that of CSI obtained by non-AI / ML methods;

[0382] The priority of the AI / ML-based CSI is equal to the priority of the CSI obtained based on the non-AI / ML method, and the priority of the AI / ML-based CSI and the priority of the CSI obtained based on the non-AI / ML method are determined according to the corresponding serving cell identification information and / or CSI reporting configuration identification information;

[0383] The priority of CSI based on AI / ML is lower than that of CSI obtained by non-AI / ML methods;

[0384] The priority of CSI based on AI / ML is lower than the priority of RSRP obtained based on non-AI / ML methods;

[0385] The priority of AI / ML-based CSI is lower than the priority of AI / ML-based RSRP;

[0386] AI / ML-based CSI has lower priority than AI / ML-based beam information;

[0387] AI / ML-based CSI has a lower priority than feedback information used for model monitoring;

[0388] Model information of the first model.

[0389] In some embodiments, the model information includes one or more of the following:

[0390] The function of the first model, the type of the first model, and identification information of the first model.

[0391] In some embodiments, the CSI obtained based on a non-AI / ML method includes:

[0392] CSI obtained based on a codebook, and / or CSI obtained based on a non-codebook approach.

[0393] In some embodiments, the second determination unit 1001 is further configured to, when the code rate of the uplink channel is higher than the maximum transmission code rate configured by the network device, and the first CSI is the CSI with the lowest priority among the CSI carried by the uplink channel, the network device determines that the terminal device compresses the first CSI.

[0394] In some embodiments, the compressed first CSI is obtained by the terminal device switching the first model to a second model and processing the channel information through the second model; the output bit number of the second model is less than the output bit number of the first model; and / or,

[0395] The compressed first CSI is obtained by the terminal device adjusting the initial output bit number of the first model to a target output bit number and processing the channel information using the first model after adjusting the output bit number; the target output bit number is less than the initial output bit number; and / or,

[0396] The compressed first CSI is obtained by the terminal device by compressing the channel information input by the first model and reducing the number of output bits of the first model.

[0397] In some embodiments, the terminal device supports multiple candidate models, and the multiple candidate models include at least the first model and the second model;

[0398] The second model is a candidate model among the multiple candidate models, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0399] In some embodiments, the second model is a candidate model among the multiple candidate models, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device and has the largest output bit number.

[0400] In some embodiments, the plurality of candidate models are predefined or configured by the network device.

[0401] In some embodiments, the first model supports multiple output bit numbers, and the multiple output bit numbers include at least the initial output bit number and the target output bit number;

[0402] The target output bit number is an output bit number among the multiple output bit numbers that makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0403] In some embodiments, the target output bit number is the maximum output bit number among the multiple output bit numbers, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

[0404] In some embodiments, the multiple output bit numbers supported by the first model are predefined or configured by the network device.

[0405] In some embodiments, the compressed first CSI is obtained by the terminal device inputting partial channel information in the channel information into the first model; the partial channel information includes one or more of the following:

[0406] Channel information of some sub-bands;

[0407] Broadband channel information;

[0408] Partial beam information in the beam information;

[0409] Partial delay information in the delay information.

[0410] In some embodiments, the compressed first CSI is sent through the uplink channel, and a code rate of the uplink channel is less than or equal to a maximum transmission code rate configured by the network device.

[0411] In some embodiments, the receiving unit 1002 is further configured to receive first information, where the first information indicates one or more of the following:

[0412] model information of the second model;

[0413] Target number of output bits;

[0414] Partial channel information;

[0415] The number of output bits after partial channel information is processed by the first model.

[0416] In some embodiments, the compressed first CSI includes CSI part 1 and CSI part 2, the number of bits of the CSI part 1 is fixed, and the number of bits of the CSI part 2 is determined according to the content indicated by the CSI part 1; the first information is carried by the CSI part 1.

[0417] In some embodiments, the second determining unit 1001 is further configured to determine that the terminal device discards the first CSI when one or more of the following conditions are met:

[0418] The output bit number of the second model is less than the first threshold;

[0419] The target number of output bits is less than the first threshold;

[0420] Any candidate model among multiple candidate models supported by the terminal device, such that the code rate of the uplink channel carrying the compressed first CSI is greater than the maximum transmission code rate configured by the network device;

[0421] The first model supports any one of a plurality of output bit numbers, so that the code rate of the uplink channel carrying the compressed first CSI is greater than the maximum transmission code rate configured for the network device;

[0422] The terminal device compresses the channel information input by the first model so that the code rate of the uplink channel carrying the compressed first CSI is greater than the maximum transmission code rate configured by the network device.

[0423] In some embodiments, the first threshold is predefined or configured by the network device.

[0424] In some embodiments, the first CSI includes CSI obtained by AI / ML-based CSI compression and / or CSI obtained by AI / ML-based CSI prediction.

[0425] Those skilled in the art should understand that the relevant description of the above-mentioned information compression device in the embodiment of the present application can be understood with reference to the relevant description of the information compression method in the embodiment of the present application.

[0426] Figure 11 is a schematic diagram of a communication device 1100 provided in an embodiment of the present application. The communication device can be a terminal device or a network device. The communication device 1100 shown in Figure 11 includes a processor 1110, which can call and execute a computer program from a memory to implement the method in the embodiment of the present application.

[0427] Optionally, as shown in FIG11 , the communication device 1100 may further include a memory 1120. The processor 1110 may call and execute a computer program from the memory 1120 to implement the method in the embodiment of the present application.

[0428] The memory 1120 may be a separate device independent of the processor 1110 , or may be integrated into the processor 1110 .

[0429] Optionally, as shown in FIG11 , the communication device 1100 may further include a transceiver 1130 , and the processor 1110 may control the transceiver 1130 to communicate with other devices, specifically, to send information or data to other devices, or to receive information or data sent by other devices.

[0430] The transceiver 1130 may include a transmitter and a receiver. The transceiver 1130 may further include an antenna, and the number of antennas may be one or more.

[0431] Optionally, the communication device 1100 may specifically be a network device in an embodiment of the present application, and the communication device 1100 may implement the corresponding processes implemented by the network device in each method in the embodiment of the present application. For the sake of brevity, they will not be repeated here.

[0432] Optionally, the communication device 1100 may specifically be a mobile terminal / terminal device of an embodiment of the present application, and the communication device 1100 may implement the corresponding processes implemented by the mobile terminal / terminal device in each method of the embodiment of the present application. For the sake of brevity, they will not be repeated here.

[0433] Figure 12 is a schematic structural diagram of a chip according to an embodiment of the present application. The chip 1200 shown in Figure 12 includes a processor 1210, which can call and run a computer program from a memory to implement the method according to the embodiment of the present application.

[0434] Optionally, as shown in FIG12 , the chip 1200 may further include a memory 1220. The processor 1210 may call and execute a computer program from the memory 1220 to implement the method in the embodiment of the present application.

[0435] The memory 1220 may be a separate device independent of the processor 1210 , or may be integrated into the processor 1210 .

[0436] Optionally, the chip 1000 may further include an input interface 1230. The processor 1210 may control the input interface 1230 to communicate with other devices or chips, and specifically, may obtain information or data sent by other devices or chips.

[0437] Optionally, the chip 1200 may further include an output interface 1240. The processor 1210 may control the output interface 1240 to communicate with other devices or chips, and specifically, may output information or data to other devices or chips.

[0438] Optionally, the chip can be applied to the network device in the embodiments of the present application, and the chip can implement the corresponding processes implemented by the network device in each method of the embodiments of the present application. For the sake of brevity, they will not be repeated here.

[0439] Optionally, the chip can be applied to the mobile terminal / terminal device in the embodiments of the present application, and the chip can implement the corresponding processes implemented by the mobile terminal / terminal device in the various methods of the embodiments of the present application. For the sake of brevity, they will not be repeated here.

[0440] 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.

[0441] FIG13 is a schematic block diagram of a communication system 1300 provided in an embodiment of the present application. As shown in FIG13 , the communication system 1300 includes a terminal device 1310 and a network device 1320 .

[0442] Among them, the terminal device 1310 can be used to implement the corresponding functions implemented by the terminal device in the above method, and the network device 1320 can be used to implement the corresponding functions implemented by the network device in the above method. For the sake of brevity, they are not repeated here.

[0443] It should be understood that the processor of the embodiments of the present application may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method embodiment can be completed by hardware integrated logic circuits in the processor or software instructions. The above processor can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The various methods, steps, and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed by a hardware decoding processor, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium mature in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0444] It is understood that the memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. 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), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct RAM bus random access memory (DR RAM). It should be noted that the memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0445] It should be understood that the above-mentioned memories are exemplary but not restrictive. For example, the memories in the embodiments of the present application may also be static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct RAM RAM (DR RAM), etc. In other words, the memories in the embodiments of the present application are intended to include, but are not limited to, these and any other suitable types of memories.

[0446] An embodiment of the present application also provides a computer-readable storage medium for storing a computer program.

[0447] Optionally, the computer-readable storage medium can be applied to the network device in the embodiments of the present application, and the computer program enables the computer to execute the corresponding processes implemented by the network device in the various methods of the embodiments of the present application. For the sake of brevity, they are not repeated here.

[0448] Optionally, the computer-readable storage medium can be applied to the mobile terminal / terminal device in the embodiments of the present application, and the computer program enables the computer to execute the corresponding processes implemented by the mobile terminal / terminal device in the various methods of the embodiments of the present application. For the sake of brevity, they will not be repeated here.

[0449] An embodiment of the present application also provides a computer program product, including computer program instructions.

[0450] Optionally, the computer program product can be applied to the network device in the embodiments of the present application, and the computer program instructions enable the computer to execute the corresponding processes implemented by the network device in the various methods of the embodiments of the present application. For the sake of brevity, they are not repeated here.

[0451] Optionally, the computer program product can be applied to the mobile terminal / terminal device in the embodiments of the present application, and the computer program instructions enable the computer to execute the corresponding processes implemented by the mobile terminal / terminal device in the various methods of the embodiments of the present application. For the sake of brevity, they will not be repeated here.

[0452] The embodiment of the present application also provides a computer program.

[0453] Optionally, the computer program can be applied to the network device in the embodiments of the present application. When the computer program runs on a computer, the computer executes the corresponding processes implemented by the network device in the various methods of the embodiments of the present application. For the sake of brevity, they are not described here.

[0454] Optionally, the computer program can be applied to the mobile terminal / terminal device in the embodiments of the present application. When the computer program runs on the computer, the computer executes the corresponding processes implemented by the mobile terminal / terminal device in the various methods of the embodiments of the present application. For the sake of brevity, they will not be repeated here.

[0455] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0456] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0457] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0458] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0459] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0460] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0461] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A method for compressing information, the method comprising: The terminal device determines, according to a priority of the first channel state information CSI in the CSI carried by the uplink channel, that the first CSI needs to be compressed, wherein the first CSI is an AI / ML-based CSI, and the uplink channel is used to carry the first CSI; The terminal device compresses the first CSI by adjusting a first model, wherein the first model is used to generate the first CSI.

2. The method according to claim 1, wherein: The priority of the first CSI in the CSI carried by the uplink channel is determined based on one or more of the following: The priority of CSI based on AI / ML is higher than that of CSI obtained by non-AI / ML methods; The priority of the AI / ML-based CSI is equal to the priority of the CSI obtained based on the non-AI / ML method, and the priority of the AI / ML-based CSI and the priority of the CSI obtained based on the non-AI / ML method are determined according to the corresponding serving cell identification information and / or CSI reporting configuration identification information; The priority of CSI based on AI / ML is lower than that of CSI obtained based on non-AI / ML methods; The priority of CSI based on AI / ML is lower than the priority of RSRP obtained based on non-AI / ML methods; The priority of AI / ML-based CSI is lower than the priority of AI / ML-based RSRP reporting; The priority of AI / ML-based CSI is lower than the priority of AI / ML-based beam information; AI / ML-based CSI has a lower priority than feedback information used for model monitoring; Model information of the first model.

3. The method according to claim 2, wherein: The model information includes one or more of the following: The function of the first model, the type of the first model, and identification information of the first model.

4. The method according to claim 2 or 3, wherein: The CSI obtained based on non-AI / ML methods includes: CSI obtained based on a codebook, and / or CSI obtained based on a non-codebook approach.

5. The method according to any one of claims 1 to 4, wherein: The terminal device determines, according to a priority of the first channel state information CSI in the CSI carried by the uplink channel, that the first CSI needs to be compressed, including: When the code rate of the uplink channel is higher than the maximum transmission code rate configured by the network device, and the first CSI is the CSI with the lowest priority among the CSIs carried by the uplink channel, the terminal device determines that the first CSI needs to be compressed.

6. The method according to any one of claims 1 to 5, wherein: The terminal device compresses the first CSI by adjusting the first model, including one or more of the following: Switching the first model to a second model, processing the channel information by the second model to obtain a compressed first CSI; the output bit number of the second model is less than the output bit number of the first model; Adjusting the initial output bit number of the first model to a target output bit number, and processing the channel information by using the first model after adjusting the output bit number to obtain a compressed first CSI; The target output bit number is less than the initial output bit number; The channel information input by the first model is compressed to reduce the number of output bits of the first model, thereby obtaining a compressed first CSI.

7. The method according to claim 6, wherein: The terminal device supports multiple candidate models, and the multiple candidate models include at least the first model and the second model; The second model is a candidate model among the multiple candidate models, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

8. The method according to claim 7, wherein: The second model is a candidate model among the multiple candidate models, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device and has the largest output bit number.

9. The method according to claim 7 or 8, wherein: The multiple candidate models are predefined or configured by the network device.

10. The method according to claim 6, wherein: The first model supports multiple output bit numbers, and the multiple output bit numbers at least include the initial output bit number and the target output bit number; The target output bit number is the output bit number among the multiple output bit numbers that makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

11. The method according to claim 10, wherein: The target output bit number is the maximum output bit number among the multiple output bit numbers, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device, and the output bit number.

12. The method according to claim 10 or 11, wherein: The multiple output bit numbers supported by the first model are predefined or configured by the network device.

13. The method according to claim 6, wherein: The compressing the channel information input by the first model to reduce the number of output bits of the first model to obtain the compressed first CSI includes: Using part of the channel information as input of the first model to obtain the compressed first CSI; the part of the channel information includes one or more of the following: Channel information of some sub-bands; Broadband channel information; Partial beam information in the beam information; Part of the delay information.

14. The method according to any one of claims 1 to 13, wherein: Also includes: The terminal device sends the compressed first CSI through the uplink channel, and the code rate of the uplink channel is less than or equal to the maximum transmission code rate configured by the network device.

15. The method according to any one of claims 1 to 14, wherein: Also includes: The terminal device sends first information, where the first information indicates one or more of the following: model information of the second model; Target number of output bits; Partial channel information; The number of output bits after part of the channel information is processed by the first model.

16. The method according to claim 15, wherein: The compressed first CSI includes a CSI part 1 and a CSI part 2, the number of bits of the CSI part 1 is fixed, and the number of bits of the CSI part 2 is determined according to the content indicated by the CSI part 1; The first information is carried by the CSI part 1.

17. The method according to any one of claims 1 to 13, wherein: Also includes: When one or more of the following conditions are met, the terminal device discards the first CSI; The output bit number of the second model is less than a first threshold; The target output bit number is less than the first threshold; Any candidate model among the multiple candidate models supported by the terminal device makes the code rate of the uplink channel carrying the compressed first CSI greater than the maximum transmission code rate configured by the network device; Any output bit number among the multiple output bit numbers supported by the first model makes the code rate of the uplink channel carrying the compressed first CSI greater than the maximum transmission code rate configured by the network device; The terminal device compresses the channel information input by the first model so that the code rate of the uplink channel carrying the compressed first CSI is greater than the maximum transmission code rate configured by the network device.

18. The method according to claim 17, wherein: The first threshold is predefined or configured by a network device.

19. The method according to any one of claims 1 to 18, wherein: The first CSI includes CSI obtained by CSI compression based on AI / ML and / or CSI obtained by CSI prediction based on AI / ML.

20. A method for compressing information, the method comprising: The network device determines, based on a priority of the first channel state information CSI in the CSI carried by the uplink channel, that the terminal device needs to compress the first CSI, wherein the first CSI is an AI / ML-based CSI, and the uplink channel is used to carry the first CSI; The network device receives a compressed first CSI; the compressed first CSI is obtained by the terminal device compressing the first CSI by adjusting a first model.

21. The method according to claim 20, wherein: The priority of the first CSI in the CSI carried by the uplink channel is determined based on one or more of the following: The priority of CSI based on AI / ML is higher than that of CSI obtained by non-AI / ML methods; The priority of the AI / ML-based CSI is equal to the priority of the CSI obtained based on the non-AI / ML method, and the priority of the AI / ML-based CSI and the priority of the CSI obtained based on the non-AI / ML method are determined according to the corresponding serving cell identification information and / or CSI reporting configuration identification information; The priority of CSI based on AI / ML is lower than that of CSI obtained based on non-AI / ML methods; The priority of CSI based on AI / ML is lower than the priority of RSRP obtained based on non-AI / ML methods; The priority of AI / ML-based CSI is lower than the priority of AI / ML-based RSRP; The priority of AI / ML-based CSI is lower than the priority of AI / ML-based beam information; AI / ML-based CSI has a lower priority than feedback information used for model monitoring; Model information of the first model.

22. The method according to claim 21, wherein: The model information includes one or more of the following: The function of the first model, the type of the first model, and identification information of the first model.

23. The method according to claim 21 or 22, wherein: The CSI obtained based on non-AI / ML methods includes: CSI obtained based on a codebook, and / or CSI obtained based on a non-codebook approach.

24. The method according to any one of claims 20 to 23, wherein: The network device determines, based on a priority of the first channel state information CSI in the CSI carried by the uplink channel, that the terminal device needs to compress the first CSI, including: When the code rate of the uplink channel is higher than the maximum transmission code rate configured by the network device and the first CSI is the CSI with the lowest priority among the CSIs carried by the uplink channel, the network device determines that the terminal device compresses the first CSI.

25. The method according to any one of claims 20 to 24, wherein: Also includes one or more of the following: The compressed first CSI is obtained by the terminal device switching the first model to a second model and processing the channel information through the second model; the output bit number of the second model is less than the output bit number of the first model; The compressed first CSI is obtained by the terminal device adjusting the initial output bit number of the first model to the target output bit number, and processing the channel information through the first model after adjusting the output bit number; The target output bit number is less than the initial output bit number; The compressed first CSI is obtained by the terminal device by compressing the channel information input by the first model and reducing the number of output bits of the first model.

26. The method according to claim 25, wherein: The terminal device supports multiple candidate models, and the multiple candidate models include at least the first model and the second model; The second model is a candidate model among the multiple candidate models, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

27. The method according to claim 26, wherein: The second model is a candidate model among the multiple candidate models, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device and has the largest output bit number.

28. The method according to claim 26 or 27, wherein: The multiple candidate models are predefined or configured by the network device.

29. The method according to claim 25, wherein: The first model supports multiple output bit numbers, and the multiple output bit numbers at least include the initial output bit number and the target output bit number; The target output bit number is an output bit number among the multiple output bit numbers that makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device.

30. The method of claim 29, wherein: The target output bit number is the maximum output bit number among the multiple output bit numbers, which makes the code rate of the uplink channel carrying the compressed first CSI less than or equal to the maximum transmission code rate configured by the network device, and the output bit number.

31. The method according to claim 29 or 30, wherein: The multiple output bit numbers supported by the first model are predefined or configured by the network device.

32. The method of claim 25, wherein: The compressing the channel information input by the first model to reduce the number of output bits of the first model to obtain the compressed first CSI includes: Inputting part of the channel information into the first model to obtain the compressed first CSI; the part of the channel information includes one or more of the following: Channel information of some sub-bands; Broadband channel information; Partial beam information in the beam information; Part of the delay information.

33. The method according to any one of claims 20 to 32, wherein: The compressed first CSI is sent through the uplink channel, and the code rate of the uplink channel is less than or equal to the maximum transmission code rate configured by the network device.

34. The method according to any one of claims 20 to 33, wherein: Also includes: The network device receives first information, where the first information indicates one or more of the following: model information of the second model; Target number of output bits; Partial channel information; The number of output bits after part of the channel information is processed by the first model.

35. The method of claim 34, wherein: The compressed first CSI includes a CSI part 1 and a CSI part 2, the number of bits of the CSI part 1 is fixed, and the number of bits of the CSI part 2 is determined according to the content indicated by the CSI part 1; The first information is carried by the CSI part 1.

36. The method according to any one of claims 20 to 32, wherein: Also includes: When one or more of the following conditions are met, the network device determines that the terminal device discards the first CSI; The output bit number of the second model is less than the first threshold; The target output bit number is less than the first threshold; Any candidate model among multiple candidate models supported by the terminal device, such that the code rate of the uplink channel carrying the compressed first CSI is greater than the maximum transmission code rate configured by the network device; The first model supports any output bit number among a plurality of output bit numbers, so that a code rate of an uplink channel carrying the compressed first CSI is greater than a maximum transmission code rate configured by the network device; The terminal device compresses the channel information input by the first model so that the code rate of the uplink channel carrying the compressed first CSI is greater than the maximum transmission code rate configured by the network device.

37. The method of claim 36, wherein: The first threshold is predefined or configured by the network device.

38. The method according to any one of claims 20 to 37, wherein: The first CSI includes CSI obtained by CSI compression based on AI / ML and / or CSI obtained by CSI prediction based on AI / ML.

39. An information compression device, applied to a terminal device, comprising: A first determining unit is configured to determine that the first channel state information CSI needs to be compressed according to a priority of the first CSI in the CSI carried by the uplink channel, wherein the first CSI is an AI / ML-based CSI, and the uplink channel is used to carry the first CSI; A compression unit is configured to compress the first CSI by adjusting a first model, wherein the first model is used to generate the first CSI.

40. An information compression device, applied to a network device, comprising: A second determining unit is configured to determine that the terminal device needs to compress the first channel state information CSI based on a priority of the first channel state information CSI in the CSI carried by the uplink channel, wherein the first CSI is an AI / ML-based CSI, and the uplink channel is used to carry the first CSI; The receiving unit is configured to receive a compressed first CSI; the compressed first CSI is obtained by the terminal device compressing the first CSI by adjusting the first model.

41. A terminal device, comprising: A memory for storing computer executable instructions; A processor, connected to the memory, configured to implement the method of any one of claims 1 to 19 by executing the computer executable instructions.

42. A network device comprising: A memory for storing computer executable instructions; A processor, connected to the memory, configured to implement the method of any one of claims 20 to 38 by executing the computer executable instructions.

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

44. A computer-readable storage medium storing a computer program, wherein the computer program, when executed by at least one processor, implements the method according to any one of claims 1 to 19, or implements the method according to any one of claims 20 to 38.

45. A computer program product, comprising a computer storage medium storing a computer program, wherein the computer program comprises instructions executable by at least one processor, and when the instructions are executed by the at least one processor, the method of any one of claims 1 to 19 is implemented, or the method of any one of claims 20 to 38 is implemented.

46. ​​A computer program, the computer program causing a computer to execute the method according to any one of claims 1 to 19, or to implement the method according to any one of claims 20 to 38.

Images