Information transmission and receiving method and apparatus, and information transmission method and apparatus

Abstract

Provided in the embodiments of the present application are an information transmission and receiving method and apparatus, and an information transmission method and apparatus. The information transmission and receiving method comprises: receiving first configuration information from a network device, wherein the first configuration information is related to AI / ML model selection of a terminal device; and reporting channel state information (CSI) feedback information to the network device, wherein the CSI feedback information is generated by the terminal device by means of using the AI / ML model selected on the basis of the first configuration information.

Description

Information transceiving method and device, information sending method and device TECHNICAL FIELD

[0001] Embodiments of the present application relate to the field of communication technology. BACKGROUND

[0002] With the development of Artificial Intelligence / Machine Learning (AI / ML) technology and the development of hardware such as Graphics Processing Unit (GPU) and Neural Processing Unit (NPU), AI / ML has made great progress in Computer View (CV), Natural Language Processing (NLP), etc. In view of the successful application of AI / ML technology in the fields of CV and NLP, mobile communication researchers have also gradually begun to focus on the application of AI / ML technology in communication systems.

[0003] At present, the design of mobile communication systems based on AI / ML technology has become a research hotspot and focus of 5G-A (5th Generation Mobile Communication Advanced) and 6G (6th Generation Mobile Communication).

[0004] In the Release 19 stage, the standardization of AI / ML models by the 3GPP standard organization is still under discussion.

[0005] It should be noted that the above introduction to the technical background is only to facilitate the clear and complete description of the technical solutions of the present application, and to facilitate the understanding of those skilled in the art. The above technical solutions cannot be considered as known to those skilled in the art merely because they are described in the background section of the present application.

[0006] SUMMARY

[0007] The inventors find that for the Case 2 use case focused on Release 19, i.e., history channel state information (CSI) assisted CSI compression, the terminal device and / or the network device utilizes the CSI information at a historical time to help compress the CSI information at a current time. When there is strong correlation in the CSI in the time domain, the CSI at the historical time helps the compression and / or recovery of the CSI at the current time. Therefore, the Case 2 use case can greatly reduce the feedback overhead or improve the feedback accuracy.

[0008] However, for the Case 2 use case, each company designs its own AI / ML model for research or performance evaluation, and there is no unified model or model structure, which affects the multi-vendor / inter-vendor collaboration of the bilateral model and increases the complexity of vendor collaboration. Currently, the standardization of the AI / ML model for the Case 2 use case is still under discussion, and there is no solution in the protocol to solve the problem.

[0009] To address at least one of the above problems, embodiments of the present disclosure provide an information receiving method and device, and an information sending method and device.

[0010] According to an aspect of embodiments of the present disclosure, an information receiving method and device are provided, configured in a terminal device, and the device includes:

[0011] a receiver configured to receive first configuration information from a network device, the first configuration information being related to AI / ML model selection of the terminal device;

[0012] a transmitter configured to report channel state information (CSI) feedback information to the network device, the CSI feedback information being generated by the terminal device using an AI / ML model selected according to the first configuration information.

[0013] According to another aspect of embodiments of the present disclosure, an information sending device is provided, configured in a network device, and the device includes:

[0014] a transmitter configured to send first configuration information to a terminal device, the first configuration information being related to AI / ML model selection of the terminal device.

[0015] One of the beneficial effects of the embodiments of the present application is that the terminal device performs function and / or model selection according to the configuration information related to the function and / or model selection of the terminal device from the network device, so that the terminal device can select the AI / ML model according to the configuration of the network device, thereby standardizing the AI / ML model used on the terminal device side and the network device side, solving the multi-vendor / inter-vendor collaboration problem of the bilateral model, reducing the complexity of vendor collaboration, and thereby solving the problems in the prior art.

[0016] Specific embodiments of the application are disclosed in detail in the following description and claims, indicating the ways in which the principles of the application can be employed. It should be understood that the embodiments of the application are not limited in scope to the specific embodiments described herein. Embodiments of the application include many alternatives, modifications and equivalents.

[0017] Features described and / or illustrated with respect to one implementation can be used in one or more other implementations in the same or similar way, in combination with or in place of features in other implementations.

[0018] It should be emphasized that the term "comprises / comprising" when used in this text is taken to mean the presence of stated features, integers, steps or components, but not to the exclusion of one or more other features, integers, steps or components, or groups thereof. BRIEF DESCRIPTION OF DRAWINGS

[0019] Elements and features depicted in one drawing or embodiment of the application can be combined with elements and features depicted in one or more other drawings or embodiments. Also, in the drawings, like reference numerals designate corresponding parts throughout the several views, and can be used to designate corresponding parts in more than one embodiment.

[0020] FIG. 1 is a schematic diagram of a communication system according to an embodiment of the present application;

[0021] FIG. 2 is a schematic diagram of terminal device and network device interaction of Case 2 according to an embodiment of the present application;

[0022] FIG. 3 is a schematic diagram of an information transceiving method according to an embodiment of the present application;

[0023] FIG. 4 is another schematic diagram of an information transceiving method according to an embodiment of the present application;

[0024] FIG. 5 is another schematic diagram of an information transceiving method according to an embodiment of the present application;

[0025] FIG. 6 is a schematic diagram of an information transmitting method according to an embodiment of the present application;

[0026] FIG. 7 is a schematic diagram of an information transceiving device according to an embodiment of the present application;

[0027] FIG. 8 is a schematic diagram of an information transmitting device according to an embodiment of the present application;

[0028] FIG. 9 is a schematic diagram of a network device according to an embodiment of the present application;

[0029] FIG. 10 is a schematic diagram of a terminal device according to an embodiment of the present application. DETAILED DESCRIPTION

[0030] The foregoing and other features of the present application will become apparent to those skilled in the art upon consideration of the following description of specific embodiments of the present application, which are not intended to be limiting. Various modifications, variations, and alternatives to those embodiments described herein will be apparent to those skilled in the art from the teachings of the present application, which are to be understood as including all such modifications, variations, and alternatives within the scope of the appended claims.

[0031] In the embodiments of the present application, the terms "first", "second", and the like are used to distinguish different elements from one another, but do not indicate spatial arrangement or temporal order of the elements, and the elements should not be limited by these terms. The term "and / or" includes any one and all combinations of the associated listed terms. The terms "comprise", "include", "have", and the like are meant to be interpreted inclusively, not exclusively, so that they indicate the presence of the stated features, elements, components, or components, but do not exclude the presence or addition of one or more other features, elements, components, or components.

[0032] In the embodiments of the present application, the singular forms "a", "an", and "the" include the plural forms, and should be broadly understood as "one" or "one type" rather than the meaning of "one", and in addition, the term "said" should be understood as including both singular and plural forms, unless the context clearly indicates otherwise. In addition, the term "according to" should be understood as "at least partially according to", and the term "based on" should be understood as "at least partially based on", unless the context clearly indicates otherwise.

[0033] In the embodiments of the present application, the term "communication network" or "wireless communication network" can refer to a network that complies with any communication standard, such as New Radio (NR), Long Term Evolution (LTE), LTE-Advanced (LTE-A), Wideband Code Division Multiple Access (WCDMA), High-Speed Packet Access (HSPA), and the like.

[0034] Also, communication between devices in a communication system can be in accordance with communication protocols of any stage, for example, can include but is not limited to the following communication protocols: 1G (generation), 2G, 2.5G, 2.75G, 3G, 4G, 4.5G and 5G, New Radio (NR), 6G and future communication, etc., and / or other currently known or to be developed in the future communication protocols.

[0035] In embodiments of the present application, the term "network device" refers to, for example, a device that accesses a terminal device to a communication network and provides services for the terminal device in a communication system. The network device can include but is not limited to the following devices: base station (BS), access point (AP), transmission reception point (TRP), broadcast transmitter, mobile management entity (MME), gateway, server, radio network controller (RNC), base station controller (BSC), etc.

[0036] Among them, the base station can include but is not limited to: node B (NodeB or NB), evolved node B (eNodeB or eNB), 5G base station (gNB), 6G base station and future base station, etc., in addition to remote radio head (RRH), remote radio unit (RRU), relay or low power node (such as femto, pico, etc.). And the term "base station" can include some or all functions of them, and each base station can provide communication coverage for a specific geographic area. The term "cell" can refer to a base station and / or its coverage area, depending on the context in which the term is used.

[0037] In embodiments of the present application, the term "user equipment" (UE) or "terminal equipment" (TE) refers to, for example, a device that accesses a communication network through a network device and receives network services. The terminal equipment can be fixed or mobile, and can also be referred to as a mobile station (MS), a terminal, a subscriber station (SS), an access terminal (AT), a station, etc.

[0038] The terminal device can include, but is not limited to, a cellular phone, a personal digital assistant (PDA), a wireless modem, a wireless communication device, a handheld device, a machine type communication device, a laptop, a cordless phone, a smartphone, a smartwatch, a digital camera, and the like.

[0039] For another example, in scenarios such as Internet of Things (IoT), the terminal device can also be a machine or device that performs monitoring or measurement, and can include, but is not limited to, a machine type communication (MTC) terminal, a vehicle-mounted communication terminal, a device to device (D2D) terminal, a machine to machine (M2M) terminal, and the like.

[0040] In addition, the term “network device” or “network device side” refers to the side of the network, which can be a certain base station, or can include one or more network devices as described above. The term “user side” or “terminal device” or “terminal device side” refers to the side of the user or terminal, which can be a certain UE, or can include one or more terminal devices as described above. In this document, “device” can refer to a network device or a terminal device unless otherwise specified.

[0041] In the following description, the terms “AI / ML model” and “AI / ML function” or “AI / ML model / function” or “AI / ML model and / or function” can be interchangeable without causing confusion.

[0042] The terms “parameters of an AI / ML model” and “weights of an AI / ML model” can be interchangeable.

[0043] The terms “size of a time window” and “length of a time window” can be interchangeable.

[0044] The AI / ML model of the present application includes, but is not limited to, an input layer (input), a plurality of convolutional layers, a concatenation layer (concat), a fully connected layer (FC), and a quantizer, etc. The processing results of the plurality of convolutional layers are merged in the concatenation layer, and the specific structure of the AI / ML model can refer to the prior art, which will not be described here.

[0045] An AI / ML function and / or model is defined by an AI / ML feature or a group of AI / ML features. In 5G and 6G systems, a very wide range of terminal device features (UE features) are defined, and not all of them are required to be supported by terminal devices. Network devices send a capability reporting request to terminal devices, and terminal devices report their capabilities, i.e., UE capabilities, according to the capability reporting request, so that network devices can understand the feature support of terminal devices and provide a basis for subsequent scheduling, configuration, and communication of terminal devices.

[0046] To support the diversity of terminal functions, the existing standard predefines a plurality of features for terminals, each feature further predefines one or more feature groups, and each feature group predefines one or more components. An identifier or serial number is defined for each feature, feature group, and component. According to the current 5G-Advanced and 6G technology trends, when the capabilities of a UE can support multiple AI / ML models and / or functions, corresponding features or feature groups can be predefined for AI / ML models and / or functions.

[0047] In the function-based life cycle management process, when a terminal device supports multiple AI / ML models and / or functions, an AI / ML model and / or function can be defined by an AI / ML feature or a group of AI / ML features.

[0048] The above-mentioned AI / ML models and / or functions and their related parameters can be specified by predefined features or feature groups for common understanding when network devices and terminal devices communicate capability information.

[0049] In the embodiments of the present application, a function or model can also be referred to as a feature or a feature group. Different features or feature groups contain corresponding parameters or application conditions. Different features or feature groups can have an identifier (ID) or a corresponding index or a logical identifier.

[0050] In this way, when a terminal supports multiple AI / ML models and / or functions, the models and / or functions can be distinguished by model and / or function identifiers or logical identifiers, or by feature or feature group identifiers or serial numbers.

[0051] In the function-based life cycle management process, to implement a certain function, a terminal device can be implemented by one AI / ML model or multiple AI / ML models, or multiple functions can be implemented by one AI / ML model. The specific implementation method does not need to be informed to the network device.

[0052] In the function-based life cycle management process, when a function is implemented by multiple models, different models can correspond to different scenarios, different sites, different cells, different configurations, different application conditions, etc., which are determined by the training and development process of the model.

[0053] The scenarios of the embodiments of the present application are described below by way of example, but the present application is not limited thereto.

[0054] FIG. 1 is a schematic diagram of a communication system according to an embodiment of the present application, which illustrates a case with a terminal device and a network device as an example. As shown in FIG. 1, the communication system 100 can include a network device 101 and terminal devices 102 and 103. For simplicity, FIG. 1 illustrates only two terminal devices and one network device as an example, but the embodiments of the present application are not limited thereto.

[0055] In the embodiments of the present application, the network device 101 and the terminal devices 102 and 103 can perform existing services or future implementable services transmission. For example, these services can include, but are not limited to, enhanced mobile broadband (eMBB), massive machine type communication (mMTC), and ultra-reliable and low-latency communication (URLLC), etc.

[0056] Among them, the terminal device 102 can send data to the network device 101, for example, using a licensed or unlicensed transmission mode. The network device 101 can receive data sent by one or more terminal devices 102, and feed back information to the terminal device 102, such as acknowledgement ACK / non-acknowledgement NACK information, etc., so that the terminal device 102 can confirm the end of the transmission process, or can further perform new data transmission, or can perform data retransmission.

[0057] It is worth noting that FIG. 1 shows that both terminal devices 102 and 103 are within the coverage of the network device 101, but the present application is not limited thereto. Both terminal devices 102 and 103 can be outside the coverage of the network device 101, or one terminal device 102 is within the coverage of the network device 101 while the other terminal device 103 is outside the coverage of the network device 101.

[0058] In the scenario of the above communication system, artificial intelligence / machine learning technology has been studied. At the RAN1#118 meeting, companies selected the new research cases (Cases 0 / 1 / 2 / 3 / 4 / 5) for Rel-19, and decided to focus on Case 2 and Case 3. For these two new research cases, the standardization of the AI / ML model structure can be as described in Table 1:

[0059] Table 1

[0060] FIG. 2 is a schematic diagram of the interaction between a terminal device and a network device of Case 2 according to an embodiment of the present application. As shown in FIG. 2, for Case 2, the terminal device and / or the network device uses the CSI information at historical time points (also referred to as “historical CSI information”, i.e., “Historical CSI info” in FIG. 2) to help compress the CSI information at the current time point, which refers to the CSI information measured at the current time point (i.e., “CSI measurement” in FIG. 2). The historical CSI information at the terminal device and the network device is different. When there is strong correlation in the time domain, the CSI information at the historical time point helps to compress and / or recover the CSI at the current time point. Therefore, the Case 2 use case can greatly reduce the feedback overhead or improve the feedback accuracy.

[0061] For example, as shown in FIG. 2, a first AI / ML model (i.e., “Encoder” in FIG. 2) is configured at the terminal device, and a second AI / ML model (i.e., “Decoder” in FIG. 2) is configured at the network device. Taking the current time point (t+d) as an example, the terminal device processes the CSI information measured at the current time point (CSI measurement) (for example, processes the format, dimension, etc. of the measured CSI information to match the input information of the first AI / ML model at the current time point), and then inputs the CSI information at the current time point into the first AI / ML model (“Encoder”) corresponding to the time point (t+d); the first AI / ML model performs time domain CSI compression on the input CSI information at the current time point based on the historical CSI information at the terminal device (i.e., “Historical CSI info” at the terminal device in FIG. 2), outputs the compression result and sends it to the network device. After receiving the compression result from the terminal device, the network device inputs it into the second AI / ML model (“Decoder”) corresponding to the time point (t+d), and the second AI / ML model performs time domain CSI decompression or reconstruction on the input CSI information at the current time point based on the historical CSI information at the network device (i.e., “Historical CSI info” at the network device in FIG. 2), and outputs the decompression or reconstruction result.

[0062] However, in the discussion of 3GPP, for Case 2 use case, each company designs its own AI / ML model for research or performance evaluation, which leads to no unified AI / ML model or model structure. In order to solve the multi-vendor / inter-vendor collaboration problem of bilateral model or reduce the complexity of vendor collaboration, 3GPP RAN1 proposes two potential solutions:

[0063] Solution 1: Standardize AI / ML model (structure of model and weight / parameter of model);

[0064] Solution 2: Only standardize the structure of AI / ML model, and the weight / parameter of model is transferred from network device to terminal device or terminal device to network device.

[0065] Both solutions involve the standardization of AI / ML model, but there is no specific solution for the standardization of AI / ML model for Case 2 use case, so further discussion is needed.

[0066] The present application is described below in conjunction with the accompanying drawings and embodiments.

[0067] Embodiments of the first aspect

[0068] The embodiments of the present application provide an information transceiving method, which is applied to a terminal device side, and described below in conjunction with the accompanying drawings.

[0069] FIG. 3 is a schematic diagram of the information transceiving method according to an embodiment of the present application, as shown in FIG. 3, the method comprises:

[0070] 301, receiving first configuration information from a network device, the first configuration information is related to AI / ML model selection of the terminal device;

[0071] 302, sending channel state information (Channel State Information, CSI) feedback information to the network device; the CSI feedback information is generated by the terminal device using the AI / ML model selected according to the first configuration information.

[0072] According to the above embodiments, the terminal device performs function and / or model selection and time-domain CSI compression according to the configuration information related to the function and / or model selection of the terminal device from the network device, thereby the terminal device can select the AI / ML model according to the configuration of the network device, so as to standardize the AI / ML model used on the terminal device side and the network device side, solve the multi-vendor / inter-vendor cooperation problem of bilateral models, reduce the complexity of vendor cooperation, and thus solve the problems in the prior art.

[0073] It is worth noting that the above FIG. 3 only schematically illustrates the embodiments of the present application, but the present application is not limited thereto. For example, the execution order between the operations can be appropriately adjusted, and in addition, some operations can be added or some operations can be reduced. Those skilled in the art can make appropriate modifications according to the above content, and the present application is not limited to the above FIG. 3.

[0074] In some embodiments, the AI / ML model includes an associated first AI / ML model and a second AI / ML model, wherein the first AI / ML model is configured or deployed on the terminal device side, and the second AI / ML model is configured or deployed on the network device side. The AI / ML model is pre-trained, the structure and parameters of the first AI / ML model in the associated AI / ML model are pre-sent to the terminal device side, and the structure and parameters of the first AI / ML model are pre-sent to the network device side. For example, the network device pre-trains at least one set of associated first AI / ML model and second AI / ML model, and sends the structure and parameters of the first AI / ML model in it to the terminal device, and vice versa, the terminal device can also train at least one set of associated first AI / ML model and second AI / ML model, and send the structure and parameters of the second AI / ML model in it to the network device. Thus, the terminal device can select the first AI / ML model with specific structure and parameters for deployment to perform time-domain CSI compression, and the network device can select the second AI / ML model associated with the first AI / ML model for deployment to perform time-domain CSI reconstruction.

[0075] In some embodiments, the first AI / ML model and the second AI / ML model are in one-to-one correspondence, and / or one first AI / ML model is associated with multiple second AI / ML models, and / or one second AI / ML model is associated with multiple first AI / ML models. For example, in the case where the structure and parameters of each first AI / ML model and second AI / ML model are fixed, one first AI / ML model is associated with one second AI / ML model. For another example, in the case where the first AI / ML model and the second AI / ML model are extensible, for example, the structure of the first AI / ML model and the second AI / ML model is fixed, but the parameters are variable, one first AI / ML model can be associated with multiple second AI / ML models, or one second AI / ML model can be associated with multiple first AI / ML models, which is not limited in the present application.

[0076] In the present application, the AI / ML model on the terminal device side can also be referred to as the first AI / ML model or the first AI / ML model on the terminal device side, and the AI / ML model on the network device side can also be referred to as the second AI / ML model or the second AI / ML model on the network device side.

[0077] In some embodiments, in 301, the first configuration information includes first indication information indicating input information of the AI / ML model on the terminal device side, and / or second indication information indicating CSI feedback overhead of the AI / ML model on the terminal device side. For example, information indicating any one of the input information of the AI / ML model on the terminal device side provided below can be included, or information indicating the CSI feedback overhead of the AI / ML model on the terminal device side provided below can be included, or information indicating any combination of any one of the input information of the AI / ML model on the terminal device side provided below and the CSI feedback overhead of the AI / ML model on the terminal device side can be included. The input information of the AI / ML model on the terminal device side and the CSI feedback overhead of the AI / ML model on the terminal device side are described in detail below.

[0078] The content related to the input information of the AI / ML model on the terminal device side is described below.

[0079] In some embodiments, the input information of the AI / ML model on the terminal device side includes the number of historical CSI information input to the AI / ML model on the terminal device side, and / or the type of historical CSI information input to the AI / ML model on the terminal device side.

[0080] In some embodiments, the first indication information indicates the number of historical CSI information input to the AI / ML model on the terminal device side at at least one time point.

[0081] For example, the network device indicates to the terminal device the number of historical CSI information input to the AI / ML model at time t.

[0082] For another example, the network device indicates to the terminal device the number of historical CSI information input to the AI / ML model at time t, time t+d, time t+2d, and time t+3d. The present application does not limit this. Wherein, d refers to a CSI reporting period or a CSI reporting interval, and d is, for example, 5 milliseconds (ms).

[0083] In some embodiments, the CSI reporting period is pre-configured, for example, configured by the CSI reporting configuration information sent by the network device to the terminal device.

[0084] In some embodiments, for different times, the number of historical CSI information input to the AI / ML model at the terminal device side is the same.

[0085] For example, in the case where the network device indicates to the terminal device the number of historical CSI information input to the AI / ML model at time t, time t+d, time t+2d, and time t+3d, it is assumed that the number of historical CSI information input to the AI / ML model at time t, time t+d, time t+2d, and time t+3d is n1, n2, n3, and n4 respectively, and the values of n1, n2, n3, and n4 are the same.

[0086] In some embodiments, for different times, the number of historical CSI information input to the AI / ML model at the terminal device side is different.

[0087] For example, there are at least two different times, and the number of historical CSI information input to the AI / ML model at the terminal device side is different. For example, it is assumed that the network device indicates to the terminal device the number of historical CSI information input to the AI / ML model at time t, time t+d, time t+2d, and time t+3d is n1, n2, n3, and n4 respectively, and the values of at least two of n1, n2, n3, and n4 are different, for example, the values of n1 and n2 are different, the values of n3 and n4 are the same as that of n1; for another example, the values of n1 and n2 are the same, the values of n3 and n4 are the same, the values of n1 and n3 are different; for another example, the values of n1, n2, and n3 are all different, and the value of n4 is the same as that of n1; for another example, the values of n1, n2, n3, and n4 are all different, that is, the number of historical CSI information input to the AI / ML model at the terminal device side is different at all times. The present application does not limit this.

[0088] In some embodiments, the first indication information indicates the number of historical CSI information of the AI / ML model on the terminal device side within at least one first time window. Here, the "first time window" is used to represent the time window configured for the terminal device side, in order to distinguish from the "second time window" configured for the network device side in the following. In addition, the "first time window" can also be replaced by "CSI reporting window" or other similar expressions. The "first time window" does not specifically refer to one time window, which can include one time window or multiple time windows. In the case of multiple time windows, i.e., multiple first time windows are configured for the terminal device side, the start position, size, etc. of each first time window can be the same or different, which will be described in subsequent embodiments. Similarly, the "second time window" does not specifically refer to one time window, which can include one time window or multiple time windows. In the case of multiple time windows, i.e., multiple second time windows are configured for the network device side, the start position, size, etc. of each second time window can be the same or different, which is not limited by the present application. In addition, the number, start position, size, etc. of the first time window configured for the terminal device can be the same as or different from the number, start position, size, etc. of the second time window configured for the network device, which will be described in subsequent embodiments.

[0089] In some embodiments, the network device indicates to the terminal device the number of historical CSI information of the AI / ML model on the terminal device side at multiple time points within a first time window. For example, assuming that there are 4 CSI reporting time points within a first time window T11, which are time point t, time point t+d, time point t+2d, and time point t+3d, the first indication information indicates the number of historical CSI information of the AI / ML model at time point t, time point t+d, time point t+2d, and time point t+3d, for example, {1, 1, 2, 3}. It should be noted that the above is an example of the first time window and CSI reporting time point provided by the embodiments of the present application, and the present application is not limited thereto.

[0090] In some embodiments, for multiple time points within a first time window, the number of historical CSI information of the AI / ML model on the terminal device side is the same, or the number of historical CSI information of the AI / ML model on the terminal device side at at least two time points is different.

[0091] For example, in the example that the first indication information respectively indicates that the number of historical CSI information inputting the AI / ML model at the four time instants t, t+d, t+2d and t+3d in the first time window T11 is {1, 1, 2, 3}, the number of historical CSI information inputting the AI / ML model at the time instant t and the time instant t+d in the first time window T11 is the same, and the number of historical CSI information inputting the AI / ML model at the time instant t+d and the time instant t+2d is different, the present application does not make any limitation.

[0092] In some embodiments, the network device indicates to the terminal device the number of historical CSI information inputting the AI / ML model at multiple time instants in multiple first time windows. For example, for N first time windows T11-T1N, the network device indicates to the terminal device the following information: the number of historical CSI information inputting the AI / ML model at a time instants in the first first time window T11 {n 11 ,…,n 1a};the number of historical CSI information inputting the AI / ML model at b time instants in the second first time window T12 {n 21 ,…,n 2b}, …, and the number of historical CSI information inputting the AI / ML model at c time instants in the Nth first time window T1N {n N1 ,…,n Nc}. Wherein, the number of first time windows N, the number of time instants a, b, …, c in the first time window are all positive integers.

[0093] In some embodiments, the number of time instants contained in different first time windows is the same, that is, the size or length of different first time windows is the same, for example, the first first time window T11 contains a time instants, the second time window contains b time instants, and the values of a and b are the same. Or, the number of time instants contained in at least two different first time windows is different, that is, the size or length of at least two different first time windows is different, for example, the first first time window T11 contains a time instants, the second time window contains b time instants, and the values of a and b are different.

[0094] In some embodiments, the number of historical CSI information input to the AI / ML model at the terminal device side is the same in different first time windows. In this case, the size of the different first time windows is the same, and the number of historical CSI information input to the AI / ML model at the corresponding time in each first time window is the same. For example, taking two first time windows as an example, the size of the first first time window T11 and the second first time window T12 is the same and each includes four time points, the number of historical CSI information input to the AI / ML model at the terminal device side in the four time points in the first first time window T11 is {1, 1, 2, 3}, and the number of historical CSI information input to the AI / ML model at the terminal device side in the four time points in the second first time window T11 is also {1, 1, 2, 3}.

[0095] In some embodiments, the number of historical CSI information input to the AI / ML model at the terminal device side is different in different first time windows. In this case, the size of the different first time windows is different, and / or the number of historical CSI information input to the AI / ML model at at least one corresponding time in each first time window is different. For example, taking two first time windows as an example, the size of the first first time window T11 and the second first time window T2 is the same and each includes four time points, the number of historical CSI information input to the AI / ML model at the terminal device side in the four time points in the first first time window T11 is {1, 1, 2, 3}, and the number of historical CSI information input to the AI / ML model at the terminal device side in the four time points in the second first time window T11 is {1, 2, 3, 1}; or, the size of the first first time window T11 and the second first time window T12 is different, the first first time window includes four time points, and the number of historical CSI information input to the AI / ML model at the terminal device side in the four time points is {1, 1, 2, 3}, and the second first time window includes five time points, and the number of historical CSI information input to the AI / ML model at the terminal device side in the five time points is {1, 1, 2, 3, 2}.

[0096] In some embodiments, the first configuration information further includes time window configuration information for configuring the first time window.

[0097] In some embodiments, the time window configuration information includes size information of the first time window, and / or start position information of the first time window.

[0098] In some embodiments, the number of historical CSI information input to the AI / ML model at the terminal device side is one or more of the number of historical CSI information supported by the terminal device for reporting to the network device for use by the AI / ML model.

[0099] For example, the terminal device reports to the network device a number of historical CSI information used by at least one AI / ML model supported by the terminal device, for example, the terminal device reports L numbers, for any one time instant, the network device can select one from the L numbers reported by the terminal device as the number of historical CSI information input to the AI / ML model at the terminal device side at the time instant. For example, in the case of indicating the number of historical CSI information input to the AI / ML model at the terminal device side at one time instant, the network device selects one from the L numbers reported by the terminal device as the number of historical CSI information input to the AI / ML model at the terminal device side at the time instant, and indicates the selected number value to the terminal device. For another example, in the case of indicating the number of historical CSI information input to the AI / ML model at the terminal device side at multiple time instants, the network device selects one from the L numbers reported by the terminal device for each of the multiple time instants, as the number of historical CSI information input to the AI / ML model at the terminal device side at the time instant, and indicates the selected multiple number values corresponding to the multiple time instants to the terminal device. For another example, in the case of indicating the number of historical CSI information input to the AI / ML model at the terminal device side within a first time window, the network device selects one from the L numbers reported by the terminal device for each of the time instants within the first time window, as the number of historical CSI information input to the AI / ML model at the terminal device side at the time instant, and indicates the selected multiple number values corresponding to the multiple time instants within the first time window to the terminal device. For another example, in the case of indicating the number of historical CSI information input to the AI / ML model at the terminal device side within multiple first time windows, the network device selects one from the L numbers reported by the terminal device for each of the time instants within each of the multiple first time windows, as the number of historical CSI information input to the AI / ML model at the terminal device side at the time instant, and indicates the selected multiple number values corresponding to the multiple time instants within the multiple first time windows to the terminal device.

[0100] In some embodiments, the number of historical CSI information input to the AI / ML model at the terminal device side is less than or equal to a maximum value of the number of historical CSI information used by the AI / ML model supported by the terminal device and reported to the network device.

[0101] For example, the terminal device reports to the network device a maximum value L_max of the number of historical CSI information used by at least one AI / ML model supported by the terminal device, for any one time instant, the network device selects any positive integer less than or equal to L_max reported by the terminal device as the number of historical CSI information input to the AI / ML model at the terminal device side at the time instant.

[0102] For example, in the case of indicating the number of historical CSI information of the input terminal device side AI / ML model at one time, the network device determines an arbitrary positive integer less than or equal to L_max as the number of historical CSI information of the input terminal device side AI / ML model at this time, and indicates the determined number value to the terminal device.

[0103] For example, in the case of indicating the number of historical CSI information of the input terminal device side AI / ML model at one time, the network device determines an arbitrary positive integer less than or equal to L_max as the number of historical CSI information of the input terminal device side AI / ML model at this time, and indicates the determined number value to the terminal device.

[0104] For example, in the case of indicating the number of historical CSI information of the input terminal device side AI / ML model at one time, the network device determines an arbitrary positive integer less than or equal to L_max as the number of historical CSI information of the input terminal device side AI / ML model at this time, and indicates the determined number value to the terminal device.

[0105] For example, in the case of indicating the number of historical CSI information of the input terminal device side AI / ML model at one time, the network device determines an arbitrary positive integer less than or equal to L_max as the number of historical CSI information of the input terminal device side AI / ML model at this time, and indicates the determined number value to the terminal device.

[0106] In some embodiments, the historical CSI information of the input terminal device side AI / ML model has multiple types, for example, including at least one of the following:

[0107] The measurement CSI information of the historical time, for example, please refer to FIG. 2, at time t+d, the historical CSI information of the input terminal device side AI / ML model can be the measurement CSI information at time t (i.e. the CSI measurement corresponding to time t in FIG. 2);

[0108] Information derived based on the measurement CSI information of the historical time;

[0109] The output information of the AI / ML model at the terminal device side at the historical moment, for example, referring to FIG. 2, the historical CSI information input into the AI / ML model at the terminal device side at moment t+d can be the time-domain CSI compression result output by the first AI / ML model at the terminal device side at moment t;

[0110] The information derived based on the output information of the AI / ML model at the terminal device side at the historical moment.

[0111] In some embodiments, the type of the historical CSI information input into the AI / ML model at the terminal device side is customized by the terminal device or pre-set, or configured by the network device, for example, similar to the number of the historical CSI information input into the AI / ML model at the terminal device side, the type of the historical CSI information input into the AI / ML model at the terminal device side is indicated by the first indication information as one of the input information.

[0112] The following describes the content related to the CSI feedback overhead of the AI / ML model at the terminal device side.

[0113] In some embodiments, the CSI feedback overhead of the AI / ML model at the terminal device side indicated by the second indication information is the CSI feedback overhead for which the AI / ML model is input. If the input of the AI / ML model is single-layer CSI, the feedback overhead is the feedback overhead of the single-layer CSI, and if the input of the AI / ML model is multi-layer CSI, the feedback overhead is the feedback overhead of the multi-layer CSI.

[0114] In addition, when the Rank value of the CSI reporting is R, R layers of CSI need to be fed back by the terminal device. R can be, for example, a positive integer such as 1, 2, 3, 4, …, and the present application does not limit this.

[0115] In some embodiments, the second indication information indicates the CSI feedback overhead of the AI / ML model at the terminal device side at at least one moment, and the unit of the feedback overhead is, for example, bit (bit).

[0116] For example, the network device indicates to the terminal device the CSI feedback overhead of the AI / ML model at the terminal device side at a moment t.

[0117] For another example, the network device indicates to the terminal device the CSI feedback overhead of the AI / ML model at the terminal device side at moments t, t+d, t+2d, and t+3d. The present application does not limit this. Wherein, d refers to the CSI reporting period or CSI reporting interval, and d is, for example, 5 milliseconds (ms). In some embodiments, the CSI reporting period is pre-configured, for example, configured by the CSI reporting configuration information sent by the network device to the terminal device.

[0118] In some embodiments, the CSI feedback overhead of the AI / ML model at the terminal device side is the same for different time instants.

[0119] For example, in the case where the network device indicates to the terminal device the CSI feedback overhead of the AI / ML model at the terminal device side at time instant t, time instant t+d, time instant t+2d, and time instant t+3d, it is assumed that the CSI feedback overhead of the AI / ML model at the terminal device side at time instant t, time instant t+d, time instant t+2d, and time instant t+3d is k1, k2, k3, and k4, respectively, and the values of k1, k2, k3, and k4 are the same.

[0120] In some embodiments, the CSI feedback overhead of the AI / ML model at the terminal device side is different for different time instants.

[0121] For example, there are at least two different time instants, and the CSI feedback overhead of the AI / ML model at the terminal device side is different. For example, it is assumed that the network device indicates to the terminal device the CSI feedback overhead of the AI / ML model at the terminal device side at time instant t, time instant t+d, time instant t+2d, and time instant t+3d, which is k1, k2, k3, and k4, respectively, and the values of at least two of k1, k2, k3, and k4 are different, for example, the values of k1 and k2 are different, the values of k3 and k4 are the same as that of k1; for another example, the values of k1 and k2 are the same, the values of k3 and k4 are the same, the values of k1 and k3 are different; for another example, the values of k1, k2, and k3 are all different, and the value of k4 is the same as that of k1; for another example, the values of k1, k2, k3, and k4 are all different, that is, the number of historical CSI information input into the AI / ML model at the terminal device side is different at all time instants. The present application does not make any limitation in this regard.

[0122] In some embodiments, the second indication information indicates the CSI feedback overhead of multiple time instants within at least one first time window. Similar to the foregoing embodiments, here the "first time window" is used to represent the time window configured at the terminal device side, in order to distinguish from the "second time window" configured at the network device side in the following description. For details, reference can be made to the description of the related embodiments herein, which will not be repeated here. In addition, the "first time window" can be replaced by "CSI reporting window" or other similar expressions.

[0123] In some embodiments, the network device indicates to the terminal device the CSI feedback overhead of the AI / ML model at the terminal device side at multiple time instants within a first time window. For example, assuming that there are 4 time instants for CSI reporting within a first time window T1, namely time instant t, time instant t+d, time instant t+2d, and time instant t+3d, the second indication information indicates the CSI feedback overhead of the AI / ML model at the terminal device side at the four time instants t, t+d, t+2d, and t+3d, for example, as {120, 30, 30, 30} (unit: bits). It should be noted that the above is an example of the first time window and the CSI reporting time instant provided by the embodiments of the present application, and the present application is not limited thereto.

[0124] In some embodiments, for multiple time instants within a first time window, the CSI feedback overhead of the AI / ML model at the terminal device side is the same, or the CSI feedback overhead of the AI / ML model at the terminal device side is different.

[0125] For example, in the example where the second indication information indicates that the CSI feedback overhead of the AI / ML model at the terminal device side at the four time instants t, t+d, t+2d, and t+3d within the first time window T1 is {120, 30, 30, 30} (unit: bits), the CSI feedback overhead of the AI / ML model at the terminal device side is the same at the time instants t+d, t+2d, and t+3d within the first time window T1, and the CSI feedback overhead of the AI / ML model at the terminal device side is different at the time instants t and t+d, and the present application is not limited thereto.

[0126] In some embodiments, the network device indicates to the terminal device the CSI feedback overhead of the AI / ML model at the terminal device side at multiple time instants within multiple first time windows. For example, for N first time windows T1-TN, the network device indicates to the terminal device the following information: the CSI feedback overhead of the AI / ML model at the terminal device side at a time instants within the first first time window T1 {k 11 ,…,k 1a};the CSI feedback overhead of the AI / ML model at the terminal device side at b time instants within the second first time window T2 {k 21 ,…,k 2b}, …, and the CSI feedback overhead of the AI / ML model at the terminal device side at c time instants within the Nth first time window TN {k T1 ,…,k Tc}. Wherein, the number of first time windows N, the number of time instants a, b, …, c within the first time window are all positive integers.

[0127] In some embodiments, the number of time instants included in different first time windows is the same, i.e., the size or length of different first time windows is the same, for example, the first first time window T1 includes a time instants, and the second first time window includes b time instants, and a and b have the same value. Alternatively, the number of time instants included in at least two different first time windows is different, i.e., the size or length of at least two different first time windows is different, for example, the first first time window T1 includes a time instants, and the second first time window includes b time instants, and a and b have different values.

[0128] In some embodiments, the CSI feedback overhead of the AI / ML model at the terminal device side is the same in different first time windows. In this case, the size of the different first time windows, and the CSI feedback overhead of the AI / ML model at the terminal device side for the corresponding time instants in each first time window are all the same. For example, taking two first time windows as an example, the first first time window T1 and the second first time window T2 have the same size and each include four time instants, the CSI feedback overhead of the AI / ML model at the terminal device side for the four time instants in the first first time window T1 is {120, 30, 30, 30} (unit: bit), and the CSI feedback overhead of the AI / ML model at the terminal device side for the four time instants in the second first time window T1 is also {120, 30, 30, 30} (unit: bit).

[0129] In some embodiments, the CSI feedback overhead of the AI / ML model at the terminal device side is different in different first time windows. In this case, the size of the different first time windows is different, and / or the CSI feedback overhead of the AI / ML model at the terminal device side is different for at least one corresponding time instant in each first time window. For example, taking two first time windows as an example, the first first time window T1 and the second first time window T2 have the same size and each include four time instants, the CSI feedback overhead of the AI / ML model at the terminal device side for the four time instants in the first first time window T1 is {120, 30, 30, 30} (unit: bit), and the CSI feedback overhead of the AI / ML model at the terminal device side for the four time instants in the second first time window T1 is also {120, 90, 60, 30} (unit: bit); or, the size of the first first time window T1 and the second first time window T2 is different, the first first time window includes four time instants, and the CSI feedback overhead of the AI / ML model at the terminal device side for the four time instants is {120, 30, 30, 30} (unit: bit), and the second first time window includes five time instants, and the CSI feedback overhead of the AI / ML model at the terminal device side for the five time instants is {120, 30, 30, 30, 20} (unit: bit).

[0130] In some embodiments, the first configuration information further comprises time window configuration information for configuring the first time window.

[0131] In some embodiments, the time window configuration information comprises size information of the first time window, and / or start position information of the first time window.

[0132] In some embodiments, the CSI feedback overhead of the AI / ML model at the terminal device side is one or more of the CSI feedback overheads of the AI / ML models supported by the terminal device and reported to the network device.

[0133] For example, the terminal device reports the CSI feedback overheads of at least one AI / ML model supported by the terminal device to the network device, for example, the terminal device reports K CSI feedback overheads, for any time instant, the network device can select one from the K CSI feedback overheads reported by the terminal device as the CSI feedback overhead of the AI / ML model at the terminal device side at that time instant.

[0134] For example, in the case of indicating the CSI feedback overhead of the AI / ML model at the terminal device side at one time instant, the network device selects one from the K CSI feedback overheads reported by the terminal device as the CSI feedback overhead of the AI / ML model at the terminal device side at that time instant, and indicates the selected CSI feedback overhead value to the terminal device.

[0135] For another example, in the case of indicating the CSI feedback overhead of the AI / ML model at the terminal device side at multiple time instants, the network device selects one from the K CSI feedback overheads reported by the terminal device for each of the multiple time instants as the CSI feedback overhead of the AI / ML model at the terminal device side at that time instant, and indicates the selected multiple CSI feedback overhead values corresponding to the multiple time instants to the terminal device.

[0136] For another example, in the case of indicating the CSI feedback overhead of the AI / ML model at the terminal device side within a first time window, the network device selects one from the K CSI feedback overheads reported by the terminal device for each time instant within the first time window as the CSI feedback overhead of the AI / ML model at the terminal device side at that time instant, and indicates the selected multiple CSI feedback overhead values corresponding to the multiple time instants within the first time window to the terminal device.

[0137] For example, in the case of indicating the CSI feedback overhead of the AI / ML model at the terminal device side in a plurality of first time windows, the network device selects one of the K CSI feedback overheads reported by the terminal device for each time in each of the plurality of first time windows, as the CSI feedback overhead of the AI / ML model at the terminal device side at the time, and indicates the selected plurality of CSI feedback overhead values corresponding to the plurality of times in the plurality of first time windows to the terminal device.

[0138] In some embodiments, the CSI feedback overhead of the AI / ML model at the terminal device side is less than or equal to the maximum value of the CSI feedback overhead of the supported AI / ML model reported by the terminal device to the network device.

[0139] For example, the terminal device reports the maximum value K_max of the CSI feedback overhead of at least one AI / ML model supported by the terminal device to the network device, and for any one time, the network device selects any positive integer less than or equal to K_max as the CSI feedback overhead of the AI / ML model at the terminal device side at the time.

[0140] For example, in the case of indicating the CSI feedback overhead of the AI / ML model at the terminal device side at one time, the network device determines any positive integer less than or equal to K_max as the CSI feedback overhead of the AI / ML model at the terminal device side at the time, and indicates the determined CSI feedback overhead value to the terminal device.

[0141] For example, in the case of indicating the CSI feedback overhead of the AI / ML model at the terminal device side at one time, the network device determines any positive integer less than or equal to K_max as the CSI feedback overhead of the AI / ML model at the terminal device side at the time, and indicates the determined CSI feedback overhead value to the terminal device.

[0142] For example, in the case of indicating the CSI feedback overhead of the AI / ML model at the terminal device side in a plurality of first time windows, the network device selects one of the K CSI feedback overheads reported by the terminal device for each time in each of the plurality of first time windows, as the CSI feedback overhead of the AI / ML model at the terminal device side at the time, and indicates the selected plurality of CSI feedback overhead values corresponding to the plurality of times in the plurality of first time windows to the terminal device.

[0143] For another example, in the case of indicating the CSI feedback overhead of the AI / ML model at the terminal device side in multiple first time windows, the network device determines, for each time instant in each of the multiple first time windows, an arbitrary positive integer less than or equal to K_max as the CSI feedback overhead of the AI / ML model at the terminal device side at the time instant, and indicates the determined multiple CSI feedback overhead values corresponding to multiple time instants in the multiple first time windows to the terminal device.

[0144] In some embodiments, in the case that the second indication information indicates the CSI feedback overhead of multiple time instants, or indicates the CSI feedback overhead in a first time window, or indicates the CSI feedback overhead in multiple time windows, the CSI feedback overhead of the first time instant is the largest, and the CSI feedback overhead of each time instant after the first time instant is smaller than the CSI feedback overhead of the first time instant. In this way, high-precision time-domain CSI compression with high overhead is performed at the first time instant, and the high-precision time-domain CSI compression result at the first time instant is used as historical CSI information for time-domain CSI compression at each time instant after the first time instant, so that the overhead can be greatly reduced while ensuring the accuracy.

[0145] In some embodiments, the CSI feedback overhead of multiple time instants in a first time window gradually decreases over time. For example, assuming that a first time window T1 includes four time instants for CSI reporting, namely time instant t, time instant t+d, time instant t+2d, and time instant t+3d, the second indication information indicates that the CSI feedback overhead at time instant t is the largest, for example, 120 bits, and the feedback overhead at time instants t+d, t+2d, and t+3d decreases successively, for example, 90 bits, 60 bits, and 30 bits, respectively. It should be noted that the above is an example of the CSI feedback overhead provided by the embodiments of the present application, and the present application is not limited thereto.

[0146] In some embodiments, the CSI feedback overhead of the first time instant in a first time window is the largest, and the CSI feedback overhead of each time instant after the first time instant is equal and smaller than the CSI feedback overhead of the first time instant. For example, assuming that a first time window T1 includes four time instants for CSI reporting, namely time instant t, time instant t+d, time instant t+2d, and time instant t+3d, the second indication information indicates that the CSI feedback overhead at time instant t is the largest, for example, 120 bits, and the feedback overhead at time instants t+d, t+2d, and t+3d is equal and smaller than the CSI feedback overhead of the first time instant, for example, the feedback overhead at time instants t+d, t+2d, and t+3d is 30 bits. It should be noted that the above is an example of the CSI feedback overhead provided by the embodiments of the present application, and the present application is not limited thereto.

[0147] In some embodiments, one AI / ML model supports one CSI feedback overhead, for example, a first AI / ML model at the terminal device side only supports a 120-bit CSI feedback overhead, or one AI / ML model supports one CSI feedback overhead, for example, a first AI / ML model at the terminal device side supports both 30-bit and 120-bit CSI feedback overheads.

[0148] In some embodiments, the selected AI / ML model is the same at different time instants with different CSI feedback overheads. For example, assuming there is a first AI / ML model that supports both 30-bit and 120-bit CSI feedback overheads, and the second indication information indicates that the CSI feedback overhead of the AI / ML model at the terminal device side is 120 bits at time instant t, 30 bits at time instant t+d, 30 bits at time instant t+2d, and 30 bits at time instant t+3d, then the first AI / ML model that supports both 30-bit and 120-bit CSI feedback overheads can be used at time instant t, time instant t+d, time instant t+2d, and time instant t+3d. It should be noted that the above is an example of the CSI feedback overhead provided by the embodiments of the present application, and the present application is not limited thereto.

[0149] In some embodiments, the selected AI / ML model is different at different time instants with different CSI feedback overheads. For example, assuming there is a first AI / ML model that only supports 30-bit CSI feedback overhead, and there is another first AI / ML model that only supports 120-bit CSI feedback overhead, and the second indication information indicates that the CSI feedback overhead of the AI / ML model at the terminal device side is 120 bits at time instant t, 30 bits at time instant t+d, 30 bits at time instant t+2d, and 30 bits at time instant t+3d, then the first AI / ML model that supports 120-bit CSI feedback overhead is used at time instant t, and the first AI / ML model that supports 30-bit CSI feedback overhead is used at time instant t+d. It should be noted that the above is an example of the CSI feedback overhead provided by the embodiments of the present application, and the present application is not limited thereto.

[0150] In some embodiments, the AI / ML model at the terminal device side (the first AI / ML model) is associated with the AI / ML model configured at the network device side (the second AI / ML model), for example, the first AI / ML model at the terminal device side that compresses the measured CSI information at time instant t is associated with the second AI / ML model at the network device side that decompresses / reconstructs the compression result of the measured CSI information at time instant t. The content related to the first AI / ML model and the second AI / ML model is described in the foregoing embodiments, which will not be repeated here.

[0151] In some embodiments, the number of historical CSI information input to the AI / ML model at the terminal device side is the same as the number of historical CSI information input to the AI / ML model at the network device side. It should be noted that the "AI / ML model at the terminal device side (first AI / ML model)" and the "AI / ML model at the network device side (second AI / ML model)" refer to a pair of associated AI / ML models. For example, at the terminal device side, the number of historical CSI information input to the first AI / ML model for compressing the measurement CSI at time t is L1, and at the network device side, the number of historical CSI information input to the second AI / ML model for reconstructing the time domain CSI compression result reported by the terminal device at time t is L2. The values of L1 and L2 can be the same or different, and the present application does not limit this.

[0152] In some embodiments, the number of historical CSI information input to the AI / ML model at the network device side (second AI / ML model) is one or more.

[0153] In some embodiments, for different time instants, the number of historical CSI information input to the AI / ML model at the network device side is the same.

[0154] For example, for different time instants for decompressing / reconstructing the compression result of the measurement CSI information reported by the terminal device at time t, time t+d, time t+2d, and time t+3d, the number of historical CSI information input to the AI / ML model at the network device side is n1, n2, n3, and n4, respectively, and the values of n1, n2, n3, and n4 are the same.

[0155] In some embodiments, for different time instants, the number of historical CSI information input to the AI / ML model at the network device side is different.

[0156] For example, there are at least two different time instants at which the number of historical CSI information input to the AI / ML model at the network device side is different. For example, at different time instants at which the compressed results of the measurement CSI information reported by the terminal device at time t, time t+d, time t+2d, and time t+3d are decompressed / reconstructed, the number of historical CSI information input to the AI / ML model at the network device side is respectively n1, n2, n3, and n4, and at least two of n1, n2, n3, and n4 have different values, for example, n1 and n2 have different values, n3 and n4 have the same value as n1; for another example, n1 and n2 have the same value, n3 and n4 have the same value, n1 and n3 have different values; for another example, n1, n2, and n3 have different values, n4 has the same value as n1; for another example, n1, n2, n3, and n4 all have different values, that is, the number of historical CSI information input to the AI / ML model at the network device side is different at all time instants. The present application does not limit this.

[0157] In some embodiments, for multiple time instants within the same second time window, the number of historical CSI information input to the AI / ML model at the network device side is the same, or the number of historical CSI information input to the AI / ML model at the network device side is different for at least two time instants.

[0158] For example, the network device decompresses / reconstructs the compressed results of the measurement CSI information reported by the terminal device at time t, time t+d, time t+2d, and time t+3d at four time instants within the second time window T21, and for the four time instants within the second time window T21, the number of historical CSI information input to the AI / ML model at the network device side is {m1, m2, m3, m4}, and the values of m1, m2, m3, and m4 can all be the same, for example, all being 2.

[0159] For another example, for at least two of the multiple time instants within the second time window T21, the number of historical CSI information input to the AI / ML model at the network device side can be different, for example, in the above example, the values of m1 and m2 are different, the values of m3 and m4 are the same as m1; for another example, the values of m1 and m2 are the same, the values of m3 and m4 are the same, the values of m1 and m3 are different; for another example, the values of m1, m2, and m3 are different, the value of m4 is the same as m1; for another example, the values of m1, m2, m3, and m4 are all different, that is, the number of historical CSI information input to the AI / ML model at the network device side is different at all time instants. The present application does not limit this. The above is only one example provided by the present application, and the present application is not limited thereto.

[0160] In some embodiments, the number of historical CSI information input to the AI / ML model at the network device side is the same in different second time windows, or the number of historical CSI information input to the AI / ML model at the network device side is different.

[0161] For example, the network device decompresses / reconstructs the compression results of the measurement CSI information reported by the terminal device at time t and time t+d at two time instants within a first second time window T21, and decompresses / reconstructs the compression results of the measurement CSI information reported by the terminal device at time t+2d and time t+3d at two time instants within a second second time window T22. For the two time instants within the first second time window T21, the number of historical CSI information input to the AI / ML model at the network device side is {m 11 ,m 12}, and for the two time instants within the second second time window T22, the number of historical CSI information input to the AI / ML model at the network device side is {m 21 ,m 22}. In this case, m 11 and m 21 have the same value, m 12 and m 22 have the same value, or m 11 and m 21 have different values, or m 12 and m 22 have different values.

[0162] In some embodiments, the network device configures a first time window for the terminal device, and the first time window is the same as the second time window, for example, the number, starting position, size, etc. of the first time window and the second time window are the same.

[0163] For example, the network device configures a first time window for the terminal device, and the first time window includes time t, time t+d, time t+2d, and time t+3d. The network device also configures a second time window, and the second time window includes time t, time t+d, time t+2d, and time t+3d.

[0164] In some embodiments, the network device configures a first time window for the terminal device, and the first time window is different from the second time window, for example, at least one of the number, starting position, size, etc. of the first time window and the second time window is different.

[0165] For example, the network device configures two first time windows for the terminal device, and the first time window includes time t and time t+d, and the second time window includes time t+2d and time t+3d, and the network device side configures one second time window, and the second time window includes time t+d, time t+2d and time t+3d.

[0166] In some embodiments, the historical CSI information input into the AI / ML model on the network device side has multiple types, for example, including at least one of the following:

[0167] The output information of the AI / ML model on the network device side at the historical time, for example, referring to FIG. 2, when the CSI information at time t+d is decompressed / reconstructed, the historical CSI information input into the AI / ML model on the network device side can be the decompression / reconstruction result of the CSI information output by the AI / ML model on the network device side at time t;

[0168] Information derived based on the output information of the AI / ML model on the network device side at the historical time;

[0169] The input information of the AI / ML model on the network device side at the historical time, for example, referring to FIG. 2, when the CSI information at time t+d is decompressed / reconstructed, the historical CSI information input into the AI / ML model on the network device side can be the compression result of the measurement CSI information reported by the terminal device at time t;

[0170] Information derived based on the input information of the AI / ML model on the network device side at the historical time.

[0171] In some embodiments, the type of the historical CSI information input into the AI / ML model on the network device side is defined by the network device or pre-set.

[0172] In some embodiments, in 302, the terminal device sends / reports the CSI feedback information to the network device. The sending / reporting manner of the CSI feedback information includes, for example, periodic, or semi-persistent, or aperiodic.

[0173] For example, in the case of periodic or semi-persistent sending / reporting of the CSI feedback information, if the network device configures the terminal device with one CSI reporting time point, and the number of historical CSI information input into the AI / ML model on the terminal device side is 2, then in the process of periodic or semi-persistent reporting, the number of historical CSI information input into the AI / ML model on the terminal device side is 2 at each time point.

[0174] For another example, in the case of periodic or semi-persistent CSI feedback information sending / reporting, if the network device configures the terminal device with multiple CSI reporting time windows containing 3 CSI reporting time instants, and the number of historical CSI information of the AI / ML model at the terminal device side input is {1, 1, 2} for the 3 CSI reporting time instants of the first CSI reporting time window, then in each time window during the periodic or semi-persistent reporting process, the number of historical CSI information of the AI / ML model at the terminal device side input is {1, 1, 2}.

[0175] For another example, in the case of aperiodic CSI feedback information sending / reporting, if the network device configures the number of historical CSI information of the AI / ML model at the terminal device side input for one or more CSI reporting time instants, then at the above one or more CSI reporting time instants, the terminal device determines the number of historical CSI information of the AI / ML model input according to the configuration of the network device.

[0176] For another example, in the case of aperiodic CSI feedback information sending / reporting, if the network device configures one or more CSI reporting time windows for the terminal device, and configures the number of historical CSI information of the AI / ML model at the terminal device side input for each CSI reporting time window, then at each time instant within the above one or more CSI reporting time windows, the terminal device determines the number of historical CSI information of the AI / ML model input according to the configuration of the network device.

[0177] In some embodiments, the CSI reporting mode is pre-configured, for example, by the CSI reporting configuration information sent by the network device to the terminal device.

[0178] In some embodiments, the number of ports of the CSI information measured or reported by the terminal device is the same. For example, at each CSI reporting time instant, the number of ports of the CSI information measured or reported by the terminal device is 32.

[0179] In some embodiments, due to energy saving considerations, it may be necessary to turn off part of the ports, or to perform RRC reconfiguration to change the number of ports, so the number of ports of the CSI information measured or reported by the terminal device may be different. For example, at time t, the number of ports of the CSI information measured or reported by the terminal device is 32, and at time t+d, the number of ports of the CSI information measured or reported by the terminal device is 16, which is not limited in the present application.

[0180] In some embodiments, the number of ports of the historical CSI information at different time instants can be the same or different.

[0181] In some embodiments, the historical CSI information of the port number X can be used to help compress the CSI information of the current port number Y.

[0182] In some embodiments, the carrier component (CC) of the CSI information measured or reported by the terminal device is the same. For example, at each CSI reporting moment, the CSI information of the terminal device is measured on the first carrier component.

[0183] In some embodiments, the carrier component of the CSI information measured or reported by the terminal device is different. For example, at time t, the CSI information of the terminal device is measured on the first carrier component, and at time t+d, the CSI information of the terminal device is measured on the second carrier component, which is not limited by the present application.

[0184] In some embodiments, the carrier component of the historical CSI information at different time can be the same or different.

[0185] In some embodiments, the historical CSI information measured on the first carrier component can be used to help compress the CSI information measured on the second carrier component.

[0186] FIG. 4 is another schematic diagram of the information transceiving method according to an embodiment of the present application. As shown in FIG. 4, the information transceiving method comprises:

[0187] 401, receiving first configuration information from a network device, the first configuration information being related to AI / ML model selection of the terminal device;

[0188] 402, selecting an AI / ML model (first AI / ML model) at least one time according to the first configuration information;

[0189] 403, reporting channel state information (CSI) feedback information to the network device; the CSI feedback information is generated by the terminal device using the AI / ML model selected according to the first configuration information.

[0190] In some embodiments, the implementation of 401 and 403 can refer to 301 and 302, and the repeated parts will not be described again.

[0191] In some embodiments, in 402, the terminal device selects an AI / ML model (first AI / ML model) at least one time according to the first configuration information.

[0192] For example, in a case where the first configuration information includes the first indication information, and the first indication information indicates that the number of historical CSI information of the AI / ML model at the terminal device side at time t is 2, the terminal device selects an AI / ML model with the number of input historical CSI information of 2 to perform time-domain CSI compression on the CSI information measured at time t.

[0193] For another example, in a case where the first configuration information includes the second indication information, and the second indication information indicates that the CSI feedback overhead of the AI / ML model at the terminal device side at time t is 120 bits, the terminal device selects an AI / ML model with the CSI feedback overhead of 120 bits to perform time-domain CSI compression on the CSI information measured at time t.

[0194] For another example, in a case where the first configuration information includes the first indication information and the second indication information, the first indication information indicates that the number of historical CSI information of the AI / ML model at the terminal device side at time t is 2, and the second indication information indicates that the CSI feedback overhead of the AI / ML model at the terminal device side at time t is 120 bits, the terminal device selects an AI / ML model with the number of input historical CSI information of 2 and the CSI feedback overhead of 120 bits to perform time-domain CSI compression on the CSI information measured at time t.

[0195] FIG. 5 is another schematic diagram of the information transceiving method according to an embodiment of the present application. As shown in FIG. 5, the information transceiving method includes:

[0196] 501, sending capability reporting information to a network device;

[0197] 502, receiving first configuration information from the network device, the first configuration information being related to AI / ML model selection of the terminal device;

[0198] 503, selecting at least one AI / ML model (first AI / ML model) at a time according to the first configuration information;

[0199] 504, reporting channel state information (CSI) feedback information to the network device, the CSI feedback information being generated by the terminal device using the AI / ML model selected according to the first configuration information.

[0200] In some embodiments, the implementation of 502 and 504 can refer to 301 and 302, and the implementation of 503 can refer to 402, and the repeated parts will not be described herein.

[0201] In some embodiments, in 501, the terminal device sends capability reporting information to the network device. The capability reporting information includes at least one of the following:

[0202] The number of supported AI / ML model using historical CSI information;

[0203] The maximum value of the number of AI / ML model using historical CSI information supported;

[0204] The CSI feedback overhead of at least one AI / ML model supported;

[0205] The maximum value of the CSI feedback overhead of AI / ML model supported.

[0206] In some embodiments, the network device determines the first configuration information according to the capability reporting information of the terminal device, for details, please refer to the related embodiments herein, which will not be repeated here.

[0207] In some embodiments, the method further includes (not shown):

[0208] Receiving the CSI resource configuration information sent by the network device, the CSI resource configuration information including information indicating the CSI feedback resource for the terminal device to report CSI feedback information.

[0209] Correspondingly, in 302, 403, 504, the terminal device reports the CSI feedback information to the network device on the CSI feedback resource.

[0210] In some embodiments, the network device determines the total feedback overhead (or total number of bits) of the CSI reporting on the terminal device side according to at least one of the CSI feedback overhead of the AI / ML model indicated by the second indication information in the first configuration information, the Rank value, and the sub-band information, and performs allocation of the CSI feedback resource for CSI feedback according to the total feedback overhead.

[0211] For example, in the case where the CSI feedback overhead of the AI / ML model single layer indicated by the second indication information is K0 bits and the Rank value is r, the total feedback overhead of the CSI reporting on the terminal device side is K0xr.

[0212] In some embodiments, the Rank value is an estimated value of the Rank value in the terminal device CSI reporting by the network device, for example, the Rank value is reported by the terminal device on the RI domain of the CSI reporting.

[0213] In some embodiments, the Rank value is estimated by the network device based on the Rank value reported by the terminal device at a historical time, or estimated by the network device based on uplink reference information (such as SRS (Sounding Reference Signal) reference signal), or obtained by the network device based on the Part I part of the CSI reported at the latest historical time, and the present application does not make any limitation in this regard.

[0214] In some embodiments, for different CSI reporting times, the network device respectively calculates the total feedback overhead for each CSI reporting time, and allocates CSI feedback resources for the terminal device to feed back CSI at the corresponding time according to the total feedback overhead at each time.

[0215] In some embodiments, for a time window (such as the first time window), the network device allocates CSI feedback resources for the terminal device to feed back CSI within the time window according to the total feedback overhead at each time within the time window.

[0216] For example, the network device allocates CSI feedback resources for the corresponding time according to the total feedback overhead at each time within the time window, or the network device allocates CSI feedback resources for each time within the time window according to the maximum value in the total feedback overhead at each time within the time window.

[0217] In some embodiments, for a time window (such as the first time window), the network device allocates the same CSI feedback resources for each time within the time window. For example, in the case where the network device allocates CSI feedback resources for each time within the time window according to the maximum value in the total feedback overhead at each time within the time window, the network device allocates the same CSI feedback resources for each time within the time window.

[0218] The above embodiments are only exemplary, but the present application is not limited thereto, and appropriate modifications can be made on the basis of the above embodiments. For example, each of the above embodiments can be used alone, or one or more of the above embodiments can be combined.

[0219] Through the above embodiments, the terminal device performs function and / or model selection according to the configuration information related to the function and / or model selection of the terminal device from the network device, so that the terminal device can select an AI / ML model according to the configuration of the network device, thereby standardizing the AI / ML model used on the terminal device side and the network device side, solving the multi-vendor / inter-vendor collaboration problem of bilateral models, reducing the complexity of vendor collaboration, and thereby solving the problems in the prior art.

[0220] Embodiments of the second aspect

[0221] Embodiments of the present application provide an information sending method, which is described from the network device side. The same content as the embodiments of the first aspect will not be described again.

[0222] FIG. 6 is a schematic diagram of information sending according to an embodiment of the present application. As shown in FIG. 6, the method comprises:

[0223] 601. Send first configuration information to the terminal device, wherein the first configuration information is related to AI / ML model selection of the terminal device.

[0224] It is worth noting that the above FIG. 6 only schematically illustrates the embodiments of the present application, but the present application is not limited thereto. For example, the execution order between each operation can be appropriately adjusted, and in addition, some operations can be added or some operations can be reduced. Those skilled in the art can make appropriate modifications according to the above content, and the above FIG. 6 is not limited thereto.

[0225] In some embodiments, in 601, the first configuration information comprises first indication information indicating input information of the AI / ML model on the terminal device side, and / or second indication information indicating CSI feedback overhead of the AI / ML model on the terminal device side. The first configuration information is as described in the embodiments of the first aspect, and its content is incorporated herein, which will not be described again.

[0226] In some embodiments, the method further comprises (not shown in the figure):

[0227] Receiving the capability reporting information sent by the terminal device.

[0228] In some embodiments, the capability reporting information comprises at least one of the following:

[0229] The number of supported at least one AI / ML model using historical CSI information;

[0230] The maximum value of the number of supported AI / ML models using historical CSI information;

[0231] The CSI feedback overhead of the supported at least one AI / ML model;

[0232] The maximum value of the CSI feedback overhead of the supported AI / ML model.

[0233] In this way, the network device can obtain the information of the AI / ML model supported by the terminal device, so as to make relevant configuration for the terminal device in accordance with its capability.

[0234] In some embodiments, the method further comprises (not shown in the figure):

[0235] The network device generates the first configuration information according to the capability reporting information sent by the terminal device.

[0236] In some embodiments, the method further comprises (not shown):

[0237] The CSI reporting configuration information and / or the CSI resource configuration information are sent to the terminal device.

[0238] In some embodiments, the CSI reporting configuration information includes, for example, a time at which the terminal device reports CSI feedback information (CSI reporting time), a period or interval at which the terminal device reports CSI feedback information (CSI reporting period or interval), and a manner in which the terminal device reports CSI feedback information (CSI reporting manner), and the CSI reporting manner includes, for example, periodic reporting, semi-persistent reporting, and aperiodic reporting.

[0239] In some embodiments, the CSI resource configuration information includes, for example, information indicating a CSI feedback resource used by the terminal device to report CSI feedback information.

[0240] In some embodiments, the method further comprises (not shown):

[0241] The network device determines a total feedback overhead (or total number of bits) of CSI reporting on the terminal device side according to at least one of the CSI feedback overhead, the Rank value, and the subband information of the AI / ML model indicated by the second indication information in the first configuration information.

[0242] The allocation of CSI feedback resources for CSI feedback is performed according to the total feedback overhead.

[0243] For example, in the case where the CSI feedback overhead of the single-layer AI / ML model indicated by the second indication information is K0 bits and the Rank value is r, the total feedback overhead of CSI reporting on the terminal device side is K0xr.

[0244] In some embodiments, the Rank value is an estimated value of the Rank value in the CSI reporting of the terminal device by the network device, which is obtained, for example, by the network device based on the Rank value reported by the terminal device at a historical time, or by the network device based on uplink reference information (such as SRS (Sounding Reference Signal) reference signal), or by the network device based on the Part I part of the CSI reporting at the nearest historical time, and the present application does not limit this.

[0245] In some embodiments, for different CSI reporting instants, the network device respectively calculates total feedback overhead for each CSI reporting instant, and allocates CSI feedback resources for CSI feedback at the corresponding instant for the terminal device according to the total feedback overhead at each instant.

[0246] In some embodiments, for a time window (for example, the first time window), the network device allocates CSI feedback resources for CSI feedback within the time window for the terminal device according to the total feedback overhead at each instant within the time window.

[0247] For example, the network device allocates CSI feedback resources for the corresponding instant respectively according to the total feedback overhead at each instant within the time window, or the network device allocates CSI feedback resources for each instant within the time window according to the maximum value in the total feedback overhead at each instant within the time window.

[0248] In some embodiments, for a time window (for example, the first time window), the network device allocates the same CSI feedback resources for each instant within the time window. For example, in the case where the network device allocates CSI feedback resources for each instant within the time window according to the maximum value in the total feedback overhead at each instant within the time window, the network device allocates the same CSI feedback resources for each instant within the time window.

[0249] The implementation manners of this embodiment can refer to the embodiments of the first aspect.

[0250] The above various embodiments are only exemplarily described, but the present application is not limited thereto, and can be appropriately modified on the basis of the above various embodiments. For example, the above various embodiments can be used alone, or one or more of the above various embodiments can be combined.

[0251] Through the above embodiments, the network device sends configuration information related to function and / or model selection of the terminal device to the terminal device, so that the terminal device can perform function and / or model selection according to the configuration information, so that the terminal device can select an AI / ML model according to the configuration of the network device, thereby standardizing the AI / ML model used on the terminal device side and the network device side, solving the multi-vendor / inter-vendor collaboration problem of bilateral models, reducing the complexity of vendor collaboration, thereby solving the problems in the prior art.

[0252] Embodiments of the third aspect

[0253] The embodiment of the application provides an information transceiving device, which can be a terminal device, or can be one or more components or assemblies arranged in the terminal device. The information transceiving device is generated based on the same inventive concept as the information transceiving method in the embodiment of the first aspect of the application, and the solving principle is similar, so the implementation of the information transceiving device is described with reference to the implementation of the information transceiving method in the embodiment of the first aspect of the application, and the repeated parts will not be described herein. The term "unit" or "module" used below can be a combination of software and / or hardware that realizes a predetermined function. Although the system described in the following embodiment is preferably realized in software, the realization of hardware or a combination of software and hardware is also possible and is conceived.

[0254] FIG. 7 is a schematic diagram of an information transceiving device according to an embodiment of the application. As shown in FIG. 7, the information transceiving device 700 includes:

[0255] a receiver 701 configured to receive first configuration information from a network device, the first configuration information being related to AI / ML model selection of the terminal device;

[0256] a transmitter 702 configured to report channel state information (CSI) feedback information to the network device, the CSI feedback information being generated by the terminal device using an AI / ML model selected according to the first configuration information.

[0257] The first configuration information includes first indication information indicating input information of the AI / ML model on the terminal device side, and / or second indication information indicating CSI feedback overhead of the AI / ML model on the terminal device side.

[0258] In some embodiments, the first indication information indicates a number of historical CSI information of the AI / ML model on the terminal device side at at least one time point.

[0259] In some embodiments, the number of historical CSI information of the AI / ML model on the terminal device side is one or more of the number of historical CSI information supported by the terminal device for use by the AI / ML model reported to the network device; or

[0260] The number of historical CSI information of the AI / ML model on the terminal device side is less than or equal to the maximum value of the number of historical CSI information supported by the terminal device for use by the AI / ML model reported to the network device.

[0261] In some embodiments, for different time points,

[0262] The number of historical CSI information of the AI / ML model on the terminal device side is the same; or

[0263] The number of historical CSI information of the AI / ML model at the input terminal device side is different.

[0264] In some embodiments, the first indication information comprises the number of historical CSI information of the AI / ML model at the input terminal device side at multiple time instants within at least one first time window.

[0265] In some embodiments, for multiple time instants within one first time window, the number of historical CSI information of the AI / ML model at the input terminal device side is the same, or the number of historical CSI information of the AI / ML model at the input terminal device side is different; and / or

[0266] In different first time windows, the number of historical CSI information of the AI / ML model at the input terminal device side is the same, or the number of historical CSI information of the AI / ML model at the input terminal device side is different.

[0267] In some embodiments, the historical CSI information comprises at least one of:

[0268] measurement CSI information at a historical time instant;

[0269] information derived based on the measurement CSI information at the historical time instant;

[0270] output information of the AI / ML model at the input terminal device side at a historical time instant;

[0271] information derived based on the output information of the AI / ML model at the input terminal device side at a historical time instant.

[0272] In some embodiments, the second indication information indicates the CSI feedback overhead of the AI / ML model at the input terminal device side at at least one time instant.

[0273] In some embodiments, the CSI feedback overhead is one or more of supported feedback overhead values reported by the terminal device to the network device; or

[0274] The CSI feedback overhead is less than or equal to a maximum value of supported feedback overhead reported by the terminal device to the network device.

[0275] In some embodiments, for different time instants,

[0276] the CSI feedback overhead is the same; or

[0277] the CSI feedback overhead is different.

[0278] In some embodiments, the second indication information indicates the CSI feedback overhead at multiple time instants within at least one first time window.

[0279] In some embodiments, the CSI feedback overheads are the same or different for multiple time instants within a first time window.

[0280] The CSI feedback overheads are the same or different in different first time windows.

[0281] In some embodiments, the CSI feedback overheads gradually decrease over time for multiple time instants within a first time window; or

[0282] The CSI feedback overhead is the largest for a first time instant within a first time window, and the CSI feedback overheads are equal and smaller than the CSI feedback overhead for the first time instant for time instants after the first time instant.

[0283] In some embodiments, one AI / ML model supports one or more CSI feedback overheads.

[0284] The selected AI / ML model is the same or different for multiple time instants with different CSI feedback overheads.

[0285] In some embodiments, the first configuration information further includes time window configuration information for configuring the first time window.

[0286] In some embodiments, the time window configuration information includes size information of the first time window, and / or starting position information of the first time window.

[0287] In some embodiments, the AI / ML model on the terminal device side is associated with an AI / ML model configured on the network device side.

[0288] The number of historical CSI information input to the AI / ML model on the terminal device side is the same as or different from the number of historical CSI information input to the AI / ML model on the network device side.

[0289] In some embodiments, the number of historical CSI information input to the AI / ML model on the network device side is one or more.

[0290] In some embodiments, the number of historical CSI information input to the AI / ML model on the network device side is the same or different for different time instants; and / or

[0291] The number of historical CSI information input to the AI / ML model on the network device side is the same or different for multiple time instants within a second time window; and / or

[0292] In different second time windows, the number of historical CSI information input to the AI / ML model on the network device side is the same, or the number of historical CSI information input to the AI / ML model on the network device side is different.

[0293] In some embodiments, the network device configures the terminal device with a first time window, and the first time window is the same as the second time window, or the first time window is different from the second time window.

[0294] In some embodiments, the historical CSI information input to the AI / ML model on the network device side includes at least one of the following:

[0295] output information of the AI / ML model on the network device side at a historical moment;

[0296] information derived based on the output information of the AI / ML model on the network device side at a historical moment;

[0297] input information of the AI / ML model on the network device side at a historical moment;

[0298] information derived based on the input information of the AI / ML model on the network device side at a historical moment.

[0299] In some embodiments, the reporting of the CSI feedback information is periodic, or semi-persistent, or aperiodic.

[0300] In some embodiments, for different moments,

[0301] the number of ports of the CSI information measured or reported by the terminal device is the same, or the number of ports of the CSI information measured or reported by the terminal device is different; and / or

[0302] the carrier frequency of the CSI information measured or reported by the terminal device is the same, or the carrier frequency of the CSI information measured or reported by the terminal device is different.

[0303] In one embodiment, the information transceiver device further comprises (not shown):

[0304] a processor configured to select an AI / ML model (first AI / ML model) at at least one moment according to the first configuration information;

[0305] In one embodiment,

[0306] the transmitter is configured to send capability reporting information to the network device.

[0307] The capability reporting information includes at least one of the following:

[0308] a number of historical CSI information used by the supported at least one AI / ML model;

[0309] a maximum value of a number of historical CSI information used by the supported AI / ML model;

[0310] a CSI feedback overhead of the supported at least one AI / ML model;

[0311] a maximum value of a CSI feedback overhead of the supported AI / ML model.

[0312] In some embodiments, the network device determines the first configuration information according to the capability reporting information of the terminal device, for details, please refer to the related embodiments herein, which are not repeated here.

[0313] The above various embodiments are only exemplarily described, but the present application is not limited thereto, and can be appropriately modified on the basis of the above various embodiments. For example, the above various embodiments can be used alone, or one or more of the above various embodiments can be combined.

[0314] It is worth noting that the above only describes the components or modules related to the present application, but the present application is not limited thereto. The information transceiver 700 can also include other components or modules, and the specific content of these components or modules can be referred to related technologies.

[0315] In addition, for the sake of simplicity, only the connection relationship or signal path between the various components or modules is exemplarily shown in FIG. 7, but it should be clear to those skilled in the art that various related technologies such as bus connection can be used. The above various components or modules can be implemented by hardware facilities such as processors, memories, transmitters, receivers, etc.; the present application is not limited thereto.

[0316] Through the above embodiments, the terminal device performs function and / or model selection according to the configuration information related to the function and / or model selection of the terminal device from the network device, so that the terminal device can select the AI / ML model according to the configuration of the network device, thereby standardizing the AI / ML model used on the terminal device side and the network device side, solving the multi-vendor / inter-vendor collaboration problem of bilateral models, reducing the complexity of vendor collaboration, thereby solving the problems in the prior art.

[0317] Embodiments of the fourth aspect

[0318] The embodiment of the present application provides an information sending device. The device can be a network device, or a component or assembly arranged in the network device. The information sending device is based on the same inventive concept as the information sending method in the embodiment of the second aspect of the present application, and the solving principle is similar. Therefore, the implementation of the information sending device is described in the embodiment of the second aspect of the present application, and the repeated part will not be described. The term "unit" or "module" used below can be a combination of software and / or hardware that can realize a predetermined function. Although the system described in the following embodiment is preferably realized by software, the realization of hardware or a combination of software and hardware is also possible and is conceived.

[0319] FIG. 8 is a schematic diagram of an information sending device according to an embodiment of the present application. As shown in FIG. 8, the information sending device 800 includes:

[0320] The transmitter 800 is configured to send first configuration information to the terminal device, and the first configuration information is related to AI / ML model selection of the terminal device.

[0321] In some embodiments, the first configuration information and the function and / or model information are as described in the first aspect embodiment, and the implementation of the transmitter 800 corresponds to 601 of the second aspect embodiment, which will not be described herein.

[0322] In some embodiments, the first configuration information includes first indication information indicating input information of the AI / ML model, and / or second indication information indicating CSI feedback overhead of the AI / ML model.

[0323] In some embodiments, the device further includes (not shown):

[0324] The processor is configured to determine the total number of bits of CSI reporting on the terminal device side according to at least one of the CSI feedback overhead of the AI / ML model, the Rank value, and the sub-band information indicated in the first configuration information;

[0325] The uplink resource allocation for CSI feedback is performed according to the total number of bits.

[0326] In some embodiments, the device further includes (not shown):

[0327] The receiver is configured to receive the capability reporting information sent by the terminal device, and determine the first configuration information according to the capability reporting information;

[0328] The capability reporting information includes at least one of:

[0329] The number of at least one AI / ML model using historical CSI information;

[0330] a maximum value of the number of supported AI / ML models using historical CSI information;

[0331] CSI feedback overhead of the at least one supported AI / ML model;

[0332] a maximum value of CSI feedback overhead of the supported AI / ML model.

[0333] The above embodiments are only exemplary for the embodiments of the present application, but the present application is not limited thereto, and can be appropriately modified on the basis of the above embodiments. For example, the above embodiments can be used alone, or one or more of the above embodiments can be combined.

[0334] It is worth noting that the above only illustrates the components or modules related to the present application, but the present application is not limited thereto. The information sending device 800 can also include other components or modules, and the specific content of these components or modules can be referred to related technologies.

[0335] In addition, for the sake of simplicity, only the connection relationship or signal path between the components or modules is exemplarily shown in FIG. 8, but it should be clear to those skilled in the art that various related technologies such as bus connection can be used. The above components or modules can be implemented by hardware facilities such as processors, memories, transmitters, receivers, etc.; the present application is not limited thereto.

[0336] Through the above embodiments, the network device sends the terminal device configuration information related to the function and / or model selection of the terminal device, so that the terminal device can select the function and / or model according to the configuration information, so that the terminal device can select the AI / ML model according to the configuration of the network device, thereby standardizing the AI / ML model used on the terminal device side and the network device side, solving the multi-vendor / inter-vendor collaboration problem of bilateral models, reducing the complexity of vendor collaboration, thereby solving the problems in the prior art.

[0337] Embodiments of the fifth aspect

[0338] The embodiments of the present application also provide a communication system, which can be referred to FIG. 1, and the same content as the embodiments of the first to fourth aspects will not be repeated.

[0339] In some embodiments, the communication system 100 can at least include: a network device 101 and / or a terminal device 102, the terminal device receives first configuration information from the network device, the first configuration information is related to AI / ML model selection of the terminal device; and reports channel state information (CSI) feedback information to the network device; the CSI feedback information is generated by the terminal device using an AI / ML model selected according to the first configuration information.

[0340] In some embodiments, the implementation of the first configuration information and the terminal setting execution operation described above can refer to the embodiments of the first aspect and the second aspect, which will not be described here.

[0341] Embodiments of the present application also provide a network device, which can be a base station for example, but the present application is not limited thereto, and can also be other network devices.

[0342] FIG. 9 is a schematic diagram of the network device according to an embodiment of the present application. As shown in FIG. 9, the network device 900 can include a processor 910 (such as a central processing unit CPU) and a memory 920; the memory 920 is coupled to the processor 910. The memory 920 can store various data; in addition, it also stores a program 930 for information processing, and executes the program 930 under the control of the processor 910.

[0343] For example, the processor 910 can be configured to execute the program to implement the information sending method as described in the embodiments of the second aspect. For example, the processor 910 can be configured to perform the following control: sending first configuration information to the terminal device, the first configuration information is related to AI / ML model selection of the terminal device.

[0344] In addition, as shown in FIG. 9, the network device 900 can also include a transceiver 940 and an antenna 950, etc.; wherein the functions of the above components are similar to those of the prior art, which will not be described here. It is worth noting that the network device 900 does not necessarily include all the components shown in FIG. 9; in addition, the network device 900 can also include components not shown in FIG. 9, which can refer to the prior art.

[0345] Embodiments of the present application also provide a terminal device, but the present application is not limited thereto, and can also be other devices.

[0346] FIG. 10 is a schematic diagram of a terminal device according to an embodiment of the present application. As shown in FIG. 10, the terminal device 1000 can include a processor 1010 and a memory 1020. The memory 1020 stores data and programs and is coupled to the processor 1010. It is worth noting that this figure is exemplary; other types of structures can also be used to supplement or replace this structure to implement telecommunications functions or other functions.

[0347] For example, the processor 1010 can be configured to execute programs to implement the information transceiving method according to the embodiments of the first aspect. For example, the processor 1010 can be configured to perform the following control:

[0348] receiving first configuration information from the network device, the first configuration information being related to AI / ML model selection of the terminal device;

[0349] reporting channel state information (CSI) feedback information to the network device, the CSI feedback information being generated by the terminal device using an AI / ML model selected according to the first configuration information.

[0350] As shown in FIG. 10, the terminal device 1010 can further include a communication module 1030, an input unit 1040, a display 1050, and a power supply 1060. The functions of the above-mentioned components are similar to those of the prior art, and will not be described here. It is worth noting that the terminal device 1000 does not necessarily include all the components shown in FIG. 10, and the above-mentioned components are not essential; in addition, the terminal device 1000 can also include components not shown in FIG. 10, which can be referred to the prior art.

[0351] The embodiments of the present application also provide a computer program, wherein when the program is executed in a terminal device, the program causes the terminal device to perform the information transceiving method according to the embodiments of the first aspect.

[0352] The embodiments of the present application also provide a storage medium storing a computer program, wherein the computer program causes a terminal device to perform the information transceiving method according to the embodiments of the first aspect.

[0353] The embodiments of the present application also provide a computer program, wherein when the program is executed in a network device, the program causes the network device to perform the information sending method according to the embodiments of the second aspect.

[0354] The embodiments of the present application also provide a storage medium storing a computer program, wherein the computer program causes a network device to perform the information sending method according to the embodiments of the second aspect.

[0355] The apparatuses and methods described above can be implemented by hardware, or by hardware combined with software. The present application relates to a computer readable program, which, when executed by a logic component, causes the logic component to implement the above-described apparatuses or components, or causes the logic component to implement the above-described various methods or steps. The present application also relates to a storage medium for storing the above program, such as a hard disk, a magnetic disk, an optical disk, a DVD, a flash memory, etc.

[0356] The methods / apparatuses described in connection with the embodiments of the present application can be directly embodied as hardware, software modules executed by a processor, or a combination thereof. For example, one or more of the functional blocks shown in the figures and / or a combination of one or more of the functional blocks can correspond to individual software modules of a computer program flow, or to individual hardware modules. The software modules can correspond to individual steps shown in the figures. The hardware modules can be implemented by, for example, fixing the software modules using a field programmable gate array (FPGA).

[0357] The software modules can be located in a RAM memory, a flash memory, a ROM memory, an EPROM memory, an EEPROM memory, a register, a hard disk, a mobile disk, a CD-ROM, or any other form of storage medium known in the art. A storage medium can be coupled to the processor, so that the processor can read information from, and write information to, the storage medium; or the storage medium can be integral to the processor. The processor and the storage medium can be located in an ASIC. The software modules can be stored in the memory of the mobile terminal, or in a memory card that can be inserted into the mobile terminal. For example, if the device (e.g., the mobile terminal) uses a MEGA-SIM card or a large-capacity flash memory device, the software modules can be stored in the MEGA-SIM card or the large-capacity flash memory device.

[0358] One or more of the functional blocks shown in the figures and / or a combination of one or more of the functional blocks can be implemented as 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 device, discrete gate or transistor logic, discrete hardware components, or any appropriate combination thereof, for performing the functions described in the present application. One or more of the functional blocks shown in the figures and / or a combination of one or more of the functional blocks can also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in communication with a DSP core, or any other such configuration.

[0359] The present application is described in detail above with reference to specific embodiments. However, it should be clear to those skilled in the art that the description is exemplary and not limiting to the scope of protection of the present application. Those skilled in the art can make various modifications and changes to the present application according to the spirit and principles of the present application, and these modifications and changes are within the scope of the present application.

[0360] In connection with the embodiments including the above embodiments, the following notes are also disclosed:

[0361] 1. An information transceiving method applied to a terminal device, the method comprising:

[0362] receiving first configuration information from a network device, the first configuration information being related to AI / ML model selection of the terminal device;

[0363] reporting channel state information (CSI) feedback information to the network device, the CSI feedback information being generated by the terminal device using an AI / ML model selected according to the first configuration information.

[0364] 2. The method of note 1, wherein the first configuration information comprises first indication information indicating input information of the AI / ML model on the terminal device side, and / or the first configuration information comprises second indication information indicating CSI feedback overhead of the AI / ML model on the terminal device side;

[0365] the first indication information comprises a number of historical CSI information input to the AI / ML model on the terminal device side at multiple time instants within at least one first time window;

[0366] the second indication information indicates the CSI feedback overhead of the AI / ML model on the terminal device side at at least one time instant.

[0367] 3. The method of note 2, wherein,

[0368] for multiple time instants within one first time window, the number of historical CSI information input to the AI / ML model on the terminal device side is the same, or the number of historical CSI information input to the AI / ML model on the terminal device side is different; and / or

[0369] in different first time windows, the number of historical CSI information input to the AI / ML model on the terminal device side is the same, or the number of historical CSI information input to the AI / ML model on the terminal device side is different.

[0370] 4. The method of note 2, wherein the historical CSI information comprises at least one of:

[0371] measurement CSI information at the historical moment;

[0372] information derived based on the measurement CSI information at the historical moment;

[0373] output information of the AI / ML model at the terminal device side at the historical moment;

[0374] information derived based on the output information of the AI / ML model at the terminal device side at the historical moment.

[0375] 5. The method of appendage 2, wherein the second indication information comprises CSI feedback overheads at multiple moments within at least one first time window;

[0376] the CSI feedback overheads at multiple moments within one first time window gradually decrease over time; or

[0377] the CSI feedback overhead at the first moment within one first time window is the largest, and the CSI feedback overheads at each moment after the first moment are equal and smaller than the CSI feedback overhead at the first moment.

[0378] 6. The method of appendage 2, wherein,

[0379] one AI / ML model supports one or more CSI feedback overheads;

[0380] the selected AI / ML models are the same or different at multiple moments with different CSI feedback overheads.

[0381] 7. The method of appendage 1, wherein the network device configures a first time window for the terminal device, and the network device is configured with a second time window, wherein the first time window is the same as the second time window, or the first time window is different from the second time window.

[0382] 8. The method of appendage 1, wherein,

[0383] the historical CSI information input to the AI / ML model at the network device side comprises at least one of:

[0384] output information of the AI / ML model at the network device side at the historical moment;

[0385] information derived based on the output information of the AI / ML model at the network device side at the historical moment;

[0386] input information of the AI / ML model at the network device side at the historical moment;

[0387] information derived based on the input information of the AI / ML model at the network device side at the historical moment.

[0388] 9. The method of clause 1, wherein the reporting of the CSI feedback information is periodic, or semi-persistent, or aperiodic.

[0389] 10. The method of clause 1, wherein, for different time instances,

[0390] the number of ports of the CSI information measured or reported by the terminal device is the same, or the number of ports of the CSI information measured or reported by the terminal device is different; and / or

[0391] the carrier frequencies of the CSI information measured or reported by the terminal device are the same, or the carrier frequencies of the CSI information measured or reported by the terminal device are different.

Claims

1. An information transceiving apparatus configured in a terminal device, characterized by comprising: The apparatus comprises: a receiver configured to receive first configuration information from a network device, the first configuration information being related to AI / ML model selection of the terminal device; a transmitter configured to report channel state information (CSI) feedback information to the network device, the CSI feedback information being generated by the terminal device using an AI / ML model selected according to the first configuration information.

2. The apparatus of claim 1, wherein, The first configuration information comprises first indication information indicating input information of the AI / ML model on the terminal device side, and / or second indication information indicating CSI feedback overhead of the AI / ML model on the terminal device side.

3. The apparatus of claim 2, wherein, The first indication information indicates a number of historical CSI information input to the AI / ML model on the terminal device side at least one time point.

4. The apparatus of claim 3, wherein The number of historical CSI information input to the AI / ML model on the terminal device side is one or more of a number of historical CSI information supported by the terminal device and reported to the network device and used by the AI / ML model; or The number of historical CSI information input to the AI / ML model on the terminal device side is less than or equal to a maximum value of a number of historical CSI information supported by the terminal device and reported to the network device and used by the AI / ML model.

5. The apparatus of claim 3, wherein, For different time points, The number of historical CSI information input to the AI / ML model on the terminal device side is the same; or The number of historical CSI information input to the AI / ML model on the terminal device side is different.

6. The apparatus of claim 3, wherein, The first indication information comprises a number of historical CSI information input to the AI / ML model on the terminal device side at a plurality of time points within at least one first time window.

7. The apparatus of claim 2, wherein, The second indication information indicates CSI feedback overhead of the AI / ML model on the terminal device side at at least one time point.

8. The apparatus of claim 7, wherein The CSI feedback overhead is one or more of a feedback overhead value supported by the terminal device and reported to the network device; or The CSI feedback overhead is less than or equal to a maximum value of a feedback overhead supported by the terminal device and reported to the network device. For different time points, 9. The apparatus of claim 7, wherein, The CSI feedback overhead is the same; or The CSI feedback overhead is different. The second indication information indicates CSI feedback overhead at a plurality of time points within at least one first time window.

10. The apparatus of claim 7, wherein, 11. The apparatus of claim 10, wherein For a plurality of time points within one first time window, the CSI feedback overhead is the same or different; or In different first time windows, the CSI feedback overhead is the same or different. The first configuration information further comprises time window configuration information for configuring the first time window.

12. The apparatus of claim 6 or 10, wherein, The time window configuration information comprises size information of the first time window, and / or starting position information of the first time window.

13. The apparatus of claim 12, wherein, 14. The apparatus of claim 2, wherein The AI / ML model on the terminal device side is associated with an AI / ML model configured on the network device side. ​ The number of historical CSI information input to the AI / ML model at the network device side is the same as or different from the number of historical CSI information input to the AI / ML model at the terminal device side.

15. The apparatus of claim 14, wherein, The number of historical CSI information input to the AI / ML model at the network device side is one or more.

16. The apparatus of claim 15, wherein, The number of historical CSI information input to the AI / ML model at the network device side is the same or different for different time instants; and / or The number of historical CSI information input to the AI / ML model at the network device side is the same or different for different time instants; and / or The number of historical CSI information input to the AI / ML model at the network device side is the same or different for different time instants; and / or 17. An information transmitting apparatus configured to a network device, characterized by comprising: The apparatus comprises: a transmitter configured to transmit first configuration information to the terminal device, the first configuration information being related to AI / ML model selection of the terminal device.

18. The apparatus of claim 17, wherein, The first configuration information comprises first indication information indicating input information of the AI / ML model, and / or second indication information indicating CSI feedback overhead of the AI / ML model.

19. The apparatus of claim 17, wherein, The apparatus further comprises: a processor configured to determine a total number of bits of CSI reporting at the terminal device according to at least one of CSI feedback overhead, Rank value, and subband information of the AI / ML model indicated in the first configuration information; perform uplink resource allocation for CSI feedback according to the total number of bits.

20. The apparatus of claim 17, wherein, The apparatus further comprises: a receiver configured to receive capability reporting information transmitted by the terminal device, and determine the first configuration information according to the capability reporting information; The capability reporting information comprises at least one of: a number of historical CSI information supported by at least one AI / ML model; a maximum value of the number of historical CSI information supported by the AI / ML model; CSI feedback overhead of at least one AI / ML model supported; a maximum value of the CSI feedback overhead of the AI / ML model supported.

Images