Information transmission method, apparatus

Abstract

The embodiments of the present application provide an information transmission method and an apparatus. The method comprises: for a terminal device having one or more AI / ML features, the terminal device receives, from a network device, first configuration information or second configuration information related to assessment or selection of AI / ML unit performances.

Description

Information transmission method and device Technical Field

[0001] The embodiments of the present application relate to the field of communication technologies. Background Art

[0002] NR Release 18 investigated artificial intelligence / machine learning (AI / ML) over the air interface. AI / ML can be applied to the following use cases: CSI feedback enhancement, beam management, and positioning enhancement. CSI feedback enhancement includes CSI prediction and CSI compression; beam management includes spatial beam prediction and temporal beam prediction; and positioning enhancement includes direct positioning and AI / ML-assisted positioning. These use cases are preliminary selections; new use cases may be added in Release 19, such as AI / ML for mobility management.

[0003] For 6G, assuming AI / ML is integrated with communications, there will be more use cases for applying AI / ML over the air interface.

[0004] With the introduction of these use cases, standardization efforts are underway to develop new protocols, processes, and signaling to support the reliable operation of AI / ML and ensure its effective benefits. These protocol-related methods and approaches are not only targeted at 5G-Advanced standards and commercial networks and devices, but can also be further applied to 6G networks and devices.

[0005] It should be noted that the above introduction to the technical background is merely intended to provide a clear and complete description of the technical solutions of this application and facilitate understanding by those skilled in the art. Simply because these solutions are described in the background technology section of this application, it should not be assumed that the above technical solutions are well known to those skilled in the art.

[0006] Summary of the Invention

[0007] NR Rel-18 does not explicitly consider and design the monitoring and assessment of inactive AI / ML models / functionalities. Without this design, there are no clear mechanisms, processes, or signaling for activating, selecting, or switching AI / ML models / functionalities over the air interface.

[0008] One way to monitor inactive AI / ML models / functionalities is to activate them and reuse the mechanisms defined for monitoring active AI / ML models / functionalities.

[0009] The inventors discovered that, however, in function-based lifecycle management (LCM), the network (NW) is generally unaware of the specific models on the terminal (UE) side. Therefore, it is impossible to activate inactive models one by one. Even if the network knows the model information, activating all inactive models will result in high costs in terms of RS overhead, signaling overhead, and reporting delay. Therefore, it is necessary to provide an effective mechanism to monitor and evaluate AI / ML models / functions in an inactive state.

[0010] To address at least one of the above problems, embodiments of the present application provide an information transmission method and apparatus.

[0011] According to one aspect of an embodiment of the present application, an information transmission method is provided, the method comprising: for a terminal device having one or more AI / ML features, the terminal device receiving first configuration information or second configuration information related to AI / ML unit performance evaluation or selection from a network device.

[0012] According to another aspect of an embodiment of the present application, an information transmission method is provided, the method comprising: a network device sending first configuration information or second configuration information related to AI / ML unit performance evaluation or selection to a terminal device.

[0013] According to another aspect of an embodiment of the present application, an information transmission device is provided, which is arranged in a terminal device having one or more AI / ML features. The device includes: a receiving unit, which is used to receive first configuration information or second configuration information related to AI / ML unit performance evaluation or selection from a network device.

[0014] One of the beneficial effects of the embodiments of the present application is that it can support and clearly define the performance monitoring / evaluation process of non-activated AI / ML models / functions; in addition, the gains from using AI / ML become reliable.

[0015] With reference to the following description and accompanying drawings, specific embodiments of the present application are disclosed in detail, indicating the manner in which the principles of the present application can be employed. It should be understood that the embodiments of the present application are not limited in scope. Within the spirit and scope of the appended claims, the embodiments of the present application include many variations, modifications and equivalents.

[0016] Features described and / or illustrated with respect to one embodiment may be used in the same or similar manner in one or more other embodiments, combined with features in other embodiments, or substituted for features in other embodiments.

[0017] It should be emphasized that the term "include / comprising" when used herein refers to the presence of features, integers, steps or components, but does not exclude the presence or addition of one or more other features, integers, steps or components. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The elements and features described in one figure or one embodiment of the present application can be combined with the elements and features shown in one or more other figures or embodiments. In addition, in the accompanying drawings, similar reference numerals represent corresponding parts in several figures and can be used to indicate corresponding parts used in more than one embodiment.

[0019] FIG1 is a schematic diagram of a communication system according to an embodiment of the present application;

[0020] FIG2 is a schematic diagram of an information transmission method according to an embodiment of the present application;

[0021] FIG3 is a schematic diagram of the AI / ML layered functional structure of an embodiment of the present application;

[0022] FIG4 is a schematic diagram of an AL / ML model / function and related information according to an embodiment of the present application;

[0023] FIG5 is an interaction diagram of an information interaction method for evaluating or monitoring AI / ML unit activation on a terminal side according to an embodiment of the present application;

[0024] FIG6 is an interaction diagram of an information interaction method for evaluating or monitoring AI / ML unit switching on the terminal side according to an embodiment of the present application;

[0025] FIG7 is an interaction diagram of an information interaction method for evaluating or monitoring AI / ML unit activation on the network side according to an embodiment of the present application;

[0026] FIG8 is an interaction diagram of an information interaction method for evaluating or monitoring AI / ML unit switching on the terminal side according to an embodiment of the present application;

[0027] FIG9 is an interactive diagram of terminal AI / ML capability query and reporting according to an embodiment of the present application;

[0028] FIG10 is a schematic diagram of AI / ML beam management for TX beam prediction according to an embodiment of the present application;

[0029] FIG11 is another schematic diagram of the information transmission method according to an embodiment of the present application;

[0030] FIG12 is a schematic diagram of an information transmission device according to an embodiment of the present application;

[0031] FIG13 is another schematic diagram of the information transmission device according to an embodiment of the present application;

[0032] FIG14 is a schematic block diagram of a system structure of a terminal device according to an embodiment of the present application;

[0033] FIG15 is a schematic block diagram of the system structure of the network device according to an embodiment of the present application. DETAILED DESCRIPTION

[0034] The above and other features of the present application will become apparent through the following description with reference to the accompanying drawings. In the description and the accompanying drawings, specific embodiments of the present application are disclosed in detail, which illustrate some embodiments in which the principles of the present application can be adopted. It should be understood that the present application is not limited to the described embodiments. On the contrary, the present application includes all modifications, variations and equivalents that fall within the scope of the appended claims.

[0035] In the embodiments of the present application, the terms "first", "second", etc. are used to distinguish different elements from the name, but do not indicate the spatial arrangement or temporal order of these elements, and these elements should not be limited by these terms. The term "and / or" includes any one and all combinations of one or more of the associated listed terms. The terms "comprising", "including", "having", etc. refer to the presence of the stated features, elements, components or components, but do not exclude the presence or addition of one or more other features, elements, components or components.

[0036] In the embodiments of this application, the singular forms "a," "the," etc. include plural forms and should be broadly understood to mean "a" or "a type" rather than being limited to "one." Furthermore, the term "said" should be understood to include both singular and plural forms, unless the context clearly indicates otherwise. Furthermore, the term "according to" should be understood to mean "at least in part based on...", and the term "based on" should be understood to mean "at least in part based on...", unless the context clearly indicates otherwise.

[0037] In the embodiments of the present application, the term "communication network" or "wireless communication network" may refer to a network that complies with any of the following communication standards, such as Long Term Evolution (LTE), enhanced Long Term Evolution (LTE-A, LTE-Advanced), Wideband Code Division Multiple Access (WCDMA), High-Speed ​​Packet Access (HSPA), etc.

[0038] Furthermore, communication between devices in the communication system may be carried out according to communication protocols of any stage, for example, including but not limited to the following communication protocols: 1G (generation), 2G, 2.5G, 2.75G, 3G, 4G, 4.5G and 5G, New Radio (NR), future 6G, etc., and / or other communication protocols currently known or to be developed in the future.

[0039] In the embodiments of the present application, the term "network device" refers to, for example, a device in a communication system that connects a terminal device to the communication network and provides services to the terminal device. Network devices may include, but are not limited to, the following devices: base station (BS), access point (AP), transmission reception point (TRP), broadcast transmitter, mobile management entity (MME), gateway, server, radio network controller (RNC), base station controller (BSC), etc.

[0040] Among them, base stations may include but are not limited to: NodeB (NodeB or NB), evolved NodeB (eNodeB or eNB) and 5G base station (gNB), DU, CU, IAB host, etc., and may also include remote radio head (RRH, Remote Radio Head), remote radio unit (RRU, Remote Radio Unit), relay (relay) or low-power node (such as femeto, pico, etc.). The term "base station" may include some or all of their functions. Each base station can provide communication coverage for a specific geographical area. The term "cell" can refer to a base station and / or its coverage area, depending on the context in which the term is used.

[0041] In the embodiments of the present application, the term "user equipment" (UE) or "terminal equipment" (TE) refers to, for example, a device that accesses a communication network through a network device and receives network services. A terminal device can be fixed or mobile and may also be referred to as a mobile station (MS), a terminal, a subscriber station (SS), an access terminal (AT), a station, and so on.

[0042] Among them, terminal devices may include but are not limited to the following devices: cellular phones, personal digital assistants (PDAs), wireless modems, wireless communication devices, handheld devices, machine-type communication devices, laptop computers, cordless phones, smart phones, smart watches, digital cameras, etc.

[0043] For another example, in scenarios such as the Internet of Things (IoT), the terminal device can also be a machine or device for monitoring or measurement, including but not limited to: machine type communication (MTC) terminal, vehicle-mounted communication terminal, device-to-device (D2D) terminal, machine-to-machine (M2M) terminal, and so on.

[0044] In addition, the term "network side" or "network device side" refers to one side of the network, which can be a base station or one or more network devices as described above. The term "user side" or "terminal side" or "terminal device side" refers to the user or terminal side, which can be a UE or one or more terminal devices as described above. Unless otherwise specified herein, "device" can refer to either network equipment or terminal equipment.

[0045] The following describes the scenarios of the embodiments of the present application through examples, but the present application is not limited thereto.

[0046] FIG1 is a schematic diagram of a communication system according to an embodiment of the present application, schematically illustrating a situation using a terminal device and a network device as an example. As shown in FIG1 , a communication system 100 may include a network device 101 and a terminal device 102. For simplicity, FIG1 illustrates only one terminal device and one network device as an example, but the embodiments of the present application are not limited thereto.

[0047] In the embodiment of the present application, existing services or future services can be transmitted between the network device 101 and the terminal devices 102 and 103. For example, these services may include but are not limited to: enhanced mobile broadband (eMBB), massive machine type communication (mMTC), and ultra-reliable and low-latency communication (URLLC), etc.

[0048] In the embodiments of the present application, the high-layer signaling may be, for example, radio resource control (RRC) signaling; for example, an RRC message, including, for example, an MIB, system information, or a dedicated RRC message; or an RRC information element (RRC IE). The high-layer signaling may also be, for example, MAC (Medium Access Control) signaling; or a MAC control element (MAC CE). However, the present application is not limited thereto.

[0049] In embodiments of the present application, one or more AI / ML models may be configured and run in a network device and / or a terminal device. The AI / ML models may be used for various signal processing functions in wireless communications, such as CSI prediction, CSI compression, beamforming, positioning management, and the like; however, the present application is not limited thereto.

[0050] Embodiments of the first aspect

[0051] The present application provides an information transmission method. FIG2 is a schematic diagram of the information transmission method according to the present application embodiment. As shown in FIG2 , the method includes:

[0052] 201: For a terminal device having one or more AI / ML features, the terminal device receives first configuration information or second configuration information related to AI / ML unit performance evaluation or selection from a network device.

[0053] It is worth noting that FIG2 above is merely a schematic illustration of an embodiment of the present application, and the present application is not limited thereto. For example, the execution order of the various operations may be appropriately adjusted, and other operations may be added or some operations may be reduced. Those skilled in the art may make appropriate modifications based on the above description, and are not limited to the description of FIG2 above.

[0054] In some embodiments, an AI / ML unit may also be referred to as an AI / ML element.

[0055] In some embodiments, the AI / ML unit is a first AI / ML unit or a second AI / ML unit, or includes a first AI / ML unit and a second AI / ML unit.

[0056] In some embodiments, the first AI / ML unit includes one or more second AI / ML units. In other words, the first AI / ML unit is implemented by one or more second AI / ML units.

[0057] In some embodiments, the AI / ML unit corresponds to an AI / ML feature and / or an AI / ML function and / or an AI / ML logical model and / or an AI / ML physical model.

[0058] In some embodiments, functionality refers to an AI / ML feature / feature group enabled by a configuration, where the configuration is supported based on conditions indicated by UE capabilities.

[0059] For example, the function can be to use AI / ML for spatial beam prediction, or to use AI / ML for time beam prediction, or to use AI / ML for CSI prediction, or to use AI / ML for direct positioning, or to use AI / ML for assisted positioning, and so on.

[0060] In some embodiments, a terminal device having AI / ML features or AI / ML functions is implemented by one or more AI / ML models. The AI / ML model can refer to a logical model or one or more physical models.

[0061] In some embodiments, the first AI / ML feature includes one or more second AI / ML features, or the first AI / ML feature includes one or more first AI / ML functions; and / or the first AI / ML function includes one or more second AI / ML functions, or the first AI / ML function includes one or more first models; and / or the first AI / ML model includes one or more second AI / ML models.

[0062] In some embodiments, the terminal capability list lists the relevant features and / or feature groups for each over-the-air AI / ML use case, along with the parameters within each feature group. Terminal devices report their support for these features and feature groups based on their capabilities. In subsequent operations of AI / ML models or capabilities, one definition is to call an enabled feature a functionality.

[0063] For example, beam management may include an AI / ML feature for beam spatial prediction, which may include two features: transmit beam prediction and beam pair prediction. When the system enables or activates the AI / ML feature for beam spatial prediction on a terminal device, the terminal device is considered to have the AI / ML function for beam spatial prediction.

[0064] In some embodiments, a terminal device may also have multiple AI / ML features, such as AI / ML features with beam spatial prediction, AI / ML features with CSI prediction, AI / ML features with direct positioning, etc.

[0065] Table 1 is an example. For AI / ML features, a list of predefined features, feature groups, feature units, conditional parameter information, auxiliary parameter information, and related indexes is provided. The network can obtain AI / ML information from the terminal through queries and responses with the terminal.

[0066] Table 1

[0067] To better understand the AI / ML units on terminal devices, network devices often require more tables, such as input and output information and auxiliary information. In particular, for the functionality-based AI / ML lifecycle management (LCM) studied in Rel-18, providing more information on the terminal device facilitates fine-grained control of the AI / ML units on terminal devices by network devices.

[0068] Figure 3 is a schematic diagram of the AI / ML hierarchical functional structure of an embodiment of the present application. As shown in Figure 3 (a), an AI / ML feature / function includes multiple AI / ML models; as shown in Figure 3 (b), a feature of an AI / ML feature group includes multiple AI / ML functions; and as shown in Figure 3 (c), a first AI / ML unit includes multiple second AI / ML units.

[0069] The above is just an example of a two-level AI / ML layered functional structure on the terminal side. A structure with more than two levels can be used when necessary. When a structure with more than two levels is used, the third-level unit, the fourth-level unit, and higher-level units can also include the above structure.

[0070] It should be noted that the following text of this application will more often use the terms AI / ML unit, first AI / ML unit, and second AI / ML unit to more clearly represent the hierarchical relationship between them. In actual network devices, terminal devices, and network systems, AI / ML features, AI / ML functionality, and AI / ML models are more often used to describe them.

[0071] In some embodiments, an AI / ML unit, such as an AI / ML model, is provided on the terminal device side, and the terminal device monitors or evaluates the performance of the AI / ML model.

[0072] For example, for an inactive AI / ML unit, the terminal device evaluates the performance of the inactive AI / ML unit; and monitors or evaluates the performance of an active AI / ML unit.

[0073] In operation 201 , for a terminal device having one or more AI / ML features, the terminal device receives first configuration information or second configuration information related to AI / ML unit performance evaluation or selection from a network device.

[0074] In some embodiments, the first configuration information includes one or more of AI / ML activation-related performance evaluation or selection-related performance information, activation trigger information, and AI / ML unit indication information.

[0075] In some embodiments, the first configuration information corresponds to a case where an AI / ML unit is activated. For example, for an AI / ML feature of a terminal device, when all AI / ML units included in the AI / ML feature are not activated, the first configuration information includes performance information related to determining activation of the AI / ML units.

[0076] In some embodiments, the performance information includes one or more of the following performance indicator-related information used for evaluation of the AI / ML unit:

[0077] Performance thresholds for evaluating AI / ML activation on the terminal side;

[0078] Performance threshold range for evaluating AI / ML activation on the terminal side;

[0079] Reporting performance thresholds for evaluating AI / ML activation on the network side;

[0080] Performance threshold range for evaluating AI / ML activation on the network side.

[0081] In some embodiments, the performance information further includes one or more of the following performance preference indication information:

[0082] Instructions for selecting the best performing AI / ML unit;

[0083] Instruction information for selecting the best AI / ML unit whose performance is greater than or equal to the performance threshold or the reported performance threshold;

[0084] Selecting an AI / ML unit having an indication that performance meets the performance threshold range or the reported performance threshold range;

[0085] Select the indication information corresponding to the AI / ML unit with the lowest input reference signal overhead;

[0086] Selecting an AI / ML unit that meets one of the performance indicators and corresponding indication information with the lowest input reference signal overhead; and

[0087] An AI / ML unit that meets one of the performance indicators and corresponding indication information with the lowest output signal overhead are selected.

[0088] In some embodiments, the activation trigger information includes one or more of the following information:

[0089] Counter parameters that trigger activation requests;

[0090] Timer parameters that trigger activation requests;

[0091] Additional event parameters that trigger the activation request.

[0092] In some embodiments, the counter or timer is typically used in conjunction with a performance threshold.

[0093] For example, if the counter parameter is n, then when an event higher than the threshold value occurs n times in a row, the counter is triggered;

[0094] For example, if the timer parameter is t, the timer triggers when the value of the event exceeds the threshold for t consecutive times.

[0095] When the performance indicators are other, the trigger conditions can be inferred in the same way.

[0096] In some embodiments, other events may correspond to the above-mentioned counters, or timers combined with trigger conditions of other AI / ML unit operating indicators, such as unit power consumption indicators, level indicators, memory indicators, processing latency indicators, etc.

[0097] In some embodiments, a trigger can only be triggered if specific indicators are met and the counter or timer indicators are satisfied.

[0098] The above triggering, when the triggering condition is met, the terminal device sends an activation request to the network device according to the configuration.

[0099] In some embodiments, the AI / ML unit indication information includes one or more of the following information:

[0100] AI / ML auxiliary information, which is used to instruct an AI / ML unit to evaluate the auxiliary information;

[0101] AI / ML input parameter or output parameter information, which is used to indicate the evaluation of the AI / ML unit with the input parameter or output parameter;

[0102] The AI / ML condition information is used to indicate an AI / ML unit that evaluates whether the condition information is satisfied.

[0103] In some embodiments, the second configuration information includes one or more of AI / ML switching-related performance evaluation or selection-related performance information, switching trigger information, and AI / ML unit indication information.

[0104] In some embodiments, the second configuration information corresponds to a case where an AI / ML unit is switched. For example, for an AI / ML feature of a terminal device, when an AI / ML unit included in the AI / ML feature is activated, the second configuration information includes performance information for determining the switching of the AI / ML unit.

[0105] In some embodiments, the performance information includes one or more of the following performance indicator-related information used for evaluation of the AI / ML unit:

[0106] Performance threshold for evaluating AI / ML switching on the terminal side;

[0107] Performance threshold range for evaluating AI / ML switching on the terminal side;

[0108] Reporting performance thresholds used to evaluate AI / ML switching on the network side;

[0109] This parameter specifies the performance threshold range for evaluating AI / ML switching on the network side.

[0110] In some embodiments, the performance-related information further includes one or more of the following performance preference indication information:

[0111] The performance of the selected AI / ML unit is higher than that of the activated AI / ML unit, and the performance difference is higher than the performance threshold or the reporting performance threshold.

[0112] Select an AI / ML unit for performance comparison with an activated AI / ML unit, and the performance difference falls within the performance threshold range or the reporting performance threshold range.

[0113] An AI / ML unit is selected for performance comparison with an activated AI / ML unit. The performance difference falls within the performance threshold range or the reported performance threshold range, and the input reference signal or output information overhead is lower than that of the activated AI / ML unit.

[0114] An AI / ML unit is selected for performance comparison with the activated AI / ML unit. The performance difference is less than the performance threshold or the reporting performance threshold, and the input reference signal or output information overhead is lower than that of the activated AI / ML unit.

[0115] In some embodiments, the switching trigger information includes one or more of the following information:

[0116] Counter parameters that trigger AI / ML switching requests;

[0117] Timer parameters for triggering AI / ML switching request;

[0118] Other event parameters that trigger the AI / ML switching request.

[0119] In some embodiments, the counter or timer is typically used in conjunction with a performance threshold.

[0120] For example, if the counter parameter is n, then when an event higher than the threshold value occurs n times in a row, the counter is triggered;

[0121] For example, if the timer parameter is t, the timer triggers when the value of the event exceeds the threshold for t consecutive times.

[0122] When the performance indicators are other, the trigger conditions can be inferred in the same way.

[0123] In some embodiments, other events may correspond to the above-mentioned counters, or timers combined with trigger conditions of other AI / ML unit operating indicators, such as unit power consumption indicators, level indicators, memory indicators, processing latency indicators, etc.

[0124] In some embodiments, a trigger can only be triggered if specific indicators are met and the counter or timer indicators are satisfied.

[0125] The above triggering, when the triggering condition is met, the terminal device sends a switching request to the network device according to the configuration.

[0126] In some embodiments, the AI / ML unit indication information includes one or more of the following information:

[0127] AI / ML auxiliary information, which is used to indicate the evaluation of the AI / ML unit with the auxiliary information compared with the activated AI / ML unit;

[0128] AI / ML input parameter or output parameter information, which is used to indicate that the AI / ML unit having the input parameter or output parameter is evaluated and compared with the activated AI / ML unit;

[0129] The AI / ML condition information is used to indicate that the AI / ML unit that evaluates the condition information satisfies the condition information is compared with the activated AI / ML unit.

[0130] In some embodiments, as shown in FIG2 , the method further includes:

[0131] 202: Corresponding to the first configuration information and / or the second configuration information, the terminal device further receives corresponding reporting configuration information from the network device.

[0132] The reported configuration information includes part of the first configuration information and / or the second configuration information.

[0133] In some embodiments, the reporting configuration information includes instructing the terminal device to perform the reporting behavior after the performance evaluation or selection of the AI / ML unit, and includes one or more of the following reporting content-related configurations:

[0134] AI / ML switching or activation request related configuration;

[0135] Information and configuration related to the AI / ML switching or activation unit;

[0136] AI / ML switching candidate units information and configuration;

[0137] AI / ML unit input configuration information related to reporting configuration;

[0138] Reporting configuration related to AI / ML unit output configuration information;

[0139] Report configuration related to AI / ML unit auxiliary configuration information;

[0140] Configuration related to reporting of auxiliary condition information of AI / ML units;

[0141] Reporting configuration related to AI / ML performance information.

[0142] In some embodiments, the reporting configuration information includes instructing the terminal device to perform the reporting behavior after the performance evaluation or selection of the AI / ML unit, and the configuration related to the reporting content includes one or more of the following:

[0143] selected AI / ML unit input configuration information and / or selected AI / ML unit performance information;

[0144] The selected AI / ML unit outputs configuration information and / or performance information of the selected AI / ML unit;

[0145] selected AI / ML unit auxiliary information and / or selected AI / ML unit performance information;

[0146] selected AI / ML unit condition information and / or selected AI / ML unit performance information;

[0147] The selected AI / ML unit identification and / or the performance information of the selected AI / ML unit.

[0148] In some embodiments, the selected AI / ML unit is one unit or multiple units.

[0149] In some embodiments, the selected AI / ML unit is in an inactive state.

[0150] In some embodiments, the selection of the content is based on the first configuration information or the second configuration information.

[0151] In some embodiments, the format of the content report is based on the reporting configuration information.

[0152] In some embodiments, the reporting configuration information includes instructing the terminal device to perform the performance evaluation or selection of the AI / ML unit and includes one or more of the following reporting methods and resource configuration information:

[0153] Periodic reporting, non-periodic reporting, and semi-continuous reporting;

[0154] Configure dedicated PRACH resources to trigger and report AI / ML switching or activation requests;

[0155] Reporting is based on dedicated PUCCH / MAC CE / RRC messages.

[0156] In some embodiments, a terminal device sending a switching request to a network device based on a configuration can use the same reporting as a terminal device sending an activation request to a network device based on a configuration. That is, the reported configuration, resources, and content can be shared. This is because both correspond to requests to activate an AI / ML unit. The difference is that during the AI / ML unit switching process, an already activated AI / ML unit needs to be deactivated or stopped. For requests to activate a new AI / ML unit, the relevant signaling and processes can use a consistent framework.

[0157] At this point, in operation 201, the terminal device may also receive third configuration information related to AI / ML unit performance evaluation or selection from the network device. This third configuration information may consistently correspond to activation and switching evaluations, and accordingly, reporting information relative to the third configuration information may be further configured. This reporting information may consistently support AI / ML activation and switching requests, as well as reporting of related content.

[0158] In this way, the third configuration information may include one or more of AI / ML unit activation / switching related performance evaluation or selection related performance information, activation / switching trigger information, and AI / ML unit indication information.

[0159] The performance information includes one or more of the following performance indicator-related information used for evaluating the AI / ML unit:

[0160] Performance thresholds for evaluating AI / ML activation / switching on the terminal side;

[0161] Performance threshold range for evaluating AI / ML activation / switching on the terminal side;

[0162] Reporting performance thresholds used to evaluate AI / ML activation / switching on the network side;

[0163] This parameter specifies the performance threshold range for evaluating AI / ML activation / switching on the network side.

[0164] Corresponding to the case where an AI / ML unit has already been activated, after the network side receives the activation request from the terminal device, if it confirms that the AI / ML unit can be activated, in addition to sending activation signaling to the terminal device, it will also send deactivation signaling to the terminal device to deactivate the activated AI / ML unit.

[0165] To avoid unclear AI / ML unit indications, separate indications are required through indication messages from the relevant AI / ML units. Alternatively, by default, for a first AI / ML unit on a terminal device, if a second AI / ML unit is already activated and the second AI / ML unit is in an activated state, and the terminal device receives an activation command from the network device for the first AI / ML unit, the terminal device will deactivate the activated AI / ML unit. The terminal device then activates the new AI / ML unit based on the activation command information.

[0166] In some embodiments, in addition to the performance indicator-related information in the performance information, other information in the third configuration information (such as performance preference indication information, activation / switching trigger information, and AI / ML unit indication information in the performance information, etc.) may be the same or similar to the relevant information in the first configuration information or the second configuration information (such as performance preference indication information, activation trigger information, switching trigger information, and AI / ML unit indication information in the performance information, etc.), and the specific content will not be repeated.

[0167] In some embodiments, as shown in FIG2 , the method further includes:

[0168] 203: The terminal device reports the capability information of the terminal device to the network device, where the capability information of the terminal device includes AI / ML unit information.

[0169] In some embodiments, the AI / ML unit information includes one or more of the following information:

[0170] Whether the first AI / ML unit and / or the second AI / ML unit are supported according to the query of the network device;

[0171] whether there is only one first AI / ML unit and / or second AI / ML unit or multiple first AI / ML units;

[0172] The number of the first AI / ML unit and / or the second AI / ML unit being queried;

[0173] the number of second AI / ML units in the queried first AI / ML unit;

[0174] The model identifier of the AI / ML unit being queried; and

[0175] Auxiliary information of the queried AI / ML unit.

[0176] Condition information of the AI / ML unit being queried.

[0177] Input and output configuration information of the AI / ML unit being queried.

[0178] In some embodiments, the content and / or format of the report is predefined and / or configured by the network device.

[0179] In some embodiments, AI / ML unit activation may be evaluated or monitored on the terminal side or on the network side.

[0180] In some embodiments, AI / ML unit switching may be evaluated or monitored on the terminal side or on the network side.

[0181] The following are detailed descriptions of each.

[0182] FIG4 is a schematic diagram of the AL / ML model / function and related information according to an embodiment of the present application.

[0183] In some embodiments, an AI / ML model or function, one way of indicating it is through an identifier (ID) of the model or function, which in this embodiment of the application can be simply expressed as an identifier of an AI / ML unit.

[0184] However, in many cases, there may be no identification information. In this case, the corresponding unit can be indicated by the input information, output information, conditional information, auxiliary information, etc. of the AI / ML unit.

[0185] The input information may refer to configuration information related to the input reference signal of the unit, and the number of input signal dimensions may be related to the overhead related to the use case, such as the input reference signal overhead in beam management.

[0186] Output information can refer to configuration information related to the unit's output information. For example, in a CSI compression use case, the number of CSI feedback bits output by the terminal is a core metric for the AI / ML unit. For the same input signal, fewer output bits indicate a higher compression ratio. This can also be considered as lower output or feedback overhead.

[0187] It should be noted that there is no clear definition between the auxiliary information and conditional information of the AI / ML unit. It can be simply considered that the conditional information is more targeted at describing the AI / ML unit itself, that is, the application conditions of the AI / ML unit (such as the application scenario, application site, application configuration, application data, etc., or it can be storage information related to the terminal AI / ML unit, power consumption information, processing latency information, etc.), while the auxiliary information is more targeted at describing the input data characteristics of the AI / ML unit.

[0188] In some embodiments, in the terminal capability reporting, the network device identifies or understands the information of the AI / ML unit in the terminal device through reporting of one or more of the above information (identification information, input and output information, auxiliary information, condition information).

[0189] With this information, network devices can instruct, control, and operate their corresponding AI / ML units.

[0190] FIG5 is an interaction diagram of an information interaction method for evaluating or monitoring AI / ML unit activation on the terminal side according to an embodiment of the present application. As shown in FIG5 , the method includes:

[0191] The network device sends an AI / ML capability query request to the terminal device;

[0192] The terminal device generates AI / ML capability information, which may include AI / ML unit information, such as information of the first AI / ML unit and / or the second AI / ML unit;

[0193] The terminal device reports terminal device capability information to the network device, including AI / ML unit information, such as one or more of identification information, input and output information, auxiliary information, and condition information;

[0194] Network devices send AI / ML performance information to terminal devices, which may include performance indicators, performance preference information, etc.

[0195] The network device sends AI / ML unit input reference signal configuration information to the terminal device;

[0196] The network device sends AI / ML unit reporting configuration information to the terminal device, which may include one or more of AI / ML switching or activation request related configuration, AI / ML switching or activation unit related information and configuration, AI / ML switching or candidate unit related information and configuration, AI / ML unit input and output configuration information related reporting configuration, AI / ML unit auxiliary information related reporting configuration, AI / ML unit condition information related reporting configuration, and AI / ML performance information related reporting configuration;

[0197] End devices conduct evaluation of inactive AI / ML units;

[0198] The terminal device reports AI / ML information or activation requests to the network device;

[0199] Activate or enable AI / ML on network devices.

[0200] FIG6 is an interaction diagram of an information interaction method for evaluating or monitoring AI / ML unit switching on the terminal side according to an embodiment of the present application. As shown in FIG6 , the method includes:

[0201] The terminal device reports AI / ML information or activation requests to the network device;

[0202] Activate or enable AI / ML on network devices;

[0203] The network device sends input reference signal configuration information for the activated AI / ML unit and the inactivated AI / ML unit to the terminal device;

[0204] Network devices send AI / ML performance information to terminal devices, which may include performance indicators, performance preference information, etc.

[0205] The network device sends AI / ML unit reporting configuration information to the terminal device, which may include one or more of AI / ML switching or activation request related configuration, AI / ML switching or activation unit related information and configuration, AI / ML switching or candidate unit related information and configuration, AI / ML unit input and output configuration information related reporting configuration, AI / ML unit auxiliary information related reporting configuration, AI / ML unit condition information related reporting configuration, and AI / ML performance information related reporting configuration;

[0206] The terminal device performs monitoring or evaluation of activated AI / ML units and evaluation of inactive AI / ML units;

[0207] The terminal device reports the evaluation results to the network device, including inactive AI / ML unit information, such as one or more of identification information, input and output information, auxiliary information, and condition information.

[0208] FIG7 is an interaction diagram of an information interaction method for evaluating or monitoring AI / ML unit activation on the network side according to an embodiment of the present application. As shown in FIG7 , the method includes:

[0209] The network device sends an AI / ML capability query request to the terminal device;

[0210] The terminal device generates AI / ML capability information, which may include AI / ML unit information, such as information of the first AI / ML unit and / or the second AI / ML unit;

[0211] The terminal device reports terminal device capability information to the network device, including AI / ML unit information, such as one or more of identification information, input and output information, auxiliary information, and condition information;

[0212] Network devices send AI / ML performance information to terminal devices, which may include performance indicators, performance preference information, etc.

[0213] The network device sends AI / ML unit input reference signal configuration information to the terminal device;

[0214] The network device sends AI / ML unit reporting configuration information to the terminal device, which may include one or more of AI / ML switching or activation request related configuration, AI / ML switching or activation unit related information and configuration, AI / ML switching or candidate unit related information and configuration, AI / ML unit input and output configuration information related reporting configuration, AI / ML unit auxiliary information related reporting configuration, AI / ML unit condition information related reporting configuration, and AI / ML performance information related reporting configuration;

[0215] End devices conduct evaluation of inactive AI / ML units;

[0216] The terminal device reports AI / ML information to the network device, including information of all inactivated or selected AI / ML units, such as one or more of identification information, input and output information, auxiliary information, and condition information;

[0217] Network devices perform evaluation or monitoring of all inactive or selected AI / ML units;

[0218] Activate or enable AI / ML on network devices.

[0219] FIG8 is an interactive diagram of an information interaction method for evaluating or monitoring AI / ML unit switching on the terminal side according to an embodiment of the present application. As shown in FIG8 , the method includes:

[0220] The terminal device reports AI / ML information or activation requests to the network device;

[0221] Activate or enable AI / ML on network devices;

[0222] The network device sends input reference signal configuration information for the activated AI / ML unit and the inactivated AI / ML unit to the terminal device;

[0223] Network devices send AI / ML performance information to terminal devices, which may include performance indicators, performance preference information, etc.

[0224] The network device sends AI / ML unit reporting configuration information to the terminal device, which may include one or more of AI / ML switching or activation request related configuration, AI / ML switching or activation unit related information and configuration, AI / ML switching or candidate unit related information and configuration, AI / ML unit input and output configuration information related reporting configuration, AI / ML unit auxiliary information related reporting configuration, AI / ML unit condition information related reporting configuration, and AI / ML performance information related reporting configuration;

[0225] The terminal device performs monitoring or evaluation of activated AI / ML units and evaluation of inactive AI / ML units;

[0226] The terminal device reports AI / ML information to the network device, including information of all inactivated or selected AI / ML units, such as one or more of identification information, input and output information, auxiliary information, and condition information;

[0227] Network devices perform evaluation or monitoring of all inactive or selected AI / ML units;

[0228] The network device sends AI / ML unit switching information to the terminal device.

[0229] The following is a detailed description of the contents involved in each of the above steps.

[0230] 1) Regarding UE capability:

[0231] Figure 9 is an interactive diagram of terminal AI / ML capability query and reporting in an embodiment of the present application. As shown in Figure 9, the network device sends a terminal device capability query to the terminal device, and the terminal device reports the terminal device capability information to the network device based on the query.

[0232] In some embodiments, a network device (e.g., a gNB) sends a terminal device capability query, which may include AI / ML unit-related capability queries, such as AI / ML feature / feature group or function or model information queries. Therefore, the terminal device sends its AI / ML capability report to the network device.

[0233] In some embodiments, specifically, the terminal device capability query may include a query related to the AI / ML first unit and / or the second AI / ML unit. Thus, in the terminal device capability report, the terminal device may report its first unit and / or second unit information.

[0234] For example, the network device may query the terminal device side for beam management related features, which may include first AI / ML functional unit information, such as support (or non-support) of spatial domain beam prediction features and / or support (or non-support) of time domain prediction features. In addition, the query may also include second unit information, such as whether there are multiple models / functions and / or how many models or functions are used for spatial / time domain beam prediction.

[0235] Thus, the terminal device sends its AI / ML capability information in the terminal device capability report. For AI / ML features / feature groups, the report can include information about its first AI / ML unit and / or second AI / ML unit, or more information, to the network device. This information can include detailed information about the first and / or second units, or simply indicate the number of first / second units, or simply indicate the presence of multiple first and / or second AI / ML units.

[0236] In some embodiments, the report content and format are configurations and / or predefined rules sent by the network device. For example, the network device can be further configured for further specific queries of the unit. The terminal device only reports such information when such specific queries are configured.

[0237] In some embodiments, the capability information of the terminal device, such as AI / ML features and parameter lists, is predefined, such as Table XXX, and the network device and the terminal device complete corresponding queries and responses based on this table entry.

[0238] For example, for spatial prediction, the network device may further inquire whether a beam information specific model is supported.

[0239] For example, as a type of unit condition information or auxiliary information, specific beam information can be a beam index, beam angle, beam codebook, beam mode, beam width, or other beam information. When queried, the terminal device will accordingly provide the auxiliary information of the AI / ML unit in its UE report. Another example of auxiliary information or condition information can be information related to the wireless scenario.

[0240] For example, different wireless scenarios may require different AI / ML units to support, such as units for indoor use and units for outdoor use. In addition, it can be information related to mobility. Different mobile speeds or mobilities may require different AI / ML units, such as units for low mobility or units for high mobility.

[0241] It should be noted that the conditional information here is more of a description of the application conditions of the AI / ML unit, while the auxiliary information is more of a description of the data features targeted by the unit. Based on different processes, the two may or may not need to be distinguished.

[0242] 2) Regarding monitoring or evaluation of inactive AI / ML units for activation:

[0243] In some embodiments, before activating or enabling the AI / ML unit, it is necessary to compare the performance of using traditional methods (non-AI methods) with the performance of using the AI / ML unit. The network device may need to notify the terminal device of performance information and / or performance indicator information and configure the reporting information required by the network accordingly.

[0244] In some embodiments, performance indicator information is related to the network device's intent to use the AI / ML unit. For example, the network device may wish to use the AI / ML unit to reduce the overhead of input reference signals or reduce the latency required to implement the unit's functionality. In this case, if necessary, the network device may send performance metric thresholds, indications serving as performance indicator information, and performance preference information to the terminal device. Reporting related to unit input parameters can be configured in the corresponding reporting configuration.

[0245] For example, Figure 10 is a schematic diagram of AI / ML beam management for TX beam prediction in an embodiment of the present application. As shown in Figure 10, in beam management, a terminal device can report its TX beam prediction capabilities, which can serve as the first AI / ML unit or the second AI / ML unit. Simply put, it is called an AI / ML beam management unit. The output of this unit is the RSRP of all beams, called SetA; the input of this unit is the RSRP of a subset of beams, called SetB.

[0246] In the traditional beam management method, the terminal device can report the maximum total number of configured NZP-CSI-RS resources supported by the terminal device to measure L1-RSRP and / or other L1-RSRP capability information related to beam management (maxNumberCSI RS Resource, maxNumberCSI RS ResourceTwoTx, maxNumber SSB CSI RS ResourceOneTx, etc.), and such information can be directly reused for SetA in this example. Alternatively, it can be reported independently in the AI / ML beam management feature capability report. It can be assumed that the maximum number of RS (CSI-RS, SSB) resources, the maximum L1-RSRP measured accordingly by the traditional method, and the corresponding total number of beams are the same as the number of beams measured based on AI / ML (number of beams of SetA / L1-RSRP).

[0247] This AI / ML unit can include multiple second AI / ML units. Each second AI / ML unit corresponds to a specific SetB. Different SetBs have different beam / L1RSRP subsets and different SetB / SetA ratios. For example, if SetA has 16 beams and SetB has 4 beams, the ratio is 0.25; if SetB has 8 beams, the ratio is 0.5.

[0248] In one scenario, network equipment wants to reduce the latency, RS consumption, and signaling overhead of AI / ML units used in beam management. There are two mechanisms: one is monitoring / evaluation of inactive AI / ML units on the terminal side, and the other is monitoring / evaluation of AI / ML units on the network side.

[0249] Similarly, in CSI prediction, different AI / ML units are needed to support different mobile speeds and / or different scenarios, such as urban, suburban, highway, high-speed rail, etc.

[0250] For inactive AI / ML units, the network device needs to configure the end device according to its preference for using the AI / ML unit and guide the end device to make AI / ML unit selection.

[0251] In this example, the input configuration may relate to the RS configuration of its measurement window, such as how many CSI RS cases will be used, the interval between CSI RS cases, etc. The auxiliary information may include mobile speed or mobility information.

[0252] In some embodiments, a common mechanism can be used for other terminal device-side AI / ML units. Similarly, the mechanisms and processes proposed in the embodiments of this application can be applied to all use cases where AI / ML units are deployed on the terminal device side.

[0253] In some embodiments, the process for monitoring / evaluating inactive AI / ML units on the terminal side includes the following steps:

[0254] Step 1: The network device configures AI / ML unit input reference signal resources to the terminal device based on the terminal device capability report related to AI / ML features;

[0255] Step 2: The network device configures the terminal device to report related configuration information, which can be configured based on an existing reporting framework, such as a CSI reporting framework;

[0256] Step 3: Based on the terminal capability information, the network device configures the terminal device for AI / ML monitoring / evaluation, including performance information, performance indicator information, and event trigger information;

[0257] Step 4: The network device configures AI / ML activation related reports for the terminal device;

[0258] Step 5: When the terminal device detects that the performance of the second AI / ML unit matches the performance information based on the performance indicator and satisfies the activation event triggering condition, the terminal device sends an activation report / request to the AI / ML unit, which may include the input configuration information / information identifier of the selected AI / ML unit, or the auxiliary information / condition information (or corresponding information identifier), or the ID of the selected unit;

[0259] Step 6: The network device responds by activating the selected AI / ML unit through input information or auxiliary information or ID.

[0260] In some embodiments, the auxiliary information or condition information may be other conditions or parameters that can differentiate the AI / ML units.

[0261] For example, different wireless scenarios may require different AI / ML units to support, such as units for indoor use and units for outdoor use. In addition, different mobile speeds or mobilities may require different AI / ML units, such as units for low mobility and units for high mobility.

[0262] This auxiliary information or conditional information can be obtained by the network device through the terminal device capability report. Through this auxiliary information or conditional information, the network device and the terminal device can have a common understanding of the selected AI / ML unit even without the unit identification.

[0263] When there is more than one unit corresponding to the auxiliary information or condition information, the terminal selects the best one among the corresponding indicators according to the performance indicator information indicated by the network, such as the one with the highest accuracy or the lowest input overhead.

[0264] In step 3, the network device may configure the terminal device to perform AI / ML evaluation / monitoring of the first unit without providing further configuration of the second unit. The performance indicators / criteria used in the evaluation / monitoring may be configured.

[0265] For the aforementioned beam management, the network device configures the terminal device to perform AI / ML evaluation / monitoring of TX beam prediction. The network device may not have information about how many and which types of SetBs are supported on the terminal device side. This is because AI / ML is based on function-based LCM. In this case, the network device has information about the first AI / ML unit (the transmit beam prediction function) but not about the second AI / ML unit (the AI / ML model for each SetB).

[0266] For example, the performance indicator can be the TOP-1 prediction accuracy of a given beam, and / or the TOP-1 beam prediction accuracy with a 1 dB margin, or other KPIs used in NR Rel-18 discussions (e.g., [RP-231766 TR 38.843 v1.0.0]). Here, prediction accuracy (%) can be defined as the percentage of "the top-1 measured beam is the top-1 predicted beam."

[0267] As mentioned above, the expected gain of using AI / ML depends on the preferences of the network device. When the network device wants to reduce reference signal overhead and beam selection delay, the network device wants to select the smallest SetB with acceptable beam prediction accuracy, that is, the AI / ML unit has the smallest input overhead.

[0268] In another case, the network device using AI / ML expects gain for higher beam prediction accuracy and may need to report the highest SetB and its beam prediction accuracy, i.e., the AI / ML unit has the largest input overhead and the highest accuracy.

[0269] When a network device configures performance indicator information to a terminal device, if the purpose is to reduce overhead, the information may instruct the terminal device to select a model with the smallest SetB, which is close to the expected prediction accuracy. If the purpose is to achieve high prediction accuracy, the information may instruct the terminal device to select a model with the highest prediction accuracy.

[0270] For example, a detailed configuration example could be:

[0271] To reduce overhead, the configuration signaling in step 3 may include:

[0272] AI / ML first unit information: the function indication of transmitting beam prediction may correspond to the function identifier determined during the UE function reporting process;

[0273] Performance information: acceptable top-1 prediction accuracy X (e.g., 80%). Possible values ​​of prediction accuracy can be predefined and represented by an indicator, such as a multi-bit message.

[0274] Performance indicator information: The smaller the number of beams representing SetB, the better, for example, using 1 bit. Alternatively, a prediction offset value, such as 5%, can be given. This means that if the prediction accuracy is within the range of prediction accuracy X - prediction offset to prediction accuracy X (75%-80%), the smallest SetB should be selected. The offset step size or value can be predefined. Here, prediction accuracy X corresponds to the basic condition for activating AI / ML, and the offset step size corresponds to the margin condition for this basic condition. The two have different uses.

[0275] To achieve high prediction accuracy, the configuration signaling in step 3 may include:

[0276] First AI / ML unit information: the function indication of transmitting beam prediction may correspond to the function identifier determined in the UE function reporting process;

[0277] Performance information: baseline top-1 prediction accuracy X (e.g., 80%). Potential values ​​of prediction accuracy can be predefined.

[0278] Performance indicator information: indicates that the higher the prediction accuracy, the better, such as using 1 bit to represent it.

[0279] In both cases, the terminal device performs AI / ML evaluation / monitoring and selects the best second AI / ML unit that meets the performance requirements for reporting in step 4. Alternatively, the terminal device can select all second AI / ML units that meet the performance requirements and report them in step 4. In this case, the final decision rests with the network.

[0280] For example, all second units in SetB that achieve a prediction accuracy within the range of prediction accuracy X - prediction offset to prediction accuracy X (75%-80%) are reported, and the network selects the unit with the smallest input overhead, or the SetB input parameters or parameter indications of all units reporting prediction accuracy higher than X are reported, and the network selects the unit with the best performance.

[0281] In step 4, the network device configures the terminal device to report the second AI / ML unit selected for AI / ML activation.

[0282] The terminal device reports when it can meet the performance conditions. The report content may include: AI / ML activation request, selected SetB information, that is, the number of SetB beams or L1-RSRP number, and unit prediction accuracy information.

[0283] Reporting methods include:

[0284] Method 1: The terminal device sends an AI / ML activation request to the network device. When the network device further responds, the terminal device sends information about the selected unit.

[0285] Method 2: The terminal device sends the AI / ML activation request and the selected unit information to the network device.

[0286] In some embodiments, the report may be sent via one or more of a dedicated PRACH, SR (via PUCCH), MAC-CE, or RRC message.

[0287] In some embodiments, the process for monitoring / evaluating inactive AI / ML units on the network side includes the following steps:

[0288] Step 1: The network device configures RS resources for management based on the terminal device capability report;

[0289] Step 2: Configure the network device-related reporting method, such as based on the existing CSI reporting framework;

[0290] Step 3: The network device can configure the AI / ML performance information and performance indicator information for evaluation to the terminal device;

[0291] Step 4: The network device configures the terminal device to report CSI based on the second AI / ML unit;

[0292] Step 5: The terminal device reports CSI based on the second AI / ML unit;

[0293] Step 6: The network device selects the best AI / ML second unit, sends the selected unit information to the terminal device, and performs AI / ML activation.

[0294] In the description of the above embodiment, the description of the parts that are consistent with the inactive AI / ML monitoring / evaluation on the terminal device side is omitted.

[0295] In the above embodiment, one approach is to require the terminal device to report all prediction accuracies of all second AI / ML units and Set B or information of each unit. This approach will result in a high reporting overhead.

[0296] Alternatively, if the network device expects high prediction accuracy, the network device can configure a basic prediction accuracy threshold to the terminal device and require the terminal device to report second AI / ML unit information that meets the accuracy requirement, and / or corresponding prediction accuracy information.

[0297] If the network device is interested in reducing AI / ML unit overhead, the network device can configure the desired unit overhead information to the terminal device and require the terminal device to report the prediction accuracy of the second AI / ML unit that can meet the overhead request.

[0298] On the network device side, the AI / ML performance is further evaluated. The network device decides to activate the second AI / ML unit selected by the terminal. Based on the evaluation results, the network device can explicitly activate the second AI / ML unit using its unit ID (if available).

[0299] Alternatively, the network device may configure the SetB information of the selected second AI / ML unit for its activation. In case of a SetB configuration, if the terminal has multiple second units, the terminal device should select the one with the highest prediction accuracy in the most recent report.

[0300] A common problem for both device-side and network-side assessments is that, in order to ensure reliable AI / ML activation and avoid ping-pong effects, an event counter or event timer can be configured during the assessment process. This means that when the activation condition is continuously met within the time of the event counter, or continuously detected within the time period of the event timer, AI / ML activation can be requested or decided.

[0301] 3) Monitoring / evaluation of inactive AI / ML units for switching:

[0302] Part 3) is very similar to part 2) above. In this case, there is an AI / ML unit that is active and running. The problem is to select an inactive AI / ML unit with better performance and require the end device to switch and use that unit.

[0303] In function-based LCM, the network device knows that a first AI / ML unit is activated, and the network device may or may not know which second AI / ML unit is running.

[0304] The network device may configure the terminal device to monitor / evaluate the inactive second AI / ML unit.

[0305] In this configuration, similar to section 2), the network device is configured with performance information, including performance indicator information. However, the performance indicator information is based on the network device's preference for using the AI / ML unit.

[0306] To reduce overhead, the network device can configure the desired overhead configuration for the terminal device and request the terminal device to feedback the prediction accuracy. Alternatively, the network device can configure the prediction accuracy and request the terminal device to feedback the minimum overhead that can meet the accuracy request.

[0307] If the goal is better performance, the network device can configure the end device to select a second AI / ML unit with better performance than the currently running AI / ML unit. To avoid a ping-pong effect, a performance offset can be configured. This means that the second AI / ML unit will only be selected if its performance is better than the currently running unit by a gain greater than the offset value. Furthermore, an event counter or event timer can be configured. This means that a switch from the end device to the network device's AI / ML unit can only be requested if a higher performance event is continuously detected at the counter number, or if the duration of the consecutive higher performance events is longer than the timer duration.

[0308] The overhead in the above embodiment may correspond to the overhead of the model input parameters, such as the number of input RSs, or may correspond to the overhead of the output parameters.

[0309] In some embodiments, regarding the monitoring / evaluation of inactive AI / ML units on the terminal device side, the process may include the following steps:

[0310] Step 1: The network device allocates RS resources to the terminal device for management based on the terminal device capability report;

[0311] Step 2: Network equipment configuration and terminal equipment related AI / ML evaluation reporting, such as based on the CSI reporting framework;

[0312] Step 3: The network device configures the terminal device to monitor / evaluate the inactive AI / ML unit, including performance information, performance indicator information, and event trigger information;

[0313] Step 4: Network device configuration terminal device AI / ML unit switching related reporting;

[0314] Step 5: When the terminal device detects that the performance of the second AI / ML unit is consistent with the performance information based on the performance indicator, and / or the event triggering condition for AI / ML unit switching is met, the terminal device reports / requests AI / ML unit switching, which may include input information of the selected AI / ML unit, auxiliary information of the selected AI / ML unit, and the ID of the selected AI / ML unit;

[0315] Step 6: The network device responds and sends a switching instruction / configuration to the terminal device based on the ID, input information, auxiliary information, and condition information.

[0316] In the above process, one issue is that if the network device has AI / ML unit ID information, the network device can configure the selected AI / ML unit ID for switching. If the network device does not have such information, the network device needs to use input / output information obtained from the terminal device report or auxiliary information and condition information of the selected unit to indicate the AI / ML unit.

[0317] In some embodiments, the input information may be input configuration information, such as SetB information in beam management, or output configuration information, such as output information in CSI compression.

[0318] In some embodiments, the auxiliary information or condition information may be other conditions or parameters that can differentiate the AI / ML units.

[0319] For example, different wireless scenarios may require different AI / ML units to support, such as units for indoor use and units for outdoor use. In addition, different mobile speeds or mobilities may require different AI / ML units, such as units for low mobility and units for high mobility.

[0320] In some embodiments, the network device may obtain auxiliary information or conditional information through the terminal device capability report. Even without the unit ID, the network device and the terminal device may have a common understanding of the selected AI / ML unit through the auxiliary information or conditional information.

[0321] In some embodiments, regarding inactive AI / ML monitoring / assessment on the network device side, the process may include the following steps:

[0322] Step 1: The network device configures RS resources for management based on the terminal device capability report;

[0323] Step 2: Reporting related to network device configuration and AI / ML evaluation, such as based on the CSI reporting framework;

[0324] Step 3: The network device can configure AI / ML performance information and performance indicator information to the terminal device;

[0325] Step 4: The network device configures the terminal device to report based on the second AI / ML unit, such as CSI reporting;

[0326] Step 5: The terminal device reports based on the second AI / ML unit, which may include input information of the selected AI / ML unit, auxiliary information of the selected AI / ML unit, condition information, and the ID of the selected AI / ML unit;

[0327] Step 6: The network device selects the best AI / ML unit based on the ID, input information, output information, auxiliary information, and condition information, and sends a switching instruction / configuration to the terminal device.

[0328] In the embodiment where the network side performs a final evaluation to determine the selected AI / ML unit, if the terminal device cannot be indicated by the AI / ML ID but is indicated by other implicit information (such as input information, output information, auxiliary information, conditional information, etc.), when there are multiple corresponding second AI / ML units, the terminal device selects the optimal second AI / ML unit under the corresponding indicator based on the performance indicator configured during reporting to activate or switch.

[0329] The above embodiments are merely exemplary of the present invention, but the present invention is not limited thereto. Appropriate modifications may be made based on the above embodiments. For example, the above embodiments may be used alone, or one or more of the above embodiments may be combined.

[0330] As can be seen from the above embodiments, the embodiments of the present application can support and clearly define the performance monitoring / evaluation process of non-activated AI / ML models / functions; in addition, the gains from using AI / ML become reliable.

[0331] Embodiments of the second aspect

[0332] The embodiment of the present application provides an information transmission method, which is described from the perspective of a network device. The embodiment of the second aspect can be combined with the embodiment of the first aspect, or implemented separately, and the same contents as the embodiment of the first aspect will not be repeated.

[0333] FIG11 is another schematic diagram of the information transmission method according to an embodiment of the present application. As shown in FIG11 , the method includes:

[0334] 1101: The network device sends first configuration information or second configuration information related to AI / ML unit performance evaluation or selection to the terminal device.

[0335] In some embodiments, the first configuration information includes one or more of AI / ML activation-related performance evaluation or selection-related performance information, activation trigger information, and AI / ML unit indication information.

[0336] In some embodiments, the second configuration information includes one or more of AI / ML switching-related performance evaluation or selection-related performance information, switching trigger information, and AI / ML unit indication information.

[0337] Regarding the specific actions in operation 1101, the specific content of the first configuration information, the second configuration information, and other behaviors of the network device or other information received / sent, please refer to the records in the embodiment of the first aspect and will not be repeated here.

[0338] The above embodiments are merely exemplary of the present invention, but the present invention is not limited thereto. Appropriate modifications may be made based on the above embodiments. For example, the above embodiments may be used alone, or one or more of the above embodiments may be combined.

[0339] As can be seen from the above embodiments, the embodiments of the present application can support and clearly define the performance monitoring / evaluation process of non-activated AI / ML models / functions; in addition, the gains from using AI / ML become reliable.

[0340] Embodiments of the third aspect

[0341] The embodiment of the present application provides an information transmission device. The device may be, for example, a terminal device, or one or more components or assemblies configured in the terminal device, which corresponds to the embodiment of the first aspect, and the same contents as the embodiment of the first aspect are not repeated here.

[0342] FIG12 is a schematic diagram of an information transmission device according to an embodiment of the present application. As shown in FIG12 , the information transmission device 1200 includes:

[0343] The first receiving unit 1201 is configured to receive first configuration information or second configuration information related to AI / ML unit performance evaluation or selection from a network device.

[0344] In some embodiments, the first configuration information includes one or more of AI / ML activation-related performance evaluation or selection-related performance information, activation trigger information, and AI / ML unit indication information.

[0345] In some embodiments, the second configuration information includes one or more of AI / ML switching-related performance evaluation or selection-related performance information, switching trigger information, and AI / ML unit indication information.

[0346] In some embodiments, the performance information includes one or more of the following performance indicator-related information used for evaluation of the AI / ML unit:

[0347] Performance thresholds for evaluating AI / ML activation on-device; or

[0348] Performance threshold range for evaluating AI / ML activation on the terminal side;

[0349] Reporting performance thresholds for evaluating AI / ML activation on the network side;

[0350] Performance threshold range for evaluating AI / ML activation on the network side.

[0351] In some embodiments, the performance information further includes one or more of the following performance preference indication information:

[0352] Instructions for selecting the best performing AI / ML unit;

[0353] Instruction information for selecting the best AI / ML unit whose performance is greater than or equal to the performance threshold or the reported performance threshold;

[0354] Selecting an AI / ML unit having an indication that performance meets the performance threshold range or the reported performance threshold range;

[0355] Select the indication information corresponding to the AI / ML unit with the lowest input reference signal overhead;

[0356] Selecting an AI / ML unit that meets one of the performance indicators and corresponding indication information with the lowest input reference signal overhead; and

[0357] An AI / ML unit that meets one of the performance indicators and corresponding indication information with the lowest output signal overhead are selected.

[0358] In some embodiments, the activation trigger information includes one or more of the following information:

[0359] Counter parameters that trigger activation requests;

[0360] Timer parameters that trigger activation requests;

[0361] Additional event parameters that trigger the activation request.

[0362] In some embodiments, the AI / ML unit indication information includes one or more of the following information:

[0363] AI / ML auxiliary information, which is used to instruct an AI / ML unit to evaluate the auxiliary information;

[0364] AI / ML input parameter or output parameter information, which is used to indicate the evaluation of the AI / ML unit with the input parameter or output parameter;

[0365] AI / ML condition information, which is used to indicate the AI / ML unit that is evaluated to meet the condition.

[0366] In some embodiments, the performance information includes one or more of the following performance indicator-related information used for evaluation of the AI / ML unit:

[0367] Performance threshold for evaluating AI / ML switching on the terminal side;

[0368] Performance threshold range for evaluating AI / ML switching on the terminal side;

[0369] Reporting performance thresholds used to evaluate AI / ML switching on the network side;

[0370] This parameter specifies the performance threshold range for evaluating AI / ML switching on the network side.

[0371] In some embodiments, the performance-related information further includes one or more of the following performance preference indication information:

[0372] The performance of the selected AI / ML unit is higher than that of the activated AI / ML unit, and the performance difference is higher than the weighted threshold.

[0373] Select the AI / ML unit and compare its performance with the activated AI / ML unit, and the performance difference falls within the threshold range of weight 9.

[0374] The AI / ML unit is selected for performance comparison with the activated AI / ML unit. The performance difference falls within the threshold range of weight 9, and the input reference signal overhead is lower than that of the activated AI / ML unit.

[0375] The AI / ML unit is selected for performance comparison with the activated AI / ML unit. The performance difference is less than the threshold value of weight 9, and the input reference signal overhead is lower than that of the activated AI / ML unit.

[0376] In some embodiments, the switching trigger information includes one or more of the following information:

[0377] Counter parameters that trigger AI / ML switching requests;

[0378] Timer parameters for triggering AI / ML switching request;

[0379] Other event parameters that trigger the AI / ML switching request.

[0380] In some embodiments, the AI / ML unit indication information includes one or more of the following information:

[0381] AI / ML auxiliary information, which is used to indicate the evaluation of the AI / ML unit with the auxiliary information compared with the activated AI / ML unit;

[0382] AI / ML input parameter or output parameter information, which is used to indicate that the AI / ML unit having the input parameter or output parameter is evaluated and compared with the activated AI / ML unit;

[0383] AI / ML condition information, which is used to indicate that the AI / ML units that are evaluated to meet the condition are compared with the activated AI / ML units.

[0384] In some embodiments, corresponding to the first configuration information and / or the second configuration information, the terminal device also receives corresponding reporting configuration information from the network device.

[0385] In some embodiments, the reporting configuration information includes instructing the terminal device to perform the reporting behavior after the performance evaluation or selection of the AI / ML unit, and includes one or more of the following reporting content-related configurations:

[0386] AI / ML switching or activation request related configuration;

[0387] Information and configuration related to the AI / ML switching or activation unit;

[0388] AI / ML switching candidate units information and configuration;

[0389] AI / ML unit input configuration information related to reporting configuration;

[0390] Report configuration related to AI / ML unit auxiliary configuration information;

[0391] Configuration related to reporting of auxiliary condition information of AI / ML units;

[0392] Reporting configuration related to AI / ML performance information.

[0393] In some embodiments, the reporting configuration information includes instructing the terminal device to perform the reporting behavior after the performance evaluation or selection of the AI / ML unit, and the configuration related to the reporting content includes one or more of the following:

[0394] selected AI / ML unit input configuration information and / or selected AI / ML unit performance information;

[0395] selected AI / ML unit auxiliary information and / or selected AI / ML unit performance information;

[0396] selected AI / ML unit condition information and / or selected AI / ML unit performance information;

[0397] The selected AI / ML unit identification and / or the performance information of the selected AI / ML unit.

[0398] In some embodiments, the selected AI / ML unit is one unit or multiple units, and / or

[0399] The selected AI / ML unit is inactive.

[0400] In some embodiments, the selection of the content is based on the first configuration information or the second configuration information, and / or,

[0401] The format of the content report is based on the reporting configuration information.

[0402] In some embodiments, the reporting configuration information includes instructing the terminal device to perform the performance evaluation or selection of the AI / ML unit and includes one or more of the following reporting methods and resource configuration information:

[0403] Periodic reporting, non-periodic reporting, and semi-continuous reporting;

[0404] Configure dedicated PRACH resources to trigger and report AI / ML switching or activation requests;

[0405] Reporting is based on dedicated PUCCH / MAC CE / RRC messages.

[0406] In some embodiments, the AI / ML unit is a first AI / ML unit or a second AI / ML unit, or includes a first AI / ML unit and a second AI / ML unit.

[0407] The first AI / ML unit includes one or more second AI / ML units.

[0408] In some embodiments, the apparatus 1200 further includes:

[0409] The first sending unit 1202 is configured to report the capability information of the terminal device to the network device.

[0410] The capability information of the terminal device includes AI / ML unit information, which includes one or more of the following information:

[0411] Whether the first AI / ML unit and / or the second AI / ML unit are supported according to the query of the network device;

[0412] whether there is only one first AI / ML unit and / or second AI / ML unit or multiple first AI / ML units;

[0413] The number of the first AI / ML unit and / or the second AI / ML unit being queried;

[0414] the number of second AI / ML units in the queried first AI / ML unit;

[0415] The model identifier of the AI / ML unit being queried; and

[0416] Auxiliary information of the queried AI / ML unit.

[0417] Condition information of the AI / ML unit being queried.

[0418] Input and output configuration information of the AI / ML unit being queried.

[0419] In some embodiments, the content and / or format of the report is predefined and / or configured by the network device.

[0420] The above embodiments are merely exemplary of the present invention, but the present invention is not limited thereto. Appropriate modifications may be made based on the above embodiments. For example, the above embodiments may be used alone, or one or more of the above embodiments may be combined.

[0421] As can be seen from the above embodiments, the embodiments of the present application can support and clearly define the performance monitoring / evaluation process of non-activated AI / ML models / functions; in addition, the gains from using AI / ML become reliable.

[0422] Embodiments of the fourth aspect

[0423] The embodiment of the present application provides an information transmission device. The device may be, for example, a network device, or one or more components or assemblies configured on the network device, which corresponds to the embodiment of the second aspect, and the same contents as the embodiment of the second aspect are not repeated here.

[0424] FIG13 is another schematic diagram of an information transmission device according to an embodiment of the present application. As shown in FIG13 , the information transmission device 1300 includes:

[0425] The second sending unit 1301 is configured to send first configuration information or second configuration information related to AI / ML unit performance evaluation or selection to the terminal device.

[0426] In some embodiments, the first configuration information includes one or more of AI / ML activation-related performance evaluation or selection-related performance information, activation trigger information, and AI / ML unit indication information.

[0427] In some embodiments, the second configuration information includes one or more of AI / ML switching-related performance evaluation or selection-related performance information, switching trigger information, and AI / ML unit indication information.

[0428] The above embodiments are merely exemplary of the present invention, but the present invention is not limited thereto. Appropriate modifications may be made based on the above embodiments. For example, the above embodiments may be used alone, or one or more of the above embodiments may be combined.

[0429] As can be seen from the above embodiments, the embodiments of the present application can support and clearly define the performance monitoring / evaluation process of non-activated AI / ML models / functions; in addition, the gains from using AI / ML become reliable.

[0430] Embodiments of the fifth aspect

[0431] An embodiment of the present application provides a terminal device, which includes the information transmission device as described in the embodiment of the third aspect.

[0432] Figure 14 is a schematic block diagram of the system architecture of a terminal device according to an embodiment of the present application. As shown in Figure 14 , terminal device 1400 may include a processor 1410 and a memory 1420; memory 1420 is coupled to processor 1410. It should be noted that this diagram is exemplary; other types of structures may be used to supplement or replace this structure to implement telecommunication or other functions.

[0433] In one embodiment, the functionality of the information transmission device may be integrated into the processor 1410 .

[0434] The processor 1410 is configured to: for a terminal device having one or more AI / ML features, receive, by the terminal device, first configuration information or second configuration information related to AI / ML unit performance evaluation or selection from a network device.

[0435] In another embodiment, the information transmission device can be configured separately from the processor 1410. For example, the transmission device of the physical random access channel can be configured as a chip connected to the processor 1410, and the function of the transmission device of the physical random access channel is realized through the control of the processor 1410.

[0436] As shown in FIG14 , terminal device 1400 may further include: a communication module 1430, an input unit 1440, a display 1450, and a power supply 1460. It is worth noting that terminal device 1400 does not necessarily include all the components shown in FIG14 ; in addition, terminal device 1400 may also include components not shown in FIG14 , and reference may be made to related art for details.

[0437] As shown in FIG. 14 , the processor 1410 is sometimes also referred to as a controller or an operation control, and may include a microprocessor or other processor device and / or logic device. The processor 1410 receives inputs and controls the operations of various components of the terminal device 1400 .

[0438] Memory 1420 may be, for example, one or more of a cache, flash memory, a hard drive, removable media, volatile memory, non-volatile memory, or other suitable devices. It may store various data and may also store programs for executing related information. Processor 1410 may execute the programs stored in memory 1420 to implement information storage or processing. The functions of other components are similar to those of existing devices and are not further described here. Each component of terminal device 1400 may be implemented using dedicated hardware, firmware, software, or a combination thereof without departing from the scope of the present invention.

[0439] The above embodiments are merely exemplary of the present invention, but the present invention is not limited thereto. Appropriate modifications may be made based on the above embodiments. For example, the above embodiments may be used alone, or one or more of the above embodiments may be combined.

[0440] As can be seen from the above embodiments, the embodiments of the present application can support and clearly define the performance monitoring / evaluation process of non-activated AI / ML models / functions; in addition, the gains from using AI / ML become reliable.

[0441] Embodiments of the sixth aspect

[0442] An embodiment of the present application provides a network device, which includes the information transmission device as described in the embodiment of the fourth aspect.

[0443] Figure 15 is a block diagram illustrating the system configuration of a network device according to an embodiment of the present application. As shown in Figure 15 , network device 1500 may include a processor 1510 and a memory 1520 ; the memory 1520 is coupled to the processor 1510 . The memory 1520 may store various data and may also store an information processing program 1530 , which is executed under the control of the processor 1510 .

[0444] In one embodiment, the functionality of the information transmission device may be integrated into the processor 1510 .

[0445] The processor 1510 may be configured to: the network device sends first configuration information or second configuration information related to the AI / ML unit performance evaluation or selection to the terminal device.

[0446] In another embodiment, the information transmission device can be configured separately from the processor 1510. For example, the information transmission device can be configured as a chip connected to the processor 1510, and the function of the transmission device of the physical random access channel is realized under the control of the processor 1510.

[0447] Furthermore, as shown in FIG15 , network device 1500 may further include a transceiver 1540 and an antenna 1550, among others. The functions of these components are similar to those in the prior art and are not further described here. It is worth noting that network device 1500 does not necessarily include all of the components shown in FIG15 . Furthermore, network device 1500 may also include components not shown in FIG15 , for which reference may be made to the prior art.

[0448] The above embodiments are merely exemplary of the present invention, but the present invention is not limited thereto. Appropriate modifications may be made based on the above embodiments. For example, the above embodiments may be used alone, or one or more of the above embodiments may be combined.

[0449] As can be seen from the above embodiments, the embodiments of the present application can support and clearly define the performance monitoring / evaluation process of non-activated AI / ML models / functions; in addition, the gains from using AI / ML become reliable.

[0450] Embodiments of the seventh aspect

[0451] An embodiment of the present application provides a communication system, including the terminal device as described in the embodiment of the eighth aspect and / or the network device as described in the embodiment of the ninth aspect.

[0452] For example, the structure of the communication system can be referred to FIG1 .

[0453] As shown in Figure 1, the communication system 100 includes a network device 101 and a terminal device 102. The network device 101 is the same as the network device recorded in the embodiment of the sixth aspect, and the terminal device 102 is the same as the terminal device recorded in the embodiment of the fifth aspect. The repeated content will not be repeated.

[0454] An embodiment of the present application also provides a computer program, wherein when the program is executed in a terminal device, the program causes the terminal device to execute the information transmission method described in the embodiment of the first aspect.

[0455] An embodiment of the present application also provides a storage medium storing a computer program, wherein the computer program enables a terminal device to execute the information transmission method described in the embodiment of the first aspect.

[0456] An embodiment of the present application also provides a computer program, wherein when the program is executed in a network device, the program causes the network device to execute the information transmission method described in the embodiment of the second aspect.

[0457] An embodiment of the present application also provides a storage medium storing a computer program, wherein the computer program enables a network device to execute the information transmission method described in the embodiment of the second aspect.

[0458] The above devices and methods of the present application can be implemented by hardware or by a combination of hardware and software. The present application relates to such a computer-readable program that, when executed by a logic component, enables the logic component to implement the devices or components described above, or enables the logic component to implement the various methods or steps described above. The present application also relates to a storage medium for storing the above program, such as a hard disk, a magnetic disk, an optical disk, a DVD, a flash memory, etc.

[0459] The method / device described in conjunction with the embodiments of the present application can be directly embodied as hardware, a software module executed by a processor, or a combination of the two. For example, one or more of the functional block diagrams shown in the figure and / or one or more combinations of functional block diagrams can correspond to various software modules of the computer program flow or to various hardware modules. These software modules can respectively correspond to the various steps shown in the figure. These hardware modules can be implemented by solidifying these software modules, for example, using a field programmable gate array (FPGA).

[0460] The software module may be located in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. A storage medium may be coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium; or the storage medium may be an integral part of the processor. The processor and the storage medium may be located in an ASIC. The software module may be stored in the memory of the mobile terminal or in a memory card that can be inserted into the mobile terminal. For example, if the device (such as a mobile terminal) uses a large-capacity MEGA-SIM card or a large-capacity flash memory device, the software module may be stored in the MEGA-SIM card or the large-capacity flash memory device.

[0461] One or more of the functional blocks and / or one or more combinations of functional blocks described in the accompanying drawings may be implemented as a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, or any appropriate combination thereof for performing the functions described in this application. One or more of the functional blocks and / or one or more combinations of functional blocks described in the accompanying drawings may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in communication with a DSP, or any other such configuration.

[0462] The present application has been described above in conjunction with specific embodiments. However, those skilled in the art should understand that these descriptions are merely illustrative and are not intended to limit the scope of protection of the present application. Those skilled in the art may make various modifications and variations to the present application based on the spirit and principles of the present application, and such modifications and variations are also within the scope of the present application.

[0463] Regarding the implementation methods including the above embodiments, the following additional notes are also disclosed:

[0464] 1. A method for transmitting information, the method comprising:

[0465] For a terminal device having one or more AI / ML features, the terminal device receives first configuration information or second configuration information related to AI / ML unit performance evaluation or selection from a network device, wherein:

[0466] For an AI / ML feature of the terminal device, when all AI / ML units included in the AI / ML feature are not activated, the first configuration information includes performance information related to determining AI / ML unit activation; when an AI / ML unit included in the AI / ML feature is activated, the second configuration information includes performance information related to determining AI / ML unit switching.

[0467] 2. The method according to Note 1, wherein, corresponding to the first configuration information and / or the second configuration information, the terminal device also receives corresponding reported configuration information from the network device, and the reported configuration information contains partial information of the first configuration information and / or the second configuration information.

[0468] 3. A method for transmitting information, the method comprising:

[0469] For a terminal device having one or more AI / ML features, the terminal device receives third configuration information related to AI / ML unit performance evaluation or selection from a network device.

[0470] 4. The method according to Note 3, wherein the third configuration information includes one or more of performance evaluation related to AI / ML unit activation / switching or performance information related to selection, activation / switching trigger information, and AI / ML unit indication information.

[0471] 5. The method according to Supplement 4, wherein the performance information includes one or more of the following performance indicator-related information used for evaluation of the AI / ML unit:

[0472] Performance thresholds for evaluating AI / ML activation / switching on the terminal side;

[0473] Performance threshold range for evaluating AI / ML activation / switching on the terminal side;

[0474] Reporting performance thresholds used to evaluate AI / ML activation / switching on the network side;

[0475] This parameter specifies the performance threshold range for evaluating AI / ML activation / switching on the network side.

[0476] 6. The method according to any one of Notes 1 to 5, wherein:

[0477] The AI / ML unit corresponds to an AI / ML feature and / or an AI / ML function and / or an AI / ML logical model and / or an AI / ML physical model.

[0478] 7. The method according to Supplementary Note 6, wherein:

[0479] The first AI / ML feature includes one or more second AI / ML features, or the first AI / ML feature includes one or more first AI / ML features; and / or

[0480] The first AI / ML function includes one or more second AI / ML functions, or the first AI / ML function includes one or more first models; and / or

[0481] The first AI / ML model includes one or more second AI / ML models.

[0482] 8. A method for transmitting information, the method comprising:

[0483] The network device sends first configuration information, second configuration information, or third configuration information related to the AI / ML unit performance evaluation or selection to the terminal device.

[0484] 9. The method according to Supplementary Note 8, wherein:

[0485] The first configuration information includes one or more of AI / ML activation-related performance evaluation or selection-related performance information, activation trigger information, and AI / ML unit indication information.

[0486] The second configuration information includes one or more of performance evaluation related to AI / ML switching or performance information related to selection, switching trigger information, and AI / ML unit indication information.

[0487] The third configuration information includes one or more of performance evaluation related to AI / ML unit activation / switching or performance information related to selection, activation / switching trigger information, and AI / ML unit indication information.

[0488] 10. A communication system comprising a terminal device and a network device,

[0489] The terminal device includes a memory and a processor, the memory storing a computer program, and the processor being configured to execute the computer program to implement the information transmission method as described in any one of Notes 1 to 7;

[0490] The network device includes a memory and a processor, the memory stores a computer program, and the processor is configured to execute the computer program to implement the information transmission method as described in any one of Notes 8-9.

Claims

1. An information transmission device, the device being arranged in a terminal device having one or more AI / ML features, the device comprising: A first receiving unit is configured to receive first configuration information or second configuration information related to AI / ML unit performance evaluation or selection from a network device.

2. The device according to claim 1, wherein: The first configuration information includes one or more of AI / ML activation-related performance evaluation or selection-related performance information, activation trigger information, and AI / ML unit indication information.

3. The device according to claim 1, wherein: The second configuration information includes one or more of performance evaluation related to AI / ML switching or performance information related to selection, switching trigger information, and AI / ML unit indication information.

4. The device according to claim 2, wherein: The performance information includes one or more of the following performance indicator related information for evaluation of the AI / ML unit: Performance thresholds for evaluating AI / ML activation on the terminal side; Performance threshold range for evaluating AI / ML activation on the terminal side; Reporting performance thresholds for evaluating AI / ML activation on the network side; This parameter specifies the performance threshold range for evaluating AI / ML activation on the network side.

5. The device according to claim 4, wherein: The performance information also includes one or more of the following performance preference indication information: Instructions for selecting the best performing AI / ML unit; an indication of selecting the best AI / ML unit having a performance greater than or equal to the performance threshold or the reported performance threshold; Selecting an AI / ML unit having an indication that the performance meets the performance threshold range or the reported performance threshold range; Select the indication information corresponding to the AI / ML unit with the lowest input reference signal overhead; Select the indication information corresponding to the AI / ML unit with the lowest output signal overhead; Selecting an AI / ML unit that meets one of the performance indicators and corresponding indication information with the lowest input reference signal overhead; as well as An AI / ML unit that meets one of the performance indicators and corresponding indication information with the lowest output signal overhead are selected.

6. The device according to claim 2, wherein: The activation trigger information includes one or more of the following information: Counter parameters that trigger the activation request; Timer parameters that trigger the activation request; Additional event parameters that trigger the activation request.

7. The device according to claim 2, wherein: The AI / ML unit indication information includes one or more of the following information: AI / ML auxiliary information, which is used to instruct the evaluation of the AI / ML unit having the auxiliary information; AI / ML input parameter or output parameter information, which is used to indicate the evaluation of the AI / ML unit having the input parameter or output parameter; The AI / ML condition information is used to indicate an AI / ML unit that evaluates the condition information.

8. The device according to claim 3, wherein: The performance information includes one or more of the following performance indicator related information for evaluation of the AI / ML unit: Performance threshold for evaluating AI / ML switching on the terminal side; Performance threshold range for evaluating AI / ML switching on the terminal side; Reporting performance thresholds for evaluating AI / ML switching on the network side; This parameter specifies the reporting performance threshold range for evaluating AI / ML switching on the network side.

9. The device according to claim 8, wherein: The performance related information also includes one or more of the following performance preference indication information: The performance of the selected AI / ML unit is higher than that of the activated AI / ML unit, and the performance difference is higher than the performance threshold or the reported performance threshold, Select an AI / ML unit to compare performance with the activated AI / ML unit, and the performance difference falls within the performance threshold range or the reported performance threshold range, The selected AI / ML unit is compared with the activated AI / ML unit for performance, the performance difference falls within the performance threshold range or the reported performance threshold range, and the input reference signal or output information overhead is lower than that of the activated AI / ML unit, The selected AI / ML unit is compared with the activated AI / ML unit for performance, the performance difference is less than the performance threshold or the reported performance threshold, and the input reference signal or output information overhead is lower than the activated AI / ML unit.

10. The device according to claim 3, wherein: The switching trigger information includes one or more of the following information: Counter parameters that trigger AI / ML switching requests; Timer parameters for triggering AI / ML switching request; Other event parameters that trigger the AI / ML switching request.

11. The device according to claim 3, wherein: The AI / ML unit indication information includes one or more of the following information: AI / ML auxiliary information, which is used to indicate that the AI / ML unit having the auxiliary information is evaluated and compared with the activated AI / ML unit; AI / ML input parameter or output parameter information, which is used to indicate that the AI / ML unit having the input parameter or output parameter is evaluated and compared with the activated AI / ML unit; The AI / ML condition information is used to indicate that the AI / ML unit that evaluates the condition information is compared with the activated AI / ML unit.

12. The device according to claim 1, wherein: Corresponding to the first configuration information and / or the second configuration information, the terminal device also receives corresponding reporting configuration information from the network device.

13. The device according to claim 12, wherein: The reporting configuration information includes instructing the terminal device to perform the reporting behavior after the performance evaluation or selection of the AI / ML unit, and includes one or more of the following reporting content-related configurations: AI / ML switching or activation request related configuration; AI / ML switching unit information and configuration; AI / ML switching candidate units information and configuration; AI / ML unit input configuration information related to reporting configuration; AI / ML unit output configuration information related to reporting configuration; Configuration of reporting of auxiliary information of AI / ML units; AI / ML unit condition information related reporting configuration; Reporting configuration related to AI / ML performance information.

14. The device according to claim 13, wherein: The reporting configuration information includes instructing the terminal device to perform the reporting behavior after the performance evaluation or selection of the AI / ML unit, and the configuration related to the reporting content includes one or more of the following contents: selected AI / ML unit input configuration information and / or selected AI / ML unit performance information; The selected AI / ML unit outputs configuration information and / or performance information of the selected AI / ML unit; selected AI / ML unit auxiliary information and / or selected AI / ML unit performance information; selected AI / ML unit condition information and / or selected AI / ML unit performance information; The selected AI / ML unit identification and / or the performance information of the selected AI / ML unit.

15. The device according to claim 14, wherein: The AI / ML unit selected is one unit or multiple units, and / or The selected AI / ML unit is inactive.

16. The device according to claim 14, wherein: The selection of the content is based on the first configuration information or the second configuration information, and / or, The format of the content report is based on the reporting configuration information.

17. The device according to claim 12, wherein: The reporting configuration information includes instructing the terminal device to perform the performance evaluation or selection on the AI / ML unit and also includes one or more of the following reporting methods and resource configuration information: Periodic reporting, non-periodic reporting, and semi-continuous reporting; Configure dedicated PRACH resources to trigger reporting of AI / ML switching or activation requests; Reporting is based on dedicated PUCCH / MAC CE / RRC messages.

18. The device according to claim 1, wherein: The AI / ML unit is the first AI / ML unit or the second AI / ML unit, or includes the first AI / ML unit and the second AI / ML unit, The first AI / ML unit includes one or more second AI / ML units.

19. The device according to claim 1, wherein: The device also includes: a first sending unit, configured to report capability information of the terminal device to the network device, The capability information of the terminal device includes AI / ML unit information, and the AI / ML unit information includes one or more of the following information: whether the first AI / ML unit and / or the second AI / ML unit are supported according to the query of the network device; Whether there is only one or more first AI / ML units and / or second AI / ML units; the ID of the first AI / ML unit and / or the second AI / ML unit being queried; the number of second AI / ML units in the queried first AI / ML unit; The model identity of the AI / ML unit being queried; and Auxiliary information of the AI / ML unit being queried; Condition information of the AI / ML unit being queried; Input and output configuration information of the AI / ML unit being queried.

20. The device according to claim 19, wherein The content and / or format of the report is predefined and / or configured by the network device.

Images