Communication method and device, terminal equipment and network equipment

Abstract

The invention discloses a communication method and device, terminal equipment and network equipment, and relates to the technical field of communication. When the AI technology is applied to the CSI framework, since the first AI model carries out CSI compression on the input information in sequence, the grouping of the input information needs to be determined. Thus, the first AI model can be used for performing CSI compression on first input information in a group, and performing CSI compression on non-first input information in the group by using historical input information in the group, thereby realizing enhancement of CSI-related processing. Or, the input information and / or the processing information of the first AI model which originally needs to be reported are / is not reported due to some factors, so that the discarding condition of the input information and / or the processing information occurs, and in the face of the discarding condition, CSI reporting of the input information and / or the processing information which originally needs to be reported but is not reported needs to be determined, so that the CSI reporting efficiency is improved. Therefore, enhancement of CSI related processing is realized.

Description

Technical Field

[0001] This application relates to the field of communication technologies, and in particular, to a communication method and apparatus, a terminal device, and a network device. Background Art

[0002] Artificial intelligence (AI) technology can solve some problems that are difficult to solve by traditional modeling methods, such as some non-linear problems, problems with overly complex parameters, etc. By training with a large amount of data, some problem-solving models are established and relatively accurate prediction results are provided. Among them, AI can include machine learning (ML), deep learning (DL), etc., and AI models can include various linear and non-linear network models, such as linear regression, support vector machine, convolutional neural network (CNN), deep neural network (DNN), etc.

[0003] With the continuous development of AI technology and the continuous evolution of wireless communication systems, currently the 3rd Generation Partnership Project (3GPP) is discussing the combination of AI technology and wireless communication systems. One of the discussion directions is to apply AI technology to the channel state information (CSI) framework / field in order to improve air interface performance and reduce air interface overhead, etc. However, when AI technology is applied to the CSI framework, how to enhance CSI-related processing still needs further research. Summary of the Invention

[0004] This application provides a communication method and apparatus, a terminal device, and a network device, in order to expect to enhance CSI-related processing when AI technology is applied to the CSI framework.

[0005] In a first aspect, a communication method of this application includes:

[0006] Determine a grouping of first input information of a first AI model, or determine a channel state information (CSI) report of the first input information and / or first processing information;

[0007] The first AI model is used for CSI compression, and the first processing information is information obtained by inputting the first input information into the first AI model or information obtained by further processing the information obtained by inputting the first input information into the first AI model.

[0008] It can be seen that when AI technology is applied to the CSI framework, in the face of the enhancement of CSI compression and CSI recovery, since the first AI model compresses the input information in sequence, the historical input information can participate in the CSI compression of the current input information. The processed information obtained first can be used as historical processed information to participate in the CSI compression of the current input information, and / or the compressed intermediate information obtained first can be used as historical compressed intermediate information to participate in the CSI compression of the current input information. Therefore, this application needs to determine the grouping of the input information. In this way, the first AI model can be used to perform CSI compression on the first input information in a group, and use the historical input information in the group to perform CSI compression on the non-first input information in the group, so as to achieve the enhancement of CSI-related processing.

[0009] Alternatively, when the terminal device needs to report / feedback the input information and / or the processed information, since some of the input information and / or the processed information that originally needed to be reported / feedback are not reported due to some factors, there will be a situation of discarding the input information and / or the processed information. In this way, in the face of the input information and / or the processed information that originally needed to be reported but were not reported, this embodiment can determine the CSI report of the input information and / or the processed information that originally needed to be reported but were not reported, so as to achieve the enhancement of CSI-related processing.

[0010] In a second aspect, a communication method of this application includes:

[0011] Sending configuration information, where the configuration information is used to configure a time window, and the time window is used to determine the grouping of the first input information of the first AI model; or,

[0012] Sending indication information, where the indication information is used to indicate the grouping of the first input information of the first AI model;

[0013] The first AI model is used for channel state information CSI compression.

[0014] It can be seen that when AI technology is applied to the CSI framework, in the face of the enhancement of CSI compression and CSI recovery, since the first AI model compresses the input information in sequence, and the historical input information can participate in the CSI compression of the current input information. The processed information obtained first can be used as historical processed information to participate in the CSI compression of the current input information, and / or the compressed intermediate information obtained first can be used as historical compressed intermediate information to participate in the CSI compression of the current input information. Therefore, this embodiment needs to determine the grouping of the input information. In this way, the first AI model can be used to perform CSI compression on the first input information in a group, and use the historical input information in the group to perform CSI compression on the non-first input information in the group, so as to achieve the enhancement of CSI-related processing.

[0015] In a third aspect, a communication method according to the present application includes:

[0016] If the first CSI report in which the first input information and / or the first processing information of the first AI model are located fails to be reported, then receive a second CSI report, where the second CSI report includes the first input information and / or the first processing information, or the second CSI report includes the first input information and / or the first processing information, as well as second input information and / or second processing information, and the first input information and the second input information are different input information; or,

[0017] If the first CSI report in which the first input information and / or the first processing information of the first AI model are located fails to be reported, then receive the first input information and / or the first processing information at an available time domain position after the time domain position of the first CSI report; or,

[0018] If the first CSI report in which the first input information and / or the first processing information of the first AI model are located fails to be reported, then receive a third CSI report, where the third CSI report includes third input information and / or third processing information, and the first input information and the third input information are different input information;

[0019] The first AI model is used for CSI compression, the first processing information is the information obtained by inputting the first input information into the first AI model or the information obtained by further processing the information obtained by inputting the first input information into the first AI model, the second processing information is the information obtained by inputting the second input information into the first AI model or the information obtained by further processing the information obtained by inputting the second input information into the first AI model, and the third processing information is the information obtained by inputting the third input information into the first AI model or the information obtained by further processing the information obtained by inputting the third input information into the first AI model.

[0020] It can be seen that when it was originally hoped to report the first input information and / or the first processing information through the first CSI report, since the first CSI report fails to be reported, this results in the failure to report the first input information and / or the first processing information that originally needed to be reported. Therefore, the second CSI report is used to carry the first input information and / or the first processing information, so as to continue reporting the first input information and / or the first processing information that originally needed to be reported but was not reported.

[0021] Alternatively, when the first input information and / or the first processing information that originally needed to be reported fails to be reported, the second CSI report is used to carry the first input information and / or the first processing information, as well as the second input information and / or the second processing information, so that the second CSI report can be used to defer the reporting of the first input information and / or the first processing information and report the second input information and / or the second processing information required to be reported by the second CSI report, avoiding the loss of the second input information and / or the second processing information.

[0022] Alternatively, when the first input information and / or the first processing information that originally needed to be reported fails to be reported, the first input information and / or the first processing information is received at an available time domain position after the time domain position of the first CSI report, so as to continue to report / defer the reporting of the first input information and / or the first processing information.

[0023] Alternatively, when the first input information and / or the first processing information that originally needed to be reported fails to be reported, the first input information and / or the first processing information is no longer reported, but the third input information and / or the third processing information is reported through the third CSI report, realizing the reporting of new input information and / or new processing information and avoiding the loss of the new input information and / or the new processing information.

[0024] In a fourth aspect, a communication device according to the present application includes:

[0025] A determination unit, configured to group the first input information of the first AI model, or determine the channel state information (CSI) report of the first input information and / or the first processing information;

[0026] The first AI model is used for CSI compression, and the first processing information is the information obtained by inputting the first input information into the first AI model or the information obtained by further processing the information obtained by inputting the first input information into the first AI model.

[0027] In a fifth aspect, a communication device according to the present application includes:

[0028] A sending unit, configured to send configuration information for configuring a time window for determining the grouping of the first input information of the first AI model; or, configured to send indication information for indicating the grouping of the first input information of the first AI model;

[0029] The first AI model is used for channel state information (CSI) compression.

[0030] In a sixth aspect, a communication device according to the present application includes a receiving unit, and the receiving unit is configured to:

[0031] If the first CSI report in which the first input information and / or the first processing information of the first AI model is located fails to be reported, then receive a second CSI report, where the second CSI report includes the first input information and / or the first processing information, or the second CSI report includes the first input information and / or the first processing information, as well as the second input information and / or the second processing information, and the first input information and the second input information are different input information; or,

[0032] If the first CSI report in which the first input information and / or the first processing information of the first AI model is located fails to be reported, then receive the first input information and / or the first processing information at an available time domain position after the time domain position of the first CSI report; or,

[0033] If the first CSI report in which the first input information and / or the first processing information of the first AI model is located fails to be reported, then receive a third CSI report, where the third CSI report includes the third input information and / or the third processing information, and the first input information and the third input information are different input information;

[0034] The first AI model is used for CSI compression. The first processing information is the information obtained by inputting the first input information into the first AI model or the information obtained by further processing the information obtained by inputting the first input information into the first AI model. The second processing information is the information obtained by inputting the second input information into the first AI model or the information obtained by further processing the information obtained by inputting the second input information into the first AI model. The third processing information is the information obtained by inputting the third input information into the first AI model or the information obtained by further processing the information obtained by inputting the third input information into the first AI model.

[0035] In a seventh aspect, the steps in the method designed in the above first aspect are applied to a terminal device or within a terminal device.

[0036] In an eighth aspect, the steps in the method designed in the above second aspect are applied to a network device or within a network device.

[0037] In a ninth aspect, a terminal device of the present application includes a processor, a memory, and a computer program or instruction stored on the memory. Among them, the processor executes the computer program or instruction to implement the steps in the method designed in the above first aspect.

[0038] Tenth aspect: A network device according to the present application includes a processor, a memory, and a computer program or instruction stored on the memory. Wherein, the processor executes the computer program or instruction to implement the steps in the method designed in the above-mentioned second aspect or third aspect.

[0039] Eleventh aspect: A chip according to the present application includes a processor. Wherein, the processor executes the steps in the method designed in the above-mentioned first aspect, second aspect, or third aspect.

[0040] Optionally, the chip further includes a communication interface, and the processor executes the sending step and / or receiving step in the method designed in the above-mentioned first aspect, second aspect, or third aspect through the communication interface.

[0041] Twelfth aspect: A chip module according to the present application includes a chip, and the chip includes a processor. Wherein, the processor executes the steps in the method designed in the above-mentioned first aspect, second aspect, or third aspect.

[0042] Optionally, the chip module further includes a transceiver component, and the processor executes the sending step and / or receiving step in the method designed in the above-mentioned first aspect, second aspect, or third aspect through the transceiver component.

[0043] Thirteenth aspect: A computer-readable storage medium according to the present application, wherein it stores a computer program or instruction, and when the computer program or instruction is executed, the steps in the method designed in the above-mentioned first aspect, second aspect, or third aspect are implemented. For example, the computer program or instruction is executed by a processor.

[0044] Fourteenth aspect: A computer program product according to the present application includes a computer program or instruction. Wherein, when the computer program or instruction is executed, the steps in the method designed in the above-mentioned first aspect, second aspect, or third aspect are executed. For example, the computer program or instruction is executed by a processor.

[0045] Fifteenth aspect: A communication system according to the present application includes a communication device (for example, a terminal device or a chip) for executing any one of the methods provided in the first aspect, and / or includes a communication device (for example, a network device or a chip) for executing any one of the methods provided in the second aspect or the third aspect.

[0046] For the beneficial effects brought by the technical solutions of the second aspect to the fifteenth aspect, reference can be made to the technical effects brought by the technical solution of the first aspect, which will not be elaborated here. Description of the Drawings

[0047] Figure 1 It is a schematic diagram of the architecture of a communication system according to an embodiment of the present application;

[0048] Figure 2 It is a schematic structural diagram of a measurement resource configuration according to an embodiment of the present application;

[0049] Figure 3 It is a schematic structural diagram of a bilateral AI model according to an embodiment of the present application;

[0050] Figure 4 It is a schematic structural diagram of an enhanced bilateral AI model according to an embodiment of the present application;

[0051] Figure 5 It is a schematic structural diagram of another enhanced bilateral AI model according to an embodiment of the present application;

[0052] Figures 6 to 11 They are respectively schematic flowcharts of a monitoring method according to an embodiment of the present application;

[0053] Figure 12 It is a block diagram of the functional units of a communication device according to an embodiment of the present application;

[0054] Figure 13 It is a block diagram of the functional units of another communication device according to an embodiment of the present application;

[0055] Figure 14 It is a block diagram of the functional units of another communication device according to an embodiment of the present application;

[0056] Figure 15 It is a schematic structural diagram of a terminal device according to an embodiment of the present application;

[0057] Figure 16 It is a schematic structural diagram of a network device according to an embodiment of the present application;

[0058] Figure 17 It is a schematic structural diagram of another network device according to an embodiment of the present application. Detailed implementation manners

[0059] It should be understood that the terms "first", "second", etc. involved in the embodiments of the present application are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, software, product or device that includes a series of steps or units is not limited to the listed steps or units, but may also include unlisted steps or units, or may also include other steps or units inherent to these processes, methods, products or devices.

[0060] In the embodiments of the present application, the term "embodiment" means that the specific features, structures or characteristics described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase at various positions in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein may be combined with other embodiments.

[0061] The "and / or" in the embodiments of the present application describes the association relationship of associated objects and indicates that three relationships may exist. For example, A and / or B may represent the following three situations: A exists alone; A and B exist simultaneously; B exists alone. Among them, A and B may be singular or plural.

[0062] In the embodiments of the present application, the symbol " / " may indicate that the associated objects before and after are in an "or" relationship.

[0063] The "at least one (piece)" or its similar expression in the embodiments of the present application refers to any combination of these items, including any combination of single item (piece) or plural items (pieces), which means one or more, and multiple means two or more. For example, at least one (piece) of a, b or c may represent the following seven situations: "a", "b", "c", "a and b", "a and c", "b and c", "a, b and c". Among them, each of a, b, and c may be an element or a set containing one or more elements.

[0064] In the embodiments of the present application, "higher than" can be expressed as the same concept as "greater than", "lower than" can be expressed as the same concept as "less than", "not lower than" can be expressed as the same concept as "higher than or equal to" or "greater than or equal to", and "not higher than" can be expressed as the same concept as "lower than or equal to" or "less than or equal to". In the present application, for the same solution, "equal to" can be used in combination with "less than" or in combination with "greater than", but not simultaneously with "less than" and "greater than". When "equal to" is used in combination with "less than", it applies to the technical solution adopted by "less than". When "equal to" is used in combination with "greater than", it applies to the technical solution adopted by "greater than".

[0065] In the embodiments of the present application, "of", "corresponding / relevant", "corresponding", "indicated", "associated", etc. can be used interchangeably with each other.

[0066] In the embodiments of the present application, terms such as "associated", "corresponding", "correspond to", "is", "of", "is", "belongs to", "as", "regarded as" may sometimes be used interchangeably with each other.

[0067] In the embodiments of the present application, the "monitoring metric / mechanism" may be used interchangeably with the "monitoring method / approach / scheme / solution", etc.

[0068] In the embodiments of the present application, "connection" refers to various connection methods such as direct connection or indirect connection to achieve communication between devices, and no specific limitation is imposed thereon.

[0069] In the embodiments of the present application, "network" and "system" may be expressed as the same concept, and a communication system is a communication network.

[0070] The following describes the relevant content, concepts, meanings, technical problems, technical solutions, and beneficial effects involved in the embodiments of the present application.

[0071] I. Communication System, Terminal Device, and Network Device

[0072]

Communication System

[0073] The technical solutions of the embodiments of the present application can be applied to various communication systems, such as: Long Term Evolution (LTE) system, Advanced Long Term Evolution (LTE-A) system, New Radio (NR) system, evolved system of NR system, LTE-based Access to Unlicensed Spectrum (LTE-U) system, NR-based Access to Unlicensed Spectrum (NR-U) system, Non-Terrestrial Networks (NTN) system, Universal Mobile Telecommunication System (UMTS), 6th-Generation (6G) communication system, or other communication systems, etc.

[0074] It should be noted that the number of connections supported by traditional communication systems is limited. With the development of communication technologies, the communication system of the present application can be not only a traditional communication system, but also, for example, device-to-device (D2D) communication, machine-to-machine (M2M) communication, machine type communication (MTC), vehicle-to-vehicle (V2V) communication, vehicle-to-everything (V2X) communication, narrow band internet of things (NB-IoT) communication, etc.

[0075] Optionally, the technical solution of the embodiments of the present application can be applied to scenarios such as beamforming, carrier aggregation (CA), dual connectivity (DC), or standalone (SA) deployment scenarios.

[0076] Optionally, the spectrum used for communication between the terminal device and the network device can be an authorized frequency band or an unauthorized frequency band, and there is no limitation in this regard. Additionally, the unauthorized frequency band can be understood as a shared frequency band, and the authorized spectrum can be understood as a non-shared frequency band.

[0077] Since the embodiments of the present application describe each embodiment in combination with the terminal device and the network device, the terminal device and the network device involved will be specifically described below.

[0078]

Terminal Device

[0079] The terminal device can be a device with transceiver functions, and can also be referred to as a terminal, user equipment (UE), remote UE, relay UE, access terminal device, user unit, user station, mobile station, mobile device, remote station, mobile equipment, user terminal device, intelligent terminal device, wireless communication device, user agent, or user device. It should be noted that a relay UE is a terminal device that can provide relay forwarding services for other terminal devices (including remote UEs).

[0080] Optionally, the terminal device is a mobile phone, a tablet computer (Pad), a computer with wireless transceiver function, a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal device in industrial control, a wireless terminal device in unmanned autonomous driving, a wireless terminal device in remote medical treatment, a wireless terminal device in smart grid, a wireless terminal device in transportation safety, a wireless terminal device in smart city, or a wireless terminal device in smart home, etc.

[0081] Optionally, the terminal device may be a cellular phone, a cordless phone, a session initiation protocol (SIP) phone, a wireless local loop (WLL) station, a personal digital assistant (PDA), a handheld device with wireless communication function, a computing device or other processing devices connected to a wireless modem, a vehicle-mounted device, a wearable device, a terminal device in a next-generation communication system (such as an NR communication system, a 6G communication system), or a terminal device in a future evolved public land mobile network (PLMN), etc., and no specific limitation is made thereto.

[0082] Optionally, the terminal device is deployed on land, including indoor or outdoor, handheld, wearable or vehicle-mounted; it can be deployed on water (such as a ship, etc.); it can be deployed in the air (such as an airplane, a balloon, a satellite, etc.).

[0083] Optionally, the terminal device includes a device with wireless communication function, such as a chip system, a chip, a chip module. By way of example, the chip system may include a chip and may also include other discrete devices.

[0084] Optionally, the terminal device may include a chip, a chip module, a device, a unit, etc., and no specific limitation is made thereto.

[0085] Optionally, the terminal device includes an intelligent module. Among them, the intelligent module may be a software unit and / or a hardware unit that uses an AI model to perform air interface transmission processing. For example, the intelligent module may be used to perform processing such as compression / decompression on channel-related information.

[0086] Optionally, the intelligent module includes a chip, a chip module, etc.

[0087]

Network device

[0088] A network device can be a device with transceiver functions and can be used for communication with a terminal device.

[0089] Optionally, the network device can be responsible for radio resource management (RRM) on the air interface side, quality of service (QoS) management, data compression and encryption, data transceiver, etc.

[0090] Optionally, the network device can include a base station (BS) in a communication system or a device deployed in a radio access network (RAN) for providing wireless communication functions, that is, the network device can include devices in the RAN.

[0091] For example, the devices in the RAN can include an evolved node B (eNB or eNodeB) in an LTE communication system, a next generation evolved node B (ng-eNB) in an NR communication system, a next generation node B (gNB) in an NR communication system, a master node (MN) in a dual-connectivity architecture, a second node or secondary node (SN) in a dual-connectivity architecture, etc., and no specific limitation is made thereto.

[0092] Optionally, the network device includes devices in a core network (CN).

[0093] For example, the devices in the CN can include an access and mobility management function (AMF), a user plane function (UPF), a session management function (SMF), etc.

[0094] Optionally, the network device further includes an access point (AP) in a Wireless Local Area Network (WLAN), a relay station, a communication device in a future evolved PLMN network, a communication device in an NTN network, etc.

[0095] Optionally, the network device may include a device that provides wireless communication functions for the terminal device, such as a chip system, a chip, or a chip module. For example, the chip system may include a chip, or may include other discrete devices.

[0096] Optionally, the network device may include a transmission and reception point (TRP).

[0097] Optionally, the network device may communicate with an Internet Protocol (IP) network. For example, the Internet, a private IP network, or other data networks, etc.

[0098] Optionally, the network device may include an independent node to implement the functions of the above base station, or may include two or more independent nodes to implement the functions of the above base station. For example, the network device includes a centralized unit (CU) and a distributed unit (DU), such as gNB-CU and gNB-DU. Further, in some other embodiments of the present application, the network device may further include an active antenna unit (AAU). Among them, the CU implements a part of the functions of the network device, and the DU implements another part of the functions of the network device. For example, the CU is responsible for processing non-real-time protocols and services, and implementing the functions of the radio resource control (RRC) layer, the service data adaptation protocol (SDAP) layer, and the packet data convergence protocol (PDCP) layer. The DU is responsible for processing physical layer protocols and real-time services, and implementing the functions of the radio link control (RLC) layer, the medium access control (MAC) layer, and the physical (PHY) layer. In addition, the AAU may implement some physical layer processing functions, radio frequency processing, and functions related to active antennas. Since the information of the RRC layer will ultimately become the information of the PHY layer, or is transformed from the information of the PHY layer, therefore, in this network deployment, high-layer signaling (such as RRC signaling) can be considered to be generated by the CU and sent by the DU, or sent jointly by the DU and the AAU. It can be understood that the network device may include at least one of the CU, DU, and AAU. In addition, the CU may be classified as a RAN device, or may also be classified as a core network device, and no specific limitation is made thereto.

[0099] Optionally, the network device may be any site in a multi-site that performs coherent joint transmission (CJT) with the terminal device, or another site outside the multi-site, or another network device that communicates with the terminal device over the network, without specific limitation thereto. Among them, multi-site coherent joint transmission may be joint coherent transmission by multiple sites, or different data belonging to the same physical downlink shared channel (PDSCH) is sent from different sites to the terminal device, or multiple sites are virtualized into one site for transmission, or other forms of cooperative transmission, or names with the same meaning specified in other standards are also applicable to this application, that is, this application does not limit the names of these parameters. The sites in multi-site coherent joint transmission may be Remote Radio Heads (RRHs), Transmission and Reception Points (TRPs), etc., without specific limitation thereto.

[0100] Optionally, the network device may be any site in a multi-site that performs non-coherent joint transmission with the terminal device, or another site outside the multi-site, or another network device that communicates with the terminal device over the network, without specific limitation thereto. Among them, multi-site non-coherent joint transmission may be joint non-coherent transmission by multiple sites, or different data belonging to the same PDSCH is sent from different sites to the terminal device, or other forms of non-cooperative transmission, or names with the same meaning specified in other standards are also applicable to this application, that is, this application does not limit the names of these parameters. The sites in multi-site non-coherent joint transmission may be RRHs, TRPs, etc., without specific limitation thereto.

[0101] It should be noted that the TRP in this application is not limited to the coherent joint transmission or non-coherent joint transmission scenarios, and may also be applicable to other scenarios, without specific limitation thereto.

[0102] Optionally, the network device may have mobility characteristics. For example, the network device may be a mobile device. Optionally, the network device may be a satellite or a balloon station. For example, the satellite may be a low earth orbit (LEO) satellite, a medium earth orbit (MEO) satellite, a geostationary earth orbit (GEO) satellite, a high elliptical orbit (HEO) satellite, etc. Optionally, the network device may also be a base station located on land, water, etc.

[0103] Optionally, a network device may provide services to a cell, and a terminal device in the cell may communicate with the network device through transmission resources (such as spectrum resources). Among them, the cell may be a macro cell, a small cell, a metro cell, a micro cell, a pico cell, a femto cell, etc.

[0104] Optionally, the network device may be a chip, a chip module, a device, a unit, etc., without specific limitations.

[0105] Optionally, the network device may include an intelligent module. Among them, the intelligent module may be a software unit and / or a hardware unit that uses an AI model to perform air interface transmission processing. For example, the intelligent module may be used to perform processing such as compression / decompression on channel-related information.

[0106] Optionally, the intelligent module may be a chip, a chip module, etc.

[0107]

Exemplary Explanation

[0108] The following gives an exemplary explanation of the communication system according to the embodiments of the present application.

[0109] Exemplarily, for the network architecture of a communication system according to an embodiment of the present application, reference may be made to Figure 1 . As Figure 1 shown, the communication system 10 may include a network device 110 and a terminal device 120. The terminal device 120 may communicate with the network device 110 wirelessly.

[0110] Figure 1 This is only an example of the network architecture of a communication system, and does not limit the network architecture of the communication system according to the embodiments of the present application.

[0111] For example, the communication system 10 may further include a server or other devices.

[0112] For another example, in addition to the network device 110, the communication system 10 may include other network devices.

[0113] For another example, in addition to the terminal device 120, the communication system 10 may include other terminal devices.

[0114] II. Combination of AI Technology and CSI Framework

[0115] 1. CSI-related Configuration

[0116]

Measurement Resources

[0117] CSI-related configurations may include CSI resource configuration and / or CSI report configuration, and may be defined by the higher layer parameter CSI-MeasConfig (CSI measurement configuration). Among them, the higher layer parameter CSI-MeasConfig may indicate / configure / include the following two higher layer parameters: CSI-ResourceConfig (CSI resource configuration) and CSI-ReportConfig (CSI report configuration).

[0118] Since the higher layer parameter CSI-ReportConfig may indicate / include the parameter CSI-ResourceConfigId (CSI resource configuration identifier), through the parameter CSI-ResourceConfigId, the higher layer parameter CSI-ResourceConfig is associated / corresponds / maps to the higher layer parameter CSI-ReportConfig.

[0119] The higher layer parameter CSI-ReportConfig can be used for the configuration of CSI reports, that is, to configure CSI reports.

[0120] The higher layer parameter CSI-ResourceConfig can be used to configure the measurement resources for CSI measurement. In addition, the higher layer parameter CSI-ResourceConfig can configure the parameter CSI-Resource-Set (CSI resource set), and the parameter CSI-Resource-Set can include the most basic channel state information reference signal (CSI-RS) resources.

[0121] The parameter CSI-Resource-Set may indicate / include at least one of the following: the parameter non-zero power CSI-RS resource set (Non-Zero-Power-CSI-RS-ResourceSet, NZP-CSI-RS-ResourceSet), the parameter CSI interference measurement resource set (CSI-Interference-Measurement-ResourceSet, CSI-IM-ResourceSet), the parameter CSI synchronization signal block resource set (CSI-SSB-ResourceSet).

[0122] The parameter NZP-CSI-RS-ResourceSet can be used for channel measurement and / or interference measurement; the parameter CSI-IM-ResourceSet can be used for interference measurement; the parameter CSI-SSB-ResourceSet can be used for channel measurement.

[0123] The parameter CSI-ResourceConfig and / or the parameter NZP-CSI-RS-ResourceSet can configure one or more NZP-CSI-RS resource sets. Each NZP-CSI-RS resource set can include K (K≥1) NZP-CSI-RS resources.

[0124] For example, the network device sends CSI resource setting to the terminal device through RRC signaling. Each CSI resource setting includes S (S≥1 and S is an integer) CSI resource sets, and each CSI resource set includes K (K≥1 and K is an integer) NZP-CSI-RS resources. The terminal device can receive CSI-RS on the K NZP-CSI-RS resources indicated by the network device.

[0125] In summary, the measurement resources can include at least one of NZP-CSI-RS resources, CSI-IM resources, and CSI-SSB resources. NZP-CSI-RS resources can be used for channel measurement and / or interference measurement; CSI-IM resources can be used for interference measurement; CSI-SSB resources can be used for channel measurement.

[0126] In addition, the measurement resources can also include channel measurement resources (CMR) and / or interference measurement resources (IMR). Among them, the channel measurement resources refer to the measurement resources used for channel measurement, such as NZP-CSI-RS resources or CSI-SSB resources; the interference measurement resources refer to the measurement resources used for interference measurement, such as NZP-CSI-RS resources and CSI-IM resources.

[0127]

Time domain behavior

[0128] The time-domain behavior can include periodic, semi-persistent, and aperiodic. Among them, the time-domain behavior includes the time-domain behavior of measurement resources and the time-domain behavior of CSI reports. Specific descriptions are given below respectively.

[0129] The time-domain behavior of measurement resources can include periodic measurement resources, aperiodic measurement resources, and semi-persistent measurement resources. Among them, the time-domain behavior of measurement resources can be indicated by high-layer parameters (such as resource type resourceType).

[0130] For periodic measurement resources, after the network device configures the periodic measurement resources, the network device will send the measurement resources periodically.

[0131] For aperiodic measurement resources, after the network device configures the aperiodic measurement resources, the network device will not send the measurement resources immediately, but trigger them through trigger signaling and then send the aperiodic measurement resources to the terminal device after the trigger.

[0132] For semi-persistent measurement resources, after the network device configures the semi-persistent measurement resources, the network device will not send CSI-RS immediately, but first send activation signaling and then send the semi-persistent measurement resources to the terminal device after activation. Among them, once the network device triggers the sending of semi-persistent measurement resources, it will send the measurement resources periodically until the network device issues deactivation signaling.

[0133] The time-domain behavior of CSI reports can include periodic CSI reports, aperiodic CSI reports, semi-persistent CSI reports carried on the physical uplink control channel (PUCCH), and semi-persistent CSI reports carried on the physical uplink shared channel (PUSCH). Among them, the time-domain behavior of CSI reports can be indicated by high-layer parameters (such as report configuration type (reportConfigType)).

[0134] For aperiodic CSI reports and semi-persistent CSI reports carried on PUSCH, the network device will also configure high-layer parameters TriggerState and high-layer parameter reportTriggerSize for use in conjunction with the CSI request field in the downlink control information (DCI).

[0135] For periodic CSI reporting, after configuring the periodic CSI-RS Resource and Report parameters via RRC signaling, it will take effect immediately without the need to activate / trigger the measurement resource transmission and CSI reporting via MAC-CE / DCI.

[0136] For semi-persistent CSI reporting carried on PUCCH, if the transmission of semi-persistent measurement resources is configured via RRC signaling, it is necessary to first activate the measurement resource transmission via MAC CE1 and then activate the CSI reporting via MAC CE2; if the transmission of periodic measurement resources is configured via RRC signaling, there is no need to activate the measurement resource transmission via MAC CE1, and only the CSI reporting needs to be activated via MAC CE2.

[0137] For semi-persistent CSI reporting carried on PUSCH, if the transmission of semi-persistent measurement resources is configured via RRC signaling, it is necessary to first activate the measurement resource transmission via MAC CE1 and then trigger the CSI reporting via DCI; if the transmission of periodic measurement resources is configured via RRC signaling, there is no need to activate the measurement resource transmission via MAC CE1, and only the CSI reporting needs to be triggered via DCI.

[0138] It should be noted that for DCI, this DCI can be DCI format 0_1 scrambled with SP-CSI-RNTI (semi-persistent CSI-RNTI), and the CSI request field in this DCI can be associated with the corresponding trigger state (TriggerState) through the setting of the codepoint. The TriggerState will define the associated CSI-ReportConfig, so that the CSI-ReportConfig associated with the semi-persistent CSI reporting carried on PUSCH can be found through this TriggerState.

[0139] For aperiodic CSI reporting, the aperiodic CSI reporting is triggered by DCI, and the process is similar to the above semi-persistent CSI reporting. When associating the corresponding TriggerState through the codepoint of the CSI request field in DCI format 0_1 / 0_2, different from the DCI trigger in the above semi-persistent CSI reporting, if the value of this CSI request field is 0, it means that triggering of semi-periodic CSI reporting is not required; if the value of the CSI request field is 1, it means that the aperiodic CSI reporting associated with TriggerState 1 is triggered, and so on.

[0140] It should be noted that the CSI report configuration is related to the measurement resource configuration, as shown in Table 1. In Table 1, for periodic CSI reports, they can only be associated with / use periodic resources; for semi-persistent CSI reports carried on PUCCH, they can be associated with / use periodic measurement resources and semi-persistent measurement resources, and are activated / deactivated through MAC CE; for semi-persistent CSI reports carried on PUSCH, they can be associated with / use periodic measurement resources and semi-persistent measurement resources, and are triggered by DCI; for aperiodic CSI reports, they can be associated with / use periodic measurement resources, semi-persistent measurement resources, and aperiodic measurement resources, and are triggered by DCI.

[0141] Table 1

[0142]

[0143]

CSI Measurement and CSI Reporting

[0144] During the CSI measurement, CSI acquisition, and CSI reporting processes, the terminal device can perform channel measurement / interference measurement on the downlink channel according to the measurement resources to obtain channel information / CSI information, and then report the channel information / CSI information to the network device.

[0145] Optionally, the CSI information includes at least one of the following: CSI reference signal resource indicator index (CSI-RS Resource indicator, CRI), synchronization signal block resource indicator index (SSB resource indicator, SSBRI), rank indicator index (rank indicator, RI), precoding matrix indicator index (precoding matrix indicator, PMI), channel quality indicator index (channel quality indicator, CQI), layer indicator index (layer indicator, LI), capability index (Capability Index), time-domain channel properties (time-domain channel properties, TDCP), etc.

[0146] It should be noted that the CRI can represent the CSI-RS resources recommended (or selected) by the terminal device, and the SSBRI can represent the SSB resources recommended (or selected) by the terminal device. Among them, one CSI-RS resource can represent a beam or antenna direction, and one SSB resource can represent a beam or antenna direction.

[0147] The RI may represent the number of layers recommended (or selected) by the terminal device, and the number of layers may determine which codebook. Among them, each number of layers corresponds to a codebook, and a codebook consists of one or more codewords. For example, the codebook with the number of layers being 2 or the codebook with the number of layers being 1.

[0148] The PMI may represent the index of the codeword in the codebook recommended (or selected) by the terminal device, or the quantized precoding information. Among them, one codeword corresponds to one precoding matrix. The RI and PMI may represent the number of layers and the precoding matrix recommended by the terminal device as a whole.

[0149] The CQI may represent the quality of the current channel feedback by the terminal device to the network device. Among them, the terminal device needs to calculate the CQI.

[0150] Optionally, the channel information may include the relevant information obtained by channel measurement according to reference signals (such as CSI-RS, DMRS, SSB, etc.).

[0151] For example, the channel information includes at least one of the following: channel matrix H, equivalent channel matrix, precoding matrix W (the precoding matrix W can be derived from the channel matrix H), right singular vector V of the channel matrix H, square matrix H T eigenvector of H (i.e., the eigenvector of the matrix obtained by multiplying the conjugate transpose H T of the channel matrix H by the channel matrix H), at least one vector associated with the channel matrix H (such as at least one vector under a certain deformation of the channel matrix H, etc.), time domain information of the channel, channel information in the delay-doppler domain, CRI, SSBRI, RI, PMI, CQI, LI, CSI, etc.

[0152]

CSI Reference Resource

[0153] It should be noted that the CSI report corresponds to a CSI reference resource. For example, for the periodic CSI report, the CSI is reported periodically, and each CSI report corresponds to a CSI reference resource.

[0154] The time domain of the CSI reference resource may be a time slot, and the frequency domain of the CSI reference resource may be the frequency resource for obtaining CSI measurement.

[0155] Optionally, when the uplink time slot where the CSI report is located is time slot n′ (n′ represents the time slot index), the downlink time slot where the CSI reference resource is located is time slot K offset represents a parameter configured by the higher layer, represents in frequency range 1 and Koffset Sub - carrier spacing configuration at μ = 0 DL Indicates the downlink sub - carrier spacing configuration.

[0156] Since the CSI reference resource is a downlink resource, the time domain of the CSI reference resource is a downlink time slot. Thus, time slot n represents the downlink time slot corresponding to time slot n′ and is calculated as follows:

[0157]

[0158] where μ UL Indicates the uplink sub - carrier spacing configuration, Downlink time slot offset, μ offset,DL Represents the maximum value among the minimum values of the downlink sub - carrier spacing configuration of the primary cell (Pcell) / primary secondary cell (PScell) and the minimum value of the downlink sub - carrier spacing configuration of the secondary cell (Scell). Uplink time slot offset, μ offset,UL Represents the maximum value among the minimum values of the uplink sub - carrier spacing configuration of the Pcell / PScell and the minimum value of the uplink sub - carrier spacing configuration of the Scell.

[0159] n CSI_ref Represents the time slot offset between time slot n and the valid downlink time slot, and the CSI reference resource is located in this valid downlink time slot. Among them, whether the downlink time slot is a "valid downlink time slot" needs to meet the following conditions: this downlink time slot includes at least one downlink symbol or flexible symbol configured by higher - layer signaling; and / or, this downlink time slot is not within the configured measurement gap.

[0160] Optionally, for periodic or semi - persistent CSI reporting, if this CSI reporting is associated with a channel measurement resource (such as a CSI - RS / SSB for channel measurement), then n CSI_ref Is the minimum value greater than or equal to in which the CSI reference resource corresponding to this CSI reporting is located in the valid downlink time slot.

[0161] For example, assume that the uplink and downlink sub - carrier spacing configurations are the same, i.e., μ DL = μ UL = 0. Starting from the minimum value greater than or equal to let n CSI_ref be Calculated as And determine whether time slot n-4 is a valid downlink time slot. If time slot n-4 is a valid downlink time slot, the CSI reference resource is located in time slot n-4; if time slot n-4 is not a valid downlink time slot, then let n CSI_ref be calculated to be And determine whether time slot n-5 is a valid downlink time slot. If time slot n-5 is a valid downlink time slot, the CSI reference resource is located in time slot n-5; if time slot n-5 is not a valid downlink time slot, continue to look forward until a valid downlink time slot is found.

[0162] Optionally, for periodic or semi-persistent CSI reporting, if the CSI report is associated with multiple channel measurement resources (such as multiple CSI-RS / SSBs for channel measurement), then n CSI_ref is the minimum value in greater than or equal to such that the CSI reference resource corresponding to the CSI report is located in a valid downlink time slot.

[0163] Optionally, for aperiodic CSI reporting, if the DCI used to trigger the CSI report and the CSI report are in the same time slot, the CSI reference resource is located in the time slot where the DCI is located; otherwise, n CSI_ref is the minimum value in greater than or equal to such that the CSI reference resource corresponding to the CSI report is located in a valid downlink time slot, where \(\mu\) represents the number of symbols contained in a time slot, and \(Z'\) represents the processing time for CSI calculation.

[0164] Optionally, the CSI reference resource can have the following two functions.

[0165] Function 1: Used to calculate the channel quality indicator (CQI)

[0166] Channel measurement essentially depends on the implementation of the terminal device. For example, some terminal devices have very strong capabilities and can achieve the target block error rate (BLER) requirement by adopting a certain modulation and coding scheme (MCS) when the SINR = 10 dB. However, some terminal devices have weaker capabilities and need to be at SINR = 12 dB to achieve the target BLER (the target BLER is configured by the network device) requirement. Therefore, if the SINR measured by two terminal devices is both 10 dB, the CQI values may be different when feeding back.

[0167] In addition, even for the same terminal device, different transport block sizes (TBS) will result in different block error rates (BLER). For example, when the signal-to-interference-plus-noise ratio (SINR) = 10 dB, if the currently transmitted TBS is 1000 bits and a specific modulation and coding scheme (MCS) is selected for transmission, the BLER is 0.01. If a TBS of 500 bits is used and the same MCS is selected for transmission, the corresponding BLER may be 0.1. The two are different.

[0168] Therefore, when quantifying the channel quality indicator (CQI), the concept of CSI reference resources is needed so that the terminal device can obtain the CQI / CSI through the CSI reference resources.

[0169] Suppose the terminal device uses CSI-RS for channel measurement and obtains the channel state, such as SINR. The terminal device will assume that a downlink data packet is transmitted on the CSI reference resources, and the corresponding channel state is the one measured by CSI-RS. At this time, the terminal device selects a CQI from the CQI table, and this CQI can make the BLER of the assumed transmitted downlink data packet less than the target BLER when transmitted in the measured channel state.

[0170] Function 2: Used to determine the measurement resources associated with the CSI report

[0171] The network device can configure one or more measurement resources for CSI measurement for the terminal device. Among them, the time domain position of the measurement resources associated with the CSI report may be before the CSI reference resources corresponding to the CSI report. That is to say, the measurement resources before the CSI reference resources corresponding to the CSI report can be used as the measurement resources associated with the CSI report, and as for which measurements can be used as the measurement resources associated with the CSI report, it can be configured through high-layer signaling, protocol regulations, pre-configuration, or determined autonomously by the terminal device, etc.

[0172] For example, taking periodic measurement resources as an example, as Figure 2 shown, the periodic measurement resources include the measurement resources associated with the CSI report, and the time domain position of the measurement resources associated with the CSI report is before the CSI reference resources corresponding to the CSI report. That is to say, only the measurement resources before the CSI reference resources corresponding to the CSI report may be used as the measurement resources associated with the CSI report, and as for which one or which resources of the measurements can be used as the measurement resources associated with the CSI report, it can be further configured by the network through high-layer signaling. For example, the network can configure the resource measurement 210 that is the closest / latest / adjacent to the CSI reference resources as the measurement resources associated with the CSI report.

[0173] It should be noted that this embodiment does not impose specific restrictions on the content of the "CSI reference resource", and this part of the content does not exclude new content included in the further evolution of the standard protocol.

[0174] 2. Enhancement of CSI-related processing

[0175]

Description

[0176] AI can include ML, DL, etc., and the AI model can have certain model parameters and model architectures, etc., and the AI model can include various linear and non-linear network models, such as linear regression, support vector machine, CNN, DNN, etc.

[0177] It should be noted that the "AI model" mentioned in this embodiment can be understood as an ML model, a DL model, an intelligent model, an AI algorithm, an AI module, or an intelligent module, etc., and no specific restrictions are imposed on this.

[0178] When AI technology is applied to the CSI framework, the AI model can be used for CSI compression and CSI resume / recovery to achieve communication intelligence and reduce air interface overhead, etc.

[0179] Of course, in addition to CSI compression and CSI recovery, it can also include other processing in the communication process, such as encoding / decoding, quantization / dequantization, etc., and no specific restrictions are imposed on this.

[0180]

Enhancement of CSI compression and CSI recovery

[0181] During the CSI measurement / CSI reporting process, the terminal device can perform channel measurement / interference measurement according to the measurement resources to obtain channel information / CSI information. However, the channel information / CSI information may have characteristics such as a large number of bits, and directly feeding back the channel information / CSI information through the CSI report will cause problems such as more resources occupied and large signaling overhead. Therefore, in order to reduce the occupied resources and signaling overhead, etc., this embodiment can perform CSI compression on the channel information / CSI information, so that the compressed information has characteristics such as a smaller number of bits. Corresponding to CSI compression is CSI recovery, so as to restore the compressed information to the original channel information / CSI information through CSI recovery.

[0182] In the traditional CSI compression and CSI recovery process, the terminal device can input the channel information / CSI information into its own AI model (of course, this AI model can be regarded as an "encoder" or a CSI compression model, etc.) for compression and other processing to obtain the processed information, and finally send the processed information to the network device. Among them, the information input into the AI model of the terminal device in this embodiment can be called "input information".

[0183] In addition, the "processed information" in this embodiment can be the information obtained by inputting the input information into the AI model or the information obtained by further processing the information obtained by inputting the input information into the AI model. Among them, the information obtained by inputting the input information into the AI model can include CSI compression information; the "further processing" here can include at least one of encoding, quantization, dequantization, or decoding, etc.

[0184] Correspondingly, the network device receives the processed information and inputs the processed information into its own AI model (of course, this AI model can be called "decoder or CSI recovery model, etc.") for decoding to recover the channel information / CSI information as much as possible. Among them, the information output by the AI model of the network device in this embodiment can be called "output information".

[0185] In this way, the CSI compression and CSI recovery processes are implemented through the AI models of the terminal device and the network device.

[0186] It should be noted that since CSI compression and CSI recovery are two mutually existing processes, the AI models of the terminal device and the network device need to appear in pairs. In this way, the AI models of the terminal device and the network device can be called "AI model pair" or "two-sided AI model". Among them, for the convenience of description, the AI model used for CSI compression in the two-sided AI model in this embodiment can be called "the first AI model", and the AI model used for CSI recovery in the two-sided AI model can be called "the second AI model".

[0187] For example, taking the two-sided AI model including an encoder and a decoder as an example, as Figure 3 shown. In Figure 3 , for the terminal device side, the input information V undergoes CSI compression through the encoder 311 to obtain the processed information V E , and then the processed information V E undergoes quantization through the quantizer 312 to obtain the quantized processed information V Q . For the network device side, the quantized processed information V Q first undergoes dequantization through the dequantizer 321 and then undergoes CSI recovery through the decoder 322 to obtain the output information

[0188] Different from the traditional CSI compression and CSI recovery processes, in order to improve the accuracy of CSI compression and CSI recovery and save the air interface overhead, this embodiment needs to consider enhancing CSI compression and CSI recovery.

[0189] For the enhancement of CSI compression, in this embodiment, it is considered that the AI model of the terminal device can use the first historical information to participate in the CSI compression of the current input information. Among them, the first historical information may include at least one of historical input information, historical processing information, or historical compression intermediate information.

[0190] The historical input information can be defined according to the order of the input information, or according to the order of the first AI model processing the input information, or according to the generation time of the input information, or according to the order of the input information entering the first AI model, and so on.

[0191] For the definition according to the order of the input information, the input information before the current input information can be called "historical input information".

[0192] For the definition according to the order of the first AI model processing the input information, when the first AI model is processing the current input information, the input information processed before the current input information can be called "historical input information".

[0193] For the definition according to the generation time of the input information, the input information generated before the current input information can be called "historical input information".

[0194] For the definition according to the order of the input information entering the first AI model, the input information that enters the first AI model before the current input information can be called "historical input information".

[0195] The historical processing information can be the information that appears in the process of the bilateral AI model from inputting historical input information to outputting historical output information, or the information obtained by inputting historical input information into the first AI model, or the information obtained by further processing the information obtained by inputting historical input information into the first AI model.

[0196] The historical compression intermediate information can be the information generated by the bilateral AI model / the first AI model in the process of compressing historical input information.

[0197] It can be understood that when the first historical information is introduced into CSI compression, since the first historical information and the current input information can reflect the time-domain change or time-domain correlation of the channel, the first AI model can consider the time-domain change or time-domain correlation of the channel when performing CSI compression on the current input information, which is beneficial to improving the accuracy of CSI compression, enhancing the performance of CSI compression, and realizing the enhancement of CSI compression.

[0198] For the enhancement of CSI recovery, this embodiment considers that the AI model of the network device can use the second historical information to participate in the CSI recovery of the current output information. Among them, the second historical information may include at least one of historical output information, historical processing information, or historical recovery intermediate information.

[0199] The historical output information can be defined according to the order of the output information, or according to the order of the second AI model's output of the output information, or according to the order of the generation time of the output information, or the output information corresponding to the historical input information, etc. It should be noted that the understanding of various definitions of historical output information can refer to the relevant descriptions of the above historical input information, which will not be elaborated here.

[0200] The historical recovery intermediate information can be the information generated by the second AI model during the process of recovering the historical output information.

[0201] It can be understood that when the second historical information is introduced into the CSI recovery, since the second historical information and the current output information can reflect the time-domain change or time-domain correlation of the channel, the second AI model can consider the time-domain change or time-domain correlation of the channel when performing CSI recovery on the current processing information, which is beneficial to improving the accuracy of CSI recovery, enhancing the performance of CSI recovery, and realizing the enhancement of CSI recovery.

[0202] In the enhancement of CSI compression and CSI recovery, the AI model of the terminal device and the AI model of the network device can be referred to as the "enhanced AI model pair" or "enhanced bilateral AI model". Among them, for the convenience of description, this embodiment can also refer to the AI model used for CSI compression in the enhanced bilateral AI model as the "first AI model", and the AI model used for CSI recovery in the enhanced bilateral AI model as the "second AI model".

[0203] For example, taking the enhanced bilateral AI model including an encoder, a quantizer, a dequantizer, and a decoder as an example, as Figure 4 shown. In Figure 4 , for the terminal device side, when the input information V1 is input into the encoder 411, the encoder 411 performs CSI compression on the input information V1 to obtain the processed information V E1 , and the quantizer 412 quantizes the processed information V E1 to obtain the quantized processed information V Q1 . In addition, the input information V1 can be stored as historical input information, and the processed information V E1 and / or the quantized processed information V Q1It can be stored as historical processing information, and the information generated during the process of compressing the input information V1 can be stored as historical compression intermediate information. For the network device side, the quantized processing information V Q1 First, it is dequantized by the dequantizer 421, and then the CSI is restored by the decoder 422 to obtain the output information In addition, the output information can be stored as historical output information, and the information generated during the process of restoring the output information can be stored as historical restoration intermediate information.

[0204] In Figure 4 , for the terminal device side, when the input information V2 is input to the encoder 411, the encoder 411 can use the first historical information to perform CSI compression on the input information V2 to obtain the processing information V E2 , and the quantizer 412 quantizes the processing information V E2 to obtain the quantized processing information V Q2 . In addition, the input information V2 can be stored as historical input information, and the processing information V E2 and / or the quantized processing information V Q2 can be stored as historical processing information, and the information generated during the process of compressing the input information V2 can be stored as historical compression intermediate information. For the network device side, the quantized processing information V Q2 is first dequantized by the dequantizer 421, and then the CSI is restored by the decoder 422 to obtain the output information In addition, the output information can be stored as historical output information, and the information generated during the process of restoring the output information can be stored as historical restoration intermediate information.

[0205] In Figure 4 , for the terminal device side, when the input information V i (i = 2, 3,...) is input to the encoder 411, the encoder 411 can use the first historical information to perform CSI compression on the input information V i to obtain the processing information V Ei , and the quantizer 412 quantizes the processing information V Ei to obtain the quantized processing information V Qi . In addition, the input information V i can be stored as historical input information, and the processing information V Ei and / or the quantized processing information V QiIt can be stored as historical processing information and the information generated during the process of compressing the input information V i The information generated during the process of i can be stored as historical compression intermediate information. For the network device side, the quantized processing information V Qi First, it goes through a dequantizer 421 for dequantization, and then through a decoder 422 for CSI recovery to obtain the output information In addition, the output information It can be stored as historical output information, and the information generated during the process of restoring the output information can be stored as historical restoration intermediate information.

[0206] It should be noted that in the temporal domain, spatial domain, and frequency domain, traditional bilateral AI models can be used for CSI compression and CSI recovery in the spatial-frequency (SF) domain, and the AI models in traditional bilateral AI models can be called "spatial-frequency (SF) domain AI models". Enhanced bilateral AI models can be used for CSI compression and CSI recovery in the spatio-temporal-frequency (TSF) domain, and the AI models in enhanced bilateral AI models can be called "spatio-temporal-frequency (TSF) domain AI models". Of course, traditional bilateral AI models and enhanced bilateral AI models can also be described by other terms, and no specific restrictions are imposed on this.

[0207]

Input Information

[0208] The input information of the enhanced bilateral first AI model can refer to the information input into the enhanced bilateral first AI model.

[0209] Regarding the input information of the enhanced bilateral AI model, the input information involves concepts such as "historical input information", "first input information", "non-first input information", "current input information", etc.

[0210] Historical input information can be defined according to the chronological order of input information, or according to the chronological order of the first AI model processing input information, or according to the chronological order of input information generation, or according to the chronological order of input information input into the first AI model, etc. For example, in Figure 4 , when input information V2 is input into the encoder 411 at the current moment, since input information V1 has been input into the encoder 411 before input information V2, input information V1 can be regarded as historical input information.

[0211] The first input information can refer to the initial / first information input into the first AI model for CSI compression within a certain time range. Since it is the initial / first input information, no first historical information has been stored yet, and thus only the first input information can be directly subjected to CSI compression. For example, Figure 4 the input information V1 in

[0212] Non-first input information can refer to information that is not the initial / first input into the first AI model for CSI compression within a certain time range, or information that is input into the first AI model for CSI compression after the first input information. At this time, since the first historical information has already been stored, the first AI model can use the first historical information to perform CSI compression on the non-first input information. For example, Figure 4 the input information V2 and the input information V in i can be regarded as non-first input information. Meanwhile, when performing CSI processing on non-first input information, the first input information can be regarded as historical input information, and the processing information corresponding to the first input information can be regarded as historical processing information. Additionally, the previous non-first input information can be regarded as the historical input information of the next non-first input information, and the processing information corresponding to the previous non-first input information can be regarded as the historical processing information of the next non-first input information.

[0213] The current input information can refer to the input information at the current moment / specific moment, such as the input information input into the first AI model at the current moment / specific moment, etc.

[0214] In summary, the first AI model can be used to perform CSI compression on the first input information and use the first historical information to perform CSI compression on non-first input information.

[0215]

Output Information

[0216] The output information of the enhanced bilateral AI model can be the information output by the second AI model, and the output information corresponds one-to-one with the input information.

[0217] Regarding the output information of the enhanced bilateral AI model, the output information involves concepts such as "historical output information", "first output information", "non-first output information", etc.

[0218] Historical output information can be defined according to the order of output information, or according to the order of the second AI model's output of the output information, or according to the order of the generation time of the output information, or the output information corresponding to historical input information, etc. For example, in Figure 4In this case, when the input information V2 is input to the encoder 411 at the current moment, since the information V1 has been input to the encoder 411 before the input information V2, the input information V1 can be regarded as historical input information, and the output information can be regarded as historical output information.

[0219] The first output information can refer to the information initially / firstly output by the second AI model within a certain time range. Since it is the information initially / firstly output, the second historical information has not been stored yet. For example, Figure 4 the output information in can be regarded as the first output information.

[0220] The non-first output information can refer to the information that is not initially / firstly output by the second AI model within a certain time range, or the information output by the second AI model after the first output information, or the output information corresponding to the non-first input information. At this time, since the second historical information has been stored, the second AI model can use the second historical information to participate in the CSI recovery of the non-first output information. For example, Figure 4 the output information in and the output information can be regarded as non-first output information. At the same time, for the CSI recovery process of the non-first output information, the first output information can be regarded as historical output information. In addition, the previous non-first output information can be regarded as the historical output information of the next non-first output information.

[0221] In summary, the second AI model can be used for the CSI recovery of the first output information and participate in the CSI recovery of the non-first output information using the second historical information.

[0222]

Differentiation of Input Information

[0223] As can be seen from the content in the above "CSI-related Configuration", the CSI resource configuration is used for the configuration of measurement resources, the CSI report configuration is used for the configuration of CSI reports, and the CSI report is associated with certain measurement resources. Therefore, in enhanced CSI measurement and CSI reporting, the terminal device can perform channel measurement / interference measurement according to the measurement resources associated with the CSI report to obtain channel information / CSI information, and then input the channel information / CSI information into the enhanced bilateral AI model for CSI compression, and the information input into the AI model of the terminal device is called "input information".

[0224] The CSI report and the measurement resources associated with the CSI report are two different concepts. In this embodiment, one of the CSI report or the measurement resources can be used to define the input information, and which one to use specifically can be determined by network configuration, the terminal device autonomously, or protocol regulations, etc., and no specific restrictions are made in this regard. When using the CSI report to define the input information, the input information in this embodiment can be the channel information / CSI information corresponding to the CSI report. When using the measurement resources to define the input information, the input information in this embodiment can be the channel information / CSI information corresponding to the measurement resources.

[0225] For the input information corresponding to the CSI report, when the network configures multiple CSI reports, this embodiment can distinguish different input information according to the differences of these CSI reports. For example, when the network device triggers multiple aperiodic CSI reports through a DCI, the terminal device can perform CSI measurements according to the measurement resources associated with these aperiodic CSI reports to obtain the input information corresponding to each of these aperiodic CSI reports.

[0226] Optionally, for distinguishing different input information according to the differences of the CSI reports, this embodiment can distinguish different input information according to the different time-domain positions of the CSI reports (such as the uplink time slot where the CSI report is located), or according to the differences of the ID / index of the CSI reports.

[0227] Optionally, since the CSI report includes periodic CSI reports, aperiodic CSI reports, and semi-persistent CSI reports, the input information corresponding to the CSI report can include at least one of the input information corresponding to the periodic CSI report, the input information corresponding to the aperiodic CSI report, or the input information corresponding to the semi-persistent CSI report, and specifically can be determined according to the time-domain behavior of the CSI report configured / indicated by the network.

[0228] For example, the terminal device can perform CSI measurements according to the measurement resources associated with the periodic CSI reports to obtain the input information corresponding to the periodic CSI reports, so as to group the input information corresponding to these periodic CSI reports.

[0229] For the input information corresponding to the measurement resources, when the network configures multiple measurement resources and these multiple measurement resources are associated with at least one CSI report, this embodiment can distinguish different input information according to the differences of the measurement resources.

[0230] Optionally, for distinguishing different input information according to the differences of the measurement resources, this embodiment can distinguish different input information according to the different time-domain positions of the measurement resources, or according to the differences of the ID / index of the measurement resources.

[0231] Optionally, since the measurement resources include periodic measurement resources, aperiodic measurement resources, and semi-persistent measurement resources, the input information corresponding to the measurement resources may include at least one of the input information corresponding to the periodic measurement resources, the input information corresponding to the aperiodic measurement resources, or the input information corresponding to the semi-persistent measurement resources, and specifically, it may be determined according to the time-domain behavior of the measurement resources configured / indicated by the network.

[0232]

Input order of input information and output order of output information

[0233] It should be noted that in the enhanced CSI compression and CSI recovery process, the enhanced bilateral AI model will perform CSI compression on the input information in order. In addition, since the input information and the output information are in one-to-one correspondence, the input order of the input information and the output order of the output information are in one-to-one correspondence.

[0234] Optionally, the input information corresponding to the CSI report may be input into the enhanced bilateral AI model in the order of the time-domain position of the CSI report (such as the uplink time slot where the CSI report is located). That is, the input information corresponding to the CSI report with an earlier time-domain position will be input into the enhanced bilateral AI model earlier for CSI compression and CSI recovery, while the input information corresponding to the CSI report with a later time-domain position will be input into the enhanced bilateral AI model later for CSI compression and CSI recovery.

[0235] In this way, the input information corresponding to the first CSI report may refer to the input information corresponding to the CSI report with the earliest time-domain position, the input information corresponding to the second CSI report may refer to the input information after the input information corresponding to the first CSI report, and so on.

[0236] Optionally, the input information corresponding to the CSI report may be input into the enhanced bilateral AI model in the order of the size of the ID / index of the CSI report. For example, the input information corresponding to the CSI report with a smaller ID / index will be input into the enhanced bilateral AI model earlier for CSI compression and CSI recovery, while the input information corresponding to the CSI report with a larger ID / index will be input into the enhanced bilateral AI model later for CSI compression and CSI recovery.

[0237] In this way, the input information corresponding to the first CSI report may refer to the input information corresponding to the CSI report with a smaller ID / index, the input information corresponding to the second CSI report may refer to the input information after the input information corresponding to the first CSI report, and so on.

[0238] Optionally, the input information corresponding to the measurement resources can be input into the enhanced bilateral AI model in the order of the time-domain positions of the measurement resources. That is, the input information corresponding to the measurement resources with earlier time-domain positions will be input into the enhanced bilateral AI model earlier for CSI compression and CSI recovery, while the input information corresponding to the measurement resources with later time-domain positions will be input into the enhanced bilateral AI model later for CSI compression and CSI recovery.

[0239] In this way, the input information corresponding to the first measurement resource can refer to the input information corresponding to the measurement resource with the earliest time-domain position, the input information corresponding to the second measurement resource can refer to the input information after the input information corresponding to the first measurement resource, and so on.

[0240] Optionally, the input information corresponding to the measurement resources can be input into the enhanced bilateral AI model in the order of the sizes of the IDs / indexes of the measurement resources. For example, the input information corresponding to the measurement resources with smaller IDs / indexes will be input into the enhanced bilateral AI model earlier for CSI compression and CSI recovery, while the input information corresponding to the measurement resources with larger IDs / indexes will be input into the enhanced bilateral AI model later for CSI compression and CSI recovery.

[0241] In this way, the input information corresponding to the first measurement resource can refer to the input information corresponding to the measurement resource with a smaller ID / index, the input information corresponding to the second measurement resource can refer to the input information after the input information corresponding to the first measurement resource, and so on.

[0242] Combining the above content, the following solutions are used in this embodiment to specifically illustrate the enhancement of CSI-related processing. Among them, these solutions can be related to each other or independent of each other, and the same content between different solutions can be referred to each other, which will not be elaborated here.

[0243]

Solution 1

[0244] In the process of enhanced CSI compression and CSI recovery, since the bilateral AI model will perform CSI compression and recovery on the input information in sequence, and the historical input information can participate in the CSI compression of the current input information, the processed information obtained earlier can be used as historical processed information to participate in the CSI compression of the current input information, and / or the intermediate compression information obtained earlier can be used as historical intermediate compression information to participate in the CSI compression of the current input information.

[0245] Based on this, in "Solution 1", this embodiment considers grouping the input information. Among them, the purpose of grouping is that the historical input information within a group can participate in the CSI compression of the non-first input information within the same group, and all the input information in a group does not participate in the CSI compression of any input information in other groups.

[0246] For example, taking the input information divided into two groups as an example, the first input information within the first group is directly subjected to CSI compression, and the historical input information within the first group only participates in the CSI compression of the non-first input information within this group, rather than participating in the CSI compression of any input information within the second group; the first input information within the second group is directly subjected to CSI compression, and the historical input information within the second group only participates in the CSI compression of the non-first input information within this group.

[0247] In this way, the enhanced bilateral AI model can be used to perform CSI compression on the first input information within a group, and use the historical input information within this group to perform CSI compression on the non-first input information within this group.

[0248] For example, when grouping the Figure 4 input information, if the input information from V1 to the input information V i-1 is divided into the same group, and the input information after the input information V i-1 is divided into another group, then the sequence of the encoder 511, quantizer 512, dequantizer 521, and decoder 522 in the enhanced bilateral AI model to process the information is as Figure 5 shown.

[0249] Optionally, the group can have a group length. Among them, the group length can be the maximum number of input information within the group.

[0250] Optionally, the group length can be finite or infinite, and specifically can be specified by network configuration or protocol.

[0251] It should be noted that when the group length is finite, this embodiment can determine one or more groups according to the group length, that is, all the input information can be in one or more groups; when the group length is infinite, it is generally considered that there can only be one group, that is, all the input information is in one group. In addition, when the group length is finite, this indicates that there is an exact maximum number of input information; when the group length is infinite, it is generally considered that the maximum number of input information is infinite.

[0252] Optionally, the group lengths of different groups can be different or the same.

[0253] It should be noted that the groups can have identifiers (IDs) or indices to distinguish different groups by the identifiers / indices. Additionally, the group lengths of different groups can be different or the same. When the network configures the group lengths, the network can flexibly configure different group lengths for different groups to improve the flexibility of group length configuration; or, the network can configure the same group length for different groups with only one configuration to save the signaling overhead required for configuration.

[0254] Combined with the content in the above "Differentiation of Input Information", in this embodiment, the input information corresponding to the CSI report or the input information corresponding to the measurement resources can be grouped.

[0255] The input information corresponding to the CSI report can be only the input information corresponding to the periodic CSI report, or only the input information corresponding to the aperiodic CSI report, or only the input information corresponding to the semi-persistent CSI report, or a combination of the input information corresponding to the periodic CSI report, the input information corresponding to the aperiodic CSI report, and / or the input information corresponding to the semi-persistent CSI report, etc., and specifically can be determined according to the time-domain behavior of the CSI report configured / indicated by the network.

[0256] The input information corresponding to the measurement resources can be only the input information corresponding to the periodic measurement resources, or only the input information corresponding to the aperiodic measurement resources, or only the input information corresponding to the semi-persistent measurement resources, or a combination of the input information corresponding to the periodic measurement resources, the input information corresponding to the aperiodic measurement resources, and / or the input information corresponding to the semi-persistent measurement resources, etc., and specifically can be determined according to the time-domain behavior of the measurement resources configured / indicated by the network.

[0257] For example, taking the CSI report to define the input information as an example, the terminal device performs CSI measurements based on the measurement resources associated with the periodic CSI report to obtain the input information corresponding to the periodic CSI report, so as to group the input information corresponding to these periodic CSI reports.

[0258] Again, taking the CSI report to define the input information as an example, the terminal device performs CSI measurements based on the measurement resources associated with the semi-persistent CSI report to obtain the input information corresponding to the semi-persistent CSI report, so as to group the input information corresponding to these semi-persistent CSI reports.

[0259] Again, taking the CSI report to define the input information as an example, the terminal device performs CSI measurements based on the measurement resources associated with multiple aperiodic CSI reports triggered by one DCI to obtain the input information corresponding to each aperiodic CSI report, so as to group the input information corresponding to these aperiodic CSI reports.

[0260] For another example, taking CSI report as an example to define input information, the terminal device performs CSI measurement according to the measurement resources associated with periodic CSI report, semi-persistent CSI report and aperiodic CSI report respectively to obtain the corresponding input information. At this time, this embodiment can group the input information corresponding to the periodic CSI report, the input information corresponding to the semi-persistent CSI report and the input information corresponding to the aperiodic CSI report together, or group them separately, which can be determined by network configuration, terminal device autonomously or stipulated by protocol, etc., and there is no specific limitation on this.

[0261] In addition, after the input information is grouped, since CSI compression and CSI recovery coexist, and the input information in a group needs to be CSI compressed in sequence, and the processing information needs to be used for CSI recovery in sequence, after CSI compression is performed on certain input information in a group to obtain processing information, the terminal device may need to report the position of the input information in the group or the position of the processing information in the group to the network device.

[0262] In this way, when the network device knows the position of the input information in the packet or the position of the processed information in the packet, the network device can determine the order of the processed information in CSI recovery according to the position of the input information or the position of the processed information in the packet.

[0263] It is worth noting that this embodiment mainly describes the grouping of input information, but since there is a one-to-one correspondence between input information, processing information and output information, the grouping of input information can also be the grouping of processing information or the grouping of output information, and the content of the grouping of input information can be adapted to the grouping of processing information or the grouping of output information, which will not be described in detail.

[0264] The present embodiment will now specifically describe how to determine the grouping of input information in the following manners, wherein these manners may be interrelated or independent of each other, and the same contents between different manners may be referenced to each other, which will not be described in detail.

[0265] Method 1

[0266] In "method 1", this embodiment can introduce a time window.

[0267] Among them, if the time domain position of at least one CSI report is within the time window, the input information corresponding to the at least one CSI report belongs to the same group; or, if the time domain position of at least one measurement resource or the time domain position of the CSI report associated with the at least one measurement resource is within the time window, the input information corresponding to the at least one measurement resource belongs to the same group.

[0268] In this way, the input information is grouped through a time window.

[0269] Of course, the time window can be described by other terms, such as window, time period, or duration, etc., and no specific restrictions are imposed on this. Optionally, the length of the time window is configured by the network. For example, the network device can configure the length of the time window for the terminal device through high-layer signaling (such as RRC signaling, MAC signaling) or DCI.

[0270] Optionally, for periodic CSI reporting or semi-persistent CSI reporting, the length of the time window can be an integer multiple of the CSI reporting period. In this way, the time-domain positions of multiple periodic CSI reports or multiple semi-persistent CSI reports can be within one time window.

[0271] Optionally, for periodic CSI reporting or CSI reporting carried on PUCCH, the starting position of the time window satisfies the following:

[0272]

[0273] where μ represents the subcarrier spacing configuration of the uplink bandwidth part (UL BWP) for transmitting the CSI report; represents the number of time slots in a frame under the subcarrier spacing configuration μ; n f represents the system frame number where the CSI report is located; represents the time slot number where the CSI report is located; T offset represents the time slot offset; N represents the maximum number of input information within a packet; T CSI represents the CSI reporting period.

[0274] Optionally, for CSI reporting carried on PUSCH, the starting position of the time window satisfies the following:

[0275]

[0276] where, represents the system frame number of the PUSCH where the first CSI report is sent after DCI activation; represents the time slot number of the PUSCH where the first CSI report is sent after DCI activation.

[0277] Optionally, the starting position of the time window is configured by the network. For example, the network device can configure an offset for the terminal device through high-layer signaling (such as RRC signaling, MAC signaling) or DCI, and this offset can determine the starting position of the time window.

[0278] Optionally, the time window is periodic, and the period of the time window is configured by the network. For example, the network device may configure the period of the time window for the terminal device through high-layer signaling (such as RRC signaling, MAC signaling) or DCI.

[0279] Taking the interaction between the network device and the terminal device as an example below, "Method 1" will be illustrated. As Figure 6 shown, Figure 6 is a schematic flowchart of a communication method according to an embodiment of the present application. Among them, the method includes the following steps:

[0280] S610. The network device sends configuration information, where the configuration information is used to configure a time window, and the time window is used to determine the grouping of the first input information of the first AI model.

[0281] Correspondingly, the terminal device receives the configuration information.

[0282] It can be seen that in this embodiment, the network configures the time window through the configuration information.

[0283] It should be noted that for the first input information, this embodiment also involves first processing information and first output information. Among them, the first processing information is the information obtained by inputting the first input information into the first AI model or the information obtained by further processing the information obtained by inputting the first input information into the first AI model; the first output information is the information obtained by inputting the first processing information into the first AI model or the information obtained by further processing the information obtained by inputting the first processing information into the first AI model.

[0284] It is worth noting that "Method 1" mainly describes the grouping of the first input information. However, since there is a one-to-one correspondence among the first input information, the first processing information, and the first output information, the grouping of the first input information can also be regarded as the grouping of the first processing information or the grouping of the first output information. At the same time, the content of the grouping of the first input information can be adapted to the grouping of the first processing information or the grouping of the first output information, which will not be elaborated here.

[0285] Optionally, the "first AI model" here refers to the AI model used for CSI compression in the enhanced bilateral AI model.

[0286] Optionally, the configuration information includes at least one of length information, start information, or period information;

[0287] The length information is used to configure the length of the time window and / or the grouping length of the input information;

[0288] The start information is used to configure the start position of the time window;

[0289] The period information is used to configure the period of the time window.

[0290] Optionally, the length of the time window is an integer multiple of the CSI reporting period.

[0291] Optionally, the starting position of the time window satisfies the following:

[0292] Or,

[0293]

[0294] Optionally, the configuration information is carried by functionality identification information or model identification information.

[0295] Optionally, the configuration information is carried by high-layer signaling (such as RRC signaling or MAC signaling) or DCI.

[0296] Optionally, the first input information is among the input information corresponding to at least one CSI report, and the time-domain behavior of the at least one CSI report is at least one of periodic, semi-persistent, or aperiodic.

[0297] It can be understood that the input information corresponding to the at least one CSI report includes at least one of the input information corresponding to a periodic CSI report, the input information corresponding to a semi-persistent CSI report, or the input information corresponding to an aperiodic CSI report. At the same time, the first input information can be the input information corresponding to a periodic CSI report, the input information corresponding to a semi-persistent CSI report, or the input information corresponding to an aperiodic CSI report.

[0298] Optionally, the first input information can be the first input information and / or non-first input information among the input information corresponding to the at least one CSI report.

[0299] Optionally, the input information corresponding to the at least one CSI report is input into the first AI model in the order of the time-domain positions of the CSI reports or the order of the sizes of the CSI report IDs / indexes.

[0300] Optionally, the at least one CSI report includes all aperiodic CSI reports triggered by a DCI.

[0301] It can be understood that the input information corresponding to the at least one CSI report includes the input information corresponding to all aperiodic CSI reports triggered by a DCI.

[0302] Optionally, the first input information is among the input information corresponding to at least one measurement resource, and the time-domain behavior of the at least one measurement resource is at least one of periodic, semi-persistent, or aperiodic.

[0303] It can be understood that the input information corresponding to the at least one measurement resource includes at least one of the input information corresponding to a periodic measurement resource, the input information corresponding to a semi-persistent measurement resource, or the input information corresponding to an aperiodic CSI report. At the same time, the first input information can be the input information corresponding to a periodic CSI report, the input information corresponding to a semi-persistent CSI report, or the input information corresponding to an aperiodic CSI report.

[0304] Optionally, the first input information can be the first input information and / or non-first input information among the input information corresponding to the at least one measurement resource.

[0305] Optionally, the input information corresponding to the at least one measurement resource is input into the first AI model in the order of the time-domain positions of the measurement resources or the order of the sizes of the IDs / indices of the measurement resources.

[0306] S620. The terminal device determines the grouping of the first input information according to the time window.

[0307] Optionally, if the first input information is among the input information corresponding to at least one CSI report and the time-domain positions of the at least one CSI report are within the time window, then the input information corresponding to the at least one CSI report belongs to the same group.

[0308] In this way, for the group to which the input information corresponding to the at least one CSI report belongs, when the first input information is the first input information in the group, the first AI model can directly perform CSI compression on the first input information; when the first input information is the non-first input information in the group, the first AI model can use the historical input information in the group to perform CSI compression on the first input information.

[0309] Optionally, if the first input information is among the input information corresponding to at least one measurement resource and the time-domain positions of the at least one measurement resource are within the time window or the time-domain positions of the CSI reports associated with the at least one measurement resource are within the time window, then the input information corresponding to the at least one measurement resource belongs to the same group.

[0310] In this way, for the group to which the input information corresponding to the at least one measurement resource belongs, when the first input information is the first input information in the group, the first AI model can directly perform CSI compression on the first input information; when the first input information is the non-first input information in the group, the first AI model can use the historical input information in the group to perform CSI compression on the first input information.

[0311] Optionally, after S620, the method may further include:

[0312] The terminal device sends first location information, which is used to indicate the position of the first input information in the packet or the position of the first processing information in the packet.

[0313] Correspondingly, the network device receives the first location information.

[0314] It can be seen that since CSI compression and CSI recovery are interdependent, and the input information in a packet needs to be subjected to CSI compression in sequence to obtain processing information, and these processing information needs to be subjected to CSI recovery in sequence. Therefore, after performing CSI compression on the first input information to obtain the first processing information, the terminal device reports the position of the first input information in the packet or the position of the first processing information in the packet to the network device through the first location information. In this way, the network device can know the order in the CSI recovery process of the first processing information according to the position of the first input information in the packet or the position of the first processing information in the packet, so as to perform CSI recovery on the first processing information in sequence.

[0315] Optionally, the first location information is carried by one of CSI reports, high-layer signaling (such as RRC signaling or MAC signaling), or UCI.

[0316] "Method 2"

[0317] In "Method 2", this embodiment may configure / indicate the packet of the input information through high-layer signaling (such as RRC signaling or MAC signaling) or DCI, etc.

[0318] For example, the network device may configure / indicate the packet ID associated with the CSI report through high-layer signaling (such as RRC signaling or MAC signaling) or DCI, etc. In this way, the input information corresponding to the CSI report is associated with the packet ID, and the input information corresponding to the CSI reports associated with the same packet ID is divided into one packet.

[0319] For another example, the network device may configure / indicate the packet ID associated with the measurement resource through high-layer signaling (such as RRC signaling or MAC signaling) or DCI, etc. In this way, the input information corresponding to the measurement resource is associated with the packet ID, and the input information corresponding to the measurement resources associated with the same packet ID is divided into one packet.

[0320] Next, taking the interaction between the network device and the terminal device as an example, an example illustration of "Method 2" is given. As Figure 7 shown, Figure 7 is a schematic flowchart of another communication method according to an embodiment of the present application. Wherein, the method includes the following steps:

[0321] S710. The network device sends indication information, where the indication information is used to indicate the grouping of the first input information of the first AI model.

[0322] Correspondingly, the terminal device receives the indication information.

[0323] S720. The terminal device determines the grouping of the first input information according to the indication information.

[0324] It can be seen that in this embodiment, the network indicates the grouping of the input information through the indication information.

[0325] Optionally, the "first AI model" here refers to the AI model used for CSI compression in the enhanced bilateral AI model.

[0326] Optionally, the indication information is carried by the function registration information or the model registration information.

[0327] Optionally, the indication information is carried by the high-layer signaling (such as RRC signaling or MAC signaling) or DCI.

[0328] Optionally, the first input information is among the input information corresponding to at least one CSI report, and the time-domain behavior of the at least one CSI report is at least one of periodic, semi-persistent, or aperiodic.

[0329] It can be understood that the input information corresponding to the at least one CSI report includes at least one of the input information corresponding to the periodic CSI report, the input information corresponding to the semi-persistent CSI report, or the input information corresponding to the aperiodic CSI report. At the same time, the first input information can be the input information corresponding to a periodic CSI report, the input information corresponding to a semi-persistent CSI report, or the input information corresponding to an aperiodic CSI report.

[0330] Optionally, the first input information can be the first input information or non-first input information among the input information corresponding to the at least one CSI report.

[0331] Optionally, the input information corresponding to the at least one CSI report is input into the first AI model in the order of the time-domain positions of the CSI reports or the order of the sizes of the CSI report IDs / indexes.

[0332] Optionally, the at least one CSI report includes all aperiodic CSI reports triggered by one DCI.

[0333] It can be understood that the input information corresponding to the at least one CSI report includes the input information corresponding to all aperiodic CSI reports triggered by one DCI.

[0334] Optionally, the first input information is among the input information corresponding to at least one measurement resource, and the time-domain behavior of the at least one measurement resource is at least one of periodic, semi-persistent, or aperiodic.

[0335] It can be understood that the input information corresponding to the at least one measurement resource includes at least one of the input information corresponding to a periodic measurement resource, the input information corresponding to a semi-persistent measurement resource, or the input information corresponding to an aperiodic CSI report. At the same time, the first input information can be the input information corresponding to a periodic CSI report, the input information corresponding to a semi-persistent CSI report, or the input information corresponding to an aperiodic CSI report.

[0336] Optionally, the first input information can be the first input information or non-first input information among the input information corresponding to the at least one measurement resource.

[0337] Optionally, the input information corresponding to the at least one measurement resource is input into the first AI model in the order of the time-domain positions of the measurement resources or the order of the sizes of the IDs / indexes of the measurement resources.

[0338] Optionally, after S720, the method may further include:

[0339] The terminal device sends first location information, and the first location information is used to indicate the position of the first input information in the packet or the position of the first processing information in the packet.

[0340] Correspondingly, the network device receives the first location information.

[0341] It can be seen that since CSI compression and CSI recovery are interdependent, and the input information in a packet needs to be sequentially CSI-compressed to obtain processing information, and these processing information needs to be sequentially CSI-recovered, therefore, after CSI-compressing the first input information to obtain the first processing information, the terminal device reports to the network device the position of the first input information in the packet or the position of the first processing information in the packet through the first location information. In this way, the network device can know the order in the CSI recovery process of the first processing information according to the position of the first input information in the packet or the position of the first processing information in the packet, so as to perform CSI recovery on the first processing information in order.

[0342]

Solution 2

[0343] During the enhanced CSI compression and CSI recovery processes, the terminal device may need to report / feedback input information and / or processed information. However, some input information and / or processed information that originally needed to be reported / feedback may not be reported due to some factors, resulting in the problem of possible dropping / omission of input information and / or processed information.

[0344] For example, when considering limited uplink resources, the input information and / or processed information that originally needed to be reported may not be reported due to insufficient uplink resources.

[0345] Based on this, in "Solution 2", this embodiment needs to determine how to perform CSI reporting on the input information and / or processed information that originally needed to be reported but was not reported. Specifically, it can be determined by network configuration, the terminal device itself, or protocol regulations, etc., and no specific restrictions are imposed on this.

[0346] Optionally, the terminal device can continue to report / postpone reporting the input information and / or processed information that originally needed to be reported but was not reported. It can be understood that when the input information and / or processed information that originally needed to be reported fails to be reported at the current moment / current CSI reporting process, the terminal device can report the input information and / or processed information at a subsequent moment / subsequent CSI reporting process. For example, carry the input information and / or processed information through a subsequent CSI report.

[0347] Optionally, the terminal device can no longer report the input information and / or processed information that originally needed to be reported but was not reported. It can be understood that when the input information and / or processed information that originally needed to be reported fails to be reported at the current moment / current CSI reporting process, the terminal device will no longer report the input information and / or processed information at a subsequent moment / subsequent CSI reporting process.

[0348] It should be noted that the input information discussed in "Solution 2" can be grouped according to the above "Solution 1" or not grouped, and no specific restrictions are imposed on this.

[0349] Next, this embodiment specifically describes how to determine the reporting of the input information and / or processed information that originally needed to be reported but was not reported in the following manner. Among them, these methods can be related to each other or independent of each other, and the same content between different methods can be referenced to each other, and no further elaboration is made on this.

[0350] "Method 1"

[0351] Since traditional bilateral AI models / enhanced bilateral AI models need to use input information and / or processed information for CSI recovery, in the face of input information and / or processed information that originally needed to be reported but was not reported, the network device cannot perform CSI recovery because it fails to receive the input information and / or processed information. Additionally, in traditional bilateral AI models / enhanced bilateral AI models, CSI recovery needs to be performed in sequence among different pieces of information. For example, in Figure 4 the processed information V after quantization Q1 needs to be sequentially input into the dequantizer 421 and the decoder 422 for CSI recovery to obtain the output information and then it is the processed information V after quantization Q2 sequentially input into the dequantizer 421 and the decoder 422 for CSI recovery to obtain the output information

[0352] Based on this, in "Mode 1", the terminal device can continue to report / postpone reporting the input information and / or processed information that originally needed to be reported but was not reported.

[0353] In this way, by continuing to report / postpone reporting, the input information and / or current processed information that originally needed to be reported but was not reported can be continuously reported, avoiding the inability to perform CSI recovery on this part of the information and subsequent information due to the network device's failure to receive the input information and / or processed information.

[0354] Optionally, if the current CSI report where the input information and / or processed information that originally needed to be reported is located fails to be reported, the terminal device can carry the input information and / or processed information in a subsequent CSI report, thereby achieving continuous reporting / postponing reporting of the input information and / or processed information that originally needed to be reported but was not reported through the subsequent CSI report.

[0355] Optionally, if the current CSI report where the input information and / or processed information that originally needed to be reported is located fails to be reported, the terminal device can send the input information and / or the processed information at an available time domain position after the time domain position of the current CSI report, thereby achieving continuous reporting / postponing reporting of the input information and / or processed information that originally needed to be reported but was not reported through the subsequent available time domain position.

[0356] For example, the terminal device can send the input information and / or the processed information at an available uplink time slot after the uplink time slot where the current CSI report is located.

[0357] Next, taking the interaction between the network device and the terminal device as an example, "Mode 1" will be illustrated. As Figure 8 shown, Figure 8It is a schematic flowchart of another communication method according to an embodiment of the present application. The method includes the following steps:

[0358] S810. If the first CSI report where the first input information and / or the first processing information is located fails to be reported, the terminal device sends a second CSI report, where the second CSI report includes the first input information and / or the first processing information, or the terminal device sends the first input information and / or the first processing information at an available time domain position after the time domain position of the first CSI report.

[0359] The first processing information is the information obtained by inputting the first input information into the first AI model or the information obtained by further processing the information obtained by inputting the first input information into the first AI model.

[0360] Correspondingly, the network device receives the second CSI report, or receives the first input information and / or the first processing information at an available time domain position after the time domain position of the first CSI report.

[0361] It can be seen that when it is originally desired to report the first input information and / or the first processing information through the first CSI report, since the first CSI report fails to be reported, this causes the first input information and / or the first processing information that originally needed to be reported to fail to be reported. Therefore, the terminal device can carry the first input information and / or the first processing information through the second CSI report, so as to continue to report the first input information and / or the first processing information that originally needed to be reported but was not reported.

[0362] Alternatively, when the first input information and / or the first processing information that originally needed to be reported fails to be reported, the terminal device can send the first input information and / or the first processing information at an available time domain position after the time domain position of the first CSI report, so as to continue to report / postpone reporting the first input information and / or the first processing information.

[0363] Optionally, the time domain position of the second CSI report is after the time domain position of the first CSI report. In this way, since the time domain position of the second CSI report is after the time domain position of the first CSI report, the first input information and / or the first processing information that originally needed to be reported but was not reported can be postponed and reported through the second CSI report.

[0364] Optionally, the second CSI report is the next CSI report adjacent to the first CSI report. In this way, the first input information and / or the first processing information can be postponed and reported through the next CSI report adjacent to the first CSI report. At the same time, since the second CSI report and the first CSI report are adjacent, the time for postponed reporting can be reduced.

[0365] Optionally, the method further includes:

[0366] The terminal device sends postponement information, where the postponement information is used to indicate that the first input information and / or the first processing information is a postponement of reporting of the first CSI report.

[0367] Correspondingly, the network device receives the deferral information.

[0368] It can be seen that the postponement information is used to indicate to the network device which CSI reporting the first input information and / or the first processing information is for, so as to know the original order of the postponed information in CSI recovery.

[0369] Optionally, the postponement information may be carried by the second CSI report. In this way, the terminal device can not only postpone the reporting of the first input information and / or the first processed information that was originally required to be reported but was not reported, but also indicate to the network device which time the first input information and / or the first processed information is postponed, through the second CSI report.

[0370] Optionally, the deferral information is carried by one of CSI report, higher layer signaling (such as RRC signaling or MAC signaling) or UCI.

[0371] "Method 2"

[0372] In "Method 2", the terminal device can postpone reporting the input information and / or processing information that was originally required to be reported but was not reported. In this way, the input information and / or processing information that was originally required to be reported but was not reported can continue to be reported by postponing the reporting, thereby avoiding the inability to perform CSI recovery on this part of information due to the network device being unable to receive the input information and / or processing information.

[0373] At the same time, when the input information and / or processing information that was originally required to be reported but was not reported is postponed to the subsequent CSI reporting, since the input information and / or processing information that was originally required to be reported but was not reported will be postponed to the subsequent CSI reporting, and the subsequent CSI reporting may have the input information and / or processing information that needs to be reported, the terminal device can report the input information and / or the processing information that was originally required to be reported in the subsequent CSI reporting in addition to reporting the input information and / or processing information that was originally required to be reported in the subsequent CSI reporting.

[0374] Optionally, the input information that was originally required to be reported in the subsequent CSI reporting may be the input information after the input information and / or processing information that was originally required to be reported but was not reported; the processing information that was originally required to be reported in the subsequent CSI reporting may be the processing information after the processing information that was originally required to be reported but was not reported.

[0375] Optionally, the input information that should have been reported in this subsequent CSI report can be the remaining input information in the group where the input information that should have been reported but was not reported originally is located.

[0376] Optionally, the input information that should have been reported in this subsequent CSI report and the input information that should have been reported but was not reported originally belong to the same group or different groups.

[0377] Optionally, if the input information that should have been reported in this subsequent CSI report and the input information that should have been reported but was not reported originally belong to different groups, then the input information that should have been reported in this subsequent CSI report can be the first input information in the new group.

[0378] Next, taking the interaction between a network device and a terminal device as an example, an example description of "Method 2" is given. As Figure 9 shown, Figure 9 is a schematic flowchart of another communication method according to an embodiment of the present application. Among them, the method includes the following steps:

[0379] S910: If the first CSI report where the first input information and / or the first processing information is located fails to be reported, the terminal device sends a second CSI report, and the second CSI report includes the first input information and / or the first processing information, and the second input information and / or the second processing information.

[0380] Among them, the second processing information is the information obtained by inputting the second input information into the first AI model or the information obtained by further processing the information obtained by inputting the second input information into the first AI model.

[0381] Among them, the first input information and the second input information are different input information.

[0382] Correspondingly, the network device receives the second CSI report.

[0383] It can be seen that when it was originally hoped to report the first input information and / or the first processing information through the first CSI report, since the first CSI report fails to be reported, this causes the first input information and / or the first processing information that should have been reported to fail to be reported. Therefore, the first input information and / or the first processing information, and the second input information and / or the second processing information are carried through the second CSI report, so that both delaying the reporting of the first input information and / or the first processing information and reporting the second input information and / or the second processing information required to be reported by the second CSI report can be achieved through the second CSI report, avoiding the loss of the second input information and / or the second processing information.

[0384] Optionally, the second input information and / or the second processing information may be the input information and / or the processing information that the second CSI report originally needs to carry. In this way, both the second input information and / or the second processing information that originally need to be reported can be reported through the second CSI report, and at the same time, the first input information and / or the first processing information can be reported along with the second CSI report.

[0385] Optionally, the second input information may be the input information after the first input information; the second processing information may be the processing information after the first processing information.

[0386] Optionally, the second input information may be the remaining input information in the group where the first input information is located.

[0387] Optionally, the first input information and the second input information are in the same group or different groups. If the first input information and the second input information are in different groups, then the second input information is the first input information in the group.

[0388] It can be seen that when delaying the reporting of the input information (i.e., the first input information) that originally needs to be reported but has not been reported in the original group, in this embodiment, the remaining input information (i.e., the second input information) after this input information in the original group may not be reconstructed into a new group, but the reporting of this input information and the remaining input information after this output information in the original group continues.

[0389] Alternatively, in this embodiment, the remaining input information after this input information (i.e., the first input information) in the original group may be reconstructed into a new group, so that the remaining input information after this input information (i.e., the second input information) is the first input information in the new group, so that the terminal device can continue to report this input information and the first input information in the new group.

[0390] Optionally, the time domain position of the second CSI report is after the time domain position of the first CSI report. In this way, since the time domain position of the second CSI report is after the time domain position of the first CSI report, the first input information and / or the first processing information that originally need to be reported but have not been reported are delayed and reported through the second CSI report.

[0391] Optionally, the second CSI report is the next CSI report adjacent to the first CSI report. In this way, the first input information and / or the first processing information are delayed and reported through the next CSI report adjacent to the first CSI report. At the same time, since the second CSI report and the first CSI report are adjacent, the time for delaying the reporting can be reduced.

[0392] Optionally, the method further includes:

[0393] The terminal device sends a postponement message, which is used to indicate that the first input information and / or the first processing information is a postponed report of the first CSI report.

[0394] Correspondingly, the network device receives the postponement message.

[0395] It can be seen that through the postponement message, the network device is indicated which CSI report the first input information and / or the first processing information is a postponed report of, so as to know the original order of the postponed report information in the CSI recovery.

[0396] Optionally, the postponement message can be carried by the second CSI report. In this way, the terminal device can not only postpone the reporting of the first input information and / or the first processing information that originally needed to be reported but was not reported, but also indicate to the network device which postponed report the first input information and / or the first processing information is.

[0397] Optionally, the postponement message is carried by one of a CSI report, a high-layer signaling (such as an RRC signaling or a MAC signaling), or a UCI.

[0398] "Mode 3"

[0399] In "Mode 3", in the face of the input information and / or the processing information that originally needed to be reported but was not reported, the terminal device can choose not to report the input information and / or the processing information that originally needed to be reported but was not reported.

[0400] Based on this, in "Mode 3", when the terminal device no longer reports the current input information and / or the current processing information, the terminal device can start reporting other input information.

[0401] Taking the interaction between the network device and the terminal device as an example, "Mode 3" is illustrated as follows. As Figure 10 shown, Figure 10 is a schematic flowchart of another communication method according to an embodiment of the present application. Among them, the method includes the following steps:

[0402] S1010. If the first CSI report in which the first input information and / or the first processing information is located fails to be reported, the terminal device no longer sends the first input information and / or the first processing information, and sends a third CSI report, where the third CSI report includes third input information and / or third processing information.

[0403] Among them, the third processing information is the information obtained by inputting the third input information into the first AI model or the information obtained by further processing the information obtained by inputting the third input information into the first AI model.

[0404] Among them, the first input information and the third input information are different input information.

[0405] Correspondingly, the network device receives a third CSI report.

[0406] It can be seen that when originally hoping to report the first input information and / or the first processing information through the first CSI report, since the first CSI report fails to be reported, this causes the first input information and / or the first processing information that originally needed to be reported to fail to be reported. Therefore, the terminal device no longer reports the first input information and / or the first processing information, but reports the third input information and / or the third processing information through the third CSI report, realizing the reporting of new input information and / or new processing information and avoiding the loss of new input information and / or new processing information.

[0407] Optionally, the third input information and / or the third processing information may be the input information and / or the processing information that the third CSI report originally needed to carry. In this way, the third input information and / or the third processing information that originally needed to be reported are reported through the third CSI report.

[0408] Optionally, the third input information may be the input information after the first input information; the third processing information may be the processing information after the first processing information.

[0409] Optionally, the third input information may be the remaining input information in the group where the first input information is located.

[0410] Optionally, the time domain position of the third CSI report is after the time domain position of the first CSI report.

[0411] Optionally, the first input information and the third input information are in different groups, and the third input information is the first input information in the group.

[0412] It can be seen that when no longer reporting the input information (i.e., the first input information) that originally needed to be reported but was not reported in the original group, this embodiment can reconstruct the group for the remaining input information after the input information in the original group, so that the remaining input information after the input information is the first input information (i.e., the third input information) in the new group, so that the terminal device can report the first input information in the new group.

[0413] Optionally, the third CSI report is the next CSI report adjacent to the first CSI report. In this way, the reporting of the third input information and / or the third processing information is postponed through the next CSI report adjacent to the first CSI report.

[0414]

Example illustration of a communication method

[0415] Combined with the above content, the following is an example illustration of another communication method of the embodiments of the present application, asFigure 11 As shown in Figure 11 , this method can be applied to a terminal device. The method includes the following steps:

[0416] S1110. Determine the grouping of the first input information of the first AI model, or determine the CSI reporting of the first input information and / or the first processing information; the first AI model is used for CSI compression, and the first processing information is the information obtained by inputting the first input information into the first AI model or the information obtained by further processing the information obtained by inputting the first input information into the first AI model.

[0417] It can be seen that when AI technology is applied to the CSI framework, in the face of the enhancement of CSI compression and CSI recovery, since the first AI model will perform CSI compression on the input information in sequence, the historical input information can participate in the CSI compression of the current input information, the previously obtained processing information can be used as historical processing information to participate in the CSI compression of the current input information, and / or the previously obtained compressed intermediate information can be used as historical compressed intermediate information to participate in the CSI compression of the current input information. Therefore, in this embodiment, it is necessary to determine the grouping of the input information. In this way, the first AI model can be used to perform CSI compression on the first input information in a group, and use the historical input information in the group to perform CSI compression on the non-first input information in the group, so as to achieve the enhancement of CSI-related processing.

[0418] Alternatively, when the terminal device may need to report / feedback the input information and / or the processing information, due to some factors, some of the originally required input information and / or processing information may not be reported, resulting in the discarding of the input information and / or the processing information. In this way, in the face of the input information and / or the processing information that originally needed to be reported but were not reported, this embodiment can determine the CSI reporting of the input information and / or the processing information that originally needed to be reported but were not reported, so as to achieve the enhancement of CSI-related processing.

[0419] Optionally, the first AI model is used to perform CSI compression on the first input information and use the first historical information to perform CSI compression on the non-first input information;

[0420] The first historical information includes at least one of historical input information, historical processing information, or historical compressed intermediate information. The historical processing information is the information obtained by inputting the historical input information into the first AI model or the information obtained by further processing the information obtained by inputting the historical input information into the first AI model. The historical compressed intermediate information is the information generated by the first AI model during the process of compressing the historical input information.

[0421] Optionally, the historical input information within a group participates in the CSI compression of the non-first input information within the same group, and all the input information in a group does not participate in the CSI compression of any input information in other groups.

[0422] Optionally, the first input information is among the input information corresponding to at least one CSI report, and the time domain behavior of at least one CSI report is at least one of periodic, semi-persistent, or aperiodic; or,

[0423] The first input information is among the input information corresponding to at least one measurement resource;

[0424] The time domain behavior of at least one measurement resource is at least one of periodic, semi-persistent, or aperiodic.

[0425] Optionally, the input information corresponding to at least one CSI report is input into the first AI model in the order of the time domain positions of the CSI reports or in the order of the magnitudes of the identification IDs of the CSI reports; or,

[0426] The input information corresponding to at least one measurement resource is input into the first AI model in the order of the time domain positions of the measurement resources or in the order of the magnitudes of the IDs of the measurement resources.

[0427] Optionally, at least one CSI report includes all aperiodic CSI reports triggered by a downlink control information DCI.

[0428] Optionally, in S1110, determining the grouping of the first input information of the first AI model includes:

[0429] Receiving configuration information, where the configuration information is used to configure a time window;

[0430] Determining the grouping of the first input information of the first AI model according to the time window.

[0431] Optionally, if the first input information is among the input information corresponding to at least one CSI report and the time domain position of at least one CSI report is within the time window, then the input information corresponding to at least one CSI report belongs to the same group.

[0432] Optionally, if the first input information is among the input information corresponding to at least one measurement resource and the time domain position of at least one measurement resource is within the time window or the time domain position of at least one CSI report associated with at least one measurement resource is within the time window, then the input information corresponding to at least one measurement resource belongs to the same group.

[0433] Optionally, the configuration information includes at least one of length information, start information, or period information;

[0434] Length information for configuring the length of the time window;

[0435] Start information for configuring the start position of the time window;

[0436] Period information for configuring the period of the time window.

[0437] Optionally, the length of the time window is an integer multiple of the CSI reporting period.

[0438] Optionally, the start position of the time window satisfies the following:

[0439] Or,

[0440]

[0441] Where μ represents the subcarrier spacing configuration of the uplink bandwidth part BWP for transmitting CSI reports; represents the number of time slots in a frame under the subcarrier spacing configuration μ; n f represents the system frame number where the CSI report is located; represents the time slot number where the CSI report is located; T offset represents the time slot offset; N represents the maximum number of input information in a packet; T CSI represents the CSI reporting period; represents the system frame number of the physical uplink shared channel PUSCH that sends the first CSI report after DCI activation; represents the time slot number of the PUSCH that sends the first CSI report after DCI activation.

[0442] Optionally, in S1110, determining the packet of the first input information of the first AI model includes:

[0443] Receiving indication information for indicating the packet of the first input information of the first AI model;

[0444] Determining the packet of the first input information according to the indication information.

[0445] Optionally, the method further includes:

[0446] Sending first position information for indicating the position of the first input information in the packet or the position of the first processing information in the packet.

[0447] Optionally, in S1110, determining the CSI reporting of the first input information and / or the first processing information includes:

[0448] If the first CSI report in which the first input information and / or the first processed information is located fails to be reported, then send a second CSI report, where the second CSI report contains the first input information and / or the first processed information, or send the first input information and / or the first processed information at an available time domain position after the time domain position of the first CSI report.

[0449] Optionally, in S1110, determining the CSI report for the first input information and / or the first processed information includes:

[0450] If the first CSI report in which the first input information and / or the first processed information is located fails to be reported, then send a second CSI report, where the second CSI report contains the first input information and / or the first processed information, as well as the second input information and / or the second processed information;

[0451] The second processed information is the information obtained by inputting the second input information into the first AI model or the information obtained by further processing the information obtained by inputting the second input information into the first AI model;

[0452] The first input information and the second input information are different input information.

[0453] Optionally, the time domain position of the second CSI report is after the time domain position of the first CSI report.

[0454] Optionally, the first input information and the second input information are in the same packet or different packets; if the first input information and the second input information are in different packets, then the second input information is the first input information in the packet.

[0455] Optionally, the method further includes:

[0456] Send a postponement message, where the postponement message is used to indicate that the first input information and / or the first processed information is a postponed report of the first CSI report.

[0457] Optionally, in S1110, determining the CSI report for the first input information and / or the first processed information includes:

[0458] If the first CSI report in which the first input information and / or the first processed information is located fails to be reported, then no longer send the first input information and / or the first processed information, and send a third CSI report, where the third CSI report contains the third input information and / or the third processed information;

[0459] The third processed information is the information obtained by inputting the third input information into the first AI model or the information obtained by further processing the information obtained by inputting the third input information into the first AI model;

[0460] The first input information and the third input information are different input information.

[0461] Optionally, the time domain position of the third CSI report is after the time domain position of the first CSI report. Optionally, the first input information and the third input information are in different groups, and the third input information is the first input information in the group.

[0462] III. Example description of a communication device

[0463]

Description

[0464] The above mainly introduces the solution of the embodiment of the present application from the perspective of the method side. It can be understood that in order for the terminal device to implement the above functions, it includes the corresponding hardware structure and / or software module for executing each function. Those skilled in the art should easily realize that, combining the units and algorithm steps of each example described in the embodiments disclosed in this article, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a certain function is executed in the way of hardware or computer software driving hardware depends on the specific application and design constraint conditions of the technical solution. Those skilled in the art can use different methods to implement the described function for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0465] The embodiment of the present application can divide the functional units of the terminal device according to the above method example. For example, each functional unit can be divided corresponding to each function, or two or more functions can be integrated into one processing unit. The above integrated unit can be implemented in the form of hardware or in the form of a software program module. It should be noted that the division of units in the embodiment of the present application is illustrative, only a logical function division, and there can be other division methods in actual implementation.

[0466] In the case of adopting an integrated unit, Figure 12 is a block diagram of the functional unit composition of a communication device according to an embodiment of the present application. Among them, the communication device 1200 includes a determination unit 1201.

[0467] Optionally, the determination unit 1201 can be a module unit for receiving and processing signals, information, etc. or determining a monitoring mechanism, and no specific limitation is made thereto.

[0468] Optionally, the communication device 1200 may further include a storage unit for storing the computer program code or instructions executed by the communication device 1200. Among them, the storage unit can be a memory.

[0469] Optionally, the communication device 1200 can be a chip or a chip module.

[0470] Optionally, the determination unit 1201 may be integrated in the communication unit. Wherein, the communication unit may be a communication interface, a transceiver, a transceiver circuit, etc.

[0471] Optionally, the determination unit 1201 may be integrated in the processing unit.

[0472] It should be noted that the processing unit may be a processor or a controller. For example, it may be a baseband processor, a baseband chip, a central processing unit (CPU), 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, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logic blocks, modules, and circuits described in connection with the disclosure of the present application. The processing unit may also be a combination that implements a computing function, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0473] Optionally, the communication device 1200 is used to execute any step performed by the terminal device / chip / chip module, etc. in the above method embodiments.

[0474] In specific implementation, the determination unit 1201 is used to execute any step in the above method embodiments, and when performing actions such as sending, etc., it can optionally call other units to complete the corresponding operations. The following is a detailed description.

[0475] The determination unit 1201 is used to determine the grouping of the first input information of the first AI model, or determine the CSI report of the first input information and / or the first processing information;

[0476] The first AI model is used for CSI compression, and the first processing information is the information obtained by inputting the first input information into the first AI model or the information obtained by further processing the information obtained by inputting the first input information into the first AI model.

[0477] It can be seen that when the AI technology is applied to the CSI framework, in the face of the enhancement of CSI compression and CSI recovery, since the first AI model compresses the input information in sequence, the historical input information can participate in the CSI compression of the current input information. The processed information obtained first can be used as historical processed information to participate in the CSI compression of the current input information, and / or the compressed intermediate information obtained first can be used as historical compressed intermediate information to participate in the CSI compression of the current input information. Therefore, in this embodiment, it is necessary to determine the grouping of the input information. In this way, the first AI model can be used to perform CSI compression on the first input information in a group, and use the historical input information in the group to perform CSI compression on the non-first input information in the group, so as to achieve the enhancement of CSI-related processing.

[0478] Alternatively, when the terminal device may need to report / feedback the input information and / or the processed information. However, some of the input information and / or the processed information that originally needed to be reported / feedback may not be reported due to some factors, resulting in the possible discarding of the input information and / or the processed information. In the face of the input information and / or the processed information that originally needed to be reported but were not reported, this embodiment needs to determine the CSI reporting of the input information and / or the processed information that originally needed to be reported but were not reported, so as to achieve the enhancement of CSI-related processing.

[0479] It should be noted that Figure 12 For the specific implementation of each operation in the above embodiments, reference can be made to the description in the method embodiments shown above, and details will not be elaborated here.

[0480]

Some possible implementation manners

[0481] Optionally, the first AI model is used to perform CSI compression on the first input information and use the first historical information to perform CSI compression on the non-first input information;

[0482] The first historical information includes at least one of historical input information, historical processed information, or historical compressed intermediate information. The historical processed information is the information obtained by inputting the historical input information into the first AI model or the information obtained by further processing the information obtained by inputting the historical input information into the first AI model. The historical compressed intermediate information is the information generated by the first AI model during the compression of the historical input information.

[0483] Optionally, the historical input information in a group participates in the CSI compression of the non-first input information in the same group, and all the input information in a group does not participate in the CSI compression of any input information in other groups.

[0484] Optionally, the first input information is among the input information corresponding to at least one CSI report, and the time-domain behavior of at least one CSI report is at least one of periodic, semi-persistent, or aperiodic; or,

[0485] The first input information is among the input information corresponding to at least one measurement resource;

[0486] The time-domain behavior of at least one measurement resource is at least one of periodic, semi-persistent, or aperiodic.

[0487] Optionally, the input information corresponding to at least one CSI report is input into the first AI model in the order of the time-domain positions of the CSI reports or in the order of the magnitudes of the identification IDs of the CSI reports; or,

[0488] The input information corresponding to at least one measurement resource is input into the first AI model in the order of the time-domain positions of the measurement resources or in the order of the magnitudes of the IDs of the measurement resources.

[0489] Optionally, at least one CSI report includes all aperiodic CSI reports triggered by a DCI.

[0490] Optionally, in determining the grouping of the first input information of the first AI model, the determining unit 1201 is configured to:

[0491] Receive configuration information for configuring a time window;

[0492] Determine the grouping of the first input information of the first AI model according to the time window.

[0493] Optionally, if the first input information is among the input information corresponding to at least one CSI report and the time-domain position of at least one CSI report is within the time window, the input information corresponding to at least one CSI report belongs to the same group.

[0494] Optionally, if the first input information is among the input information corresponding to at least one measurement resource and the time-domain position of at least one measurement resource is within the time window or the time-domain position of at least one CSI report associated with at least one measurement resource is within the time window, the input information corresponding to at least one measurement resource belongs to the same group.

[0495] Optionally, the configuration information includes at least one of length information, start information, or period information;

[0496] The length information is used to configure the length of the time window;

[0497] The start information is used to configure the start position of the time window;

[0498] The period information is used to configure the period of the time window.

[0499] Optionally, the length of the time window is an integer multiple of the CSI reporting period.

[0500] Optionally, the starting position of the time window satisfies the following:

[0501] Or,

[0502]

[0503] where μ represents the subcarrier spacing configuration of the uplink bandwidth part BWP for transmitting the CSI report; represents the number of time slots in a frame under the subcarrier spacing configuration μ; n f represents the system frame number where the CSI report is located; represents the time slot number where the CSI report is located; T offset represents the time slot offset; N represents the maximum number of input information in a packet; T CSI represents the CSI reporting period; represents the system frame number of the physical uplink shared channel PUSCH that sends the first CSI report after DCI activation; represents the time slot number of the PUSCH that sends the first CSI report after DCI activation.

[0504] Optionally, in determining the packet of the first input information of the first AI model, the determining unit 1201 is configured to:

[0505] Receive indication information, where the indication information is used to indicate the packet of the first input information of the first AI model;

[0506] Determine the packet of the first input information according to the indication information.

[0507] Optionally, the communication device 1200 further includes a sending unit;

[0508] The sending unit is configured to send first location information, where the first location information is used to indicate the location of the first input information in the packet or the location of the first processing information in the packet.

[0509] Optionally, in determining the CSI reporting of the first input information and / or the first processing information, the determining unit 1201 is configured to:

[0510] If the first CSI report where the first input information and / or the first processing information is located fails to be reported, then send a second CSI report, where the second CSI report includes the first input information and / or the first processing information, or send the first input information and / or the first processing information at an available time domain position after the time domain position of the first CSI report.

[0511] Optionally, in determining the CSI report of the first input information and / or the first processing information, the determining unit 1201 is configured to:

[0512] If the first CSI report where the first input information and / or the first processing information is located fails to be reported, then send a second CSI report, where the second CSI report includes the first input information and / or the first processing information, and the second input information and / or the second processing information;

[0513] The second processing information is the information obtained by inputting the second input information into the first AI model or the information obtained by further processing the information obtained by inputting the second input information into the first AI model;

[0514] The first input information and the second input information are different input information.

[0515] Optionally, the time domain position of the second CSI report is after the time domain position of the first CSI report.

[0516] Optionally, the first input information and the second input information are in the same packet or different packets; if the first input information and the second input information are in different packets, then the second input information is the first input information in the packet.

[0517] Optionally, the communication device 1200 further includes a sending unit;

[0518] The sending unit is configured to send a postponement message, where the postponement message is used to indicate that the first input information and / or the first processing information is a postponed report of the first CSI report.

[0519] Optionally, in determining the CSI report of the first input information and / or the first processing information, the determining unit 1201 is configured to:

[0520] If the first CSI report where the first input information and / or the first processing information is located fails to be reported, then no longer send the first input information and / or the first processing information, and send a third CSI report, where the third CSI report includes the third input information and / or the third processing information;

[0521] The third processing information is the information obtained by inputting the third input information into the first AI model or the information obtained by further processing the information obtained by inputting the third input information into the first AI model;

[0522] The first input information and the third input information are different input information.

[0523] Optionally, the time domain position of the third CSI report is after the time domain position of the first CSI report.

[0524] Optionally, the first input information and the third input information are in different groups, and the third input information is the first input information in the group.

[0525] IV. Another example illustration of a communication device

[0526]

Description

[0527] The above mainly introduces the solutions of the embodiments of the present application from the perspective of the method side. It can be understood that in order for the network device to implement the above functions, it includes the corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should easily realize that, in combination with the units and algorithm steps of each example described in the embodiments disclosed in this article, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a certain function is executed in the way of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0528] The embodiments of the present application can divide the functional units of the network device according to the above method examples. For example, each functional unit can be divided corresponding to each function, or two or more functions can be integrated into one processing unit. The above integrated unit can be implemented in the form of hardware or in the form of a software program module. It should be noted that the division of units in the embodiments of the present application is illustrative, only a logical function division, and there can be other division methods in actual implementation.

[0529] In the case of adopting an integrated unit, Figure 13 is a block diagram of the functional unit composition of another communication device in the embodiments of the present application. Among them, the communication device 1300 includes a sending unit 1301.

[0530] Optionally, the sending unit 1301 can be a module unit for performing sending processing on signals, information, etc., and no specific limitation is imposed thereon.

[0531] Optionally, the communication device 1300 can further include a receiving unit. Among them, the receiving unit can be a module unit for performing receiving processing on signals, information, etc., and no specific limitation is imposed thereon.

[0532] Optionally, the communication device 1300 can further include a storage unit for storing the computer program code or instructions executed by the communication device 1300. Among them, the storage unit can be a memory.

[0533] Optionally, the communication device 1300 can be a chip or a chip module.

[0534] Optionally, the sending unit 1301 may be integrated in the communication unit. Herein, the communication unit may be a communication interface, a transceiver, a transceiver circuit, etc.

[0535] Optionally, the sending unit 1301 may be integrated in the processing unit.

[0536] It should be noted that the processing unit may be a processor or a controller. For example, it may be a baseband processor, a baseband chip, a central processing unit (CPU), 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, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logical blocks, modules, and circuits described in connection with the disclosure of this application. The processing unit may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0537] Optionally, the communication device 1300 is used to execute any step performed by a chip / chip module / network device, etc. in the above method embodiments.

[0538] In specific implementation, the sending unit 1301 is used to execute any step in the above method embodiments, and when performing actions such as receiving, other units may be selectively called to complete the corresponding operations. Details are described below.

[0539] The sending unit 1301 is used to send configuration information, where the configuration information is used to configure a time window, and the time window is used to determine the grouping of the first input information of the first AI model; or, it is used to send indication information, where the indication information is used to indicate the grouping of the first input information of the first AI model; the first AI model is used for channel state information CSI compression.

[0540] It can be seen that when AI technology is applied to the CSI framework, in the face of the enhancement of CSI compression and CSI recovery, since the first AI model will perform CSI compression on the input information in sequence, the historical input information can participate in the CSI compression of the current input information. The processed information obtained first can be used as historical processed information to participate in the CSI compression of the current input information, and / or the compressed intermediate information obtained first can be used as historical compressed intermediate information to participate in the CSI compression of the current input information. Therefore, in this embodiment, it is necessary to determine the grouping of the input information. In this way, the first AI model can be used to perform CSI compression on the first input information in a group, and use the historical input information in the group to perform CSI compression on the non-first input information in the group, so as to achieve the enhancement of CSI-related processing.

[0541] It should be noted that Figure 13 For the specific implementation of each operation in the above embodiment, reference can be made to the description in the method embodiment shown above, and details will not be repeated here.

[0542]

Some possible implementation manners

[0543] Optionally, the first AI model is used to perform CSI compression on the first input information, and use the first historical information to perform CSI compression on the non-first input information;

[0544] The first historical information includes at least one of historical input information, historical processed information, or historical compressed intermediate information. The historical processed information is the information obtained by inputting the historical input information into the first AI model or the information obtained by further processing the information obtained by inputting the historical input information into the first AI model. The historical compressed intermediate information is the information generated by the first AI model during the compression of the historical input information.

[0545] Optionally, the historical input information in a group participates in the CSI compression of the non-first input information in the same group, and all the input information in a group does not participate in the CSI compression of any input information in other groups.

[0546] Optionally, if the first input information is among the input information corresponding to at least one CSI report and the time domain position of at least one CSI report is within the time window, then the input information corresponding to at least one CSI report belongs to the same group; or,

[0547] If the first input information is among the input information corresponding to at least one measurement resource and the time domain position of at least one measurement resource is within the time window or the time domain position of the CSI report associated with at least one measurement resource is within the time window, then the input information corresponding to at least one measurement resource belongs to the same group.

[0548] Optionally, the configuration information includes at least one of length information, start information, or period information;

[0549] The length information is used to configure the length of the time window;

[0550] The start information is used to configure the start position of the time window;

[0551] The period information is used to configure the period of the time window.

[0552] Optionally, the length of the time window is an integer multiple of the CSI reporting period; or,

[0553] The start position of the time window satisfies the following:

[0554] Or,

[0555]

[0556] where μ represents the subcarrier spacing configuration of the uplink bandwidth part BWP for transmitting CSI reports; represents the number of time slots in a frame under the subcarrier spacing configuration μ; n f represents the system frame number where the CSI report is located; represents the time slot number where the CSI report is located; T offset represents the time slot offset; N represents the maximum number of input messages in a packet; T CSI represents the CSI reporting period; represents the system frame number of the physical uplink shared channel PUSCH that sends the first CSI report after DCI activation; represents the time slot number of the PUSCH that sends the first CSI report after DCI activation.

[0557] Optionally, the communication device 1300 further includes a receiving unit;

[0558] The receiving unit is configured to receive first position information, where the first position information is used to indicate the position of the first input information in the packet or the position of the first processing information in the packet.

[0559] V. Another Example Illustration of a Communication Device

[0560]

Description

[0561] In the case of adopting an integrated unit, Figure 14 is a functional unit composition block diagram of another communication device according to an embodiment of the present application. Among them, the communication device 1400 includes a receiving unit 1401.

[0562] Optionally, the receiving unit 1401 may be a module unit for receiving and processing signals, information, etc., and no specific limitation is imposed thereon.

[0563] Optionally, the communication device 1400 may further include a receiving unit. The receiving unit may be a module unit for receiving and processing signals, information, etc., and no specific limitation is imposed thereon.

[0564] Optionally, the communication device 1400 may further include a storage unit for storing the computer program code or instructions executed by the communication device 1400. The storage unit may be a memory.

[0565] Optionally, the communication device 1400 may be a chip or a chip module.

[0566] Optionally, the receiving unit 1401 may be integrated in the communication unit. The communication unit may be a communication interface, a transceiver, a transceiver circuit, etc.

[0567] Optionally, the receiving unit 1401 may be integrated in the processing unit.

[0568] It should be noted that the processing unit may be a processor or a controller. For example, it may be a baseband processor, a baseband chip, a central processing unit (CPU), 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, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logical blocks, modules, and circuits described in connection with the disclosure of the present application. The processing unit may also be a combination that implements a computing function, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0569] Optionally, the communication device 1400 is used to execute any step performed by a chip / chip module / network device, etc. in the above method embodiment.

[0570] In specific implementation, the receiving unit 1401 is used to execute any step in the above method embodiment, and when performing actions such as receiving, other units may be selectively called to complete the corresponding operations. Details are described below.

[0571] A receiving unit 1401, configured to receive a second CSI report if a first CSI report in which first input information and / or first processing information of a first AI model is located fails to be reported, where the second CSI report includes the first input information and / or the first processing information, or the second CSI report includes the first input information and / or the first processing information and second input information and / or second processing information, and the first input information and the second input information are different input information; or,

[0572] if a first CSI report in which first input information and / or first processing information of a first AI model is located fails to be reported, receive the first input information and / or the first processing information at an available time domain position after the time domain position of the first CSI report; or,

[0573] if a first CSI report in which first input information and / or first processing information of a first AI model is located fails to be reported, receive a third CSI report, where the third CSI report includes third input information and / or third processing information, and the first input information and the third input information are different input information;

[0574] The first AI model is used for CSI compression, and the first processing information is information obtained by inputting the first input information into the first AI model or information obtained by further processing the information obtained by inputting the first input information into the first AI model.

[0575] It can be seen that when it was originally hoped to report the first input information and / or the first processing information through the first CSI report, since the first CSI report fails to be reported, this causes the first input information and / or the first processing information that originally needed to be reported to fail to be reported. Therefore, the second CSI report is used to carry the first input information and / or the first processing information, so as to continue to report the first input information and / or the first processing information that originally needed to be reported but was not reported.

[0576] Or, when the first input information and / or the first processing information that originally needed to be reported fails to be reported, the second CSI report is used to carry the first input information and / or the first processing information and the second input information and / or the second processing information, so as to realize both delaying the reporting of the first input information and / or the first processing information and reporting the second input information and / or the second processing information that the second CSI report needs to report through the second CSI report, and avoiding the loss of the second input information and / or the second processing information.

[0577] Or, when the first input information and / or the first processing information that originally needed to be reported fails to be reported, receive the first input information and / or the first processing information at an available time domain position after the time domain position of the first CSI report, so as to realize continuing to report / delaying the reporting of the first input information and / or the first processing information.

[0578] Alternatively, when the first input information and / or the first processing information that originally needed to be reported fails to be reported, the first input information and / or the first processing information is received at an available time domain position after the time domain position of the first CSI report, so as to continue reporting / postpone reporting the first input information and / or the first processing information.

[0579] It should be noted that Figure 14 For the specific implementation of each operation in the above embodiments, reference can be made to the description in the method embodiments shown above, and details will not be elaborated here.

[0580]

Some possible implementation manners

[0581] Optionally, the time domain position of the second CSI report is after the time domain position of the first CSI report.

[0582] Optionally, the time domain position of the third CSI report is after the time domain position of the first CSI report.

[0583] Optionally, the first input information and the second input information are in the same packet or different packets; if the first input information and the second input information are in different packets, the second input information is the first input information in the packet; or

[0584] The first input information and the third input information are in different packets, and the third input information is the first input information in the packet.

[0585] Optionally, the receiving unit 1401 is further configured to receive postponement information, where the postponement information is used to indicate that the first input information and / or the first processing information is a postponed report of the first CSI report.

[0586] VI. Example description of a terminal device

[0587] Please refer to Figure 15 , Figure 15 which is a schematic structural diagram of a terminal device according to an embodiment of the present application. Among them, the terminal device 1500 may include a processor 1510, a memory 1520, and a communication bus for connecting the processor 1510 and the memory 1520.

[0588] Optionally, the memory 1520 includes, but is not limited to, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), or a compact disc read-only memory (CD-ROM). The memory 1520 is used to store the program code executed by the terminal device 1500 and the transmitted data.

[0589] Optionally, the terminal device 1500 further includes a communication interface for receiving and sending data.

[0590] Optionally, the terminal device 1500 may be the first terminal device described above.

[0591] Optionally, the processor 1510 may be one or more CPUs. When the processor 1510 is a single CPU, the CPU may be a single-core CPU or a multi-core CPU.

[0592] Optionally, the processor 1510 may be a baseband chip, a chip, a CPU, a general-purpose processor, a DSP, an ASIC, an FPGA, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof.

[0593] In specific implementation, the processor 1510 in the terminal device 1500 is used to execute the computer program or instruction 1521 stored in the memory 1520 and perform the following operations:

[0594] Determine the grouping of the first input information of the first AI model, or determine the CSI report of the first input information and / or the first processing information;

[0595] The first AI model is used for CSI compression, and the first processing information is the information obtained by inputting the first input information into the first AI model or the information obtained by further processing the information obtained by inputting the first input information into the first AI model.

[0596] It can be seen that when the AI technology is applied to the CSI framework, in the face of the enhancement of CSI compression and CSI recovery, since the first AI model compresses the input information in sequence, the historical input information can participate in the CSI compression of the current input information. The processed information obtained first can be used as historical processed information to participate in the CSI compression of the current input information, and / or the compressed intermediate information obtained first can be used as historical compressed intermediate information to participate in the CSI compression of the current input information. Therefore, in this embodiment, it is necessary to determine the grouping of the input information. In this way, the first AI model can be used to perform CSI compression on the first input information in a group, and use the historical input information in the group to perform CSI compression on the non-first input information in the group, so as to achieve the enhancement of CSI-related processing.

[0597] Alternatively, when the terminal device needs to report / feedback the input information and / or the processed information, since some of the input information and / or the processed information that originally needed to be reported / feedback are not reported due to some factors, there will be a situation of discarding the input information and / or the processed information. In this way, in the face of the input information and / or the processed information that originally needed to be reported but were not reported, this embodiment can determine the CSI reporting of the input information and / or the processed information that originally needed to be reported but were not reported, so as to achieve the enhancement of CSI-related processing.

[0598] It should be noted that the specific implementation of each operation can adopt the corresponding description of the method embodiment shown above. The terminal device 1500 can be used to execute the method embodiment of the present application, and details are not described herein again.

[0599] VII. Example Illustration of a Network Device

[0600] Please refer to Figure 16 , Figure 16 which is a schematic structural diagram of a network device provided by an embodiment of the present application. Among them, the network device 1600 includes a processor 1610, a memory 1620, and a communication bus for connecting the processor 1610 and the memory 1620.

[0601] Optionally, the memory 1620 includes but is not limited to RAM, ROM, EPROM, or CD-ROM, and the memory 1620 is used to store relevant instructions and data.

[0602] Optionally, the network device 1600 further includes a communication interface, which is used to receive and send data.

[0603] Optionally, the processor 1610 can be one or more CPUs. In the case where the processor 1610 is a single CPU, the CPU can be a single-core CPU or a multi-core CPU.

[0604] Optionally, the processor 1610 may be a baseband chip, a chip, a CPU, a general-purpose processor, a DSP, an ASIC, an FPGA, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof.

[0605] Optionally, the processor 1610 in the network device 1600 is configured to execute the computer program or instruction 1621 stored in the memory 1620 and perform the following operations:

[0606] Send configuration information for configuring a time window, where the time window is used to determine the grouping of the first input information of the artificial intelligence first AI model; or, send indication information for indicating the grouping of the first input information of the first AI model; the first AI model is used for channel state information CSI compression.

[0607] It can be seen that when AI technology is applied to the CSI framework, in the face of the enhancement of CSI compression and CSI recovery, since the first AI model will perform CSI compression on the input information in sequence, the historical input information can participate in the CSI compression of the current input information, the processed information obtained first can be used as historical processed information to participate in the CSI compression of the current input information, and / or the compressed intermediate information obtained first can be used as historical compressed intermediate information to participate in the CSI compression of the current input information. Therefore, this embodiment needs to determine the grouping of the input information. In this way, the first AI model can be used to perform CSI compression on the first input information in a group and use the historical input information in the group to perform CSI compression on the non-first input information in the group, thereby achieving the enhancement of CSI-related processing.

[0608] It should be noted that the specific implementation of each operation can adopt the corresponding description of the method embodiment shown above. The network device 1600 can be used to execute the method embodiment of the present application, which will not be elaborated herein.

[0609] VIII. Another Example Illustration of a Network Device

[0610] Please refer to Figure 17 , Figure 17 which is a schematic structural diagram of a network device provided by an embodiment of the present application. Among them, the network device 1700 includes a processor 1710, a memory 1720, and a communication bus for connecting the processor 1710 and the memory 1720.

[0611] Optionally, the memory 1720 includes but is not limited to RAM, ROM, EPROM, or CD-ROM, and the memory 1720 is used to store relevant instructions and data.

[0612] Optionally, the network device 1700 further includes a communication interface for receiving and sending data.

[0613] Optionally, the processor 1710 may be one or more CPUs. When the processor 1710 is a single CPU, the CPU may be a single-core CPU or a multi-core CPU.

[0614] Optionally, the processor 1710 may be a baseband chip, a chip, a CPU, a general-purpose processor, a DSP, an ASIC, an FPGA, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof.

[0615] Optionally, the processor 1710 in the network device 1700 is configured to execute the computer program or instruction 1721 stored in the memory 1720 and perform the following operations:

[0616] If the first CSI report in which the first input information and / or the first processing information of the first AI model are located fails to be reported, then receive a second CSI report. The second CSI report includes the first input information and / or the first processing information, or the second CSI report includes the first input information and / or the first processing information and the second input information and / or the second processing information. The time domain position of the second CSI report is after the time domain position of the first CSI report, and the first input information and the second input information are different input information; or,

[0617] If the first CSI report in which the first input information and / or the first processing information of the first AI model are located fails to be reported, then receive the first input information and / or the first processing information at an available time domain position after the time domain position of the first CSI report; or,

[0618] If the first CSI report in which the first input information and / or the first processing information of the first AI model are located fails to be reported, then receive a third CSI report. The third CSI report includes the third input information and / or the third processing information. The time domain position of the third CSI report is after the time domain position of the first CSI report; the first input information and the third input information are different input information;

[0619] The first AI model is used for CSI compression, and the first processing information is the information obtained by inputting the first input information into the first AI model or the information obtained by further processing the information obtained by inputting the first input information into the first AI model.

[0620] It can be seen that when it was originally hoped to report the first input information and / or the first processing information through the first CSI report, since the first CSI report fails to be reported, this results in the first input information and / or the first processing information that originally needed to be reported not being reported. Therefore, the second CSI report is used to carry the first input information and / or the first processing information, so as to continue reporting the first input information and / or the first processing information that originally needed to be reported but was not reported.

[0621] Alternatively, when the first input information and / or the first processing information that originally needed to be reported fails to be reported, the second CSI report is used to carry the first input information and / or the first processing information, as well as the second input information and / or the second processing information, so that the second CSI report can be used to delay the reporting of the first input information and / or the first processing information and report the second input information and / or the second processing information required to be reported by the second CSI report, avoiding the loss of the second input information and / or the second processing information.

[0622] Alternatively, when the first input information and / or the first processing information that originally needed to be reported fails to be reported, the first input information and / or the first processing information is received at an available time domain position after the time domain position of the first CSI report, so as to continue to report / delay the reporting of the first input information and / or the first processing information.

[0623] Alternatively, when the first input information and / or the first processing information that originally needed to be reported fails to be reported, the first input information and / or the first processing information is no longer reported, but the third input information and / or the third processing information is reported through the third CSI report, so as to report new input information and / or new processing information and avoid the loss of the new input information and / or the new processing information.

[0624] It should be noted that the specific implementation of each operation can adopt the corresponding description of the method embodiments shown above. The network device 1700 can be used to execute the method embodiments of the present application, which will not be elaborated herein.

[0625] IX. Other Related Example Descriptions

[0626] Optionally, the above method embodiments can be applied to a terminal device or in a terminal device. That is to say, the execution subject of the above method embodiments can be a terminal device, a chip, a chip module or a module, etc., and no specific limitation is made thereto.

[0627] Optionally, the above method embodiments can be applied to a network device or in a network device. That is to say, the execution subject of the above method embodiments can be a network device, a chip, a chip module or a module, etc., and no specific limitation is made thereto.

[0628] The embodiments of the present application further provide a chip, including a processor, a memory, and a computer program or instruction stored in the memory. The processor executes the computer program or instruction to implement the steps described in the above method embodiments.

[0629] An embodiment of the present application further provides a chip module, including a transceiver component and a chip. The chip includes a processor, a memory, and a computer program or instruction stored on the memory. Wherein, the processor executes the computer program or instruction to implement the steps described in the above method embodiment.

[0630] An embodiment of the present application further provides a computer-readable storage medium, which stores a computer program or instruction. When the computer program or instruction is executed, it implements the steps described in the above method embodiment.

[0631] An embodiment of the present application further provides a computer program product, including a computer program or instruction. When the computer program or instruction is executed, it implements the steps described in the above method embodiment.

[0632] An embodiment of the present application further provides a communication system, including the above terminal device and the above network device.

[0633] It should be noted that, for the above embodiments, for the sake of simple description, they are all expressed as a series of action combinations. Those skilled in the art should know that the present application is not limited by the described action sequence, because some steps in the embodiments of the present application can be in other sequences or performed simultaneously. In addition, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions, steps, modules or units involved are not necessarily essential to the embodiments of the present application.

[0634] In the above embodiments, the embodiments of the present application have different emphases in the description of each embodiment. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0635] The steps of the method or algorithm described in the embodiments of the present application can be implemented in a hardware manner, or can be implemented by a processor executing software instructions. The software instructions can be composed of corresponding software modules. The software modules can be stored in RAM, flash memory, ROM, EPROM, electrically erasable programmable read-only memory (EEPROM), registers, hard disk, removable hard disk, CD-ROM or any other form of storage medium well known in the art. An exemplary storage medium is coupled to the processor, so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an ASIC. In addition, the ASIC can be located in a terminal device or a management device. Of course, the processor and the storage medium can also exist as discrete components in a terminal device or a management device.

[0636] Those skilled in the art should be able to realize that in one or more of the above examples, the functions described in the embodiments of the present application can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as coaxial cable, fiber optic, digital subscriber line (DSL)) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a digital video disc (DVD)), or a semiconductor medium (such as a solid state disk (SSD)), etc.

[0637] Each device and product described in the above embodiments includes various modules / units, which can be software modules / units, hardware modules / units, or can be partly software modules / units and partly hardware modules / units. For example, for each device and product applied to or integrated into a chip, each of the modules / units it includes can be implemented in the form of hardware such as circuits. Or, at least some of the modules / units can be implemented in the form of software programs that run on the processor integrated inside the chip, and the remaining (if any) part of the modules / units can be implemented in the form of hardware such as circuits. For each device and product applied to or integrated into a chip module, each of the modules / units it includes can be implemented in the form of hardware such as circuits. Different modules / units can be located in the same component (such as a chip, a circuit module, etc.) or different components of the chip module. Or, at least some of the modules / units can be implemented in the form of software programs that run on the processor integrated inside the chip module, and the remaining (if any) part of the modules / units can be implemented in the form of hardware such as circuits. For each device and product applied to or integrated into a terminal device, each of the modules / units it includes can be implemented in the form of hardware such as circuits. Different modules / units can be located in the same component (such as a chip, a circuit module, etc.) or different components inside the terminal device. Or, at least some of the modules / units can be implemented in the form of software programs that run on the processor integrated inside the terminal device, and the remaining (if any) part of the modules / units can be implemented in the form of hardware such as circuits.

[0638] The specific embodiments described above further elaborate on the purpose, technical solutions, and beneficial effects of the embodiments of the present application. It should be understood that the above description is only the specific embodiments of the embodiments of the present application and is not used to limit the protection scope of the embodiments of the present application. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solutions of the embodiments of the present application should be included in the protection scope of the embodiments of the present application.

Claims

1. A communication method, characterized in that, Including: Determining a grouping of first input information of an artificial intelligence first AI model, or determining channel state information CSI reporting of the first input information and / or first processing information; The first AI model is used for CSI compression, and the first processing information is information obtained by inputting the first input information into the first AI model or information obtained after further processing of the information obtained by inputting the first input information into the first AI model.

2. The method according to claim 1, characterized in that, The first AI model is used for performing CSI compression on the first input information and using first historical information to perform CSI compression on non-first input information; The first historical information includes at least one of historical input information, historical processing information, or historical compression intermediate information. The historical processing information is information obtained by inputting the historical input information into the first AI model or information obtained after further processing of the information obtained by inputting the historical input information into the first AI model. The historical compression intermediate information is information generated by the first AI model during the process of compressing the historical input information.

3. The method according to claim 1, characterized in that, The historical input information within a grouping participates in the CSI compression of non-first input information within the same grouping, and all input information in a grouping does not participate in the CSI compression of any input information in other groupings.

4. The method according to claim 1, characterized in that The first input information is among the input information corresponding to at least one CSI report, and the time domain behavior of the at least one CSI report is at least one of periodic, semi-persistent, or aperiodic; Or, The first input information is among the input information corresponding to at least one measurement resource; The time domain behavior of the at least one measurement resource is at least one of periodic, semi-persistent, or aperiodic.

5. The method according to claim 4, wherein The input information corresponding to the at least one CSI report is input into the first AI model in the order of the time domain positions of the CSI reports or in the order of the magnitudes of the identification IDs of the CSI reports; or, The input information corresponding to the at least one measurement resource is input into the first AI model in the order of the time domain positions of the measurement resources or in the order of the magnitudes of the IDs of the measurement resources.

6. The method according to any one of claims 1-5, characterized in that, The determining the grouping of the first input information of the first AI model includes: Receiving configuration information, where the configuration information is used to configure a time window; Determining the grouping of the first input information of the first AI model according to the time window.

7. The method according to claim 6, wherein If the first input information is among the input information corresponding to at least one CSI report and the time domain position of the at least one CSI report is within the time window, then The input information corresponding to the at least one CSI report belongs to the same grouping.

8. The method according to claim 6, wherein If the first input information is among the input information corresponding to at least one measurement resource and the time domain position of the at least one measurement resource is within the time window or the time domain position of the CSI report associated with the at least one measurement resource is within the time window, then The input information corresponding to the at least one measurement resource belongs to the same grouping.

9. The method according to any one of claims 6 - 8, characterized in that, The configuration information includes at least one of length information, start information, or period information; The length information is used to configure the length of the time window; The starting information is used to configure the starting position of the time window; The period information is used to configure the period of the time window.

10. The method according to any one of claims 1-5, characterized in that, The grouping of the first input information for determining the first AI model includes: Receiving indication information, where the indication information is used to indicate the grouping of the first input information of the first AI model; Determining the grouping of the first input information according to the indication information.

11. The method according to any one of claims 6 - 10, characterized in that, It further includes: Sending first position information, where the first position information is used to indicate the position of the first input information in the grouping or the position of the first processing information in the grouping.

12. The method according to any one of claims 1-5, characterized in that, The CSI reporting for determining the first input information and / or the first processing information includes: If the first CSI report where the first input information and / or the first processing information is located fails to be reported, then sending a second CSI report, where the second CSI report includes the first input information and / or the first processing information; or, If the first CSI report where the first input information and / or the first processing information is located fails to be reported, then sending the first input information and / or the first processing information at an available time domain position after the time domain position of the first CSI report.

13. The method according to any one of claims 1-5, characterized in that, The CSI reporting for determining the first input information and / or the first processing information includes: If the first CSI report where the first input information and / or the first processing information is located fails to be reported, then sending a second CSI report, where the second CSI report includes the first input information and / or the first processing information and the second input information and / or the second processing information; The second processing information is the information obtained by inputting the second input information into the first AI model or the information obtained by further processing the information obtained by inputting the second input information into the first AI model; The first input information and the second input information are different input information.

14. The method according to claim 13, wherein The first input information and the second input information are in the same grouping or different groupings; if the first input information and the second input information are in different groupings, then the second input information is the first input information in the grouping.

15. The method according to any one of claims 12 - 14, characterized in that It further includes: Sending postponement information, where the postponement information is used to indicate that the first input information and / or the first processing information is a postponed report of the first CSI report.

16. The method according to any one of claims 1-5, characterized in that, The CSI reporting for determining the first input information and / or the first processing information includes: If the first CSI report where the first input information and / or the first processing information is located fails to be reported, then no longer sending the first input information and / or the first processing information, and sending a third CSI report, where the third CSI report includes the third input information and / or the third processing information; The third processing information is the information obtained by inputting the third input information into the first AI model or the information obtained by further processing the information obtained by inputting the third input information into the first AI model; The first input information and the third input information are different input information.

17. The method according to claim 16, wherein The first input information and the third input information are in different groupings, and the third input information is the first input information in the grouping.

18. A communication method, characterized in that, It includes: Send configuration information, where the configuration information is used to configure a time window, and the time window is used to determine the grouping of the first input information of the first AI model; Or, Send indication information, where the indication information is used to indicate the grouping of the first input information of the first AI model; The first AI model is used for channel state information (CSI) compression.

19. The method according to claim 18, wherein The first AI model is used to perform CSI compression on the first input information and use first historical information to perform CSI compression on non-first input information; The first historical information includes at least one of historical input information, historical processing information, or historical compression intermediate information. The historical processing information is the information obtained by inputting the historical input information into the first AI model or the information obtained by further processing the information obtained by inputting the historical input information into the first AI model. The historical compression intermediate information is the information generated by the first AI model during the process of compressing the historical input information.

20. The method according to claim 18, wherein The historical input information within a grouping participates in the CSI compression of the non-first input information within the same grouping, and all the input information in a grouping does not participate in the CSI compression of any input information in other groupings.

21. The method according to any one of claims 18 - 20, characterized in that, If the first input information of the first AI model is among the input information corresponding to at least one CSI report and the time domain position of the at least one CSI report is within the time window, then the input information corresponding to the at least one CSI report belongs to the same grouping; or, If the first input information is among the input information corresponding to at least one measurement resource and the time domain position of the at least one measurement resource is within the time window or the time domain position of the CSI report associated with the at least one measurement resource is within the time window, then the input information corresponding to the at least one measurement resource belongs to the same grouping.

22. The method according to any one of claims 18-21, characterized in that, The configuration information includes at least one of length information, start information, or period information; The length information is used to configure the length of the time window; The start information is used to configure the start position of the time window; The period information is used to configure the period of the time window.

23. A communication method, characterized in that, Include: If the first CSI report where the first input information and / or the first processing information of the first AI model is located fails to be reported, then receive a second CSI report, where the second CSI report contains the first input information and / or the first processing information, or the second CSI report contains the first input information and / or the first processing information and the second input information and / or the second processing information, and the first input information and the second input information are different input information; Or, If the first CSI report where the first input information and / or the first processing information of the first AI model is located fails to be reported, then receive the first input information and / or the first processing information at an available time domain position after the time domain position of the first CSI report; Or, If the first CSI report in which the first input information and / or the first processing information of the first AI model is located fails to be reported, then a third CSI report is received, where the third CSI report includes third input information and / or third processing information, and the first input information and the third input information are different input information; The first AI model is used for CSI compression. The first processing information is the information obtained by inputting the first input information into the first AI model or the information obtained after further processing the information obtained by inputting the first input information into the first AI model. The second processing information is the information obtained by inputting the second input information into the first AI model or the information obtained after further processing the information obtained by inputting the second input information into the first AI model. The third processing information is the information obtained by inputting the third input information into the first AI model or the information obtained after further processing the information obtained by inputting the third input information into the first AI model.

24. The method according to claim 23, wherein The first input information and the second input information are in the same packet or different packets; if the first input information and the second input information are in different packets, then the second input information is the first input information in the packet; Or, The first input information and the third input information are in different packets, and the third input information is the first input information in the packet.

25. The method according to claim 23 or 24, characterized in that, It further includes: Receiving delay information, where the delay information is used to indicate that the first input information and / or the first processing information is a delayed report of the first CSI report.

26. A communication device, characterized in that, It includes: A determination unit, which is used for grouping the first input information of the first AI model, or determining the channel state information CSI report of the first input information and / or the first processing information; The first AI model is used for CSI compression. The first processing information is the information obtained by inputting the first input information into the first AI model or the information obtained after further processing the information obtained by inputting the first input information into the first AI model.

27. A communication device, characterized in that, It includes: A sending unit, which is used to send configuration information, where the configuration information is used to configure a time window, and the time window is used to determine the grouping of the first input information of the first AI model; or, used to send indication information, where the indication information is used to indicate the grouping of the first input information of the first AI model; The first AI model is used for channel state information CSI compression.

28. A communication device, characterized in that, It includes a receiving unit, and the receiving unit is used for: If the first CSI report in which the first input information and / or the first processing information of the first AI model is located fails to be reported, then a second CSI report is received, where the second CSI report includes the first input information and / or the first processing information, or the second CSI report includes the first input information and / or the first processing information and the second input information and / or the second processing information, and the first input information and the second input information are different input information; Or, If the first CSI report in which the first input information and / or the first processing information of the first AI model is located fails to be reported, then receive the first input information and / or the first processing information at an available time-domain position after the time-domain position of the first CSI report; Or, If the first CSI report in which the first input information and / or the first processing information of the first AI model is located fails to be reported, then receive a third CSI report, where the third CSI report includes third input information and / or third processing information, and the first input information and the third input information are different input information; The first AI model is used for CSI compression. The first processing information is the information obtained by inputting the first input information into the first AI model or the information obtained by further processing the information obtained by inputting the first input information into the first AI model. The second processing information is the information obtained by inputting the second input information into the first AI model or the information obtained by further processing the information obtained by inputting the second input information into the first AI model. The third processing information is the information obtained by inputting the third input information into the first AI model or the information obtained by further processing the information obtained by inputting the third input information into the first AI model.

29. A terminal device, comprising a processor, a memory, and a computer program or instruction stored on the memory, characterized in that, The processor executes the computer program or instruction to implement the steps of the method according to any one of claims 1-17.

30. A network device, comprising a processor, a memory, and a computer program or instruction stored on the memory, characterized in that, The processor executes the computer program or instruction to implement the steps of the method according to any one of claims 18-22 or 23-25.

31. A chip, comprising a processor and a communication interface, characterized in that, The processor executes the steps of the method according to any one of claims 1-25 through the communication interface.

32. A computer-readable storage medium, characterized in that, It stores a computer program or instruction, and when the computer program or instruction is executed, the steps of the method according to any one of claims 1-25 are executed.

Images