Model identification method and device, medium, communication equipment and communication system

Abstract

A model identification method and apparatus, a storage medium, a communication device and a communication system, the method comprising: sending first information to a network device, the first information being used by the network device to determine an AI / ML function supported by a terminal and / or an AI / ML model supported by the terminal. According to the invention, the network equipment can complete model identification of the terminal.

Description

Model recognition method, device, medium, communication equipment and communication system Technical Field

[0001] The present disclosure relates to the field of communication technology, and in particular to a model recognition method, apparatus, medium, communication equipment, and communication system. Background Art

[0002] In communication systems, artificial intelligence (AI) / machine learning (ML) models can be deployed in network nodes to perform prediction and inference through AI / ML models to improve the performance of communication systems.

[0003] Summary of the Invention

[0004] Embodiments of the present disclosure provide a model recognition method, apparatus, storage medium, communication equipment, and communication system.

[0005] According to a first aspect of an embodiment of the present disclosure, a model recognition method is provided, which is executed by a terminal, and the method includes:

[0006] Sending first information to a network device, where the first information is used by the network device to determine an AI / ML function supported by the terminal and / or an AI / ML model supported by the terminal.

[0007] According to a second aspect of an embodiment of the present disclosure, a model recognition method is provided, which is performed by a network device and includes:

[0008] receiving first information sent by a terminal;

[0009] Determine, based on the first information, an AI / ML function supported by the terminal and / or an AI / ML model supported by the terminal.

[0010] According to a third aspect of an embodiment of the present disclosure, a model recognition method is provided, which is performed by a communication system, the communication system including a terminal and a network device, and the method includes:

[0011] first information sent by the terminal to the network device;

[0012] The network device determines, based on the first information, an AI / ML function supported by the terminal and / or an AI / ML model supported by the terminal.

[0013] According to a fourth aspect of an embodiment of the present disclosure, a model recognition device is provided, and the method includes:

[0014] The first sending module is configured to send first information to the network device, where the first information is used by the network device to determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal.

[0015] According to a fifth aspect of an embodiment of the present disclosure, there is provided a model recognition device, comprising:

[0016] A receiving module, configured to receive first information sent by a terminal;

[0017] The first determining module is configured to determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal based on the first information.

[0018] According to a sixth aspect of an embodiment of the present disclosure, a terminal is provided, including:

[0019] a first processor;

[0020] a first memory for storing first processor executable instructions;

[0021] The first processor is configured to execute the executable instructions to implement the model recognition method provided in the first aspect of the present disclosure.

[0022] According to a seventh aspect of an embodiment of the present disclosure, a network device is provided, including:

[0023] a second processor;

[0024] a second memory for storing second processor executable instructions;

[0025] The second processor is configured to execute the executable instructions to implement the model recognition method provided in the second aspect of the present disclosure.

[0026] According to an eighth aspect of an embodiment of the present disclosure, a storage medium is provided, on which computer program instructions are stored. When the program instructions are executed by a processor, the steps of the model recognition method provided by the first aspect or the second aspect of the present disclosure are implemented.

[0027] According to the ninth aspect of the embodiment of the present disclosure, the embodiment of the present disclosure provides a program product, which, when executed by a terminal, enables the terminal to execute the model recognition method provided in the first aspect, or when executed by a network device, enables the network device to execute the second aspect.

[0028] According to a tenth aspect of the embodiments of the present disclosure, the embodiments of the present disclosure provide a computer program, which, when executed on a computer, enables the computer to execute the steps of the model recognition method provided in the first aspect or the second aspect.

[0029] The technical solution provided by the embodiments of the present disclosure may include the following beneficial effects: the terminal can send first information to the network device, so that the network device can determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal based on the first information, thereby completing model recognition of the terminal.

[0030] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The above and / or additional aspects and advantages of the present disclosure will become apparent and readily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0032] FIG1a is a schematic diagram showing a communication system according to some embodiments.

[0033] FIG1 b is a schematic diagram showing CSI compression feedback and recovery according to some embodiments.

[0034] FIG2 a is a flowchart illustrating a model identification method according to some embodiments.

[0035] FIG2 b is a flowchart illustrating a model identification method according to some embodiments.

[0036] FIG2 c is a flowchart illustrating a model identification method according to some embodiments.

[0037] FIG3 a is a flowchart illustrating a model identification method according to some embodiments.

[0038] FIG3 b is a flowchart illustrating a model identification method according to some embodiments.

[0039] FIG3 c is a flowchart illustrating a model identification method according to some embodiments.

[0040] FIG3 d is a flowchart illustrating a model identification method according to some embodiments.

[0041] FIG3e is a flowchart illustrating a model identification method according to some embodiments.

[0042] FIG3 f is a flowchart illustrating a model identification method according to some embodiments.

[0043] FIG4 a is a flowchart illustrating a model identification method according to some embodiments.

[0044] FIG4 b is a flowchart illustrating a model identification method according to some embodiments.

[0045] FIG4 c is a flowchart illustrating a model identification method according to some embodiments.

[0046] FIG4 d is a flowchart illustrating a model identification method according to some embodiments.

[0047] FIG5 a is a flowchart illustrating a model identification method according to some embodiments.

[0048] FIG5 b is a flowchart illustrating a model identification method according to some embodiments.

[0049] FIG5 c is a flowchart illustrating a model identification method according to some embodiments.

[0050] FIG5 d is a flowchart illustrating a model identification method according to some embodiments.

[0051] FIG5e is a flowchart illustrating a model identification method according to some embodiments.

[0052] FIG5 f is a flowchart illustrating a model identification method according to some embodiments.

[0053] FIG6 is a flowchart showing a model identification method according to some embodiments.

[0054] FIG7 is a schematic diagram illustrating the functions and model ID of AI-based CSI compression feedback according to some embodiments.

[0055] FIG8 is a block diagram showing a model recognition device according to some embodiments.

[0056] FIG9 is a block diagram showing a model recognition device according to some embodiments.

[0057] FIG10 is a block diagram of a terminal according to some embodiments.

[0058] FIG11 is a block diagram of a network device according to some embodiments. DETAILED DESCRIPTION

[0059] The present disclosure provides a "model recognition method." In some embodiments, the terms "model recognition method" and "information processing method" and "communication method" are interchangeable; the terms "model recognition device" and "information processing device" and "communication device" are interchangeable; and the terms "communication system" and "information processing system" and "communication system" are interchangeable.

[0060] The embodiments of the present disclosure are not exhaustive and are merely illustrative of some embodiments, and are not intended to be a specific limitation on the scope of protection of the present disclosure. In the absence of contradiction, each step in a certain embodiment can be implemented as an independent embodiment, and the steps can be arbitrarily combined. For example, a solution after removing some steps in a certain embodiment can also be implemented as an independent embodiment, and the order of the steps in a certain embodiment can be arbitrarily exchanged. In addition, the optional implementation methods in a certain embodiment can be arbitrarily combined; in addition, the embodiments can be arbitrarily combined. For example, some or all steps of different embodiments can be arbitrarily combined, and a certain embodiment can be arbitrarily combined with the optional implementation methods of other embodiments.

[0061] The terms used in the embodiments of the present disclosure are only for the purpose of describing specific embodiments and are not intended to limit the present disclosure. The singular expressions "a," "an," "the," "above," "said," "aforementioned," "this," etc. in the embodiments of the present disclosure also include plural expressions, unless there is a clear contrary indication in the context. The predefined in the embodiments of the present disclosure can be understood as definition, pre-definition, storage, pre-storage, pre-negotiation, pre-configuration, curing, or pre-firing, etc.

[0062] Prefixes such as "first" and "second" in the embodiments of the present disclosure are only used to distinguish different description objects, and have no limiting effect on the position, order, priority, quantity or content of the described objects. For example, if the described object is a "field", then the ordinal number before the "field" in the "first field" and the "second field" does not limit the position or order between the "fields", and "first" and "second" do not limit whether the "fields" they modify are in the same message, nor do they limit the order of the "first field" and the "second field". For another example, if the described object is a "level", then the ordinal number before the "level" in the "first level" and the "second level" does not limit the priority between the "levels". For another example, the number of described objects is not limited by the ordinal number and can be one or more. Taking the "first device" as an example, the number of "devices" can be one or more. Furthermore, the objects modified by different prefixes can be the same or different. For example, if the object being described is a "device," then the "first device" and the "second device" can be devices of the same type or different types. For another example, if the object being described is "information," then the "first information" and the "second information" can be information of the same content or information of different contents. In summary, the use of prefixes such as ordinal numbers in the embodiments of the present disclosure to distinguish between the described objects does not constitute a limitation on the described objects. For a description of the described objects, refer to the contextual description in the claims or embodiments, and the use of such prefixes should not constitute an unnecessary limitation.

[0063] In the embodiments of the present disclosure, "multiple" refers to two or more. In the embodiments of the present disclosure, "and / or" is used to describe the association relationship of associated objects, indicating three relationships that can exist independently. For example, A and / or B can mean: A exists alone, B exists alone, and A and B exist at the same time. In the embodiments of the present disclosure, descriptions such as "at least one of Al, A2, ..., An (or at least one, at least one)" include the situation where any one of Al, A2, ..., An exists alone, and also include any combination of any multiple of Al, A2, ..., An, each of which can exist alone; for example, the description of "at least one of A, B, C" includes the situation where A exists alone, B exists alone, C exists alone, A and B are combined, A and C are combined, B and C are combined, and A and B and C are combined.

[0064] In some embodiments, descriptions such as "in one case A, in another case B," or "in response to one case A, in response to another case B," may include the following technical solutions depending on the situation: executing A independently of B (in some embodiments, A); executing B independently of A (in some embodiments, B); selectively executing A and B (in some embodiments, selecting between A and B); and executing both A and B (in some embodiments, A and B). The same applies when there are more branches, such as A, B, and C.

[0065] In some embodiments, terms such as "in response to...", "in response to determining...", "in the case of...", "at the time of...", "when...", "if...", "if...", etc. can be used interchangeably.

[0066] In some embodiments, “including A,” “comprising A,” “used to indicate A,” and “carrying A” can be interpreted as directly carrying A or indirectly indicating A.

[0067] In some embodiments, terms such as "greater than", "greater than or equal to", "above", "higher than", and "not less than" can be replaced with each other, and terms such as "less than", "less than or equal to", "below", "lower than", and "not greater than" can be replaced with each other.

[0068] In some embodiments, terms such as "network element", "node", "function", "device", "equipment", "system", "chip", and "chip system" can be used interchangeably.

[0069] In some embodiments, the terms terminal, terminal device, user equipment (UE), user terminal, mobile station (MS), mobile terminal (MT), subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, mobile device, wireless device, wireless communication device, remote device, mobile subscriber station, access terminal, mobile terminal, wireless terminal, remote terminal, handset, user agent, mobile client, client, etc. can be used interchangeably.

[0070] In addition, each element, each row, or each column in the table of the embodiment of the present disclosure can be implemented as an independent embodiment, and the combination of any elements, any rows, and any columns can also be implemented as an independent embodiment.

[0071] It should be noted that all actions of acquiring signals, information or data in the present disclosure are carried out in compliance with the corresponding data protection laws and policies of the country where they are located and with the authorization given by the owner of the corresponding device.

[0072] Although operations are described in a particular order in the accompanying drawings in the disclosed embodiments, this should not be construed as requiring that the operations be performed in the particular order shown or in a serial order, or that all of the operations shown be performed to obtain the desired result. In certain circumstances, multitasking and parallel processing may be advantageous. Furthermore, sending multiple messages via the same message may also be advantageous.

[0073] The following describes an application environment of a model recognition method provided by this embodiment.

[0074] FIG1a is a schematic structural diagram of a communication system according to an embodiment of the present disclosure.

[0075] As shown in FIG. 1 a , a communication system 100 includes a terminal 101 and a network device 102 .

[0076] In some embodiments, the terminal 101 includes, for example, a mobile phone, a wearable device, an Internet of Things device, a car with communication function, a smart car, a tablet computer, a computer with wireless transceiver function, a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal device in industrial control, a wireless terminal device in self-driving, a wireless terminal device in remote medical surgery, a wireless terminal device in a smart grid, a wireless terminal device in transportation safety, a wireless terminal device in a smart city, and at least one of a wireless terminal device in a smart home, but is not limited thereto.

[0077] In some embodiments, the network device 102 may include at least one of an access network device and a core network device.

[0078] In some embodiments, the access network device is, for example, a node or device that accesses a terminal to a wireless network, such as a radio access network (RAN). The access network device may include an evolved NodeB (eNB), a next generation evolved NodeB (ng-eNB), a next generation NodeB (gNB), a node B (NB), a home node B (HNB), a home evolved nodeB (HeNB), a wireless backhaul device, a radio network controller (RNC), a base station controller (BSC), a base transceiver station (BTS), a base band unit (BBU), a mobile switching center, a base station in a 6G communication system, an open base station (Open RAN), a cloud base station (Cloud RAN), a base station in other communication systems, and at least one of an access node in a wireless fidelity (WiFi) system, but is not limited thereto.

[0079] In some embodiments, the technical solution of the present disclosure can be applied to the Open RAN architecture. In this case, the interfaces between or within the access network devices involved in the embodiments of the present disclosure can be transformed into internal interfaces of the Open RAN, and the processes and information interactions between these internal interfaces can be implemented through software or programs.

[0080] In some embodiments, the access network device can be composed of a centralized unit (CU) and a distributed unit (DU), where the CU can also be called a control unit. The CU-DU structure can be used to split the protocol layer of the access network device, with the functions of some protocol layers centrally controlled by the CU, and the functions of the remaining part or all of the protocol layers distributed in the DU, which is centrally controlled by the CU, but is not limited to this.

[0081] In some embodiments, the core network device may be a single device including one or more network elements, or may be multiple devices or device groups each including all or part of the one or more network elements. The network elements may be virtual or physical.

[0082] It can be understood that the communication system described in the embodiment of the present disclosure is for the purpose of more clearly illustrating the technical solution of the embodiment of the present disclosure, and does not constitute a limitation on the technical solution provided by the embodiment of the present disclosure. Ordinary technicians in this field can know that with the evolution of the system architecture and the emergence of new business scenarios, the technical solution provided by the embodiment of the present disclosure is also applicable to similar technical problems.

[0083] The following embodiments of the present disclosure may be applied to the communication system 100 shown in FIG1a, or a portion thereof, but are not limited thereto. The entities shown in FIG1a are illustrative only. The communication system may include all or a portion of the entities shown in FIG1a, or may include other entities other than those shown in FIG1a. The number and form of the entities may be arbitrary. The connection relationship between the entities is illustrative only. The entities may be connected or disconnected, and the connection may be in any manner, including direct or indirect, wired or wireless.

[0084] The embodiments of the present disclosure can be applied to Long Term Evolution (LTE), LTE-Advanced (LTE-A), LTE-Beyond (LTE-B), SUPER 3G, IMT-Advanced, 4th generation mobile communication system (4G), 5th generation mobile communication system (5G), 5G new radio (NR), future radio access (FRA), new radio access technology (RAT), new radio (NR), new radio access (NX), future generation radio access (FX), Global System for Mobile communications (GSM (registered trademark)), CDMA2000, Ultra Mobile Broadband (UMB), IEEE 802.11 (Wi-Fi (registered trademark)), IEEE 802.16 (WiMAX (registered trademark)), IEEE 802.20, Ultra-WideBand (UWB), Bluetooth (registered trademark), Public Land Mobile Network (PLMN) networks, Device-to-Device (D2D) systems, Machine-to-Machine (M2M) systems, Internet of Things (IoT) systems, Vehicle-to-Everything (V2X), systems utilizing other communication methods, and next-generation systems based on and extending these methods. Furthermore, multiple systems may be combined (for example, a combination of LTE or LTE-A with 5G).

[0085] In some embodiments of the present disclosure, AI / ML models may be deployed in network nodes, and prediction and inference may be performed through the AI / ML models to improve the performance of the communication system.

[0086] Optionally, the network node may be a terminal or a network device.

[0087] Optionally, AI technology can be used to reduce terminal feedback overhead or improve the accuracy of channel state information (CSI) feedback. For example, a bilateral AI / ML model based on a terminal-side CSI generation model and a network-side CSI recovery model can be used to implement compressed CSI feedback and recovery, respectively. For another example, time-domain CSI prediction can be implemented based on the UE-side AI / ML model.

[0088] For example, Figure 1b below shows a schematic diagram of implementing CSI compression feedback and recovery based on a bilateral AI / ML model. The UE compresses the downlink channel information H through the CSI generation model, quantizes it into a binary bit stream s, and sends it to the gNB. The gNB recovers H', which is approximately the original downlink information, through the CSI recovery model.

[0089] For the AI / ML model on the UE side or some models on the UE side in the bilateral model, the network and the UE can reach a consensus through AI / ML function identification and / or AI / ML model identification, thereby managing the life cycle of the model. Among them, AI / ML function identification and / or AI / ML model identification can be referred to as identifying the model.

[0090] Figure 2a is a flow chart of a model recognition method according to some embodiments. As shown in Figure 2a, the model recognition method can be used in a terminal and includes the following steps.

[0091] In step S210a, first information is sent to the network device.

[0092] In some embodiments, the first information is used by the network device to determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal.

[0093] In some embodiments, the network device may receive first information sent by the terminal, and determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal based on the first information, thereby completing the process of identifying the models supported by the terminal.

[0094] In some embodiments, the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal may be all AI / ML functions and / or AI / ML models deployed and available in the terminal, or may be the AI / ML functions and / or AI / ML models currently used by the terminal, which may be determined based on the first information sent by the terminal, or based on a policy for sending the first information by the terminal.

[0095] In some embodiments, the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal may also be referred to as AI / ML functions corresponding to the terminal and / or AI / ML models corresponding to the terminal.

[0096] In some embodiments, the first information may separately include information used by the network device to determine the AI / ML functions supported by the terminal, or may separately include information used by the network device to determine the AI / ML models supported by the terminal, or may simultaneously include information used by the network device to determine the AI / ML functions supported by the terminal, and information used by the network device to determine the AI / ML models supported by the terminal.

[0097] The embodiments of the present disclosure provide that the terminal may send first information to the network device by way of information reporting, so that the network device may determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal based on the first information, thereby completing model recognition of the terminal.

[0098] In some embodiments, after determining the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal, the network device can perform lifecycle management on the terminal model based on the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal.

[0099] In some embodiments, the network device performs lifecycle management of the terminal's model based on the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal, which may include management of activation, deactivation, selection, switching, etc. of the terminal's AI / ML model.

[0100] In some implementations, the network device may instruct the terminal to activate the AI / ML function supported by the terminal based on the AI / ML function supported by the terminal, thereby explicitly activating the AI / ML function supported by the terminal. For example, the network device may instruct the terminal to activate the AI / ML function supported by the terminal through an indication message.

[0101] In some implementations, the network device may configure the terminal with configuration parameters corresponding to the AI / ML functions supported by the terminal based on the AI / ML functions supported by the terminal, thereby implicitly activating the AI / ML functions supported by the terminal.

[0102] In some implementations, the network device may instruct the terminal to activate the AI / ML model supported by the terminal based on the AI / ML model supported by the terminal, thereby activating the AI / ML model supported by the terminal. For example, the network device may instruct the terminal to activate the AI / ML model supported by the terminal through an instruction message.

[0103] In some implementations, the network device may configure the terminal with configuration parameters corresponding to the AI / ML model supported by the terminal based on the AI / ML model supported by the terminal, thereby activating the AI / ML model supported by the terminal.

[0104] In some implementations, the first information may include one or more of the following information:

[0105] First function information, second function information, third function information, first model identifier, second model identifier, third model identifier, terminal-associated meta information, and feature type information.

[0106] In some implementations, the feature type information is used to indicate the feature types supported by the terminal.

[0107] Thus, in the embodiments of the present disclosure, the feature type information included in the first information can be used to identify the feature types supported by the terminal. Optionally, the feature type information can be combined with other types of first information, for example, with at least one of the first function information, the second function information, the third function information, the first model identifier, the second model identifier, the third model identifier, and terminal-associated meta-information, so that the network device can determine the AI / ML function and / or AI / ML model corresponding to the feature type information.

[0108] In some embodiments, the feature types may include AI-based CSI compression feedback and AI-based time-domain CSI prediction.

[0109] For example, when the feature type supported by the terminal is AI-based CSI compression feedback, the first information used to determine the AI / ML function supported by the terminal may be selected for reporting. For another example, when the feature type supported by the terminal is AI-based time-domain CSI prediction, the first information used to determine the AI / ML model supported by the terminal may be selected for reporting.

[0110] It should be noted that in the above examples, the correspondence between the feature types supported by the terminal and the first information used to determine the AI / ML functions and / or AI / ML models supported by the terminal can be swapped according to actual needs. For example, when the feature type supported by the terminal is AI-based time-domain CSI prediction, the first information used to determine the AI / ML functions supported by the terminal can be selected for reporting.

[0111] In some implementations, the first capability information can be understood as fine-grained capability information that describes all capability types corresponding to a complete set of AI / ML capabilities supported by the terminal. Thus, the network device can determine the AI / ML capabilities supported by the terminal based on the received first capability information.

[0112] Optionally, the first function information may describe all function types corresponding to a complete set of AI / ML functions supported by the terminal in the form of a function identifier. In this case, the correspondence between the function identifier and the fine-grained function information may be pre-stored. For example, in a mobile scenario, assuming that function 1 corresponds to the terminal being able to predict channel information on a maximum of 10 slots, and the interval between adjacent slots is 2, which can be used for reasoning below 30 km / h, function 2 is defined as the terminal being able to predict channel information on a maximum of 20 slots, and the interval between adjacent slots is 5, which can be used for reasoning below 60 km / h. If the first function information includes function 1, the network device may determine, based on the first function information and the preset correspondence, that the AI / ML function supported by the terminal is the channel information that can be predicted on a maximum of 10 slots, and the interval between adjacent slots is 2, which can be used for reasoning below 30 km / h.

[0113] Optionally, the first function information can describe all function types corresponding to a complete set of AI / ML functions supported by the terminal in the form of a combination of multiple function types. In this case, the first function information can be, for example, channel information on a maximum of 10 slots that can be predicted, and the interval between adjacent slots is 2, which can be used for reasoning below 30 km / h. Thus, the network device can determine, based on the first function information, that the AI / ML function supported by the terminal is channel information on a maximum of 10 slots that can be predicted, and the interval between adjacent slots is 2, which can be used for reasoning below 30 km / h.

[0114] Among them, the channel information on a maximum of 10 slots and the interval between adjacent slots can be predicted. The reasoning used for speeds below 30 km / h can be understood as part of the functional types corresponding to a complete set of AI / ML functions.

[0115] In some embodiments, a complete set of AI / ML functions may be all function types supported by a model. Therefore, each function type corresponding to a complete set of AI / ML functions may be determined based on the actual situation of the terminal model.

[0116] Continuing with the previous example, a model in the terminal supports the function of predicting channel information on a maximum of 10 slots, with an interval of 2 between adjacent slots, and can be used for inference below 30 km / h. Then, each function type corresponding to a complete set of AI / ML functions may include the channel information on the maximum preset number of slots that can be predicted, the interval between adjacent slots, and the inferred movement speed.

[0117] In some embodiments, the second function information can be understood as coarse-grained function information, which is used to describe the partial function types corresponding to a complete set of AI / ML functions supported by the terminal. The first model identifier is used by the network device to determine the partial function types corresponding to a complete set of AI / ML functions supported by the terminal based on a preset correspondence.

[0118] Thus, the network device can jointly determine the AI / ML function supported by the terminal based on the received second function information and the first model identifier.

[0119] Exemplarily, assuming that all function types corresponding to a complete set of AI / ML functions can include channel scenarios and the payload of the model output, at this time, the second function information can, for example, support the Uma (Urban Macro, urban macrocell) / UMi (Urban Micro, urban microcell) scenario in the channel scenario, and support the medium and high-speed mobile scenario in the channel scenario. In the Uma / UMi scenario, the first model identifier can, for example, include ID (Identity document) 1 and ID2, which respectively correspond to the AI / ML function with a maximum payload size of 120 bits after the maximum compressed information quantization and the AI / ML function with a maximum payload size of 60 bits. In the medium and high-speed mobile scenario, the first model identifier can, for example, include ID3 and ID4, which respectively correspond to the AI / ML model with a maximum payload size of 120 bits after the maximum compressed information quantization and the AI / ML model with a maximum payload size of 60 bits. In this case, if the second function information includes supporting the Uma / UMi scenario in the channel scenario, and the first model identifier includes ID1, the network device can determine that the AI / ML function supported by the terminal is to support the Uma / UMi scenario based on the second function information and the first model identifier, and supports the AI / ML function with a maximum payload size of 120 bits after quantization of the maximum compressed information.

[0120] In addition, considering that the functional information may correspond to the model, for example, some functions are only supported by a certain model, therefore, after determining the AI / ML functions supported by the terminal, the corresponding AI / ML model can be obtained. Therefore, the network device can determine the AI / ML model supported by the terminal based on the functional information and the model identifier. Therefore, in some embodiments, the first information may include third functional information and a third model identifier, and the third functional information and the third model identifier are used by the network device to determine the AI / ML model supported by the terminal.

[0121] In some embodiments, the third function information can be understood as coarse-grained function information that describes some function types corresponding to a complete set of AI / ML functions supported by the terminal. The third model identifier is used by the network device to determine some function types corresponding to a complete set of AI / ML functions supported by the terminal based on a preset correspondence.

[0122] In some implementations, the metadata associated with the terminal is used by the network device to determine parameter configurations of AI / ML functions supported by the terminal.

[0123] Considering that the terminal model may be a model with strong generalization capabilities, it can be used for reasoning or prediction of various different parameters. For example, it can be used for reasoning or prediction of different numbers of ports. In this case, in order to fully determine the AI / ML functions supported by the terminal, such as accurately determining the number of ports currently used by the terminal, after determining that the AI / ML function supported by the terminal is to predict channel information on a maximum of 10 slots, and the interval between adjacent slots is 2, and it can be used for reasoning below 30 km / h, the number of ports supported by the terminal can be further determined in combination with the metadata associated with the terminal.

[0124] In some implementations, the metadata associated with the terminal is used by the network device to determine parameter configurations of the AI / ML model supported by the terminal.

[0125] Similarly, considering that the model supported by the terminal may be a model with strong generalization capabilities, it can be used for reasoning or prediction of a variety of different parameters, for example, it can be used for reasoning or prediction of different numbers of ports. In this case, in order to fully determine the AI / ML model supported by the terminal, such as accurately determining the number of ports currently used by the terminal, it can be further determined in combination with the metadata associated with the terminal.

[0126] In some embodiments, the first information may separately include terminal-associated meta-information, that is, the network device may determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal through the terminal-associated meta-information.

[0127] In some embodiments, the terminal-associated meta-information may be used in combination with other information used to determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal. For example, the first information may include first function information and terminal-associated meta-information, the first information may include second function information, a first model identifier, and terminal-associated meta-information, the first information may include a second model identifier and terminal-associated meta-information, and the first information may include third function information, a third model identifier, and terminal-associated meta-information.

[0128] In some implementations, one piece of second identification information may correspond to a unique model, so that the network device may determine the AI / ML model supported by the terminal through the second identification information.

[0129] In some implementations, one second identification information may correspond to multiple models, so that the network device may determine multiple AI / ML models supported by the terminal through the second identification information.

[0130] In some embodiments, when the first information is used by a network device to determine an AI / ML function supported by a terminal, the first information may include:

[0131] First function information is used to describe the AI / ML functions supported by the terminal; or

[0132] The second function information and the first model identifier are used by the network device to determine the AI / ML function supported by the terminal; or

[0133] A second model identifier, where the second model identifier is used by the network device to determine the AI / ML function supported by the terminal based on a preset correspondence; or

[0134] Meta information associated with the terminal is used by network devices to determine the parameter configuration of AI / ML functions supported by the terminal.

[0135] In an embodiment of the present disclosure, it is stipulated that the terminal may send first function information to the network device so that the network may determine the AI / ML function supported by the terminal based on the first function information. The terminal may also send second function information and a first model identifier to the network device so that the network may determine the AI / ML function supported by the terminal based on the second function information and the first model identifier. The terminal may also send a second model identifier to the network device so that the network may determine the AI / ML function supported by the terminal based on the second model identifier. The terminal may also send terminal-associated metadata to the network device so that the network may determine the AI / ML function supported by the terminal based on the terminal-associated metadata.

[0136] In some embodiments, when the first information is used by a network device to determine an AI / ML model supported by a terminal, the first information includes:

[0137] The second model identifier is used by the network device to determine the AI / ML model supported by the terminal; or

[0138] Third function information and a third model identifier, where the third function information and the third model identifier are used by the network device to determine the AI / ML model supported by the terminal; or

[0139] Meta information associated with the terminal is used by network devices to determine the parameter configuration of the AI / ML model supported by the terminal.

[0140] The embodiments of the present disclosure provide that the terminal may send third function information and a third model identifier to the network device, so that the network may determine the AI / ML model supported by the terminal based on the third function information and the third model identifier. The terminal may also send a second model identifier to the network device, so that the network may determine the AI / ML model supported by the terminal based on the second model identifier. The terminal may also send terminal-associated meta-information to the network device, so that the network may determine the AI / ML model supported by the terminal based on the terminal-associated meta-information.

[0141] In some implementations, the terminal-associated meta-information may include one or more of the following information:

[0142] The transmitting antenna port corresponding to the terminal, the receiving antenna port corresponding to the terminal, the carrier information of the terminal, the number of subbands supported by the terminal, and the location information of the terminal.

[0143] In some implementations, the location information of the terminal may be the ID information of the cell where the terminal is located.

[0144] In some implementations, the value range of the second model identifier is divided according to one or more of the following information: feature types supported by the terminal, function information supported by the terminal, and meta information associated with the terminal.

[0145] In the embodiment of the present disclosure, the second model identifier can be divided into different value ranges in advance according to the feature types supported by the terminal, the second model identifier can be divided into different value ranges according to the functional information supported by the terminal, the second model identifier can be divided into different value ranges according to the meta-information associated with the terminal, the second model identifier can be divided into different value ranges according to the feature types and functional information supported by the terminal, and the second model identifier can be divided into different value ranges according to the feature types, functional information and meta-information supported by the terminal, etc.

[0146] For example, following the above example, assuming that the terminal supports feature types including AI-based CSI compression feedback and AI-based time-domain CSI prediction, and there can be 1000 models for AI-based CSI compression feedback and 1000 models for AI-based time-domain CSI prediction, then the identifiers corresponding to the 1000 models of AI-based CSI compression feedback can be numbered 1-1000, and the identifiers corresponding to the 1000 models of AI-based time-domain CSI prediction can be numbered 1001-2000. Thus, assuming that the second model identifier received by the network device is 500, the AI ​​model numbered 500 can be determined from the models corresponding to the AI-based CSI compression feedback, thereby accurately determining the AI / ML model supported by the terminal, and the subsequent network device can perform lifecycle management on the model.

[0147] In some implementations, when the feature type supported by the terminal includes AI-based CSI compression feedback, the first capability information includes one or more of the following information:

[0148] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0149] In some implementations, when the feature type supported by the terminal includes AI-based time-domain CSI prediction, the first capability information includes one or more of the following information:

[0150] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0151] In some implementations, when the feature type supported by the terminal includes AI-based CSI compression feedback, the second function information includes any one of the following information:

[0152] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0153] The third function information includes any of the following information:

[0154] Model output includes payload, channel scenario, network configuration parameters, and mobile speed.

[0155] In some implementations, when the feature type supported by the terminal includes AI-based time-domain CSI prediction, the second function information includes any one of the following information:

[0156] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed;

[0157] The third function information includes any of the following information:

[0158] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0159] In some implementations, the predicted time domain condition requirement may include the number of predicted CSIs and the time interval between two adjacent predicted CSIs.

[0160] In some embodiments, when the feature type supported by the terminal includes AI-based CSI compression feedback, combined with the foregoing content, it can be seen that the generation part model in the terminal and the recovery part model in the network device constitute a bilateral AI / ML model. In this case, the first model identifier may include a model ID and / or a model pairing ID, the second model identifier includes a model ID and / or a model pairing ID, and the third model identifier includes a model ID and / or a model pairing ID. Among them, the model pairing ID can be called a Paring model ID.

[0161] In some embodiments, when the feature type supported by the terminal includes AI-based time domain prediction, the first model identifier includes a model ID, the second model identifier includes a model ID, and the third model identifier includes a model ID.

[0162] Figure 2b is a flow chart of a model identification method according to some embodiments. As shown in Figure 2b, the model identification method can be used in a terminal and includes the following steps.

[0163] In step S210b, a first instruction sent by a network device is received.

[0164] In step S220b, first information is sent to the network device according to the first instruction.

[0165] In some embodiments, the first information is used by the network device to determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal.

[0166] In some embodiments, the first instruction is used to request AI / ML functions supported by the terminal and / or AI / ML models supported by the terminal.

[0167] In some implementations, the first instruction is used to request the terminal to return first information.

[0168] In some embodiments, the first information may include a feature type, so that the network device can determine the AI / ML functions and / or AI / ML models supported by the terminal of the corresponding feature type based on the feature type.

[0169] In some implementations, the first information may include one or more of the following information:

[0170] First function information, second function information, third function information, first model identifier, second model identifier, third model identifier, terminal-associated meta information, and feature type information.

[0171] In some implementations, the first information is used by the network device to determine the AI / ML functions supported by the terminal, and the first information includes:

[0172] First function information is used to describe the AI / ML functions supported by the terminal; or

[0173] The second function information and the first model identifier are used by the network device to determine the AI / ML function supported by the terminal; or

[0174] A second model identifier, where the second model identifier is used by the network device to determine the AI / ML function supported by the terminal based on a preset correspondence; or

[0175] Meta information associated with the terminal is used by network devices to determine the parameter configuration of AI / ML functions supported by the terminal.

[0176] In some implementations, the first information is used by the network device to determine an AI / ML model supported by the terminal, and the first information includes:

[0177] The second model identifier is used by the network device to determine the AI / ML model supported by the terminal; or

[0178] Third function information and a third model identifier, where the third function information and the third model identifier are used by the network device to determine the AI / ML model supported by the terminal; or

[0179] Meta information associated with the terminal is used by network devices to determine the parameter configuration of the AI / ML model supported by the terminal.

[0180] In some implementations, the terminal-associated meta-information includes one or more of the following information:

[0181] The transmitting antenna port corresponding to the terminal, the receiving antenna port corresponding to the terminal, the carrier information of the terminal, the number of subbands supported by the terminal, and the location information of the terminal.

[0182] In some implementations, the value range of the second model identifier is divided according to one or more of the following information: feature types supported by the terminal, function information supported by the terminal, and meta information associated with the terminal.

[0183] In some implementations, the feature type supported by the terminal includes AI-based CSI compression feedback, and the first function information includes one or more of the following information:

[0184] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0185] The feature types supported by the terminal include AI-based time-domain CSI prediction, and the first function information includes one or more of the following information:

[0186] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0187] In some implementations, when the feature type supported by the terminal includes AI-based CSI compression feedback, the second function information includes any one of the following information:

[0188] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0189] The third function information includes any of the following information:

[0190] Model output includes payload, channel scenario, network configuration parameters, and mobile speed.

[0191] In some implementations, when the feature type supported by the terminal includes AI-based time-domain CSI prediction, the second function information includes any one of the following information:

[0192] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed;

[0193] The third function information includes any of the following information:

[0194] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0195] In some embodiments, the feature type supported by the terminal includes AI-based CSI compression feedback, the first model identifier includes a model ID and / or a model pairing ID, the second model identifier includes a model ID and / or a model pairing ID, and the third model identifier includes a model ID and / or a model pairing ID;

[0196] The feature types supported by the terminal include AI-based time domain prediction, the first model identifier includes a model ID, the second model identifier includes a model ID, and the third model identifier includes a model ID.

[0197] The embodiment of the present disclosure provides that the network device can send a first instruction to the terminal, and the terminal can send first information to the network device according to the first instruction after receiving the first instruction. In other words, the embodiment of the present disclosure provides the conditions or timing for the terminal to send the first instruction to the network device.

[0198] Among them, the detailed description of the relevant steps or parameter definitions in the embodiments of the present disclosure can be referred to the above content and will not be repeated here.

[0199] Figure 2c is a flow chart of a model recognition method according to some embodiments. As shown in Figure 2c, the model recognition method can be used in a terminal and includes the following steps.

[0200] In step S210c, a second instruction sent by the network device is received.

[0201] In step S220c, according to the second instruction, first information corresponding to the characteristic type is sent to the network device.

[0202] In some implementations, the second instruction includes a feature type.

[0203] In some embodiments, the first information of the corresponding feature type is used to determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal of the corresponding feature type.

[0204] In some embodiments, the second instruction is used to request AI / ML functions supported by the terminal and / or AI / ML models supported by the terminal of the corresponding feature type.

[0205] In some implementations, the first instruction is used to request the terminal to return first information corresponding to the feature type.

[0206] In some implementations, the first information may include one or more of the following information:

[0207] First function information, second function information, third function information, first model identifier, second model identifier, third model identifier, terminal-associated meta information, and feature type information.

[0208] In some implementations, the first information is used by the network device to determine the AI / ML functions supported by the terminal, and the first information includes:

[0209] First function information is used to describe the AI / ML functions supported by the terminal; or

[0210] The second function information and the first model identifier are used by the network device to determine the AI / ML function supported by the terminal; or

[0211] A second model identifier, where the second model identifier is used by the network device to determine the AI / ML function supported by the terminal based on a preset correspondence; or

[0212] Meta information associated with the terminal is used by network devices to determine the parameter configuration of AI / ML functions supported by the terminal.

[0213] In some implementations, the first information is used by the network device to determine an AI / ML model supported by the terminal, and the first information includes:

[0214] The second model identifier is used by the network device to determine the AI / ML model supported by the terminal; or

[0215] Third function information and a third model identifier, where the third function information and the third model identifier are used by the network device to determine the AI / ML model supported by the terminal; or

[0216] Meta information associated with the terminal is used by network devices to determine the parameter configuration of the AI / ML model supported by the terminal.

[0217] In some implementations, the terminal-associated meta-information includes one or more of the following information:

[0218] The transmitting antenna port corresponding to the terminal, the receiving antenna port corresponding to the terminal, the carrier information of the terminal, the number of subbands supported by the terminal, and the location information of the terminal.

[0219] In some implementations, the value range of the second model identifier is divided according to one or more of the following information: feature types supported by the terminal, function information supported by the terminal, and meta information associated with the terminal.

[0220] In some implementations, the feature type supported by the terminal includes AI-based CSI compression feedback, and the first function information includes one or more of the following information:

[0221] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0222] The feature types supported by the terminal include AI-based time-domain CSI prediction, and the first function information includes one or more of the following information:

[0223] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0224] In some implementations, when the feature type supported by the terminal includes AI-based CSI compression feedback, the second function information includes any one of the following information:

[0225] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0226] The third function information includes any of the following information:

[0227] Model output includes payload, channel scenario, network configuration parameters, and mobile speed.

[0228] In some implementations, when the feature type supported by the terminal includes AI-based time-domain CSI prediction, the second function information includes any one of the following information:

[0229] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed;

[0230] The third function information includes any of the following information:

[0231] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0232] In some embodiments, the feature type supported by the terminal includes AI-based CSI compression feedback, the first model identifier includes a model ID and / or a model pairing ID, the second model identifier includes a model ID and / or a model pairing ID, and the third model identifier includes a model ID and / or a model pairing ID;

[0233] The feature types supported by the terminal include AI-based time domain prediction, the first model identifier includes a model ID, the second model identifier includes a model ID, and the third model identifier includes a model ID.

[0234] The disclosed embodiments provide that a network device may issue a second instruction to a terminal, and after receiving the second instruction, the terminal may, in accordance with the second instruction, send first information of a corresponding feature type to the network device. In other words, the disclosed embodiments provide that a network device may specify that a terminal send first information of a specified feature type, so that the network device can determine the AI / ML functions and / or AI / ML models supported by the terminal of the specified feature type.

[0235] Among them, the detailed description of the relevant steps or parameter definitions in the embodiments of the present disclosure can be referred to the above content and will not be repeated here.

[0236] Figure 3a is a flow chart of a model recognition method according to some other embodiments. As shown in Figure 3a, the model recognition method can be used in a terminal and includes the following steps.

[0237] In step S310a, first information is sent to a network device, where the first information includes first function information.

[0238] In some implementations, the first function information is used by the network device to determine the AI / ML functions supported by the terminal.

[0239] In some embodiments, in addition to the first function information, the first information may also include terminal-associated meta-information, and the first function information and the terminal-associated meta-information are used by the network device to determine the AI / ML function supported by the terminal.

[0240] The detailed description of step S310a can refer to the above embodiment and will not be repeated here.

[0241] The embodiments of the present disclosure provide that the terminal can send first function information to the network device, so that the network device can determine the AI / ML function supported by the terminal based on the first function information.

[0242] Figure 3b is a flow chart of a model identification method according to some other embodiments. As shown in Figure 3b, the model identification method can be used in a terminal and includes the following steps.

[0243] In step S310b, first information is sent to the network device, where the first information includes second function information and a first model identifier.

[0244] In some embodiments, the second function information and the first model identifier included in the first information are used by the network device to determine the AI / ML function supported by the terminal.

[0245] In some embodiments, in addition to the second function information and the first model identifier, the first information may also include terminal-associated metadata. The second function information, the first model identifier, and the terminal-associated metadata are used by the network device to determine the AI / ML functions supported by the terminal.

[0246] The detailed description of step S310b can refer to the above embodiment and will not be repeated here.

[0247] The embodiment of the present disclosure stipulates that the terminal can send second function information and a first model identifier to the network device, so that the network device can determine the AI / ML function supported by the terminal based on the second function information and the first model identifier.

[0248] Figure 3c is a flow chart of a model identification method according to some other embodiments. As shown in Figure 3c, the model identification method can be used in a terminal and includes the following steps.

[0249] In step S310c, first information is sent to the network device, where the first information includes a second model identifier.

[0250] In some implementations, the second model identifier included in the first information is used by the network device to determine the AI / ML function supported by the terminal.

[0251] In some embodiments, in addition to the second model identifier, the first information may also include terminal-associated meta-information, and the second model identifier and the terminal-associated meta-information are used by the network device to determine the AI / ML functions supported by the terminal.

[0252] The detailed description of step S310c can be referred to the aforementioned embodiment and will not be repeated here.

[0253] The embodiment of the present disclosure stipulates that the terminal can send a second model identifier to the network device, so that the network device can determine the AI / ML function supported by the terminal based on the second model identifier.

[0254] Figure 3d is a flow chart of a model identification method according to some other embodiments. As shown in Figure 3d, the model identification method can be used in a terminal and includes the following steps.

[0255] In step S310d, first information is sent to the network device, where the first information includes meta-information associated with the terminal.

[0256] In some implementations, the terminal-associated metadata included in the first information is used by the network device to determine the AI / ML functions supported by the terminal.

[0257] In some embodiments, the terminal-associated metadata included in the first information is used by the network device to determine the AI / ML model supported by the terminal.

[0258] The detailed description of step S310d can refer to the above embodiment and will not be repeated here.

[0259] The embodiments of the present disclosure provide that a terminal may send terminal-associated meta-information to a network device, so that the network device may determine the AI / ML functions and / or AI / ML models supported by the terminal based on the terminal-associated meta-information.

[0260] Figure 3e is a flow chart of a model identification method according to some other embodiments. As shown in Figure 3e, the model identification method can be used in a terminal and includes the following steps.

[0261] In step S310e, first information is sent to the network device, where the first information includes a second model identifier.

[0262] In some embodiments, the second model identifier included in the first information is used by the network device to determine the AI / ML model supported by the terminal.

[0263] In some embodiments, in addition to the second model identifier, the first information may also include terminal-associated meta-information, and the second model identifier and the terminal-associated meta-information are used by the network device to determine the AI / ML model supported by the terminal.

[0264] The detailed description of step S310e can refer to the above embodiment and will not be repeated here.

[0265] The embodiment of the present disclosure stipulates that the terminal can send a second model identifier to the network device, so that the network device can determine the AI / ML model supported by the terminal based on the second model identifier.

[0266] Figure 3f is a flow chart of a model identification method according to some other embodiments. As shown in Figure 3f, the model identification method can be used in a terminal and includes the following steps.

[0267] In step S310f, first information is sent to the network device, where the first information includes third function information and a third model identifier.

[0268] In some embodiments, the third function information and the third model identifier included in the first information are used by the network device to determine the AI / ML model supported by the terminal.

[0269] In some embodiments, in addition to the third function information and the third model identifier, the first information may also include terminal-associated metadata. The third function information, the third model identifier, and the terminal-associated metadata are used by the network device to determine the AI / ML model supported by the terminal.

[0270] The detailed description of step S310f can be referred to the aforementioned embodiment and will not be repeated here.

[0271] The embodiment of the present disclosure stipulates that the terminal can send third function information and a third model identifier to the network device, so that the network device can determine the AI / ML model supported by the terminal based on the third function information and the third model identifier.

[0272] Figure 4a is a flow chart of a model identification method according to some other embodiments. As shown in Figure 4a, the model identification method can be used in a network device and includes the following steps.

[0273] In step S410a, first information sent by a terminal is received.

[0274] In step S420a, the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal are determined according to the first information.

[0275] In some embodiments, the terminal may send first information to the network device, so that the network device may determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal based on the first information.

[0276] In some implementations, the first information may include one or more of the following information:

[0277] First function information, second function information, third function information, first model identifier, second model identifier, third model identifier, terminal-associated meta information, and feature type information.

[0278] In some implementations, the first information is used by the network device to determine the AI / ML functions supported by the terminal, and the first information includes:

[0279] First function information is used to describe the AI / ML functions supported by the terminal; or

[0280] The second function information and the first model identifier are used by the network device to determine the AI / ML function supported by the terminal; or

[0281] A second model identifier, where the second model identifier is used by the network device to determine the AI / ML function supported by the terminal based on a preset correspondence; or

[0282] Meta information associated with the terminal is used by network devices to determine the parameter configuration of AI / ML functions supported by the terminal.

[0283] In some implementations, the first information is used by the network device to determine an AI / ML model supported by the terminal, and the first information includes:

[0284] The second model identifier is used by the network device to determine the AI / ML model supported by the terminal; or

[0285] Third function information and a third model identifier, where the third function information and the third model identifier are used by the network device to determine the AI / ML model supported by the terminal; or

[0286] Meta information associated with the terminal is used by network devices to determine the parameter configuration of the AI / ML model supported by the terminal.

[0287] In some implementations, the terminal-associated meta-information includes one or more of the following information:

[0288] The transmitting antenna port corresponding to the terminal, the receiving antenna port corresponding to the terminal, the carrier information of the terminal, the number of subbands supported by the terminal, and the location information of the terminal.

[0289] In some implementations, the value range of the second model identifier is divided according to one or more of the following information: feature types supported by the terminal, function information supported by the terminal, and meta information associated with the terminal.

[0290] In some implementations, the feature type supported by the terminal includes AI-based CSI compression feedback, and the first function information includes one or more of the following information:

[0291] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0292] The feature types supported by the terminal include AI-based time-domain CSI prediction, and the first function information includes one or more of the following information:

[0293] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0294] In some implementations, when the feature type supported by the terminal includes AI-based CSI compression feedback, the second function information includes any one of the following information:

[0295] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0296] The third function information includes any of the following information:

[0297] Model output includes payload, channel scenario, network configuration parameters, and mobile speed.

[0298] In some implementations, when the feature type supported by the terminal includes AI-based time-domain CSI prediction, the second function information includes any one of the following information:

[0299] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed;

[0300] The third function information includes any of the following information:

[0301] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0302] In some embodiments, the feature type supported by the terminal includes AI-based CSI compression feedback, the first model identifier includes a model ID and / or a model pairing ID, the second model identifier includes a model ID and / or a model pairing ID, and the third model identifier includes a model ID and / or a model pairing ID;

[0303] The feature types supported by the terminal include AI-based time domain prediction, the first model identifier includes a model ID, the second model identifier includes a model ID, and the third model identifier includes a model ID.

[0304] The embodiments of the present disclosure provide that the terminal may send first information to the network device by way of information reporting, so that the network device may determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal based on the first information, thereby completing model recognition of the terminal.

[0305] Among them, the detailed description of the relevant steps or parameter definitions in the embodiments of the present disclosure can be referred to the above content and will not be repeated here.

[0306] Figure 4b is a flow chart of a model identification method according to some other embodiments. As shown in Figure 4b, the model identification method can be used in a network device and includes the following steps.

[0307] In step S410b, a first instruction is sent to the terminal.

[0308] In some implementations, the first instruction is used by the terminal to send the first information.

[0309] In step S420b, the first information sent by the terminal is received.

[0310] In step S430b, the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal are determined according to the first information.

[0311] In some implementations, the first information may include one or more of the following information:

[0312] First function information, second function information, third function information, first model identifier, second model identifier, third model identifier, terminal-associated meta information, and feature type information.

[0313] In some implementations, the first information is used by the network device to determine the AI / ML functions supported by the terminal, and the first information includes:

[0314] First function information is used to describe the AI / ML functions supported by the terminal; or

[0315] The second function information and the first model identifier are used by the network device to determine the AI / ML function supported by the terminal; or

[0316] A second model identifier, where the second model identifier is used by the network device to determine the AI / ML function supported by the terminal based on a preset correspondence; or

[0317] Meta information associated with the terminal is used by network devices to determine the parameter configuration of AI / ML functions supported by the terminal.

[0318] In some implementations, the first information is used by the network device to determine an AI / ML model supported by the terminal, and the first information includes:

[0319] The second model identifier is used by the network device to determine the AI / ML model supported by the terminal; or

[0320] Third function information and a third model identifier, where the third function information and the third model identifier are used by the network device to determine the AI / ML model supported by the terminal; or

[0321] Meta information associated with the terminal is used by network devices to determine the parameter configuration of the AI / ML model supported by the terminal.

[0322] In some implementations, the terminal-associated meta-information includes one or more of the following information:

[0323] The transmitting antenna port corresponding to the terminal, the receiving antenna port corresponding to the terminal, the carrier information of the terminal, the number of subbands supported by the terminal, and the location information of the terminal.

[0324] In some implementations, the value range of the second model identifier is divided according to one or more of the following information: feature types supported by the terminal, function information supported by the terminal, and meta information associated with the terminal.

[0325] In some implementations, the feature type supported by the terminal includes AI-based CSI compression feedback, and the first function information includes one or more of the following information:

[0326] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0327] The feature types supported by the terminal include AI-based time-domain CSI prediction, and the first function information includes one or more of the following information:

[0328] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0329] In some implementations, when the feature type supported by the terminal includes AI-based CSI compression feedback, the second function information includes any one of the following information:

[0330] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0331] The third function information includes any of the following information:

[0332] Model output includes payload, channel scenario, network configuration parameters, and mobile speed.

[0333] In some implementations, when the feature type supported by the terminal includes AI-based time-domain CSI prediction, the second function information includes any one of the following information:

[0334] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed;

[0335] The third function information includes any of the following information:

[0336] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0337] In some embodiments, the feature type supported by the terminal includes AI-based CSI compression feedback, the first model identifier includes a model ID and / or a model pairing ID, the second model identifier includes a model ID and / or a model pairing ID, and the third model identifier includes a model ID and / or a model pairing ID;

[0338] The feature types supported by the terminal include AI-based time domain prediction, the first model identifier includes a model ID, the second model identifier includes a model ID, and the third model identifier includes a model ID.

[0339] The embodiment of the present disclosure provides that the network device can send a first instruction to the terminal, and the terminal can send first information to the network device according to the first instruction after receiving the first instruction. In other words, the embodiment of the present disclosure provides the conditions or timing for the terminal to send the first instruction to the network device.

[0340] Among them, the detailed description of the relevant steps or parameter definitions in the embodiments of the present disclosure can be referred to the above content and will not be repeated here.

[0341] Figure 4c is a flow chart of a model identification method according to some other embodiments. As shown in Figure 4b, the model identification method can be used in a network device and includes the following steps.

[0342] In step S410c, a second instruction is sent to the terminal.

[0343] In some implementations, the second instruction includes a feature type.

[0344] In some implementations, the second instruction is used by the terminal to send first information corresponding to the characteristic type.

[0345] In step S420c, first information of a corresponding feature type sent by the terminal is received.

[0346] In step S430c, the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal of the corresponding feature type are determined according to the first information of the corresponding feature type.

[0347] In some implementations, the first information may include one or more of the following information:

[0348] First function information, second function information, third function information, first model identifier, second model identifier, third model identifier, terminal-associated meta information, and feature type information.

[0349] In some implementations, the first information is used by the network device to determine the AI / ML functions supported by the terminal, and the first information includes:

[0350] First function information is used to describe the AI / ML functions supported by the terminal; or

[0351] The second function information and the first model identifier are used by the network device to determine the AI / ML function supported by the terminal; or

[0352] A second model identifier, where the second model identifier is used by the network device to determine the AI / ML function supported by the terminal based on a preset correspondence; or

[0353] Meta information associated with the terminal is used by network devices to determine the parameter configuration of AI / ML functions supported by the terminal.

[0354] In some implementations, the first information is used by the network device to determine an AI / ML model supported by the terminal, and the first information includes:

[0355] The second model identifier is used by the network device to determine the AI / ML model supported by the terminal; or

[0356] Third function information and a third model identifier, where the third function information and the third model identifier are used by the network device to determine the AI / ML model supported by the terminal; or

[0357] Meta information associated with the terminal is used by network devices to determine the parameter configuration of the AI / ML model supported by the terminal.

[0358] In some implementations, the terminal-associated meta-information includes one or more of the following information:

[0359] The transmitting antenna port corresponding to the terminal, the receiving antenna port corresponding to the terminal, the carrier information of the terminal, the number of subbands supported by the terminal, and the location information of the terminal.

[0360] In some implementations, the value range of the second model identifier is divided according to one or more of the following information: feature types supported by the terminal, function information supported by the terminal, and meta information associated with the terminal.

[0361] In some implementations, the feature type supported by the terminal includes AI-based CSI compression feedback, and the first function information includes one or more of the following information:

[0362] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0363] The feature types supported by the terminal include AI-based time-domain CSI prediction, and the first function information includes one or more of the following information:

[0364] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0365] In some implementations, when the feature type supported by the terminal includes AI-based CSI compression feedback, the second function information includes any one of the following information:

[0366] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0367] The third function information includes any of the following information:

[0368] Model output includes payload, channel scenario, network configuration parameters, and mobile speed.

[0369] In some implementations, when the feature type supported by the terminal includes AI-based time-domain CSI prediction, the second function information includes any one of the following information:

[0370] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed;

[0371] The third function information includes any of the following information:

[0372] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0373] In some embodiments, the feature type supported by the terminal includes AI-based CSI compression feedback, the first model identifier includes a model ID and / or a model pairing ID, the second model identifier includes a model ID and / or a model pairing ID, and the third model identifier includes a model ID and / or a model pairing ID;

[0374] The feature types supported by the terminal include AI-based time domain prediction, the first model identifier includes a model ID, the second model identifier includes a model ID, and the third model identifier includes a model ID.

[0375] The disclosed embodiments provide that a network device may issue a second instruction to a terminal, and after receiving the second instruction, the terminal may, in accordance with the second instruction, send first information of a corresponding feature type to the network device. In other words, the disclosed embodiments provide that a network device may specify that a terminal send first information of a specified feature type, so that the network device can determine the AI / ML functions and / or AI / ML models supported by the terminal of the specified feature type.

[0376] Among them, the detailed description of the relevant steps or parameter definitions in the embodiments of the present disclosure can be referred to the above content and will not be repeated here.

[0377] Figure 4d is a flow chart of a model identification method according to some other embodiments. As shown in Figure 4d, the model identification method can be used in a network device and includes the following steps.

[0378] In step S410d, the first information sent by the terminal is received.

[0379] In step S420d, the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal are determined based on the first information.

[0380] In step S430d, information for lifecycle management of the model on the terminal side is determined based on the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal.

[0381] In some implementations, the first information may include one or more of the following information:

[0382] First function information, second function information, third function information, first model identifier, second model identifier, third model identifier, terminal-associated meta information, and feature type information.

[0383] In some implementations, the first information is used by the network device to determine the AI / ML functions supported by the terminal, and the first information includes:

[0384] First function information is used to describe the AI / ML functions supported by the terminal; or

[0385] The second function information and the first model identifier are used by the network device to determine the AI / ML function supported by the terminal; or

[0386] A second model identifier, where the second model identifier is used by the network device to determine the AI / ML function supported by the terminal based on a preset correspondence; or

[0387] Meta information associated with the terminal is used by network devices to determine the parameter configuration of AI / ML functions supported by the terminal.

[0388] In some implementations, the first information is used by the network device to determine an AI / ML model supported by the terminal, and the first information includes:

[0389] The second model identifier is used by the network device to determine the AI / ML model supported by the terminal; or

[0390] Third function information and a third model identifier, where the third function information and the third model identifier are used by the network device to determine the AI / ML model supported by the terminal; or

[0391] Meta information associated with the terminal is used by network devices to determine the parameter configuration of the AI / ML model supported by the terminal.

[0392] In some implementations, the terminal-associated meta-information includes one or more of the following information:

[0393] The transmitting antenna port corresponding to the terminal, the receiving antenna port corresponding to the terminal, the carrier information of the terminal, the number of subbands supported by the terminal, and the location information of the terminal.

[0394] In some implementations, the value range of the second model identifier is divided according to one or more of the following information: feature types supported by the terminal, function information supported by the terminal, and meta information associated with the terminal.

[0395] In some implementations, the feature type supported by the terminal includes AI-based CSI compression feedback, and the first function information includes one or more of the following information:

[0396] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0397] The feature types supported by the terminal include AI-based time-domain CSI prediction, and the first function information includes one or more of the following information:

[0398] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0399] In some implementations, when the feature type supported by the terminal includes AI-based CSI compression feedback, the second function information includes any one of the following information:

[0400] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0401] The third function information includes any of the following information:

[0402] Model output includes payload, channel scenario, network configuration parameters, and mobile speed.

[0403] In some implementations, when the feature type supported by the terminal includes AI-based time-domain CSI prediction, the second function information includes any one of the following information:

[0404] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed;

[0405] The third function information includes any of the following information:

[0406] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0407] In some embodiments, the feature type supported by the terminal includes AI-based CSI compression feedback, the first model identifier includes a model ID and / or a model pairing ID, the second model identifier includes a model ID and / or a model pairing ID, and the third model identifier includes a model ID and / or a model pairing ID;

[0408] The feature types supported by the terminal include AI-based time domain prediction, the first model identifier includes a model ID, the second model identifier includes a model ID, and the third model identifier includes a model ID.

[0409] The embodiments of the present disclosure provide that the terminal may send first information to the network device by way of information reporting, so that the network device may determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal based on the first information, and further perform lifecycle management of the model on the terminal side based on the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal.

[0410] Among them, the detailed description of the relevant steps or parameter definitions in the embodiments of the present disclosure can be referred to the above content and will not be repeated here.

[0411] Figure 5a is a flow chart of a model identification method according to some other embodiments. As shown in Figure 5a, the model identification method can be used in a network device and includes the following steps.

[0412] In step S510a, first information sent by a terminal is received, where the first information includes first function information.

[0413] In step S520a, the AI / ML function supported by the terminal is determined according to the first function information.

[0414] In some embodiments, in addition to the first function information, the first information may also include terminal-associated meta-information, so that the network device can determine the AI / ML function supported by the terminal based on the first function information and the terminal-associated meta-information.

[0415] The detailed description of steps S510a and S520a can refer to the above embodiments and will not be repeated here.

[0416] The embodiments of the present disclosure provide that the terminal can send first function information to the network device, so that the network device can determine the AI / ML function supported by the terminal based on the first function information.

[0417] Figure 5b is a flow chart of a model identification method according to some other embodiments. As shown in Figure 5b, the model identification method can be used in a network device and includes the following steps.

[0418] In step S510b, first information sent by the terminal is received, where the first information includes second function information and a first model identifier.

[0419] In step S520b, the AI / ML function supported by the terminal is determined according to the second function information and the first model identifier.

[0420] In some embodiments, in addition to the second function information and the first model identifier, the first information may also include terminal-associated meta-information, so that the network device can determine the AI / ML function supported by the terminal based on the second function information, the first model identifier, and the terminal-associated meta-information.

[0421] The detailed description of steps S510b and S520b can be referred to the aforementioned embodiment and will not be repeated here.

[0422] The embodiment of the present disclosure stipulates that the terminal can send second function information and a first model identifier to the network device, so that the network device can determine the AI / ML function supported by the terminal based on the second function information and the first model identifier.

[0423] Figure 5c is a flow chart of a model identification method according to some other embodiments. As shown in Figure 5c, the model identification method can be used in a network device and includes the following steps.

[0424] In step S510c, first information sent by the terminal is received, where the first information includes a second model identifier.

[0425] In step S520c, the AI / ML function supported by the terminal is determined according to the second model identifier.

[0426] In some embodiments, in addition to the second model identifier, the first information may also include terminal-associated meta-information, so that the network device can determine the AI / ML functions supported by the terminal based on the second model identifier and the terminal-associated meta-information.

[0427] The detailed description of steps S510c and S520c can be referred to the aforementioned embodiment and will not be repeated here.

[0428] The embodiment of the present disclosure stipulates that the terminal can send a second model identifier to the network device, so that the network device can determine the AI / ML function supported by the terminal based on the second model identifier.

[0429] Figure 5d is a flow chart of a model identification method according to some other embodiments. As shown in Figure 5d, the model identification method can be used in a network device and includes the following steps.

[0430] In step S510d, first information sent by the terminal is received, where the first information includes meta-information associated with the terminal.

[0431] In step S520d, the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal are determined based on the meta-information associated with the terminal.

[0432] The detailed description of steps S510d and S520d can be referred to the aforementioned embodiment and will not be repeated here.

[0433] The embodiments of the present disclosure provide that a terminal may send terminal-associated meta-information to a network device, so that the network device may determine the AI / ML functions and / or AI / ML models supported by the terminal based on the terminal-associated meta-information.

[0434] Figure 5e is a flow chart of a model identification method according to some other embodiments. As shown in Figure 5e, the model identification method can be used in a network device and includes the following steps.

[0435] In step S510e, first information sent by the terminal is received, where the first information includes a second model identifier.

[0436] In step S520e, the AI / ML model supported by the terminal is determined according to the second model identifier.

[0437] In some embodiments, in addition to the second model identifier, the first information may also include terminal-associated meta-information, so that the network device can determine the AI / ML model supported by the terminal based on the second model identifier and the terminal-associated meta-information.

[0438] The detailed description of steps S510e and S520e can refer to the aforementioned embodiment and will not be repeated here.

[0439] The embodiment of the present disclosure stipulates that the terminal can send a second model identifier to the network device, so that the network device can determine the AI / ML model supported by the terminal based on the second model identifier.

[0440] Figure 5f is a flow chart of a model identification method according to some other embodiments. As shown in Figure 5f, the model identification method can be used in a network device and includes the following steps.

[0441] In step S510f, first information sent by the terminal is received, where the first information includes third function information and a third model identifier.

[0442] In step S520f, the AI / ML model supported by the terminal is determined according to the third function information and the third model identifier.

[0443] In some embodiments, in addition to the third function information and the third model identifier, the first information may also include terminal-associated meta-information, so that the network device can determine the AI / ML model supported by the terminal based on the third function information, the third model identifier and the terminal-associated meta-information.

[0444] The detailed description of steps S510f and S520f can be referred to the aforementioned embodiment and will not be repeated here.

[0445] The embodiment of the present disclosure stipulates that the terminal can send third function information and a third model identifier to the network device, so that the network device can determine the AI / ML model supported by the terminal based on the third function information and the third model identifier.

[0446] FIG6 is a flow chart of a model identification method according to some further embodiments. As shown in FIG6 , the model identification method can be used in a communication system including a terminal and a network device, and the method includes the following steps:

[0447] In step S610, the terminal sends first information to the network device.

[0448] In step S620, the network device determines the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal based on the first information.

[0449] The detailed description of steps S610 and S620 can refer to the aforementioned embodiment and will not be repeated here.

[0450] The following is another embodiment provided by the present disclosure, in which:

[0451] The base station requires the UE to report the AI / ML functions supported by the AI-based CSI compression feedback, so the base station sends a UE capability query instruction to the UE, such as the UECapabilityEnquiry instruction, to query which functions the UE supports and / or which AI / ML functions it supports. After receiving the query command, the UE reports UEcapabilityInformation to the base station, indicating which AI / ML functions the UE supports, and the UE will also indicate the model ID (model ID) contained in the function to the base station. For example, the AI-based CSI compression feedback corresponding to the UE is a feature type feature of the UE. In the capability report, the UE indicates that it supports the AI-based CSI compression feedback function in the UMa or UMi scenario. In addition, it also supports the AI-based CSI compression feedback function in medium and high-speed mobile scenarios. These two functions are defined as AI function ID 1 (functionality ID 1) and AI function ID 2 (functionality ID 2), respectively. Furthermore, the CSI compression feedback function supporting UMa or UMi scenarios includes an AI / ML model that supports a maximum payload size of 120 bits after inference of the maximum compressed information quantization and an AI / ML model with a maximum payload size of 60 bits, and defines the IDs of the two models as model ID 1 and model ID 2. Accordingly, the AI-based CSI compression feedback function supporting high-speed mobility scenarios also includes an AI / ML model that supports a maximum payload size of 120 bits after inference of the maximum compressed information quantization and two AI / ML models with a maximum payload size of 60 bits. It is assumed that the IDs of the two models are defined as model ID 3 and model ID 4, as shown in Figure 7. The base station can determine the AI / ML functions supported by the UE based on the supported AI / ML functions and AI / ML model IDs reported by the UE.

[0452] In some implementations, after determining the AI / ML functions supported by the UE, the base station activates a model ID under one of the functionalities through RRC signaling based on the current environment or application conditions of the UE. Alternatively, the base station first activates one of the functionalities through RRC signaling, and then activates an AI / ML model ID under the functionality through other signaling such as MAC-CE or DCI.

[0453] The following is another embodiment provided by the present disclosure, in which:

[0454] The base station sends a UE capability query instruction to the UE, such as the UECapabilityEnquiry instruction, to directly query which AI / ML models the UE supports. Each model is associated with a mode ID. The model ID is in turn associated with a certain function. Assume that each model ID has been determined to be associated with a certain function in a predefined manner. In this case, the UE does not need to report the functionality, but only reports the supported model IDs. For example, the UE only reports model ID1 and model ID4 through capability reporting, and as shown in Figure 7, the predefined configuration has indicated that model ID1 and model ID4 correspond to AI reasoning with a maximum payload of 120 bits in UMa / UMi scenarios and AI reasoning with a maximum payload of 60 bits in medium and high-speed mobile scenarios, respectively. The base station knows the specific AI reasoning supported by the UE through the model ID reported by the UE, combined with the predefined description, so that it can configure the UE with appropriate parameters for AI reasoning.

[0455] The following is another embodiment provided by the present disclosure, in which:

[0456] The base station requires the UE to report the AI / ML functions supported by the AI-based time-domain CSI prediction. The base station sends a UE capability query command to the UE, such as the UECapabilityEnquiry command, to query which AI / ML functions the UE supports. After receiving the query command, the UE reports UEcapabilityInformation to the base station, indicating which AI / ML functions the UE supports. For example, the UE reports AI / ML function 1 and AI / ML function 2. Assume that the definition of AI / ML function 1 is that the UE can predict the channel information on a maximum of 10 slots, and the interval between adjacent slots is 2, which can be used for reasoning below 30 km / h; the definition of AI / ML function 2 is that the UE can reason about speeds below 60 km / h, and the definition is that the UE can predict the channel information on a maximum of 20 slots, and the interval between adjacent slots is 5. Therefore, the base station can determine the AI / ML functions supported by the UE based on the supported AI / ML functions reported by the UE.

[0457] In some implementations, the base station indicates which AI / ML function to activate based on the UE's speed and time-domain CSI prediction requirements. Assuming the UE's current speed is 15 km / h and requires prediction of a time slot on the 5th slot, with two adjacent time slots, the base station can directly and explicitly instruct the UE to activate AI / ML function 1 via RRC signaling. The RRC information can configure the number of CSI-RS resources or the interval between adjacent time slots to 2 slots for the base station to the UE, implicitly instructing the UE to activate AI / ML function 1.

[0458] It should be noted that, even if some of the features of the multiple embodiments or features separated by "or" are not feasible, it will not affect other solutions. In addition, the embodiments of the present disclosure can be combined with other embodiments or implementations of the model recognition method and their various optional solutions, unless there is any contradiction, which will not be repeated here.

[0459] FIG8 is a block diagram of a model recognition device 800 according to some embodiments. Referring to FIG8 , the model recognition device 800 is applied to a terminal, and the model recognition device 800 may include:

[0460] The first sending module 810 is configured to send first information to the network device, where the first information is used by the network device to determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal.

[0461] In some implementations, the first sending module 810 includes:

[0462] A first receiving submodule is configured to receive a first instruction sent by the network device;

[0463] The first sending submodule is configured to send first information to the network device according to the first instruction.

[0464] In some implementations, the first sending module 810 includes:

[0465] A second receiving submodule is configured to receive a second instruction sent by the network device, where the second instruction includes a feature type;

[0466] The second sending submodule is configured to send first information corresponding to the feature type to the network device according to the second instruction, where the first information corresponding to the feature type is used to determine the AI / ML function supported by the terminal corresponding to the feature type and / or the AI / ML model supported by the terminal.

[0467] In some implementations, the first information is used by the network device to determine the AI / ML function supported by the terminal, and the first information includes:

[0468] First function information is used to describe the AI / ML function supported by the terminal; or

[0469] Second function information and a first model identifier, where the second function information and the first model identifier are used by the network device to determine the AI / ML function supported by the terminal; or

[0470] a second model identifier, where the second model identifier is used by the network device to determine the AI / ML function supported by the terminal based on a preset correspondence; or

[0471] The meta-information associated with the terminal is used by the network device to determine the parameter configuration of the AI / ML function supported by the terminal.

[0472] In some implementations, the first information is used by the network device to determine an AI / ML model supported by the terminal, and the first information includes:

[0473] A second model identifier is used by the network device to determine the AI / ML model supported by the terminal; or

[0474] third function information and a third model identifier, where the third function information and the third model identifier are used by the network device to determine an AI / ML model supported by the terminal; or

[0475] The meta-information associated with the terminal is used by the network device to determine the parameter configuration of the AI / ML model supported by the terminal.

[0476] In some implementations, the terminal-associated meta-information includes one or more of the following information:

[0477] The transmitting antenna port corresponding to the terminal, the receiving antenna port corresponding to the terminal, the carrier information of the terminal, the number of subbands supported by the terminal, and the location information of the terminal.

[0478] In some implementations, the value range of the second model identifier is divided according to one or more of the following information: feature types supported by the terminal, function information supported by the terminal, and meta information associated with the terminal.

[0479] In some implementations, the feature type supported by the terminal includes AI-based CSI compression feedback, and the first function information includes one or more of the following information:

[0480] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0481] The feature types supported by the terminal include AI-based time domain CSI prediction, and the first function information includes one or more of the following information:

[0482] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0483] In some implementations, the feature type supported by the terminal includes AI-based CSI compression feedback, and the second function information includes any one of the following information:

[0484] The model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0485] The third function information includes any of the following information:

[0486] The model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0487] The feature types supported by the terminal include AI-based time-domain CSI prediction, and the second function information includes any of the following information:

[0488] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed;

[0489] The third function information includes any of the following information:

[0490] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0491] In some embodiments, the feature type supported by the terminal includes AI-based CSI compression feedback, the first model identifier includes a model ID and / or a model pairing ID, the second model identifier includes a model ID and / or a model pairing ID, and the third model identifier includes a model ID and / or a model pairing ID;

[0492] The feature types supported by the terminal include AI-based time domain prediction, the first model identifier includes a model ID, the second model identifier includes a model ID, and the third model identifier includes a model ID.

[0493] FIG9 is a block diagram of a model recognition apparatus 900 according to some embodiments. Referring to FIG9 , the model recognition apparatus 900 is applied to a network device, and the model recognition apparatus 900 may include:

[0494] The receiving module 910 is configured to receive first information sent by the terminal;

[0495] The first determining module 920 is configured to determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal according to the first information.

[0496] In some embodiments, the apparatus 900 further comprises:

[0497] The second sending module is configured to send a first instruction to the terminal, where the first instruction is used by the terminal to send the first information.

[0498] In some embodiments, the apparatus 900 further comprises:

[0499] The third sending module is configured to send a second instruction to the terminal, where the second instruction includes a feature type, and the second instruction is used by the terminal to send first information corresponding to the feature type.

[0500] Correspondingly, the receiving module 910 is further configured to receive first information corresponding to the feature type sent by the terminal.

[0501] In some embodiments, the apparatus 900 further comprises:

[0502] The second determination module is configured to determine information for lifecycle management of the model on the terminal side based on the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal.

[0503] In some implementations, the first information is used by the network device to determine the AI / ML function supported by the terminal, and the first information includes:

[0504] First function information is used to describe the AI / ML function supported by the terminal; or

[0505] Second function information and a first model identifier, where the second function information and the first model identifier are used by the network device to determine the AI / ML function supported by the terminal; or

[0506] a second model identifier, where the second model identifier is used by the network device to determine the AI / ML function supported by the terminal based on a preset correspondence; or

[0507] The meta-information associated with the terminal is used by the network device to determine the parameter configuration of the AI / ML function supported by the terminal.

[0508] In some implementations, the first information is used by the network device to determine an AI / ML model supported by the terminal, and the first information includes:

[0509] A second model identifier is used by the network device to determine the AI / ML model supported by the terminal; or

[0510] third function information and a third model identifier, where the third function information and the third model identifier are used by the network device to determine an AI / ML model supported by the terminal; or

[0511] The meta-information associated with the terminal is used by the network device to determine the parameter configuration of the AI / ML model supported by the terminal.

[0512] In some implementations, the terminal-associated meta-information includes one or more of the following information:

[0513] The transmitting antenna port corresponding to the terminal, the receiving antenna port corresponding to the terminal, the carrier information of the terminal, the number of subbands supported by the terminal, and the location information of the terminal.

[0514] In some implementations, the value range of the second model identifier is divided according to one or more of the following information: feature types supported by the terminal, function information supported by the terminal, and meta information associated with the terminal.

[0515] In some implementations, the feature type supported by the terminal includes AI-based CSI compression feedback, and the first function information includes one or more of the following information:

[0516] Model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0517] The feature types supported by the terminal include AI-based time domain CSI prediction, and the first function information includes one or more of the following information:

[0518] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0519] In some implementations, the feature type supported by the terminal includes AI-based CSI compression feedback, and the second function information includes any one of the following information:

[0520] The model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0521] The third function information includes any of the following information:

[0522] The model outputs payload, channel scenario, network configuration parameters, and mobile speed;

[0523] The feature types supported by the terminal include AI-based time-domain CSI prediction, and the second function information includes any of the following information:

[0524] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed;

[0525] The third function information includes any of the following information:

[0526] Predicted time domain conditions, channel scenarios, network configuration parameters, and mobile speed.

[0527] In some embodiments, the feature type supported by the terminal includes AI-based CSI compression feedback, the first model identifier includes a model ID and / or a model pairing ID, the second model identifier includes a model ID and / or a model pairing ID, and the third model identifier includes a model ID and / or a model pairing ID;

[0528] The feature types supported by the terminal include AI-based time domain prediction, the first model identifier includes a model ID, the second model identifier includes a model ID, and the third model identifier includes a model ID.

[0529] Regarding the apparatus in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0530] The present disclosure also provides a computer-readable storage medium having computer program instructions stored thereon, which implement the steps of the model recognition method provided by the present disclosure when the program instructions are executed by a processor.

[0531] Figure 10 is a block diagram of a terminal according to some embodiments. For example, terminal 1000 can be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, a smart car, etc.

[0532] 10 , terminal 1000 may include one or more of the following components: a first processing component 1002 , a first memory 1004 , a first power supply component 1006 , a multimedia component 1008 , an audio component 1010 , a first input / output interface 1012 , a sensor component 1014 , and a communication component 1016 .

[0533] The first processing component 1002 generally controls the overall operation of the terminal 1000, such as operations associated with display, phone calls, data communications, camera operation, and recording operations. The first processing component 1002 may include one or more first processors 1020 to execute instructions to perform all or part of the steps of the above-described model recognition method. In addition, the first processing component 1002 may include one or more modules to facilitate interaction between the first processing component 1002 and other components. For example, the first processing component 1002 may include a multimedia module to facilitate interaction between the multimedia component 1008 and the first processing component 1002.

[0534] The first memory 1004 is configured to store various types of data to support operations on the terminal 1000. Examples of such data include instructions for any application or method operating on the terminal 1000, contact data, phone book data, messages, pictures, videos, etc. The first memory 1004 can be implemented by any type of volatile or non-volatile memory device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0535] First power supply component 1006 provides power to various components of terminal 1000. First power supply component 1006 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to terminal 1000.

[0536] The multimedia component 1008 includes a screen that provides an output interface between the terminal 1000 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor can not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 1008 includes a front camera and / or a rear camera. When the terminal 1000 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera can receive external multimedia data. Each front camera and rear camera can be a fixed optical lens system or have a focal length and optical zoom capability.

[0537] The audio component 1010 is configured to output and / or input audio signals. For example, the audio component 1010 includes a microphone (MIC), which is configured to receive external audio signals when the terminal 1000 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signal can be further stored in the first memory 1004 or transmitted via the communication component 1016. In some embodiments, the audio component 1010 also includes a speaker for outputting audio signals.

[0538] The first input / output interface 1012 provides an interface between the first processing component 1002 and peripheral interface modules, such as a keyboard, a click wheel, buttons, etc. These buttons may include but are not limited to: a home button, a volume button, a start button, and a lock button.

[0539] The sensor assembly 1014 includes one or more sensors for providing various aspects of the terminal 1000's status assessment. For example, the sensor assembly 1014 can detect the open / closed state of the terminal 1000, the relative positioning of components, such as the display and keypad of the terminal 1000. The sensor assembly 1014 can also detect changes in the position of the terminal 1000 or a component of the terminal 1000, the presence or absence of user contact with the terminal 1000, the orientation or acceleration / deceleration of the terminal 1000, and changes in the temperature of the terminal 1000. The sensor assembly 1014 can include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 1014 can also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 1014 can also include an accelerometer, a gyroscope, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0540] The communication component 1016 is configured to facilitate wired or wireless communication between the terminal 1000 and other devices. The terminal 1000 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 1016 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 1016 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0541] In an exemplary embodiment, the terminal 1000 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-mentioned model recognition method.

[0542] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is further provided, such as a first memory 1004 including instructions. The instructions can be executed by the first processor 1020 of the terminal 1000 to perform the above-mentioned model recognition method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0543] In addition to being an independent electronic device, the above-mentioned device can also be part of an independent electronic device. For example, in one embodiment, the device can be an integrated circuit (IC) or a chip, wherein the integrated circuit can be an IC or a collection of multiple ICs; the chip can include but is not limited to the following types: GPU (Graphics Processing Unit, graphics processor), CPU (Central Processing Unit, central processing unit), FPGA (Field Programmable Gate Array, programmable logic array), DSP (Digital Signal Processor, digital signal processor), ASIC (Application Specific Integrated Circuit, application-specific integrated circuit), SOC (System on Chip, SoC, system on chip or system-level chip), etc. The above-mentioned integrated circuit or chip can be used to execute executable instructions (or code) to implement the above-mentioned model recognition method. The executable instructions can be stored in the integrated circuit or chip, or can be obtained from other devices or equipment, for example, the integrated circuit or chip includes a processor, a memory, and an interface for communicating with other devices. The executable instruction can be stored in the memory, and when the executable instruction is executed by the processor, the above-mentioned model recognition method is implemented; alternatively, the integrated circuit or chip can receive the executable instruction through the interface and transmit it to the processor for execution to implement the above-mentioned model recognition method.

[0544] In another exemplary embodiment, a computer program product is further provided. The computer program product includes a computer program executable by a programmable device, and has a code portion for executing the above-mentioned model identification method when executed by the programmable device.

[0545] FIG11 is a block diagram of a network device according to an exemplary embodiment. For example, network device 1100 may be provided as other network logical entities in a core network or an access network. Referring to FIG11 , network device 1100 includes a processing component 1122, which further includes one or more processors, and memory resources represented by memory 1132 for storing instructions executable by processing component 1122, such as applications. The application stored in memory 1132 may include one or more modules, each of which corresponds to a set of instructions. In addition, processing component 1122 is configured to execute instructions to perform the steps of the communication method provided in the above-described method embodiment.

[0546] The network device 1100 may also include a power supply component 1126 configured to perform power management of the device 1100, a wired or wireless network interface 1150 configured to connect the network device 1100 to a network, and an input / output interface 1158. The network device 1100 may operate based on an operating system stored in the memory 1132, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, or the like.

[0547] Other embodiments of the present disclosure will readily occur to those skilled in the art after considering the specification and practicing the present disclosure. This disclosure is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the present disclosure being indicated by the following claims.

[0548] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A model identification method, It is characterized in that Executed by a terminal, the method includes: Sending first information to a network device, where the first information is used by the network device to determine an AI / ML function supported by the terminal and / or an AI / ML model supported by the terminal.

2. The method according to claim 1, It is characterized in that The sending of the first information to the network device comprises: receiving a first instruction sent by the network device; According to the first instruction, first information is sent to the network device.

3. The method according to claim 1, It is characterized in that The sending of the first information to the network device comprises: receiving a second instruction sent by the network device, wherein the second instruction includes a feature type; According to the second instruction, first information corresponding to the feature type is sent to the network device, and the first information corresponding to the feature type is used to determine the AI / ML functions supported by the terminal corresponding to the feature type and / or the AI / ML models supported by the terminal.

4. The method according to any one of claims 1 to 3, It is characterized in that The first information is used by the network device to determine the AI / ML function supported by the terminal, and the first information includes: First function information, used to describe the AI / ML function supported by the terminal; or, second function information and a first model identifier, where the second function information and the first model identifier are used by the network device to determine the AI / ML function supported by the terminal; or, a second model identifier, where the second model identifier is used by the network device to determine the AI / ML function supported by the terminal based on a preset corresponding relationship; or, The meta information associated with the terminal is used by the network device to determine the parameter configuration of the AI / ML function supported by the terminal.

5. The method according to any one of claims 1 to 3, It is characterized in that The first information is used by the network device to determine the AI / ML model supported by the terminal, and the first information includes: A second model identifier is used by the network device to determine an AI / ML model supported by the terminal; or third function information and a third model identifier, where the third function information and the third model identifier are used by the network device to determine an AI / ML model supported by the terminal; or The meta information associated with the terminal is used by the network device to determine the parameter configuration of the AI / ML model supported by the terminal.

6. The method according to claim 4 or 5, It is characterized in that The terminal-associated meta-information includes one or more of the following information: The transmitting antenna port corresponding to the terminal, the receiving antenna port corresponding to the terminal, the carrier information of the terminal, the number of subbands supported by the terminal, and the location information of the terminal.

7. The method according to claim 4 or 5, It is characterized in that The value range of the second model identifier is divided according to one or more of the following information: feature types supported by the terminal, function information supported by the terminal, and meta information associated with the terminal.

8. The method according to claim 4, It is characterized in that The feature types supported by the terminal include AI-based CSI compression feedback, and the first function information includes one or more of the following information: Model outputs payload, channel scenarios, network configuration parameters, and mobile speed; The feature types supported by the terminal include AI-based time domain CSI prediction, and the first function information includes one or more of the following information: Predicted time domain condition requirements, channel scenarios, network configuration parameters, and mobile speed.

9. The method according to claim 4 or 5, It is characterized in that The feature types supported by the terminal include AI-based CSI compression feedback, and the second function information includes any one of the following information: The model outputs payload, channel scenario, network configuration parameters, and mobile speed; The third function information includes any one of the following information: The model outputs payload, channel scenario, network configuration parameters, and mobile speed; The feature types supported by the terminal include AI-based time domain CSI prediction, and the second function information includes any one of the following information: Predicted time domain condition requirements, channel scenarios, network configuration parameters, and mobile speed; The third function information includes any one of the following information: Predicted time domain condition requirements, channel scenarios, network configuration parameters, and mobile speed.

10. The method according to claim 4 or 5, It is characterized in that The feature types supported by the terminal include AI-based CSI compression feedback, the first model identifier includes a model ID and / or a model pairing ID, the second model identifier includes a model ID and / or a model pairing ID, and the third model identifier includes a model ID and / or a model pairing ID; The feature types supported by the terminal include AI-based time domain prediction, the first model identifier includes a model ID, the second model identifier includes a model ID, and the third model identifier includes a model ID.

11. A model identification method, It is characterized in that Executed by a network device, the method includes: receiving first information sent by a terminal; Determine, according to the first information, an AI / ML function supported by the terminal and / or an AI / ML model supported by the terminal.

12. The method according to claim 11, It is characterized in that The method further comprises: A first instruction is sent to the terminal, where the first instruction is used by the terminal to send the first information.

13. The method according to claim 11, It is characterized in that The method further comprises: a second instruction sent to the terminal, the second instruction including a feature type, the second instruction being used by the terminal to send first information corresponding to the feature type; The first information sent by the receiving terminal includes: Receive first information corresponding to the feature type sent by the terminal.

14. The method according to any one of claims 11 to 13, It is characterized in that The method further comprises: According to the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal, determine information for lifecycle management of the model on the terminal side.

15. The method according to any one of claims 11 to 14, It is characterized in that The first information is used by the network device to determine the AI / ML function supported by the terminal, and the first information includes: First function information, used to describe the AI / ML function supported by the terminal; or, second function information and a first model identifier, where the second function information and the first model identifier are used by the network device to determine the AI / ML function supported by the terminal; or, a second model identifier, where the second model identifier is used by the network device to determine the AI / ML function supported by the terminal based on a preset corresponding relationship; or, The meta information associated with the terminal is used to characterize the parameter configuration of the AI / ML function supported by the terminal.

16. The method according to any one of claims 11 to 14, It is characterized in that The first information is used by the network device to determine the AI / ML model supported by the terminal, and the first information includes: A second model identifier is used by the network device to determine an AI / ML model supported by the terminal; or third function information and a third model identifier, where the third function information and the third model identifier are used by the network device to determine an AI / ML model supported by the terminal; or The meta information associated with the terminal is used by the network device to determine the parameter configuration of the AI / ML model supported by the terminal.

17. The method according to claim 15 or 16, It is characterized in that The terminal-associated meta-information includes one or more of the following information: The transmitting antenna port corresponding to the terminal, the receiving antenna port corresponding to the terminal, the carrier information of the terminal, the number of subbands supported by the terminal, and the location information of the terminal.

18. The method according to claim 15 or 16, It is characterized in that The value range of the second model identifier is divided according to one or more of the following information: feature types supported by the terminal, function information supported by the terminal, and meta information associated with the terminal.

19. The method according to claim 15, It is characterized in that The feature types supported by the terminal include AI-based CSI compression feedback, and the first function information includes one or more of the following information: Model outputs payload, channel scenarios, network configuration parameters, and mobile speed; The feature types supported by the terminal include AI-based time domain CSI prediction, and the first function information includes one or more of the following information: Predicted time domain condition requirements, channel scenarios, network configuration parameters, and mobile speed.

20. The method according to claim 15 or 16, It is characterized in that The feature types supported by the terminal include AI-based CSI compression feedback, and the second function information includes any of the following information: The model outputs payload, channel scenario, network configuration parameters, and mobile speed; The third function information includes any of the following information: The model outputs payload, channel scenario, network configuration parameters, and mobile speed; The feature types supported by the terminal include AI-based time domain CSI prediction, and the second function information includes any of the following information: Predicted time domain condition requirements, channel scenarios, network configuration parameters, and mobile speed; The third function information includes any of the following information: Predicted time domain condition requirements, channel scenarios, network configuration parameters, and mobile speed.

21. The method according to claim 15 or 16, It is characterized in that The feature types supported by the terminal include AI-based CSI compression feedback, the first model identifier includes a model ID and / or a model pairing ID, the second model identifier includes a model ID and / or a model pairing ID, and the third model identifier includes a model ID and / or a model pairing ID; The feature types supported by the terminal include AI-based time domain prediction, the first model identifier includes a model ID, the second model identifier includes a model ID, and the third model identifier includes a model ID.

22. A model management method, It is characterized in that Executed by a communication system, the communication system comprising a terminal and a network device, the method comprising: first information sent by the terminal to the network device; The network device determines, based on the first information, an AI / ML function supported by the terminal and / or an AI / ML model supported by the terminal.

23. A model recognition device, It is characterized in that include: The first sending module is configured to send first information to the network device, where the first information is used by the network device to determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal.

24. A model recognition device, It is characterized in that include: A receiving module, configured to receive first information sent by a terminal; The first determination module is configured to determine the AI / ML functions supported by the terminal and / or the AI / ML models supported by the terminal according to the first information.

25. A communication device, It is characterized in that include: one or more processors; The processor is used to call computer instructions so that the communication device executes the model recognition method according to any one of claims 1 to 21.

26. A communication system, It is characterized in that It comprises a terminal and a network device, wherein the terminal is configured to implement the model recognition method according to any one of claims 1 to 10, and the network device is configured to implement the model recognition method according to any one of claims 11 to 21.

27. A computer-readable storage medium storing computer instructions. It is characterized in that When the computer instructions are executed on a communication device, the communication device is caused to execute the model recognition method according to any one of claims 1 to 21.

Images