Information management method and device, terminal, network equipment, storage medium and computer program product

Abstract

The invention discloses an information management method and device, a terminal, network equipment, a storage medium and a computer program product, and the method comprises the steps that the terminal sends first information to the network equipment and receives second information sent by the network equipment, the first information indicates AI functions and / or models available for the terminal, and the second information indicates AI functions and / or models available for the terminal; the second information is used for indicating activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to an information management method, apparatus, terminal, network equipment, storage medium, and computer program product. Background Technology

[0002] Related technologies provide a basic framework for artificial intelligence (AI) or machine learning (ML), including the entire lifecycle management (LCM) process for AI / ML functions or models. However, the lack of effective management for AI functions or models on the terminal side affects the performance of the terminal or the accuracy of the output results. Summary of the Invention

[0003] To address the related technical issues, embodiments of this application provide an information management method, apparatus, terminal, network device, storage medium, and computer program product.

[0004] The technical solution of this application embodiment is implemented as follows:

[0005] This application provides an information management method applied to a terminal, the method comprising:

[0006] Send a first message to the network device, the first message indicating the artificial intelligence (AI) functions and / or models available to the terminal;

[0007] Receive second information sent by the network device, the second information being used to instruct the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

[0008] This application also provides an information management method applied to a network device, the method comprising:

[0009] The receiving terminal sends first information, which indicates the AI ​​functions and / or models available to the terminal.

[0010] Send a second message to the terminal, the second message being used to instruct the activation or deactivation, selection, switching, or rollback of one or more AI functions and / or models.

[0011] This application also provides an information management device, including:

[0012] The first sending unit is configured to send first information to the network device, wherein the first information indicates the AI ​​functions and / or models available to the terminal;

[0013] The first receiving unit is configured to receive second information sent by the network device, the second information being used to indicate the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

[0014] This application also provides an information management device, including:

[0015] The second receiving unit is used to receive first information sent by the terminal, the first information indicating the AI ​​functions and / or models available to the terminal;

[0016] The second sending unit is used to send second information to the terminal, the second information being used to indicate the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

[0017] This application also provides a terminal, including: a first processor and a first communication interface; wherein,

[0018] The first communication interface is used to send first information to the network device and receive second information sent by the network device. The first information indicates the AI ​​functions and / or models available to the terminal, and the second information is used to indicate the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

[0019] This application also provides a network device, including: a second processor and a second communication interface; wherein,

[0020] The second communication interface is used to receive first information sent by the terminal and send second information to the terminal. The first information indicates the AI ​​functions and / or models available to the terminal, and the second information is used to indicate the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

[0021] This application also provides a terminal, including a first processor and a first memory for storing computer programs that can run on the first processor.

[0022] Wherein, the first processor is used to execute the steps of any method on the terminal side when running the computer program.

[0023] This application also provides a network device, including a second processor and a second memory for storing computer programs that can run on the second processor.

[0024] The second processor is used to execute the steps of any method on the network device side when running the computer program.

[0025] This application also provides a storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of any method on the terminal side or the steps of any method on the network device side.

[0026] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of any of the above methods.

[0027] In the information management method, apparatus, terminal, network device, storage medium, and computer program product provided in the embodiments of this application, the terminal sends first information to the network device, the first information indicating the AI ​​functions and / or models available to the terminal; the network device receives the first information and sends second information to the terminal, the terminal receives the second information, the second information being used to indicate the activation or deactivation, selection, switching, or rollback of one or more AI functions and / or models. It can be seen that in the embodiments of this application, the network device can instruct the terminal to activate or deactivate, select, switch, or rollback of one or more AI functions and / or models based on the AI ​​functions and / or models reported by the terminal that are available to the terminal. The above scheme enables the network device to uniformly manage the AI ​​functions and / or models available to the terminal. Since the AI ​​functions and / or models supported by the terminal are not necessarily available, or not all are applicable in the current application scenario, the terminal reporting the available AI functions and / or models to the network device allows the network device to know the currently available AI functions and / or models of the terminal, ensuring that the currently activated or selected AI function or model is available to the terminal, thereby ensuring that the terminal can use the appropriate AI function and / or model to perform corresponding operations, which can improve the performance of the terminal or the accuracy of the output results. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of an information management method according to an embodiment of this application;

[0029] Figure 2 This is an example bitmap diagram of an embodiment of this application;

[0030] Figure 3 This is an example diagram of a MAC CE according to an embodiment of this application;

[0031] Figure 4 This is an example diagram of a MAC CE according to an embodiment of this application;

[0032] Figure 5 This is an example diagram illustrating how second information can be carried through an RRC signaling field or MAC CE according to an embodiment of this application.

[0033] Figure 6 This is an example diagram illustrating how second information can be carried through an RRC signaling field or MAC CE according to an embodiment of this application.

[0034] Figure 7 This is an example diagram illustrating how second information can be carried through an RRC signaling field or MAC CE according to an embodiment of this application.

[0035] Figure 8 This is an example diagram illustrating how second information can be carried through an RRC signaling field or MAC CE according to an embodiment of this application.

[0036] Figure 9 This is an example diagram illustrating how second information can be carried through an RRC signaling field or MAC CE according to an embodiment of this application.

[0037] Figure 10 This is an example diagram illustrating how second information can be carried through an RRC signaling field or MAC CE according to an embodiment of this application.

[0038] Figure 11 This is a schematic diagram of an information management method according to an embodiment of this application;

[0039] Figure 12 This is a schematic diagram of the structure of an information management device according to an embodiment of this application;

[0040] Figure 13 This is a schematic diagram of the structure of an information management device according to an embodiment of this application;

[0041] Figure 14 This is a schematic diagram of the terminal structure according to an embodiment of this application;

[0042] Figure 15 This is a schematic diagram of the network device structure according to an embodiment of this application. Detailed Implementation

[0043] AI functionality and model are two different granularities. Generally, functionality can be understood as coarser-grained, such as spatial beam prediction, temporal beam prediction, AI-based direct localization, and AI-assisted localization. Model, on the other hand, is finer-grained. In this case, one functionality can correspond to multiple models. Of course, if a model has good generalization ability, it can also be applied to multiple functionalities.

[0044] Management / LCM of AI functions or models, including operations such as activation, deactivation, selection, switching, and fallback.

[0045] Currently, terminal or user equipment (UE) needs to report the supported models or AI functions and the applicable models or AI functions to the network so that the network (base station) knows the UE's capabilities and from which AI functions or models to select, activate, and deactivate.

[0046] In related technologies, there are no specific solutions provided for how to effectively manage AI functions or models on the UE side, such as how to activate, deactivate, or switch an AI function or model.

[0047] Based on this, in various embodiments of this application, the terminal sends first information to the network device, the first information indicating the AI ​​functions and / or models available to the terminal; the network device receives the first information and sends second information to the terminal, the terminal receives the second information, the second information being used to indicate the activation or deactivation, selection, switching, or rollback of one or more AI functions and / or models. It can be seen that in the embodiments of this application, the network device can instruct the terminal to activate or deactivate, select, switch, or rollback of one or more AI functions and / or models based on the AI ​​functions and / or models reported by the terminal. The above scheme enables the network device to uniformly manage the AI ​​functions and / or models available to the terminal. Since the AI ​​functions and / or models supported by the terminal may not necessarily be available, or may not all be applicable in the current application scenario, the terminal reporting the available AI functions and / or models to the network device allows the network device to know the currently available AI functions and / or models of the terminal, ensuring that the currently activated or selected AI function or model is available to the terminal, thereby ensuring that the terminal can use the appropriate AI function and / or model to perform corresponding operations, which can improve the performance of the terminal or the accuracy of the output results.

[0048] The present application will now be described in further detail with reference to the accompanying drawings and embodiments.

[0049] This application provides an information management method applied to a terminal, also known as a user interface (UE), which includes mobile phones, tablets, wearable devices, laptops, etc. Figure 1 As shown, the method includes:

[0050] Step 101: Send first information to the network device, the first information indicating the AI ​​functions and / or models available to the terminal.

[0051] Here, one AI function can correspond to one or more models; that is, an AI function can be implemented through one or more models. Models include AI models and / or ML models. An AI function can be described as an AI functionality, or simply a function or functionality. The AI ​​functions available to the terminal are determined from the AI ​​functions supported by the terminal, and the models available to the terminal are determined from the models supported by the terminal. When the number of available AI functions and / or models on the terminal is greater than or equal to one, the terminal can report the priority information of multiple available AI functions and / or models to the network device. The network device can include base stations, core networks, or new network elements used for AI management. Models include, but are not limited to, AI models and / or ML models.

[0052] To ensure that network devices accurately know the AI ​​functions and / or models available to the terminal, in one embodiment, the first information includes the identifier or index of the AI ​​functions and / or models available to the terminal; and / or,

[0053] The first information includes priority information of the AI ​​functions and / or models available to the terminal.

[0054] Here, the first information can explicitly indicate the AI ​​functions and / or models available to the terminal. For example, the first information includes identifiers of the AI ​​functions and / or models available to the terminal. Alternatively, the first information can implicitly indicate the AI ​​functions and / or models available to the terminal. For example, the first information includes an index of the AI ​​functions and / or models available to the terminal, which can be determined based on the supported AI functions and / or models reported by the terminal. For instance, if the terminal reports supported AI functions #1, #5, #6, and #8, and the available AI functions are represented by indices 2 and 3, then functions #5 and #6 are implicitly indicated. The identifiers or indexes of AI functions and / or models can be locally unique or globally unique, which helps the terminal and network devices to have a unified understanding of the AI ​​functions and / or models available to the terminal.

[0055] Priority information can explicitly indicate the AI ​​functions and / or models available on the terminal. For example, the first piece of information carries a priority value, where a larger priority value indicates higher priority, or a smaller priority value indicates higher priority.

[0056] Priority information can also implicitly indicate the AI ​​functions and / or models available to the terminal, and send the available AI functions and / or models to the terminal in descending or ascending order of priority. The first message can be one or multiple messages.

[0057] It should be noted that since not all AI functions and / or models supported by the terminal may be applicable in the current scenario, there is little need to activate or deactivate AI functions and / or models that are not applicable to the current scenario. It is more accurate for network devices to activate or deactivate based on the AI ​​functions and / or models available to the terminal. Therefore, the terminal reports the AI ​​functions and / or models available to the terminal to the network device. This not only saves signaling overhead, but also saves the resources consumed by the terminal in activating or deactivating inapplicable AI functions and / or models.

[0058] To reduce complexity and signaling load, the number of available AI functions and / or models reported by the terminal can be limited. This can be specified by protocols or configured by network devices, such as through broadcast messages, Radio Resource Control (RRC) configuration messages, or RRC reconfiguration messages. Based on this, in one embodiment, the method further includes:

[0059] The terminal receives a fourth message sent by the network device, the fourth message indicating the maximum number of available AI functions and / or models to be reported by the terminal, and / or indicating the priority information of available AI functions and / or models to be reported by the terminal.

[0060] Here, before sending the first information to the network device, the terminal can also receive the fourth information sent by the network device, and then send the first information based on the fourth information, so that the number of AI functions and / or models reported by the terminal is less than or equal to the maximum number indicated by the fourth information, and / or so that the terminal reports the AI ​​functions and / or models available to the terminal according to the priority information of the AI ​​functions and / or models available to the terminal.

[0061] For example, the fourth piece of information indicates that the terminal can report a maximum of 6 available models. Even if the terminal has 10 available models internally, it will only report 6 of them. Which 6 are reported depends on the specific implementation of the terminal. For example, the terminal may report 6 available models in descending order of priority. The maximum number of available AI functions and / or models can include: the maximum number of available AI functions, and / or the maximum number of available models. The maximum number of available models refers to the maximum number of all available models, not the maximum number of models specifically used for a particular AI function.

[0062] Step 102: Receive second information sent by the network device, the second information being used to indicate activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

[0063] Here, the second piece of information is sent by the network device based on the first piece of information. The network device can perform LCM (Localization Management) on one AI function and / or one model at a time, or it can perform LCM on multiple AI functions and / or multiple models simultaneously. In other words, the network device can activate multiple AI functions and / or multiple models simultaneously. For example, considering that the terminal side may execute multiple AI functions and / or multiple models simultaneously, such as simultaneously executing AI-based localization and AI-based temporal beam prediction, the network device's simultaneous activation of multiple AI functions and / or multiple models facilitates the coexistence of multiple AI use cases, enhancing the intelligence capabilities of both the terminal and the network in various ways.

[0064] The second piece of information can explicitly indicate whether to activate or deactivate, select, switch, or roll back one or more AI functions and / or models, or it can implicitly indicate whether to activate or deactivate, select, switch, or roll back one or more AI functions and / or models. For example, the network device implicitly indicates whether to activate or deactivate, select, switch, or roll back one or more AI functions and / or models according to the reporting order of the AI ​​functions and / or models available to the terminal; as another example, the network device sends an index of one or more AI functions and / or models to the terminal according to the reporting order of the AI ​​functions and / or models available to the terminal, in order to indicate whether to activate or deactivate, select, switch, or roll back one or more AI functions and / or models. For example, if the protocol stipulates that a network device can only activate one AI function and / or one model at a time, then if the network device activates another AI function and / or model, it implicitly instructs the current AI function and / or model to be activated. For instance, if AI function 1 is currently active, and the network device instructs the terminal to activate AI function 2, then it means to activate AI function 1. If the network device instructs a switch, it can implicitly instruct that the AI ​​function and / or model to be switched to needs to be activated, thus activating the current AI function and / or model. For instance, if the network device instructs a switch to AI function 2, then it means to activate both the current AI function 1 and AI function 2. This simplifies the model management process.

[0065] The second piece of information includes the identifier or index of one or more AI functions and / or models, used to indicate whether to activate or deactivate, select or switch one or more AI functions and / or models. For example, a network device can directly send the identifier or index of an AI function and / or model to the terminal to explicitly indicate whether to activate or deactivate, select or switch or roll back one or more AI functions and / or models, so that the terminal can quickly know which AI function and / or model is being LCMed.

[0066] It's important to clarify that rolling back one or more AI functions can be understood as reverting to non-AI. Non-AI can be described as non-AI operation or non-AI mode; these terms are interchangeable. Non-AI can also be understood as not using all AI functions, not using this one AI function, or not using all AI functions involved in this use case (e.g., beam management, positioning) or application scenario. Similarly, rolling back one or more models can be understood as reverting to non-AI, or not applying these one or more models, or not using all AI models involved in this use case or application scenario.

[0067] It should be noted that there are two schemes for fallback: (1) an identifier, an indication, an RRC signaling field, a MAC CE, or a DCI to indicate that the terminal completely falls back to non-AI; (2) to indicate different use cases / AI functions / models. For example, network devices can indicate that beam management should fall back to non-AI, but positioning can still use AI. Schemes (1) and (2) can be supported simultaneously, and the priority of scheme (1) can be higher than that of scheme (2). Scheme (1) can use the least amount of signaling to indicate that the terminal completely falls back to non-AI, without having to indicate the current AI operation to fall back to non-AI operation one by one, thus reducing signaling load and complexity. Scheme (2) can reduce the impact between different AI operations or use cases. Even if use case A cannot use AI, use case B can still be executed based on AI. For different use cases (such as AI-based beam management, AI-based positioning, AI-based CSI prediction, and AI-based mobility), it means that multiple or a group of AI functions and / or models can be simultaneously instructed to fall back. For example, for AI-based beam management, if there are currently three AI functions (1 to 3) active, the network device can instruct the AI-based beam management to fall back to non-AI, that is, AI functions 1 to 3 fall back simultaneously.

[0068] One or more AI functions and / or models can be activated, deactivated, selected, switched, or rolled back using a bitmap. Based on this, in one embodiment, the second information includes one or more bitmaps; wherein a single bit in one bitmap indicates activation, deactivation, selection, or switching of an AI function and / or model, and / or a change in a bit in one bitmap indicates switching an AI function and / or model, and / or the values ​​of all bits in the bitmap indicate deactivation, indicating a rollback to not using the AI ​​function and / or model.

[0069] Here, a bitmap can be used for the LCM of one or more AI functions, or for the LCM of one or more models, or for the LCM of one or more AI functions and one or more models. Each bit in each bitmap corresponds to one AI function or one model; for example, if the terminal reports available AI functions as #1, #2, #3, #4, #5, or if the terminal reports 5 available AI functions, the first bit in the bitmap corresponds to AI function #1 or the first available AI function, the second bit corresponds to AI function #2 or the second available AI function, and so on.

[0070] The value of each bit in the bitmap indicates whether the corresponding AI function or model is activated, deactivated, selected, or switched. For example, 1 represents activation and 0 represents deactivation; or 1 represents deactivation and 0 represents activation. Changes in the state of bits within the same bitmap can implicitly indicate a switch. For instance, if the network device first sends a bitmap of 01000 (using 5 bits as an example), and then sends 00100 the second time, the second AI function is deactivated, and the third AI function is activated, meaning a switch from the second available AI function to the third available AI function. If all bits in the bitmap indicate deactivation, it indicates a fallback to non-AI operation, where no AI function and / or model is used.

[0071] It should be noted that bitmaps can be sent via RRC signaling. For example, a new RRC signaling field can be introduced: `functionalityManagementBitmap IE BIT STRING(SIZE(8))`. The field name is just an example and is not limited. `SIZE(8)` is based on 8 bits and can be other values. The value in `SIZE()` can be set according to the actual situation. A separate RRC signaling field can be introduced for AI functions and models, using an additional identifier to indicate whether it is used for AI functions or models; or two separate RRC signaling fields can be introduced, namely, one IE for the LCM of AI functions and one IE for the LCM of models. For example, separate RRC signaling fields can be introduced for AI functions and models respectively:

[0072] functionalityManagementBitmap IE BIT STRING(SIZE(8));

[0073] modelManagementBitmap IE BIT STRING(SIZE(8)).

[0074] For example, introduce an RRC signaling field for AI functions and models:

[0075] ManagementBitmap IE BIT STRING(SIZE(9)) / / For example, the first bit indicates whether it is used for AI function or model, for example, 1 represents AI function and 0 represents model;

[0076] or,

[0077] ManagementType IE ENUMERATED{functionality,model} / / This IE indicates whether it is used for the activation of AI functions or models. If functionality represents the LCM used for AI functions, model represents the LCM used for AI models; {functionality,model} can also be represented by {0,1}, such as 0 representing AI functions and 1 representing models, or vice versa.

[0078] ManagementBitmap IE BIT STRING(SIZE(9)).

[0079] It should be noted that bitmaps can also be distributed via Media Access Control (MAC) layer control elements (MAC CEs). Two separate MAC CEs can be introduced for AI functions and models, for example... Figure 2 As shown; a MAC CE can also be introduced for the LCM of AI functions and models, for example Figure 3 and Figure 4 As shown. It should be noted that, Figures 2 to 4 This illustration uses a bitmap containing 8 bits (8 available AI functions and / or models) as an example. This application does not limit the number of available AI functions and / or models.

[0080] To facilitate the terminal's understanding that the bitmap is an LCM for AI functions and / or models, in one embodiment, the bitmap includes a first identifier indicating that the bitmap is used as an LCM for AI functions and / or models; or, the bitmap includes a second identifier and a third identifier, the second identifier indicating that the bitmap is used as an LCM for models, and the third identifier indicating that the bitmap is used as an LCM for AI functions.

[0081] Here, when using a bitmap for LCM of AI functions or models, one or two bits can be used to describe whether the bitmap is used for LCM of AI functions or for LCM of models.

[0082] When a bitmap is described as an LCM for an AI function or model through a field, the bitmap contains a first identifier that indicates that the bitmap is an LCM for an AI function or model; the first identifier can be 0 or 1, for example, 0 represents an AI function-based LCM and 1 represents a model-based LCM; or, 1 represents an AI function-based LCM and 0 represents a model-based LCM.

[0083] When a bitmap is used to describe the LCM for an AI function or model using two bits, the bitmap contains a second identifier and a third identifier. The second identifier indicates that the bitmap is used for the LCM for the model, and the third identifier indicates that the bitmap is used for the LCM for the AI ​​function. For example, a second identifier of 1 represents an AI function-based LCM, and a third identifier of 1 represents a model-based LCM; or, a second identifier of 0 represents an AI function-based LCM, and a third identifier of 1 represents a model-based LCM.

[0084] The above describes the implementation of sending bitmaps via RRC signaling or MAC CE, where the second information includes one or more bitmaps. Next, we will introduce the implementation of carrying the second information via RRC signaling or MAC CE, where the identification or index of AI functions or models is carried via RRC signaling or MAC CE.

[0085] To improve the flexibility and diversity of the second information delivery, the second information can be carried in RRC signaling. A new field can be designed to carry the second information, or existing available fields in RRC signaling can be reused. Based on this, in one embodiment, the second information is carried in an RRC signaling field, where one information element (IE) corresponds to an indication of activating, deactivating, selecting, switching, or reverting one or more AI functions or models.

[0086] Here, in one embodiment, the second information is sent via RRC signaling. Taking the RRC signaling containing an AI functionality identifier as an example, a new field is introduced as follows:

[0087] For activation:

[0088] functionalityActivationList IE SEQUENCE(SIZE(1..maxNrofFunctionality))OF functionalityId;

[0089] Regarding the selection:

[0090] functionalitySelectionList IE SEQUENCE(SIZE(1..maxNrofFunctionality))OF functionalityId;

[0091] For deactivation:

[0092] functionalityDeactivationList IE SEQUENCE(SIZE(1..maxNrofFunctionality))OF functionalityId;

[0093] Regarding rollback:

[0094] 1) An indication method: fallbackToNonAI IE ENUMERATED{true};

[0095] 2) A list is provided for different AI function indication methods:

[0096] FallbackToNonAIList IE SEQUENCE(SIZE(1..maxNrofFunctionality))OFfunctionalityId;

[0097] 3) Methods of indicating different use cases:

[0098] FallbackToNonAI IE SEQUENCE

[0099] {useCase1(e.g., beamManagement)ENUMERATED{true} / / This setting to true indicates that all beam management functionalities are rolled back.

[0100] useCase2(such as positioning)ENUMERATED{true}

[0101] useCase3(such as csiPredition)ENUMERATED{true} ...

[0103] };

[0104] Regarding the switch:

[0105] functionalitySwitchList IE SEQUENCE(SIZE(1..maxNrofFunctionality))OFfunctionalityId;

[0106] functionalityId BIT STRING(SIZE(XX)).

[0107] It should be noted that maxNrofFunctionality represents the maximum number of AI functions or the maximum number of available AI functions; functionalityId represents the identifier or index of the AI ​​function. For example, the first available AI function reported by the corresponding terminal is index1, and so on.

[0108] In another embodiment, the second information is sent via RRC signaling, introducing a common field for both AI functions and models. An additional IE (flag IE in the example below) is needed to indicate whether it's for the AI ​​function or the model's LCM. The following example uses activation; other LCM processes are similar, as shown in the example below:

[0109] functionalityModelActivationList IE;

[0110] {flag ENUMERATED{functionality,model} / / This IE indicator is used for the activation of AI functions or models. If functionality represents the LCM used for AI functions, then model represents the LCM used for AI models. {functionality,model} can also be represented by {0,1}, such as 0 representing AI functions and 1 representing models, or vice versa.

[0111] IdList SEQUENCE(SIZE(1..maxNrof)) OF Id / / Identifier of the activated AI function or model

[0112] }

[0113] To improve the flexibility and diversity of the second message delivery, the second message can be carried in a MAC CE. A new MAC CE can be designed to carry the second message, or an existing MAC CE can be reused. Based on this, the second message is carried in one or more MAC CEs. A non-empty MAC CE is used to indicate activation or deactivation, selection, switching, or rollback of one or more AI functions and / or models, and / or an empty MAC CE is used to indicate rollback to not using all AI functions and / or models.

[0114] Here, at least two independent MAC CEs can be introduced for the AI ​​function and the model respectively. One MAC CE is used for the LCM of the AI ​​function, containing the identifier or index of the AI ​​function; the other MAC CE is used for the LCM of the AI ​​model, containing the identifier or index of the model. For example... Figure 5 As shown; of course, separate MAC CEs can also be introduced for activation, deactivation, selection, switching, and rollback. The MAC CE used to switch AI functions and / or models contains identifiers for the AI ​​functions and / or models, indicating the AI ​​function and / or model to which it is switched. A MAC CE can be introduced for each AI function and / or model, containing the identifier or index of the AI ​​function, and / or, the identifier or index of the model, for example... Figures 6 to 9 As shown.

[0115] For rollback, a separate MAC CE indicator can be introduced, such as an empty MAC CE. This empty MAC CE does not need to carry specific AI function identifiers and / or model identifiers. It is used to indicate that all AI functions and / or models should revert to non-AI. An empty MAC CE means that the MAC CE does not contain any content or data. When the terminal receives an empty MAC CE, all AI operations cease, and the system reverts to non-AI, i.e., the traditional operation or processing method that does not use any AI functions or models.

[0116] For the rollback of AI functions and / or models, a non-empty MAC CE can also be used as an indication. A non-empty MAC CE contains at least one identifier indicating the rollback of an AI function or model. The non-empty MAC CE may also contain an identifier of the specific AI function and / or model to be rolled back. A non-empty MAC CE may also contain the application scenario or use case indicating the rollback (such as beam management, positioning, etc.), meaning that the AI ​​function and / or model is not used in that application scenario or use case.

[0117] To facilitate terminal understanding of which operation in the LCM of an AI function and / or model is being indicated by the RRC signaling or MAC CE, thereby enabling flexible management of AI functions and / or models, in one embodiment, the RRC signaling field or a non-empty MAC CE includes one or more of the following:

[0118] A fourth identifier, which indicates the LCM used for AI functions or models;

[0119] The fifth identifier indicates the LCM used for AI functions;

[0120] The sixth identifier indicates the LCM used for the model;

[0121] The seventh identifier indicates whether to activate, deactivate, select, switch, or roll back.

[0122] The eighth identifier is used to indicate whether to activate or deactivate;

[0123] The ninth identifier indicates activation;

[0124] The tenth identifier indicates that it is used for deactivation;

[0125] The eleventh identifier indicates a fallback to not using AI features or models;

[0126] The twelfth identifier indicates a fallback to a model that is not in use;

[0127] The thirteenth identifier indicates a fallback to not using AI features.

[0128] Here, when the second information is carried in RRC signaling or MAC CE, that is, the identification or index of AI function or model is carried in RRC signaling or MAC CE, the RRC signaling field or MAC CE includes one or more of the fourth to thirteenth identifiers.

[0129] When an RRC signaling field or a non-empty MAC CE is used to indicate activation or deactivation, selection or switching of one or more AI functions and / or models, the RRC signaling field or the non-empty MAC CE shall include at least one or more of the fourth, fifth, and sixth identifiers, and may also include one or more of the seventh, eighth, ninth, and tenth identifiers; of course, the RRC signaling field or the non-empty MAC CE may also include the identifier or index of the AI ​​function and / or model. When an RRC signaling field or a non-empty MAC CE is used to indicate rollback of one or more AI functions and / or models, the RRC signaling field or the non-empty MAC CE shall include at least one or more of the eleventh, twelfth, and thirteenth identifiers, and may also include the identifier or index of the AI ​​function and / or model.

[0130] When an AI function or model is LCM indicated by an RRC signaling or MAC CE, at least one field can indicate whether the RRC signaling or MAC CE is for LCM of the AI ​​function or for LCM of the model.

[0131] For example, such as Figure 6As shown, a field F indicates whether the RRC signaling or the MAC CE is an LCM for AI functions or an LCM for the model. A fourth identifier is carried in field F. For example, a fourth identifier of 0 indicates an LCM for AI functions, and a fourth identifier of 1 indicates an LCM for the model; or, a fourth identifier of 1 indicates an LCM for AI functions, and a fourth identifier of 0 indicates an LCM for the model.

[0132] For example, such as Figure 7 As shown, two fields (field F1 and field F2) indicate whether the RRC signaling or the MAC CE is an LCM for AI functions or an LCM for the model. The RRC signaling field or the non-empty MAC CE may also include a fifth identifier and / or a sixth identifier, with the fifth identifier carried in field F1 and the sixth identifier carried in field F2; for example, a value of 1 in field F1 indicates an LCM for AI functions, in which case 1 is the fifth identifier; a value of 1 in field F2 indicates an LCM for the model, in which case 1 is the sixth identifier.

[0133] When performing LCM on an AI function or model via an RRC signaling or MAC CE indication, one field can indicate whether the RRC signaling field or MAC CE is for the AI ​​function's LCM or the model's LCM, while one or more other fields indicate which LCM process it is used for (selection, switching, activation, deactivation, rollback, etc.). If separate MAC CEs are introduced for different LCM processes, for example, MAC CE 1 for the activation process and MAC CE 2 for the deactivation process, then one or more fields are not needed to indicate which LCM process it is used for.

[0134] For example, such as Figure 7 As shown, field F1 indicates the LCM used for AI functions or models. Field F1 can occupy 1 bit, and the fourth, fifth, or sixth identifier is carried in character F1. The fourth, fifth, or sixth identifier can be 1 or 0. Field F2 indicates activation, deactivation, selection, toggling, or rollback. Field F2 occupies at least 2 bits, and the seventh identifier is carried in field F2. The seventh identifier can be 00, 01, 10, or 11. For example, a value of 1 for field F1 (fourth or fifth identifier is 1) indicates LCM used for AI functions; a value of 0 for field F1 (fourth or sixth identifier is 0) indicates LCM used for models; a value of 00 for field F2 (seventh identifier is 00) indicates selection; a value of 01 for field F2 (seventh identifier is 01) indicates activation; a value of 10 for field F2 (seventh identifier is 10) indicates deactivation; and a value of 11 for field F2 (seventh identifier is 11) indicates rollback.

[0135] For example, if only activation and deactivation are considered, then the RRC signaling field or a non-empty MAC CE includes the fourth and eighth identifiers, or includes the fifth and / or sixth identifiers, and the eighth identifier, or includes the fourth, ninth, and / or tenth identifiers; or includes the fifth and / or sixth identifiers, and the ninth and / or tenth identifiers. In the case where the RRC signaling field or a non-empty MAC CE includes the fourth and eighth identifiers, such as... Figure 7 As shown, the fourth identifier is carried in character F1, and the eighth identifier is carried in field F2. Both the fourth and eighth identifiers can have values ​​of 0 or 1 to represent different meanings. For example, a fourth identifier of 1 indicates LCM used for AI functions, a fourth identifier of 0 indicates LCM used for the model, or vice versa; a eighth identifier of 1 indicates activation, an eighth identifier of 0 indicates deactivation, or vice versa. When the fourth, ninth, and tenth identifiers are included in the RRC signaling field or a non-empty MAC CE, such as... Figure 8 As shown, the fourth identifier is carried by the character F3, the ninth identifier by the character F4, and the tenth identifier by the character F5. For example, a fourth identifier of 1 indicates LCM used for AI functions, a fourth identifier of 0 indicates LCM used for the model, or vice versa; a ninth identifier of 1 indicates activation, a tenth identifier of 1 indicates deactivation, or vice versa. The ninth and tenth identifiers cannot both be 1.

[0136] For example, two fields can be used to indicate RRC signaling or MAC CE for LCM of AI functions or models, and four fields can be used to indicate selection, activation, deactivation, and rollback. In this case, the number of fourth identifiers is two. Alternatively, the RRC signaling field or a non-empty MAC CE may include a fifth and a sixth identifier, and the number of seventh identifiers is four. Figure 9 As shown, field F6 indicates the LCM used for AI functions, and field F7 indicates the LCM used for the model. That is, the fourth identifier is carried in fields F6 and F7, or the fifth identifier is carried in field F6 and the sixth identifier is carried in field F7. Field F8 indicates selection, field F9 indicates activation, field F10 indicates deactivation, and field F11 indicates fallback. That is, the four seventh identifiers are carried in fields F8 through F11 respectively. For example, a value of 1 in field F6 indicates the LCM used for AI functions, a value of 1 in field F7 indicates the LCM used for the model, a value of 1 in field F8 indicates selection, a value of 1 in field F9 indicates activation, a value of 1 in field F10 indicates deactivation, and a value of 1 in field F11 indicates MAC CE used for fallback.

[0137] When RRC signaling or a non-empty MAC CE is used to indicate the rollback of one or more AI functions and / or models, the indication can be provided by one or two fields in the RRC signaling field or the non-empty MAC CE. For example, an eleventh identifier can be carried in one field to indicate that the terminal has rolled back to a completely non-AI state, without using any AI functions and models. Another example is... Figure 10 As shown, field F indicates the rollback of one or more AI functions and / or models. The twelfth or thirteenth identifier is carried in field F, with the twelfth and thirteenth identifiers being 0 and 1 respectively. For example, a value of 0 in field F (thirteenth identifier 0) indicates a rollback to not using AI functions, or a rollback of AI functions to non-AI; a value of 1 in field F (twelfth identifier 1) indicates a rollback to not using models, or a rollback of models to non-AI. As another example, two fields (field F1 and field F2) indicate the rollback of one or more AI functions and / or models, with the twelfth and thirteenth identifiers carried in fields F1 and F2 respectively. For example, field F1 indicates not using AI functions or a rollback of AI functions to non-AI, and field F2 indicates not using models or a rollback of models to non-AI. As yet another example, one or more fields indicate which application scenario or use case is rolled back to non-AI. For example, field F1 indicates a rollback to non-AI for beam management use cases, and field F2 indicates a rollback to non-AI for positioning use cases.

[0138] Based on the second information carried in the RRC signaling field or MAC CE, if the second information is used to indicate the rollback of one or more AI functions and / or models, in order to flexibly indicate the rollback operation of AI functions and / or models, in one embodiment, the second information includes one or more of the following:

[0139] The fourteenth identifier indicates a fallback to not using any AI features and / or models;

[0140] The fifteenth identifier indicates that the corresponding use case or application scenario has reverted to not using AI functions and / or models;

[0141] Identifiers or indexes of one or more models;

[0142] An identifier or index for one or more AI functions.

[0143] Here, where the second information is used to indicate the rollback of one or more AI functions and / or models, the second information may only include the fourteenth identifier, that is, an identifier or indication that the terminal does not use any AI function or model (a complete rollback to non-AI). The second information may only include the fifteenth identifier, or only include identifiers of one or more models, or only include identifiers of one or more AI functions; the model identifier indicates a rollback to not using the corresponding model, and the AI ​​function identifier indicates a rollback to not using the corresponding AI function. The second information may also include the fifteenth identifier, as well as identifiers of one or more models, and / or, identifiers of one or more AI functions.

[0144] To achieve rapid management of model / AI functions and reduce signaling load, the terminal also supports LCM (Local Management Module) of AI functions / models based on network-side configuration conditions. The condition-based AI functions / models can be AI functions / models available to the terminal, and may also consider AI functions / models supported by the terminal but not currently reported as available. Therefore, in one embodiment, the method further includes:

[0145] The terminal receives a fifth message and / or a sixth message sent by the network device, wherein the fifth message indicates the conditions for performing LCM on one or more AI functions and / or models, and the sixth message indicates the terminal to perform LCM on the conditions.

[0146] Here, upon receiving the fifth and / or sixth information, the terminal determines whether the conditions configured on the network side are met. If any condition is met, LCM is executed accordingly. For example, if the terminal determines that the conditions for activating a certain AI function and / or model are met, then the terminal activates that AI function and / or model and performs AI inference.

[0147] To enable rapid management of model / AI functions, the fifth piece of information may include network-side configured conditions. Based on this, in one embodiment, the fifth piece of information includes one or more of the following:

[0148] The threshold for the model's output accuracy;

[0149] Thresholds for network performance metrics;

[0150] Information about the application scenario;

[0151] Signal quality threshold;

[0152] Identification of candidate communities;

[0153] Time information;

[0154] Additional network-side conditions;

[0155] Conditions for activating AI functions and / or models;

[0156] Conditions for activating AI features and / or models;

[0157] Execute the conditions for selecting AI features and / or models;

[0158] Conditions for performing a rollback operation.

[0159] To enable rapid management of model / AI functions and reduce signaling overhead, in one embodiment, the sixth information includes one or more of the following:

[0160] The eighteenth identifier indicates the LCM of the AI ​​function and / or model of the terminal execution condition;

[0161] One or more nineteenth identifiers, one nineteenth identifier indicating one or more operations in the LCM of the terminal performing the conditional AI function and / or model.

[0162] Here, the model's output accuracy includes, but is not limited to, one or more of the following: the mean square error of the beam's reference signal received power (RSRP), positioning accuracy, channel state information (CSI) prediction accuracy, cell-level RSRP mean square error, measurement event prediction accuracy, radio link failure (RLF) prediction accuracy, and handover failure (HOF) prediction accuracy. If a higher output accuracy of any model is better, then if the model's output accuracy is greater than or equal to the corresponding threshold, it indicates that the model meets the activation condition; if the model's output accuracy is less than the corresponding threshold, it indicates that the model meets the deactivation condition. If a lower output accuracy of any model is better, such as the RSRP mean square error, then if the model's output accuracy is less than or equal to the corresponding threshold, it indicates that the model meets the activation condition; if the model's output accuracy is greater than the corresponding threshold, it indicates that the model meets the deactivation condition and / or the backoff condition.

[0163] Network performance metrics can be described as system performance metrics, including but not limited to one or more of throughput, latency, and HOF rate.

[0164] The application scenario information includes the name or identifier of the application scenario. Application scenarios include, but are not limited to, one or more of the following: indoor scenario, outdoor scenario, number of beams in Set A (e.g., 128), number of beams in Set B (e.g., 8), number of beams in Set A (e.g., 64), and number of beams in Set B (e.g., 8), terminal speed (e.g., 60 km / h), and non-line-of-sight (NLOS) scenario. If the fifth information includes application scenario information, the terminal is currently located in any of the application scenarios indicated by the fifth information, indicating that the condition is met.

[0165] The signal quality threshold may include, but is not limited to, one or more of the following: RSRP, Reference Signal Receiving Quality (RSRQ), and Signal-to-Interference plus Noise Ratio (SINR). For example, if the current signal quality of the terminal is greater than or equal to the corresponding threshold, it represents the activation condition of the AI ​​function and / or model; if the current signal quality of the terminal is less than the corresponding threshold, it represents the deactivation condition and / or fallback condition of the AI ​​function and / or model.

[0166] Candidate cell identifiers can be distributed in the form of a list, such as {cell 1,2,3}. If the fifth information includes candidate cell identifiers, and the terminal is located in any of the candidate cells indicated by the fifth information, the condition is met, and LCM is performed on the AI ​​function and / or model.

[0167] Time information can be understood as the time or time period for performing LCM on AI functions and / or models, such as the time period {T1, T2}. When the fifth piece of information includes time information, if the current time matches that time information, it indicates that the condition is met.

[0168] Network-side additional conditions can be described as including, but not limited to, associated IDs.

[0169] It should be noted that network devices can use one or more indicators to instruct terminals to perform operations such as selection, activation, deactivation, and rollback of conditional AI functions / models. Specifically, a network device can use a single indicator to instruct the execution of all LCM-related operations under a condition to save signaling overhead. For example, `conditionalManagement IEENUMERATED{true}` indicates that all operations are performed according to the condition; in this case, the sixth information includes the eighteenth identifier. Network devices can also set different indicators for different operations; in this case, the sixth information includes one or more nineteenth identifiers. Executing one or more operations in the LCM of a conditional AI function and / or model refers to operations such as selection, activation, deactivation, and rollback.

[0170] It should be noted that network devices can be configured with different conditions for different operations (selection, activation, deactivation, rollback, switching, etc.), i.e., different types of configuration parameters or specific parameter values; or different conditions (e.g., different parameters or specific parameter values) can be configured for different AI functions and / or models; both of the above can be configured simultaneously; no additional indication is required in the fifth information.

[0171] Among these features, network devices can also introduce separate Internet Explorer instances for different operations, for example:

[0172] conditionalActivationConfig IE / / Indicates conditional activation configuration

[0173] {Signal strength threshold (e.g., activated if greater than the threshold) {List of activated AI functions and / or models}};

[0174] conditionalDeactivationConfig IE / / Indicates conditional deactivation configuration

[0175] {Model output precision (e.g., deactivation if less than the threshold)};

[0176] conditionalFallbackConfig IE / / Indicates conditional fallback configuration

[0177] {Model output accuracy (e.g., back off if less than threshold) {List of models / functionalities to back off}}.

[0178] Furthermore, network devices can also introduce separate IEs for different AI functions and / or models; for example, AI Functionality #1.

[0179] {

[0180] conditionalActivationConfig IE / / Indicates conditional activation configuration

[0181] {Signal strength threshold (activate if greater than the threshold)}

[0182] conditionalDeactivationConfig IE / / Indicates conditional deactivation configuration

[0183] {Model output precision (e.g., deactivation if less than the threshold)}

[0184] conditionalFallbackConfig IE / / Indicates conditional fallback configuration

[0185] {Model output accuracy (e.g., back off if less than the threshold)}

[0186] }

[0187] The terminal can request the network side to configure condition-based LCM for AI functions / models and report auxiliary information so that the network side can configure the conditions related to the LCM of the AI ​​functions and / or models. Based on this, in one embodiment, the method further includes:

[0188] Send a seventh message and / or an eighth message to the network device, wherein the seventh message is used to request the network device to configure the AI ​​function and / or LCM of the model, and the eighth message is used to assist the network device in sending the configuration.

[0189] Here, the seventh piece of information can be an indication used to request the conditional AI feature / model's LCM. The eighth piece of information can be understood as auxiliary information.

[0190] To match the network device configuration conditions with the capabilities or requirements of the terminal, in one embodiment, the eighth information includes one or more of the following:

[0191] The terminal supports AI functions and / or model identifiers or indexes;

[0192] The terminal's available AI functions and / or model identifiers or indexes;

[0193] Application scenarios or use cases for AI functions or models;

[0194] Model output accuracy;

[0195] Additional terminal-side conditions associated with AI functions or models;

[0196] Additional network-side conditions associated with AI functions or models.

[0197] Here, the additional terminal-side conditions associated with AI functions or models include, but are not limited to, one or more of the following: terminal speed, computing power, storage capacity, and power consumption. The additional network-side conditions associated with AI functions or models include, but are not limited to, associated ID.

[0198] Upon receiving the second information, the terminal stops using the deactivated AI functions and / or models, and performs corresponding AI operations or AI inference using the activated, selected, or switched AI functions and / or models to obtain inference results. Upon receiving the second information, and the fifth and / or sixth information, the terminal assesses whether the network device configuration conditions are met. If any condition is met, the terminal performs the corresponding operation (e.g., activation, deactivation, selection, switching, rollback) on the AI ​​functions and / or models; stops using the deactivated AI functions and / or models, and performs corresponding AI operations or AI inference using the activated, selected, or switched AI functions and / or models to obtain inference results. If the second information indicates rollback of AI functions and / or models, and / or if the rollback execution conditions are met, if all AI functions and models are rolled back (no AI functions and models are used), then all AI inference is stopped; if one or more AI functions and / or models are rolled back (rollback indicated for different AI functions / models), then the corresponding AI functions / models are stopped from being used for AI inference, and traditional non-AI operations are performed to obtain the results of non-AI operations.

[0199] Upon obtaining inference results and / or measurement results (results of non-AI operations, actual measurement results), the terminal reports the inference results and / or measurement results to the network device, and indicates that the inference results are based on the output of a new AI function / model. When performing conditional LCM, the terminal can also report its current state to the network device, such as having reverted to non-AI operations or switched to a specific AI function and / or model. Based on this, in one embodiment, the method further includes:

[0200] Send a third message and / or a sixteenth identifier to the network device; wherein the sixteenth identifier indicates that the third message is the output of a new AI function and / or model, and / or indicates that the terminal has fallen back to non-AI operation or has switched to a new AI function and / or model.

[0201] Here, the third information includes inference results and / or measurement results, where measurement results include the results of non-AI operations and / or actual measurement results. When the third information includes inference results, the sixteenth identifier indicates that the third information is the output of a new AI function and / or model, and / or indicates that a switch to a new AI function and / or model has been made. When the third information includes measurement results, the sixteenth identifier indicates that the terminal has reverted to non-AI operations.

[0202] Based on the sixteenth identifier indicating that the third information is the output of a new AI function and / or model, the network device can also be informed which AI function and / or model the third information is the output of. Accordingly, in one embodiment, the third information includes a reasoning result and a seventeenth identifier, the seventeenth identifier indicating the AI ​​function and / or model associated with the reasoning result; or

[0203] The third information includes the inference result and the identifier or index of the AI ​​function and / or model associated with the inference result; or

[0204] The third information includes the reasoning results, which implicitly indicate the AI ​​functions and / or models associated with the reasoning results in the order in which the AI ​​functions and / or models are activated.

[0205] Here, the seventeenth identifier includes the identifier or index of the model associated with the inference result, and / or the identifier or index of the AI ​​function associated with the inference result. The implicit indication of the AI ​​function and / or model associated with the inference result according to the order in which the AI ​​function and / or model is activated can be understood as: reporting the inference results associated with each activated AI function and / or model in the order of activation. For example, if the AI ​​functions activated in the order of function #1, function #2, and function #3, the inference results are also reported in this order, i.e., the inference results of function #1, function #2, and function #3 are reported.

[0206] It should be noted that if the third information includes the measurement result, and the measurement result is the result of non-AI operation, the third information can also indicate that the result was generated by non-AI, or indicate that the terminal has reverted to non-AI operation.

[0207] Correspondingly, embodiments of this application provide an information management method applied to network devices, including base stations, which can be described as network-side devices or network-side equipment. For example... Figure 11 As shown, the method includes:

[0208] Step 1101: Receive first information sent by the terminal, the first information indicating the AI ​​functions and / or models available to the terminal.

[0209] To ensure that network devices accurately know the available AI functions and / or models supported by the terminal, in one embodiment, the first information includes the identifier or index of the AI ​​functions and / or models available to the terminal; and / or,

[0210] The first information includes priority information of the AI ​​functions and / or models available to the terminal.

[0211] To reduce complexity and signaling load, network devices can limit the number of available AI functions and / or models that a terminal can report. Based on this, in one embodiment, a fourth message is sent to the terminal, indicating the maximum number of available AI functions and / or models the terminal can report, and / or indicating priority information for the available AI functions and / or models the terminal can report.

[0212] Step 1102: Send second information to the terminal, the second information being used to instruct the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

[0213] To ensure that the AI ​​functions and / or models activated on the terminal are usable on the terminal, in one embodiment, sending the second information to the terminal includes:

[0214] Based on the first information, the second information is sent to the terminal.

[0215] To facilitate the terminal in quickly identifying which AI function and / or model is being LCMed, in one embodiment, the second information includes the identifier or index of one or more AI functions and / or models.

[0216] Network devices can use bitmaps to indicate the activation, deactivation, selection, switching, or rollback of one or more AI functions and / or models. Based on this, in one embodiment, the second information includes one or more bitmaps; wherein a bit in one bitmap indicates the activation, deactivation, selection, or switching of an AI function and / or model, and / or a change in a bit in one bitmap indicates switching an AI function and / or model, and / or the values ​​of all bits in the bitmap indicate deactivation to indicate rollback to not using the AI ​​function and / or model.

[0217] To facilitate the terminal's understanding that the bitmap is an LCM for AI functions and / or models, in one embodiment, the bitmap includes a first identifier indicating that the bitmap is used as an LCM for AI functions and / or models; or

[0218] The bitmap contains a second identifier and a third identifier, the second identifier indicating that the bitmap is used for the LCM of the model, and the third identifier indicating that the bitmap is used for the LCM of the AI ​​function.

[0219] To improve the flexibility and diversity of the second information delivery, the second information can be carried in RRC signaling. A new field can be designed to carry the second information, or existing available fields in RRC signaling can be reused. Based on this, in one embodiment, the second information is carried in an RRC signaling field, where one IE (Interface Entity) corresponds to an indication of activating, deactivating, selecting, switching, or reverting one or more AI functions or models.

[0220] To improve the flexibility and diversity of the second information delivery, the second information can be carried in a MAC CE. That is, the RRC signaling or MAC CE contains the identifier or index of AI functions and / or models. A new MAC CE can be designed to carry the second information, or an existing MAC CE can be reused. Based on this, in one embodiment, the second information is carried in one or more MAC CEs. A non-empty MAC CE is used to indicate activation or deactivation, selection, switching, or rollback of one or more AI functions and / or models, and / or an empty MAC CE is used to indicate rollback to not using all AI functions and / or models.

[0221] Here, network devices can configure one or more MAC CEs for each LCM operation (activation, deactivation, selection, switching, rollback), or multiple MAC CEs can be configured separately for AI functions and models. One MAC CE is used for an AI function or model. An empty MAC CE is used to instruct the terminal to roll back to non-AI, without using any AI functions and / or models.

[0222] To facilitate terminal understanding of which operation in the LCM of an AI function and / or model is being indicated by the RRC signaling or MAC CE, thereby enabling flexible management of AI functions and / or models, in one embodiment, the RRC signaling field or a non-empty MAC CE includes one or more of the following:

[0223] A fourth identifier, which indicates the LCM used for AI functions or models;

[0224] The fifth identifier indicates the LCM used for AI functions;

[0225] The sixth identifier indicates the LCM used for the model;

[0226] The seventh identifier indicates whether to activate, deactivate, select, switch, or roll back.

[0227] The eighth identifier is used to indicate whether to activate or deactivate;

[0228] The ninth identifier indicates activation;

[0229] The tenth identifier indicates that it is used for deactivation;

[0230] The eleventh identifier indicates a fallback to not using AI features or models;

[0231] The twelfth identifier indicates a fallback to a model that is not in use;

[0232] The thirteenth identifier indicates a fallback to not using AI features.

[0233] Based on the second information carried in the RRC field and / or MAC CE, if the second information is used to indicate the rollback of one or more AI functions and / or models, in order to flexibly indicate the rollback operation of AI functions and / or models, in one embodiment, the second information includes one or more of the following:

[0234] The fourteenth identifier indicates a fallback to not using any AI features and / or models;

[0235] The fifteenth identifier indicates that the corresponding use case or application scenario has reverted to not using AI functions and / or models;

[0236] Identifiers or indexes of one or more models;

[0237] An identifier or index for one or more AI functions.

[0238] To enable rapid management of model / AI functions and reduce signaling load, the terminal also supports conditional LCM (Limited Management Module) for AI functions / models. Network devices need to configure LCM-related conditions or execution conditions for the terminal. Based on this, in one embodiment, the method further includes:

[0239] Send a fifth message and / or a sixth message to the terminal, wherein the fifth message indicates the conditions for performing LCM on one or more AI functions and / or models, and the sixth message indicates that the terminal performs LCM on the conditions.

[0240] Here, the network device may send the fifth and / or sixth information to the terminal before or after sending the second information, or the network device may send the second information, the fifth information, and / or the sixth information to the terminal in parallel.

[0241] To enable rapid management of model / AI functions, the fifth piece of information may include network-side configured conditions. Based on this, in one embodiment, the fifth piece of information includes one or more of the following:

[0242] The threshold for the model's output accuracy;

[0243] Thresholds for network performance metrics;

[0244] Information about the application scenario;

[0245] Signal quality threshold;

[0246] Identification of candidate communities;

[0247] Time information;

[0248] Additional network-side conditions;

[0249] Conditions for activating AI functions and / or models;

[0250] Conditions for activating AI features and / or models;

[0251] Execute the conditions for selecting AI features and / or models;

[0252] Conditions for performing a rollback operation.

[0253] To enable rapid management of model / AI functions and reduce signaling overhead, in one embodiment, the sixth information includes one or more of the following:

[0254] The eighteenth identifier indicates the LCM of the AI ​​function and / or model of the terminal execution condition;

[0255] One or more nineteenth identifiers, one nineteenth identifier indicating one or more operations in the LCM of the terminal performing the conditional AI function and / or model.

[0256] In one embodiment, before sending the fifth and / or sixth information to the terminal, the method further includes:

[0257] The terminal sends a seventh message and / or an eighth message, wherein the seventh message is used to request the network device to configure the AI ​​function and / or the LCM of the model, and the eighth message is used to assist the network device in sending the configuration.

[0258] Here, upon receiving the seventh information, the network device determines the fifth and / or sixth information. These conditions can be dynamically configured by the network device for the terminal, or they can be conditions agreed upon between the network device and the terminal. Upon receiving the eighth information, or both the seventh and eighth information, the network device determines the fifth and / or sixth information based on the eighth information. The eighth information assists the network device in determining the execution conditions related to LCM.

[0259] To match the network device configuration conditions with the capabilities or requirements of the terminal, in one embodiment, the eighth information includes one or more of the following:

[0260] The terminal supports AI functions and / or model identifiers or indexes;

[0261] The terminal's available AI functions and / or model identifiers or indexes;

[0262] Application scenarios or use cases for AI functions or models;

[0263] Model output accuracy;

[0264] Additional terminal-side conditions associated with AI functions or models;

[0265] Additional network-side conditions associated with AI functions or models.

[0266] When the terminal activates AI functions and / or models, it performs AI inference using the activated AI functions and / or models to obtain inference results; alternatively, when the terminal reverts to non-AI mode, it performs traditional non-AI operations to obtain the results of the non-AI operations. Upon obtaining the inference results or the results of the non-AI operations, the terminal needs to report the obtained results to the network device. Based on this, in one embodiment, the method further includes:

[0267] Receive third information and / or a sixteenth identifier sent by the terminal; wherein the sixteenth identifier indicates that the third information is the output of a new AI function and / or model, and / or indicates that the terminal has fallen back to non-AI operation or has switched to a new AI function and / or model.

[0268] Based on the sixteenth identifier indicating that the third information is the output of a new AI function and / or model, to facilitate network devices in knowing which AI function and / or model the third information represents, so that the network devices can decide whether to instruct the terminal to switch AI functions and / or models, the third information includes, in one embodiment, a reasoning result and a seventeenth identifier, the seventeenth identifier indicating the AI ​​function and / or model associated with the reasoning result; or...

[0269] The third information includes the inference result and the identifier or index of the AI ​​function and / or model associated with the inference result; or

[0270] The third information includes the reasoning result and implicitly indicates the AI ​​function and / or model associated with the reasoning result in the order in which the AI ​​function and / or model is activated.

[0271] To implement the terminal-side method of this application embodiment, this application embodiment also provides an information management device, which is installed on the terminal, such as... Figure 12 As shown, the device includes:

[0272] The first sending unit 1201 is used to send first information to the network device, the first information indicating the AI ​​functions and / or models available to the terminal;

[0273] The first receiving unit 1202 is used to receive second information sent by the network device, the second information being used to indicate the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

[0274] In one embodiment, the second information includes one or more bitmaps; wherein a bit in one bitmap indicates activation or deactivation or selection or switching of an AI function and / or model, and / or a change in a bit in one bitmap indicates switching of an AI function and / or model, and / or the values ​​of all bits in the bitmap indicate deactivation indicating a fallback to not using an AI function and / or model.

[0275] In one embodiment, the bitmap includes a first identifier indicating that the bitmap is used for the LCM of AI functions and / or models; or

[0276] The bitmap contains a second identifier and a third identifier, the second identifier indicating that the bitmap is used for the LCM of the model, and the third identifier indicating that the bitmap is used for the LCM of the AI ​​function.

[0277] In one embodiment, the second information is carried in an RRC signaling field, where an IE in the RRC signaling field corresponds to an indication of activating or deactivating, selecting, switching, or rolling back one or more AI functions or models.

[0278] In one embodiment, the second information is carried in one or more MAC CEs, a non-empty MAC CE being used to indicate activation or deactivation or selection or switching or rollback of one or more AI functions and / or models, and / or an empty MAC CE being used to indicate rollback to not using all AI functions and / or models.

[0279] In one embodiment, the RRC signaling field or a non-empty MAC CE includes one or more of the following:

[0280] A fourth identifier, which indicates the LCM used for AI functions or models;

[0281] The fifth identifier indicates the LCM used for AI functions;

[0282] The sixth identifier indicates the LCM used for the model;

[0283] The seventh identifier indicates whether to activate, deactivate, select, switch, or roll back.

[0284] The eighth identifier is used to indicate whether to activate or deactivate;

[0285] The ninth identifier indicates activation;

[0286] The tenth identifier indicates that it is used for deactivation;

[0287] The eleventh identifier indicates a fallback to not using AI features or models;

[0288] The twelfth identifier indicates a fallback to a model that is not in use;

[0289] The thirteenth identifier indicates a fallback to not using AI features.

[0290] In one embodiment, the first information includes the identifier or index of the AI ​​functions and / or models available to the terminal; and / or,

[0291] The first information includes priority information of the AI ​​functions and / or models available to the terminal.

[0292] In one embodiment, the second information includes one or more of the following:

[0293] The fourteenth identifier indicates a fallback to not using any AI features and / or models;

[0294] The fifteenth identifier indicates that the corresponding use case or application scenario has reverted to not using AI functions and / or models;

[0295] Identifiers or indexes of one or more models;

[0296] An identifier or index for one or more AI functions.

[0297] In one embodiment, the device further includes:

[0298] The third sending unit is used to send third information and / or a sixteenth identifier to the network device; wherein the sixteenth identifier indicates that the third information is the output of a new AI function and / or model, and / or indicates that the terminal has fallen back to non-AI operation or has switched to a new AI function and / or model.

[0299] In one embodiment, the third information includes a reasoning result and a seventeenth identifier, the seventeenth identifier indicating the AI ​​function and / or model associated with the reasoning result; or

[0300] The third information includes the inference result and the identifier or index of the AI ​​function and / or model associated with the inference result; or

[0301] The third information includes the reasoning results, which implicitly indicate the AI ​​functions and / or models associated with the reasoning results in the order in which the AI ​​functions and / or models are activated.

[0302] In one embodiment, the device further includes:

[0303] The third receiving unit is used to receive fourth information sent by the network device, the fourth information indicating the maximum number of available AI functions and / or models that the terminal can report, and / or indicating the priority information of available AI functions and / or models that the terminal can report.

[0304] In one embodiment, the device further includes:

[0305] The fourth receiving unit is configured to receive fifth and / or sixth information sent by the network device, wherein the fifth information indicates the conditions for performing LCM on one or more AI functions and / or models, and the sixth information indicates the conditions for the terminal to perform LCM.

[0306] In one embodiment, the device further includes:

[0307] The fourth sending unit is used to send seventh information and / or eighth information to the network device. The seventh information is used to request the network device to configure the AI ​​function and / or model LCM of the configuration conditions, and the eighth information is used to assist the network device in sending the configuration.

[0308] In one embodiment, the fifth information includes one or more of the following:

[0309] The threshold for the model's output accuracy;

[0310] Thresholds for network performance metrics;

[0311] Information about the application scenario;

[0312] Signal quality threshold;

[0313] Identification of candidate communities;

[0314] Time information;

[0315] Additional network-side conditions;

[0316] Conditions for activating AI functions and / or models;

[0317] Conditions for activating AI features and / or models;

[0318] Execute the conditions for selecting AI features and / or models;

[0319] Conditions for performing a rollback operation.

[0320] In one embodiment, the sixth information includes one or more of the following:

[0321] The eighteenth identifier indicates the LCM of the AI ​​function and / or model of the terminal execution condition;

[0322] One or more nineteenth identifiers, one nineteenth identifier indicating one or more operations in the LCM of the terminal performing the conditional AI function and / or model.

[0323] In one embodiment, the eighth information includes one or more of the following:

[0324] The terminal supports AI functions and / or model identifiers or indexes;

[0325] The terminal's available AI functions and / or model identifiers or indexes;

[0326] Application scenarios or use cases for AI functions or models;

[0327] Model output accuracy;

[0328] Additional terminal-side conditions associated with AI functions or models;

[0329] Additional network-side conditions associated with AI functions or models.

[0330] In practical applications, the first sending unit 1201, the first receiving unit 1202, the third sending unit, the third receiving unit, the fourth receiving unit, and the fourth sending unit can be implemented by a processor in the information management device.

[0331] To implement the network device-side method of this application embodiment, this application embodiment also provides an information management device, which is installed on the network device, such as... Figure 13 As shown, the device includes:

[0332] The second receiving unit 1301 is used to receive first information sent by the terminal, the first information indicating the AI ​​functions and / or models available to the terminal;

[0333] The second sending unit 1302 is used to send second information to the terminal, the second information being used to indicate the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

[0334] In one embodiment, the second information includes one or more bitmaps; wherein a bit in one bitmap indicates activation or deactivation or selection or switching of an AI function and / or model, and / or a change in a bit in one bitmap indicates switching of an AI function and / or model, and / or the values ​​of all bits in the bitmap indicate deactivation, indicating a fallback to not using the AI ​​function and / or model.

[0335] In one embodiment, the bitmap includes a first identifier indicating that the bitmap is used for the LCM of AI functions and / or models; or

[0336] The bitmap contains a second identifier and a third identifier, the second identifier indicating that the bitmap is used for the LCM of the model, and the third identifier indicating that the bitmap is used for the LCM of the AI ​​function.

[0337] In one embodiment, the second information is carried in an RRC signaling field, where an IE in the RRC signaling field corresponds to an indication of activating or deactivating, selecting, switching, or rolling back one or more AI functions or models.

[0338] In one embodiment, the second information is carried in one or more MAC CEs, a non-empty MAC CE being used to indicate activation or deactivation or selection or switching or rollback of one or more AI functions and / or models, and / or an empty MAC CE being used to indicate rollback to not using all AI functions and / or models.

[0339] In one embodiment, the RRC signaling field or a non-empty MAC CE includes one or more of the following:

[0340] A fourth identifier, which indicates the LCM used for AI functions or models;

[0341] The fifth identifier indicates the LCM used for AI functions;

[0342] The sixth identifier indicates the LCM used for the model;

[0343] The seventh identifier indicates whether to activate, deactivate, select, switch, or roll back.

[0344] The eighth identifier is used to indicate whether to activate or deactivate;

[0345] The ninth identifier indicates activation;

[0346] The tenth identifier indicates that it is used for deactivation;

[0347] The eleventh identifier indicates a fallback to not using AI features or models;

[0348] The twelfth identifier indicates a fallback to a model that is not in use;

[0349] The thirteenth identifier indicates a fallback to not using AI features.

[0350] In one embodiment, the first information includes the identifier or index of the AI ​​functions and / or models available to the terminal; and / or,

[0351] The first information includes priority information of the AI ​​functions and / or models available to the terminal.

[0352] In one embodiment, the second information includes one or more of the following:

[0353] The fourteenth identifier indicates a fallback to not using any AI features and / or models;

[0354] The fifteenth identifier indicates that the corresponding use case or application scenario has reverted to not using AI functions and / or models;

[0355] Identifiers or indexes of one or more models;

[0356] An identifier or index for one or more AI functions.

[0357] In one embodiment, the second sending unit is specifically used to send the second information to the terminal based on the first information.

[0358] In one embodiment, the device further includes:

[0359] The fifth sending unit is used to send fourth information to the terminal, the fourth information indicating the maximum number of available AI functions and / or models that the terminal can report, and / or indicating the priority information of available AI functions and / or models that the terminal can report.

[0360] In one embodiment, the device further includes:

[0361] The fifth receiving unit is configured to receive the third information and / or the sixteenth identifier sent by the terminal; wherein the sixteenth identifier indicates that the third information is the output of a new AI function and / or model, and / or indicates that the terminal has reverted to non-AI operation or has switched to a new AI function and / or model.

[0362] In one embodiment, the third information includes a reasoning result and a seventeenth identifier, the seventeenth identifier indicating the AI ​​function and / or model associated with the reasoning result; or

[0363] The third information includes the inference result and the identifier or index of the AI ​​function and / or model associated with the inference result; or

[0364] The third information includes the reasoning results, which implicitly indicate the AI ​​functions and / or models associated with the reasoning results in the order in which the AI ​​functions and / or models are activated.

[0365] In one embodiment, the device further includes:

[0366] The sixth sending unit is used to send fifth information and / or sixth information to the terminal, wherein the fifth information indicates the conditions for performing LCM on one or more AI functions and / or models, and the sixth information indicates that the terminal performs LCM under the conditions.

[0367] In one embodiment, the device further includes:

[0368] The sixth receiving unit is used to receive the seventh information and / or the eighth information sent by the terminal. The seventh information is used to request the AI ​​function and / or LCM model of the network device configuration conditions, and the eighth information is used to assist the network device in sending the configuration.

[0369] In one embodiment, the fifth information includes one or more of the following:

[0370] The threshold for the model's output accuracy;

[0371] Thresholds for network performance metrics;

[0372] Information about the application scenario;

[0373] Signal quality threshold;

[0374] Identification of candidate communities;

[0375] Time information;

[0376] Additional network-side conditions;

[0377] Conditions for activating AI functions and / or models;

[0378] Conditions for activating AI features and / or models;

[0379] Execute the conditions for selecting AI features and / or models;

[0380] Conditions for performing a rollback operation.

[0381] In one embodiment, the sixth information includes one or more of the following:

[0382] The eighteenth identifier indicates the LCM of the AI ​​function and / or model of the terminal execution condition;

[0383] One or more nineteenth identifiers, one nineteenth identifier indicating one or more operations in the LCM of the terminal performing the conditional AI function and / or model.

[0384] In one embodiment, the eighth information includes one or more of the following:

[0385] The terminal supports AI functions and / or model identifiers or indexes;

[0386] The terminal's available AI functions and / or model identifiers or indexes;

[0387] Application scenarios or use cases for AI functions or models;

[0388] Model output accuracy;

[0389] Additional terminal-side conditions associated with AI functions or models;

[0390] Additional network-side conditions associated with AI functions or models.

[0391] In practical applications, the second receiving unit 1301, the second sending unit 1302, the fifth sending unit, the fifth receiving unit, the sixth sending unit, and the sixth receiving unit can be implemented by a processor in the information management device.

[0392] It should be noted that the above embodiments of the information management device are only illustrative examples of the division of program modules when performing information management. In practical applications, the above processing can be assigned to different program modules as needed, that is, the internal structure of the device can be divided into different program modules to complete all or part of the processing described above. In addition, the information management device and the information management method embodiments provided in the above embodiments belong to the same concept, and their specific implementation process can be found in the method embodiments, which will not be repeated here.

[0393] Based on the hardware implementation of the above program modules, and in order to implement the terminal-side method of the embodiments of this application, the embodiments of this application also provide a terminal, such as... Figure 14 As shown, terminal 1400 includes:

[0394] The first communication interface 1401 is capable of exchanging information with other network nodes;

[0395] The first processor 1402 is connected to the first communication interface 1401 to enable information interaction with other network nodes and to execute the methods provided by one or more of the aforementioned terminal-side technical solutions when running a computer program. The computer program is stored in the first memory 1403.

[0396] Specifically, the first communication interface 1401 is used to send first information to the network device and receive second information sent by the network device; wherein, the first information indicates the AI ​​functions and / or models available to the terminal, and the second information is used to indicate the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

[0397] In one embodiment, the second information includes one or more bitmaps; wherein a bit in one bitmap indicates activation or deactivation or selection or switching of an AI function and / or model, and / or a change in a bit in one bitmap indicates switching of an AI function and / or model, and / or the values ​​of all bits in the bitmap indicate deactivation indicating a fallback to not using an AI function and / or model.

[0398] In one embodiment, the bitmap includes a first identifier indicating that the bitmap is used for the LCM of AI functions and / or models; or

[0399] The bitmap contains a second identifier and a third identifier, the second identifier indicating that the bitmap is used for the LCM of the model, and the third identifier indicating that the bitmap is used for the LCM of the AI ​​function.

[0400] In one embodiment, the second information is carried in an RRC signaling field, where an IE in the RRC signaling field corresponds to an indication of activating or deactivating, selecting, switching, or rolling back one or more AI functions or models.

[0401] In one embodiment, the second information is carried in one or more MAC CEs, a non-empty MAC CE being used to indicate activation or deactivation or selection or switching or rollback of one or more AI functions and / or models, and / or an empty MAC CE being used to indicate rollback to not using all AI functions and / or models.

[0402] In one embodiment, the RRC signaling field or a non-empty MAC CE includes one or more of the following:

[0403] A fourth identifier, which indicates the LCM used for AI functions or models;

[0404] The fifth identifier indicates the LCM used for AI functions;

[0405] The sixth identifier indicates the LCM used for the model;

[0406] The seventh identifier indicates whether to activate, deactivate, select, switch, or roll back.

[0407] The eighth identifier is used to indicate whether to activate or deactivate;

[0408] The ninth identifier indicates activation;

[0409] The tenth identifier indicates that it is used for deactivation;

[0410] The eleventh identifier indicates a fallback to not using AI features or models;

[0411] The twelfth identifier indicates a fallback to a model that is not in use;

[0412] The thirteenth identifier indicates a fallback to not using AI features.

[0413] In one embodiment, the first information includes the identifier or index of the AI ​​functions and / or models available to the terminal; and / or,

[0414] The first information includes priority information of the AI ​​functions and / or models available to the terminal.

[0415] In one embodiment, the second information includes one or more of the following:

[0416] The fourteenth identifier indicates a fallback to not using any AI features and / or models;

[0417] The fifteenth identifier indicates that the corresponding use case or application scenario has reverted to not using AI functions and / or models;

[0418] Identifiers or indexes of one or more models;

[0419] An identifier or index for one or more AI functions.

[0420] In one embodiment, the first communication interface 1401 is further configured to send third information and / or a sixteenth identifier to the network device; wherein the sixteenth identifier indicates that the third information is the output of a new AI function and / or model, and / or indicates that the terminal has fallen back to non-AI operation or has switched to a new AI function and / or model.

[0421] In one embodiment, the third information includes a reasoning result and a seventeenth identifier, the seventeenth identifier indicating the AI ​​function and / or model associated with the reasoning result; or

[0422] The third information includes the inference result and the identifier or index of the AI ​​function and / or model associated with the inference result; or

[0423] The third information includes the reasoning result and implicitly indicates the AI ​​function and / or model associated with the reasoning result in the order in which the AI ​​function and / or model is activated.

[0424] In one embodiment, the first communication interface 1401 is further configured to receive fourth information sent by the network device, the fourth information indicating the maximum number of available AI functions and / or models that the terminal can report, and / or indicating priority information of available AI functions and / or models that the terminal can report.

[0425] In one embodiment, the first communication interface 1401 is further configured to receive fifth information and / or sixth information sent by the network device, wherein the fifth information indicates conditions for performing LCM on one or more AI functions and / or models, and the sixth information indicates that the terminal performs LCM under the conditions.

[0426] In one embodiment, the first communication interface 1401 is further configured to send a seventh message and / or an eighth message to the network device, wherein the seventh message is used to request the network device to configure AI functions and / or LCM models, and the eighth message is used to assist the network device in sending configurations.

[0427] In one embodiment, the fifth information includes one or more of the following:

[0428] The threshold for the model's output accuracy;

[0429] Thresholds for network performance metrics;

[0430] Information about the application scenario;

[0431] Signal quality threshold;

[0432] Identification of candidate communities;

[0433] Time information;

[0434] Additional network-side conditions;

[0435] Conditions for activating AI functions and / or models;

[0436] Conditions for activating AI features and / or models;

[0437] Execute the conditions for selecting AI features and / or models;

[0438] Conditions for performing a rollback operation.

[0439] In one embodiment, the sixth information includes one or more of the following:

[0440] The eighteenth identifier indicates the LCM of the AI ​​function and / or model of the terminal execution condition;

[0441] One or more nineteenth identifiers, one nineteenth identifier indicating one or more operations in the LCM of the terminal performing the conditional AI function and / or model.

[0442] In one embodiment, the eighth information includes one or more of the following:

[0443] The terminal supports AI functions and / or model identifiers or indexes;

[0444] The terminal's available AI functions and / or model identifiers or indexes;

[0445] Application scenarios or use cases for AI functions or models;

[0446] Model output accuracy;

[0447] Additional terminal-side conditions associated with AI functions or models;

[0448] Additional network-side conditions associated with AI functions or models.

[0449] It should be noted that the specific processing procedures of the first processor 1402 and the first communication interface 1401 can be understood by referring to the above method.

[0450] Of course, in practical applications, the various components in terminal 1400 are coupled together through bus system 1404. It can be understood that bus system 1404 is used to implement communication between these components. In addition to a data bus, bus system 1404 also includes a power bus, a control bus, and a status signal bus. However, for clarity, in... Figure 14 The general designated all buses as Bus System 1404.

[0451] The first memory 1403 in this embodiment is used to store various types of data to support the operation of the terminal 1400. Examples of such data include any computer program used to operate on the terminal 1400.

[0452] The methods disclosed in the embodiments of this application can be applied to the first processor 1402, or implemented by the first processor 1402. The first processor 1402 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware or by instructions in the form of software in the first processor 1402. The first processor 1402 may be a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The first processor 1402 can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the methods disclosed in the embodiments of this application can be directly reflected as being executed by a hardware decoding processor, or being executed by a combination of hardware and software modules in the decoding processor. The software modules may be located in a storage medium, which is located in the first memory 1403. The first processor 1402 reads the information in the first memory 1403 and completes the steps of the aforementioned method in combination with its hardware.

[0453] In an exemplary embodiment, terminal 1400 may be implemented by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers (MCUs), microprocessors, or other electronic components to perform the aforementioned method.

[0454] Based on the hardware implementation of the above program modules, and in order to implement the method on the network device side of the embodiments of this application, the embodiments of this application also provide a network device, such as... Figure 15 As shown, the network device 1500 includes:

[0455] The second communication interface 1501 is capable of exchanging information with other network nodes;

[0456] The second processor 1502 is connected to the second communication interface 1501 to enable information interaction with other network nodes. When running a computer program, it executes the methods provided by one or more technical solutions on the network device side. The computer program is stored in the second memory 1503.

[0457] Specifically, the second communication interface 1501 is used to receive first information sent by the terminal and to send second information to the terminal. The first information indicates the AI ​​functions and / or models available to the terminal, and the second information is used to indicate the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

[0458] In one embodiment, the second information includes one or more bitmaps; wherein a bit in one bitmap indicates activation or deactivation or selection or switching of an AI function and / or model, and / or a change in a bit in one bitmap indicates switching of an AI function and / or model, and / or the values ​​of all bits in the bitmap indicate deactivation indicating a fallback to not using an AI function and / or model.

[0459] In one embodiment, the bitmap includes a first identifier indicating that the bitmap is used for the LCM of AI functions and / or models; or

[0460] The bitmap contains a second identifier and a third identifier, the second identifier indicating that the bitmap is used for the LCM of the model, and the third identifier indicating that the bitmap is used for the LCM of the AI ​​function.

[0461] In one embodiment, the second information is carried in an RRC signaling field, where an IE in the RRC signaling field corresponds to an indication of activating or deactivating, selecting, switching, or rolling back one or more AI functions or models.

[0462] In one embodiment, the second information is carried in one or more MAC CEs, a non-empty MAC CE being used to indicate activation or deactivation or selection or switching or rollback of one or more AI functions and / or models, and / or an empty MAC CE being used to indicate rollback to not using all AI functions and / or models.

[0463] In one embodiment, the RRC signaling field or a non-empty MAC CE includes one or more of the following:

[0464] A fourth identifier, which indicates the LCM used for AI functions or models;

[0465] The fifth identifier indicates the LCM used for AI functions;

[0466] The sixth identifier indicates the LCM used for the model;

[0467] The seventh identifier indicates whether to activate, deactivate, select, switch, or roll back.

[0468] The eighth identifier is used to indicate whether to activate or deactivate;

[0469] The ninth identifier indicates activation;

[0470] The tenth identifier indicates that it is used for deactivation;

[0471] The eleventh identifier indicates a fallback to not using AI features or models;

[0472] The twelfth identifier indicates a fallback to a model that is not in use;

[0473] The thirteenth identifier indicates a fallback to not using AI features.

[0474] In one embodiment, the first information includes the identifier or index of the AI ​​functions and / or models available to the terminal; and / or

[0475] The first information includes priority information of the AI ​​functions and / or models available to the terminal.

[0476] In one embodiment, the second information includes one or more of the following:

[0477] The fourteenth identifier indicates a fallback to not using any AI features and / or models;

[0478] The fifteenth identifier indicates that the corresponding use case or application scenario has reverted to not using AI functions and / or models;

[0479] Identifiers or indexes of one or more models;

[0480] An identifier or index for one or more AI functions.

[0481] In one embodiment, the second communication interface 1501 is specifically used to send the second information to the terminal based on the first information.

[0482] In one embodiment, the second communication interface 1501 is further configured to send fourth information to the terminal, the fourth information indicating the maximum number of available AI functions and / or models that the terminal can report, and / or indicating the priority information of available AI functions and / or models that the terminal can report.

[0483] In one embodiment, the second communication interface 1501 is further configured to receive third information and / or a sixteenth identifier sent by the terminal; wherein the sixteenth identifier indicates that the third information is the output of a new AI function and / or model, and / or indicates that the terminal has reverted to non-AI operation or has switched to a new AI function and / or model.

[0484] In one embodiment, the third information includes a reasoning result and a seventeenth identifier, the seventeenth identifier indicating the AI ​​function and / or model associated with the reasoning result; or

[0485] The third information includes the inference result and the identifier or index of the AI ​​function and / or model associated with the inference result; or

[0486] The third information includes the reasoning results, which implicitly indicate the AI ​​functions and / or models associated with the reasoning results in the order in which the AI ​​functions and / or models are activated.

[0487] In one embodiment, the second communication interface 1501 is further configured to send fifth information and / or sixth information to the terminal, the fifth information indicating conditions for performing LCM on one or more AI functions and / or models, and the sixth information indicating that the terminal performs LCM on the conditions.

[0488] In one embodiment, the second communication interface 1501 is further configured to receive a seventh message and / or an eighth message sent by the terminal, wherein the seventh message is used to request the network device to configure AI functions and / or model LCM, and the eighth message is used to assist the network device in sending configuration.

[0489] In one embodiment, the fifth information includes one or more of the following:

[0490] The threshold for the model's output accuracy;

[0491] Thresholds for network performance metrics;

[0492] Information about the application scenario;

[0493] Signal quality threshold;

[0494] Identification of candidate communities;

[0495] Time information;

[0496] Additional network-side conditions;

[0497] Conditions for activating AI functions and / or models;

[0498] Conditions for activating AI features and / or models;

[0499] Execute the conditions for selecting AI features and / or models;

[0500] Conditions for performing a rollback operation.

[0501] In one embodiment, the sixth information includes one or more of the following:

[0502] The eighteenth identifier indicates the LCM of the AI ​​function and / or model of the terminal execution condition;

[0503] One or more nineteenth identifiers, one nineteenth identifier indicating one or more operations in the LCM of the terminal performing the conditional AI function and / or model.

[0504] In one embodiment, the eighth information includes one or more of the following:

[0505] The terminal supports AI functions and / or model identifiers or indexes;

[0506] The terminal's available AI functions and / or model identifiers or indexes;

[0507] Application scenarios or use cases for AI functions or models;

[0508] Model output accuracy;

[0509] Additional terminal-side conditions associated with AI functions or models;

[0510] Additional network-side conditions associated with AI functions or models.

[0511] It should be noted that the specific processing procedures of the second processor 1502 and the second communication interface 1501 can be understood by referring to the above method.

[0512] Of course, in practical applications, the various components in network device 1500 are coupled together through bus system 1504. It can be understood that bus system 1504 is used to implement communication between these components. In addition to a data bus, bus system 1504 also includes a power bus, a control bus, and a status signal bus. However, for clarity, in... Figure 15 The general labeled all buses as Bus System 1504.

[0513] The second memory 1503 in this embodiment is used to store various types of data to support the operation of the network device 1500. Examples of such data include any computer program used to operate on the network device 1500.

[0514] The methods disclosed in the embodiments of this application can be applied to, or implemented by, the second processor 1502. The second processor 1502 may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by the integrated logic circuitry of the hardware or by instructions in the form of software within the second processor 1502. The second processor 1502 may be a general-purpose processor, a DSP, or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The second processor 1502 can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the methods disclosed in the embodiments of this application can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software modules may be located in a storage medium, specifically a second memory 1503. The second processor 1502 reads information from the second memory 1503 and, in conjunction with its hardware, completes the steps of the aforementioned method.

[0515] In an exemplary embodiment, the network device 1500 may be implemented by one or more ASICs, DSPs, PLDs, CPLDs, FPGAs, general-purpose processors, controllers, MCUs, microprocessors, or other electronic components to perform the aforementioned method.

[0516] It is understood that the memories (first memory 1403 and second memory 1503) in the embodiments of this application can be volatile memory or non-volatile memory, or both. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), magnetic random access memory (FRAM), flash memory, magnetic surface memory, optical disc, or compact disc read-only memory (CD-ROM); magnetic surface memory can be disk storage or magnetic tape storage. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Synchronous Static Random Access Memory (SSRAM), Dynamic Random Access Memory (DRAM), Synchronous Dynamic Random Access Memory (SDRAM), Double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM), Enhanced Synchronous Dynamic Random Access Memory (ESDRAM), SyncLink Dynamic Random Access Memory (SLDRAM), and Direct Rambus Random Access Memory (DRRAM).The memories described in the embodiments of this application are intended to include, but are not limited to, these and any other suitable types of memories.

[0517] In an exemplary embodiment, this application also provides a storage medium, namely a computer storage medium, specifically a computer-readable storage medium, such as a first memory 1403 storing a computer program, which can be executed by a first processor 1402 of a terminal 1400 to complete the steps described in the aforementioned terminal-side method. Another example is a second memory 1503 storing a computer program, which can be executed by a second processor 1502 of a network device 1500 to complete the steps described in the aforementioned network device-side method. The computer-readable storage medium can be a memory such as FRAM, ROM, PROM, EPROM, EEPROM, Flash Memory, magnetic surface memory, optical disc, or CD-ROM.

[0518] By way of example, this application also provides a computer program product, including a computer program that can be executed by a first processor 1402 of a terminal 1400 to complete the steps described in the aforementioned terminal-side method. The computer program can also be executed by a second processor 1502 of a network device 1500 to complete the steps described in the aforementioned network device-side method.

[0519] It should be noted that terms such as "first" and "second" are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. "Multiple" can refer to two or more items, and "multiple" can refer to two or more items. The term "and / or" in this document merely describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, the term "one or more" in this document refers to any combination of at least two of the multiple elements. For example, including one or more of A, B, and C can represent including any one or at least two or more elements selected from the set consisting of A, B, and C.

[0520] Furthermore, the technical solutions described in the embodiments of this application can be combined arbitrarily without conflict.

[0521] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application.

Claims

1. An information management method, characterized in that, Applied to a terminal, the method includes: Send a first message to the network device, the first message indicating the artificial intelligence (AI) functions and / or models available to the terminal; Receive second information sent by the network device, the second information being used to instruct the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

2. The method according to claim 1, characterized in that, The second information includes one or more bitmaps; wherein, a bit in a bitmap indicates activation or deactivation or selection or toggling of an AI function and / or model, and / or, a change in a bit in a bitmap indicates toggling of an AI function and / or model, and / or, the values ​​of all bits in a bitmap indicate deactivation indicating a fallback to not using an AI function and / or model.

3. The method according to claim 2, characterized in that, The bitmap contains a first identifier indicating that the bitmap is used for lifecycle management (LCM) of AI functions and / or models; or The bitmap contains a second identifier and a third identifier, the second identifier indicating that the bitmap is used for the LCM of the model, and the third identifier indicating that the bitmap is used for the LCM of the AI ​​function.

4. The method according to claim 1, characterized in that, The second information is carried in the Radio Resource Control (RRC) signaling field. One information element (IE) in the RRC signaling field corresponds to an indication to activate or deactivate, select, switch, or roll back one or more AI functions or models.

5. The method according to claim 1, characterized in that, The second information is carried in one or more Media Access Control Layer Control Cells (MAC CEs). A non-empty MAC CE is used to indicate activation or deactivation or selection or switching or rollback of one or more AI functions and / or models, and / or an empty MAC CE is used to indicate rollback to not using all AI functions and / or models.

6. The method according to claim 4 or 5, characterized in that, The RRC signaling field or a non-empty MAC CE includes one or more of the following: A fourth identifier, which indicates the LCM used for AI functions or models; The fifth identifier indicates the LCM used for AI functions; The sixth identifier indicates the LCM used for the model; The seventh identifier indicates whether to activate, deactivate, select, switch, or roll back. The eighth identifier is used to indicate whether to activate or deactivate; The ninth identifier indicates activation; The tenth identifier indicates that it is used for deactivation; The eleventh identifier indicates a fallback to not using AI features or models; The twelfth identifier indicates a fallback to a model that is not in use; The thirteenth identifier indicates a fallback to not using AI features.

7. The method according to claim 1, characterized in that, The first information includes the identifiers or indexes of the AI ​​functions and / or models available to the terminal; and / or, The first information includes priority information of the AI ​​functions and / or models available to the terminal.

8. The method according to any one of claims 1, 4 to 5, and 7, characterized in that, The second information includes one or more of the following: The fourteenth identifier indicates a fallback to not using any AI features and / or models; The fifteenth identifier indicates that the corresponding use case or application scenario has reverted to not using AI functions and / or models; Identifiers or indexes of one or more models; An identifier or index for one or more AI functions.

9. The method according to any one of claims 1 to 5, 7, characterized in that, The method further includes: Send a third message and / or a sixteenth identifier to the network device; wherein the sixteenth identifier indicates that the third message is the output of a new AI function and / or model, and / or indicates that the terminal has fallen back to non-AI operation or has switched to a new AI function and / or model.

10. The method according to claim 9, characterized in that, The third piece of information includes the reasoning result and the seventeenth identifier, the seventeenth identifier indicating the AI ​​function and / or model associated with the reasoning result; or The third information includes the inference result and the identifier or index of the AI ​​function and / or model associated with the inference result; or The third information includes the reasoning results, which implicitly indicate the AI ​​functions and / or models associated with the reasoning results in the order in which the AI ​​functions and / or models are activated.

11. The method according to claim 1, characterized in that, The method further includes: The terminal receives a fourth message sent by the network device, the fourth message indicating the maximum number of available AI functions and / or models to be reported by the terminal, and / or indicating the priority information of available AI functions and / or models to be reported by the terminal.

12. The method according to any one of claims 1 to 5, 7, and 10 to 11, characterized in that, The method further includes: The terminal receives a fifth message and / or a sixth message sent by the network device, wherein the fifth message indicates the conditions for performing LCM on one or more models and / or AI functions, and the sixth message indicates the terminal to perform LCM under the conditions.

13. The method according to claim 12, characterized in that, The method further includes: Send a seventh message and / or an eighth message to the network device, wherein the seventh message is used to request the network device to configure the AI ​​function and / or LCM of the model, and the eighth message is used to assist the network device in sending the configuration.

14. The method according to claim 12, characterized in that, The fifth piece of information includes one or more of the following: The threshold for the model's output accuracy; Thresholds for network performance metrics; Information about the application scenario; Signal quality threshold; Identification of candidate communities; Time information; Additional network-side conditions; Conditions for activating model and / or AI functions; Conditions for deactivating model and / or AI functions; Conditions for performing the selection of models and / or AI functions; Conditions for performing a rollback operation.

15. The method according to claim 12, characterized in that, The sixth piece of information includes one or more of the following: The eighteenth identifier indicates the LCM of the AI ​​function and / or model of the terminal execution condition; One or more nineteenth identifiers, one nineteenth identifier indicating one or more operations in the LCM of the terminal performing the conditional AI function and / or model.

16. The method according to claim 13, characterized in that, The eighth piece of information includes one or more of the following: The terminal supports AI functions and / or model identifiers or indexes; The terminal's available AI functions and / or model identifiers or indexes; Application scenarios or use cases for AI functions or models; Model output accuracy; Additional terminal-side conditions associated with AI functions or models; Additional network-side conditions associated with AI functions or models.

17. An information management method, characterized in that, Applied to network devices, the method includes: The receiving terminal sends first information, which indicates the AI ​​functions and / or models available to the terminal. Send a second message to the terminal, the second message being used to instruct the activation or deactivation, selection, switching, or rollback of one or more AI functions and / or models.

18. The method according to claim 17, characterized in that, The second information includes one or more bitmaps; wherein, a bit in a bitmap indicates activation or deactivation or selection or toggling of an AI function and / or model, and / or, a change in a bit in a bitmap indicates toggling of an AI function and / or model, and / or, the values ​​of all bits in a bitmap indicate deactivation indicating a fallback to not using an AI function and / or model.

19. The method according to claim 18, characterized in that, The bitmap contains a first identifier indicating that the bitmap is used for the LCM of AI functions and / or models; or The bitmap contains a second identifier and a third identifier, the second identifier indicating that the bitmap is used for the LCM of the model, and the third identifier indicating that the bitmap is used for the LCM of the AI ​​function.

20. The method according to claim 17, characterized in that, The second information is carried in the RRC signaling field. One IE in the RRC signaling field corresponds to an indication to activate or deactivate, select, switch, or roll back one or more AI functions or models.

21. The method according to claim 17, characterized in that, The second information is carried in one or more MACCEs. A non-empty MACCE is used to indicate activation or deactivation or selection or toggling or rollback of one or more AI functions and / or models, and / or an empty MACCE is used to indicate rollback to not using all AI functions and / or models.

22. The method according to claim 20 or 21, characterized in that, The RRC signaling field or a non-empty MAC CE includes one or more of the following: A fourth identifier, which indicates the LCM used for AI functions or models; The fifth identifier indicates the LCM used for AI functions; The sixth identifier indicates the LCM used for the model; The seventh identifier indicates whether to activate, deactivate, select, switch, or roll back. The eighth identifier is used to indicate whether to activate or deactivate; The ninth identifier indicates activation; The tenth identifier indicates that it is used for deactivation; The eleventh identifier indicates a fallback to not using AI features or models; The twelfth identifier indicates a fallback to a model that is not in use; The thirteenth identifier indicates a fallback to not using AI features.

23. The method according to claim 17, characterized in that, The first information includes the identifiers or indexes of the AI ​​functions and / or models available to the terminal; and / or The first information includes priority information of the AI ​​functions and / or models available to the terminal.

24. The method according to any one of claims 17, 20 to 21, and 23, characterized in that, The second information includes one or more of the following: The fourteenth identifier indicates a fallback to not using any AI features and / or models; The fifteenth identifier indicates that the corresponding use case or application scenario has reverted to not using AI functions and / or models; Identifiers or indexes of one or more models; An identifier or index for one or more AI functions.

25. The method according to any one of claims 17 to 21, 23, characterized in that, Sending the second information to the terminal includes: Based on the first information, the second information is sent to the terminal.

26. The method according to claim 25, characterized in that, The method further includes: Send a fourth message to the terminal, the fourth message instructing the terminal to report the maximum number of available AI functions and / or models, and / or instructing the terminal to report priority information of available AI functions and / or models.

27. The method according to any one of claims 17 to 21, 23, and 26, characterized in that, The method further includes: Receive third information and / or a sixteenth identifier sent by the terminal; wherein the sixteenth identifier indicates that the third information is the output of a new AI function and / or model, and / or indicates that the terminal has fallen back to non-AI operation or has switched to a new AI function and / or model.

28. The method according to claim 27, characterized in that, The third piece of information includes the reasoning result and the seventeenth identifier, the seventeenth identifier indicating the AI ​​function and / or model associated with the reasoning result; or The third information includes the inference result and the identifier or index of the AI ​​function and / or model associated with the inference result; or The third information includes the reasoning results, which implicitly indicate the AI ​​functions and / or models associated with the reasoning results in the order in which the AI ​​functions and / or models are activated.

29. The method according to any one of claims 17 to 21, 23, and 28, characterized in that, The method further includes: Send a fifth message and / or a sixth message to the terminal, wherein the fifth message indicates the conditions for performing LCM on one or more AI functions and / or models, and the sixth message indicates that the terminal performs LCM on the conditions.

30. The method according to claim 29, characterized in that, The method further includes: The terminal sends a seventh message and / or an eighth message, wherein the seventh message is used to request the network device to configure the AI ​​function and / or the LCM of the model, and the eighth message is used to assist the network device in sending the configuration.

31. The method according to claim 29, characterized in that, The fifth piece of information includes one or more of the following: The threshold for the model's output accuracy; Thresholds for network performance metrics; Information about the application scenario; Signal quality threshold; Identification of candidate communities; Time information; Additional network-side conditions; Conditions for activating AI functions and / or models; Conditions for activating AI features and / or models; Execute the conditions for selecting AI features and / or models; Conditions for performing a rollback operation.

32. The method according to claim 29, characterized in that, The sixth piece of information includes one or more of the following: The eighteenth identifier indicates the LCM of the AI ​​function and / or model of the terminal execution condition; One or more nineteenth identifiers, one nineteenth identifier indicating one or more operations in the LCM of the terminal performing the conditional AI function and / or model.

33. The method according to claim 30, characterized in that, The eighth piece of information includes one or more of the following: The terminal supports AI functions and / or model identifiers or indexes; The terminal's available AI functions and / or model identifiers or indexes; Application scenarios or use cases for AI functions or models; Model output accuracy; Additional terminal-side conditions associated with AI functions or models; Additional network-side conditions associated with AI functions or models.

34. An information management device, characterized in that, include: The first sending unit is configured to send first information to the network device, wherein the first information indicates the AI ​​functions and / or models available to the terminal; The first receiving unit is configured to receive second information sent by the network device, the second information being used to indicate the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

35. An information management device, characterized in that, include: The second receiving unit is used to receive first information sent by the terminal, the first information indicating the AI ​​functions and / or models available to the terminal; The second sending unit is used to send second information to the terminal, the second information being used to indicate the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

36. A terminal, characterized in that, include: A first processor and a first communication interface; wherein... The first communication interface is used to send first information to the network device and receive second information sent by the network device. The first information indicates the AI ​​functions and / or models available to the terminal, and the second information is used to indicate the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

37. A network device, characterized in that, include: A second processor and a second communication interface; wherein... The second communication interface is used to receive first information sent by the terminal and send second information to the terminal. The first information indicates the AI ​​functions and / or models available to the terminal, and the second information is used to indicate the activation or deactivation or selection or switching or rollback of one or more AI functions and / or models.

38. A terminal, characterized in that, It includes a first processor and a first memory for storing computer programs that can run on the first processor. Wherein, when the first processor is used to run the computer program, it performs the steps of the method according to any one of claims 1 to 16.

39. A network device, characterized in that, It includes a second processor and a second memory for storing computer programs that can run on the second processor. Wherein, when the second processor is used to run the computer program, it performs the steps of the method according to any one of claims 17 to 33.

40. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 16, or the steps of the method according to any one of claims 17 to 33.

41. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 33.

Images