Reactive and proactive additional condition reporting by user equipment

Abstract

An apparatus is configured to process a list of supported artificial intelligence (AI) features or AI models based on a signal received from a network, determine one or more AI features or AI models from the list that are applicable for a current operation, generate a message indicating the one or more AI features or AI models that are applicable for the current operation for transmission to the network, and start a timer when the message has been transmitted to the network.

Description

Background Technology

[0001] A user equipment (UE) can establish connections with at least one of several different networks or network types. The UE may support artificial intelligence (AI) / machine learning (ML) algorithms for certain use cases, which can be customized and adapted to specific scenarios, locations, configurations, deployments, etc. AI / ML algorithms can be updated, for example, through model changes. The UE should notify the radio access network (RAN) of the applicability conditions of the AI / ML algorithms that the UE can use to support RAN control (e.g., activation, deactivation, handover, etc.). This report may be referred to as an additional conditions report performed by the UE. However, currently, there is no process for the UE to perform such an additional conditions report. Summary of the Invention

[0002] Some example implementations involve an apparatus having processing circuitry configured to: process a list of supported artificial intelligence (AI) features or AI models based on signals received from a network; determine one or more AI features or AI models from the list that are suitable for the current operation; generate a message indicating that one or more AI features or AI models are suitable for the current operation for transmission to the network; and start a timer when the message has been transmitted to the network.

[0003] Other example implementations relate to an apparatus having processing circuitry configured to: determine one or more artificial intelligence (AI) features or AI models applicable to the current operation; generate a message indicating the one or more AI features or AI models applicable to the current operation for transmission to a network; and start a timer when the message has been transmitted to the network.

[0004] Other example implementations involve an apparatus having processing circuitry configured to: process a message based on a signal received from a network, the message including a list of supported artificial intelligence (AI) features or AI models and the network, as well as one or more conditions; determine that one or more conditions are met; and generate a message indicating one or more AI features or AI models on the list and applicable to the current operation for transmission to the network. Attached Figure Description

[0005] Figure 1 Example network layouts based on various example implementation schemes are shown.

[0006] Figure 2 Example user equipment (UE) is shown according to various example implementation schemes.

[0007] Figure 3Example base stations based on various example implementation schemes are shown.

[0008] Figure 4 Example call flows for reactive condition reporting are shown according to various example implementations.

[0009] Figure 5 Example call flows for proactive condition reporting are shown according to various example implementation schemes.

[0010] Figure 6 An example call flow for mixed additional condition reporting is shown according to various example implementation schemes. Detailed Implementation

[0011] The example implementation can be further understood by referring to the following description and related figures, in which the same elements have the same reference numerals. The example implementation involves an additional condition report from the UE to the network, which relates to the applicability conditions of the UE for use in supporting artificial intelligence (AI) / machine learning (ML) algorithms for RAN control.

[0012] Example embodiments are described with reference to user equipment (UE). However, references to the UE are provided for illustrative purposes only. The example embodiments may be used with any electronic components capable of establishing a connection to a network and configured with hardware, software, and / or firmware for exchanging information and data with the network. Therefore, the UE described herein is used to represent any suitable electronic component.

[0013] Example implementations are also described with reference to 5G New Radio (NR) networks. However, the reference to 5G NR networks is provided for illustrative purposes only. The example implementations can be used with any network capable of establishing a connection with the UE and exchanging information and data with the UE (e.g., 5G-Advanced networks, 6G networks, etc.).

[0014] Throughout the manual, the terms AI and ML can be used interchangeably to refer to any type of learning model, such as classifiers, regression models, etc.

[0015] The example implementation describes various reporting mechanisms applicable to conditions related to AI / ML algorithms. The example implementation covers two main categories of reporting: reactive reporting and proactive reporting. Reactive reporting refers to the UE providing information to the network when it receives an action from the network. Proactive reporting refers to the UE providing information to the network when it does not receive an action from the network.

[0016] Example implementations related to reactive condition reporting use Radio Resource Control (RRC) signaling (e.g., RRCReconfigurationComplete) to report details of the report content. In these example implementations, the UE may have a timer for implementing reactive condition reporting and related UE behaviors.

[0017] Example implementations related to proactive condition reporting use UEAssistanceInformation (UAI) to report details of the report content. In these example implementations, the UE may also have a timer for implementing proactive condition reporting and related UE behaviors.

[0018] Example implementations also include solutions that are a hybrid of reactive and proactive conditional reporting. In these example implementations, the network configures a list of AI / ML features / models via RRC signaling and associated conditions. When the conditions are met, the UE reports a set of updated AI / ML features / models via a UAI message.

[0019] Each of these example implementation schemes will be described in more detail below.

[0020] Figure 1 An example network arrangement 100 according to various example implementations is shown. The example network arrangement 100 includes a UE 110. The UE 110 can be any type of electronic component configured to communicate via a network, such as a mobile phone, tablet computer, desktop computer, smartphone, phablet, embedded device, wearable device, Internet of Things (IoT) device, etc. A real network arrangement can include any number of UEs used by any number of users. Therefore, the example of a single UE 110 is provided for illustrative purposes only.

[0021] UE 110 can be configured to communicate with one or more networks. In the example of network deployment 100, the network with which UE 110 can wirelessly communicate is the 5G NR Radio Access Network (RAN) 120. However, UE 110 can also communicate with other types of networks (e.g., 5G cloud RAN, Next Generation RAN (NG-RAN), LTE RAN, legacy cellular networks, WLAN, etc.), and UE 110 can also communicate with the network via a wired connection. Referring to the example implementation, UE 110 can establish a connection with 5G NR RAN 120. Therefore, UE 110 may have a 5G NR chipset to communicate with NR RAN 120.

[0022] The 5G NR RAN 120 can be part of a public terrestrial mobile network (PLMN) that can be deployed by network operators (e.g., Verizon, AT&T, T-Mobile, etc.). The 5G NR RAN 120 may include, for example, cells or base stations (Node B, eNodeB, HeNB, eNB, gNB, gNodeB, macro cell, micro cell, small cell, femtocell, etc.) configured to transmit and receive services from UEs equipped with appropriate cellular chipsets.

[0023] In network deployment 100, 5G NR RAN 120 includes base stations (e.g., gNB 120A) that can be deployed in a terrestrial network (TN) or a non-terrestrial network (NTN). For example, satellite-based systems can be integrated with 5G NR RAN 120 to provide network access to UE 110 in an NTN deployment, and in some cases, the base stations can be located on non-terrestrial components (e.g., satellites). References will follow below. Figure 4 A more detailed description of the example NTN network architecture is provided.

[0024] return Figure 1 The network deployment 100 and gNB 120A may include one or more communication interfaces to exchange data and / or information with UE 110, the corresponding 5G NR RAN 120, cellular core network 130, Internet 140, etc.

[0025] UE 110 can connect to 5G NR-RAN 120 via gNB 120A. Any association procedure can be performed to connect UE 110 to 5G NR-RAN 120. For example, as discussed above, 5G NR-RAN 120 can be associated with a specific cell provider where UE 110 and / or its user have protocol and credential information (e.g., stored on a SIM card). Upon detecting the presence of 5G NR-RAN 120, UE 110 can send the corresponding credential information to associate with 5G NR-RAN 120. More specifically, UE 110 can be associated with a specific cell (e.g., gNB 120A). However, as mentioned above, the reference to 5G NR-RAN 120 is for illustrative purposes only, and any suitable type of RAN can be used.

[0026] In addition to the 5G NR RAN 120, network deployment 100 also includes a cellular core network 130, an Internet 140, an IP Multimedia Subsystem (IMS) 150, and a network service backbone 160. The cellular core network 130 can be viewed as a set of interconnected components that manage the operation and traffic of the cellular network. The cellular core network 130 also manages the traffic flowing between the cellular network and the Internet 140.

[0027] IMS 150 can generally be described as an architecture for delivering multimedia services to UE 110 using the IP protocol. IMS 150 can communicate with cellular core network 130 and the Internet 140 to provide multimedia services to UE 110. Network service backbone 160 communicates directly or indirectly with the Internet 140 and cellular core network 130. Network service backbone 160 can generally be described as a collection of components (e.g., servers, network storage deployments, etc.) that implement a set of services that can be used to extend the functionality of UE 110 in communicating with various networks.

[0028] Figure 2 Example UE 110 according to various example implementations is shown. (Refer to...) Figure 1 The network layout 100 is used to describe UE 110. UE 110 may include a processor 205, a memory layout 210, a display device 215, an input / output (I / O) device 220, a transceiver 225, and other components 230. Other components 230 may include, for example, audio input devices, audio output devices, power supplies, data acquisition devices, ports for electrically connecting UE 110 to other electronic devices, etc.

[0029] Processor 205 may be configured to execute multiple engines of UE 110. For example, an engine may include an additional condition reporting engine 235. The additional condition reporting engine 235 may perform various operations related to reporting from the UE to the network, relating to the applicability conditions of artificial intelligence (AI) / machine learning (ML) algorithms that the UE can use to support RAN control. Specifically, the additional condition reporting engine 235 may perform operations such as, but not limited to, providing AI / ML features and models to the network, determining applicable AI / ML features and models, and reporting applicable AI / ML features and models to the network. These and other operations are described in more detail below.

[0030] The engine 235 referenced above is provided as an application (e.g., a program) executed by processor 205 for illustrative purposes only. The functionality associated with engine 235 may also be represented as a separate, integrated component of UE 110, or as a modular component coupled to UE 110, such as an integrated circuit with or without firmware. For example, the integrated circuit may include input circuitry for receiving signals and processing circuitry for processing signals and other information. The engine may also be embodied as one application or multiple independent applications. Furthermore, in some UEs, the functionality described for processor 205 is split between two or more processors, such as a baseband processor and an application processor. Example implementations may be implemented according to any of these or other configurations of the UE.

[0031] Memory arrangement 210 may be a hardware component configured to store data related to operations performed by UE 110. Display device 215 may be a hardware component configured to display data to a user, while I / O device 220 may be a hardware component enabling a user to input data. Display device 215 and I / O device 220 may be separate components or may be integrated together (such as a touchscreen).

[0032] Transceiver 225 may be a hardware component configured to establish connections with 5G NR-RAN 120, LTE-RAN (not shown), legacy RAN (not shown), WLAN (not shown), etc. Accordingly, transceiver 225 may operate on a variety of different frequencies or channels (e.g., a continuous set of frequencies). Transceiver 225 includes circuitry configured to transmit and / or receive signals (e.g., control signals, data signals). Such signals may be encoded using information used to implement any of the methods described herein. Processor 205 may be operatively coupled to transceiver 225 and configured to receive signals from and / or transmit signals to transceiver 225. Processor 205 may be configured to encode and / or decode signals (e.g., signaling from a base station in the network) for use in implementing any of the methods described herein.

[0033] Figure 3 An example base station 300 is shown according to various example implementation schemes. Base station 300 may represent a gNB 120A or any other type of access node that UE 110 can use to establish connections and manage network operations.

[0034] Base station 300 may include processor 305, memory arrangement 310, input / output (I / O) devices 315, transceiver 320, and other components 325. Other components 325 may include, for example, audio input devices, audio output devices, batteries, data acquisition devices, ports for electrically connecting base station 300 to other electronic devices and / or power sources, TxRU, transceiver chains, antenna elements, antenna panels, etc.

[0035] Processor 305 may be configured to execute multiple engines of base station 300. For example, an engine may include an additional conditions engine 330. Additional conditions engine 330 may perform various operations related to receiving additional conditions from the UE, which are associated with AI / ML models applicable to the current operation of the UE. To provide some general examples, additional conditions engine 330 may perform operations such as, but not limited to, receiving supported AI / ML features and models from the UE, generating a list of publicly available AI / ML features and models supported by the network and the UE, and receiving AI / ML features and models applicable to the current operation of the UE. These and other operations are described in more detail below.

[0036] The engine 330 described above is merely an example of an application (e.g., a program) executed by the processor 305. The functionality associated with the engine 330 can also be represented as a separate integrated component of the base station 300, or as a modular component coupled to the base station 300, such as an integrated circuit with or without firmware. For example, the integrated circuit may include input circuitry for receiving signals and processing circuitry for processing signals and other information. Furthermore, in some base stations, the functionality described for the processor 305 is split among multiple processors (e.g., a baseband processor, an application processor, etc.). Example implementations can be implemented according to any of these or other configurations of the base station.

[0037] The memory arrangement 310 may be a hardware component configured to store data related to operations performed by the base station 300. The I / O device 315 may be a hardware component or port that enables a user to interact with the base station 300.

[0038] Transceiver 320 may be a hardware component configured to exchange data with UE 110 and any other UE in network arrangement 100. Transceiver 320 may operate on a variety of different frequencies or channels (e.g., a continuous set of frequencies). Therefore, transceiver 320 may include one or more components to enable data exchange with various networks and UEs. Transceiver 320 includes circuitry configured to transmit and / or receive signals (e.g., control signals, data signals). Such signals may be encoded using information used to implement any of the methods described herein. Processor 305 may be operatively coupled to transceiver 320 and configured to receive signals from and / or transmit signals to transceiver 320. Processor 305 may be configured to encode and / or decode signals (e.g., signaling from a UE) for use in implementing any of the methods described herein.

[0039] Figure 4An example call flow 400 for reactive condition reporting is shown according to various example implementations. Call flow 400 can be executed between UE 110 and base station 300 (e.g., gNB 120A).

[0040] In 410, UE 110 uses UE capability signaling (e.g., UECapabilityEnquiry / UECapabilityInformation messages) to report all supported AI / ML features / feature groups (FG).

[0041] In 420, the network (via base station 300) configures a list of AI features or AI models, for example, via an RRCReconfiguration message. This list may be a subset of the features / models reported by UE 110 in its capability report in 410. The list may be a common set of AI features or AI models supported by both UE 110 and the network. The list received in 420 can be considered an action by the network that triggers UE 110 to perform reactive conditional reporting. Therefore, the following operations can be considered part of reactive conditional reporting.

[0042] In section 430, UE 110 determines whether each listed AI feature or model is applicable to the current environment / scenario that UE 110 is experiencing. For example, although the UE supports all reported features / models, some of these features / models may not be applicable to the UE's current environment (e.g., indoor or outdoor) or the UE's internal state (e.g., memory, compute load, battery life, etc.).

[0043] In 440, UE 110 reports the applicability results in an RRC message (e.g., an RRCReconfigurationComplete message). In some example implementations, the report in 440 may include a bitmap containing a 1-bit indication of whether the AI ​​feature / model configured in 420 is applicable. In other example implementations, the report in 440 may include an index of inapplicable AI features / models configured in 420. In still other example implementations, the report in 440 may include an index of applicable AI features / models configured in 420.

[0044] UE 110 may also include a list of preferred AI features / models. In some example implementations, these preferred AI features / models may come from a list configured in 420. This may indicate that UE 110 has a preference for configuring these models due to memory limitations. In other example implementations, these preferred AI features / models may not be in the list configured in 420. This may indicate that UE 110 has available memory to use other AI features / models outside the configured list.

[0045] In 450, after sending the RRCReconfigurationComplete message, UE 110 can start a timer. The timer helps UE 110 apply AI features / models because it allows the network time to successfully receive the RRCReconfigurationComplete message, time to apply the AI ​​features / models reported by UE 110, or time for the network to respond to UE 110 with an updated / new list of AI features / models in a separate RRCReconfiguration message, as described below. The duration of the timer can be set to a value that achieves the above-described purposes.

[0046] If an RRCReconfiguration message with a new list of AI features / models is received while the timer is running (e.g., ... Figure 4 (As shown by the dashed line 460 in the image), then UE 110 can stop the timer and apply the updated list of AI features / models in a new RRC message.

[0047] However, if the timer expires in 470 if no RRC message is received during the timer duration (e.g., no RRCReconfiguration message 460 is received during the timer duration), UE 110 can apply various AI features / models. In some example implementations, UE 110 can apply confirmed AI features / models configured in 420. Confirmed AI features / models can be considered those configured in 420 and confirmed in 440 as being applicable by UE 110 in the RRCReconfigurationComplete message.

[0048] In other example implementations, UE 110 may apply the confirmed AI features / models configured in 420 and the preferred AI features / models reported in 440, such as additional preferred AI features / models in addition to those configured in 420 reported in 440.

[0049] In another example implementation, UE 110 may apply the confirmed AI features / models configured in 420 in conjunction with the preferred AI features / models reported in 440. That is, the applied AI features / models are those AI features / models that have been confirmed and indicated as preferred in 440.

[0050] As described above, the call flow is used for reactive conditional reporting. Below are some examples of proactive conditional reporting.

[0051] Figure 5 An example call flow 500 for proactive additional condition reporting is shown according to various example implementations. Call flow 500 can be executed between UE 110 and base station 300 (e.g., gNB 120A).

[0052] In 510, UE 110 uses UE capability signaling (e.g., UECapabilityEnquiry / UECapabilityInformation messages) to report all supported AI / ML features / feature groups (FG).

[0053] In step 520, UE 110 determines an initial set of preferred AI features / models. This set can be determined based on the current environment UE 110 is experiencing (e.g., outdoors or indoors) and the internal state of UE 110 (e.g., memory state, compute load, battery life, etc.). In this example, it can be seen that UE 110 is performing this operation without receiving any action from the network (e.g., the additional condition report is initiated by UE 110).

[0054] In 530, UE 510 reports the determined initial set of AI features / models in the UAI message. The report content may be the identifier (ID) of the applicable AI features / models. The ID may be, for example, an AI / ML model ID, a functional ID, an index of an FG, etc.

[0055] In 540, after successfully sending the UAI message, UE 110 starts a timer. The timer in this example implementation can be used for a different purpose than the timer in previous example implementations. For example, the timer can allow UE 110 to avoid transmitting UAI too frequently; for instance, while the timer is running, UE 110 will not transmit new UAI messages. Specifically, because this is an active reporting not in response to network triggering, UE 110 and the network may want to avoid UE 110 constantly updating AI / ML features / models. Therefore, the timer can be set to prevent UAI messages with updated AI / ML features / models from being transmitted too frequently.

[0056] In step 550, UE 110 can determine the set of updates to AI / ML features / models. For example, the scenario or environment in which UE 110 is operating may have changed since the UAI message was transmitted in step 530. Similarly, the set of updates to AI / ML features / models can be determined based on UE 110's current environment (e.g., outdoors or indoors) or UE 110's internal state (e.g., memory, compute load, battery life, etc.). In some example implementations, this update is based on the specific UE implementation, for example, where no specific triggering condition exists. In other example implementations, the triggering condition may be defined in a standard (e.g., a 3GPP Technical Specification (TS)).

[0057] In 560, the timer expires. When the timer expires, if UE 110 has determined the set of updated AI / ML features / models, then in 570, UE 110 reports the set of updated AI features / models to the network in a separate UAI message.

[0058] The following provides some examples of hybrid approaches between reactive and proactive condition reporting.

[0059] Figure 6 An example call flow 600 for hybrid condition reporting is shown according to various example implementations. Call flow 600 can be executed between UE 110 and base station 300 (e.g., gNB 120A).

[0060] In 610, UE 110 uses UE capability signaling (e.g., UECapabilityEnquiry / UECapabilityInformation messages) to report all supported AI / ML features / feature groups (FG).

[0061] In 620, the network (via base station 300) configures a list of AI features or AI models, for example, via an RRCReconfiguration message. This list may be a subset of the features / models reported by UE 110 in its capability report in 610. The list may be a common set of AI features or AI models supported by both UE 110 and the network. The list in 620 can be similar to... Figure 4 The list provided in the reactive reporting operation shown in 420.

[0062] However, in addition to the list, in this example implementation, the RRCReconfiguration message may also include one or more associated conditions. When UE 110 receives a condition, UE 110 can monitor the condition and perform one or more actions when the condition is met. These actions will be described in more detail below. However, before describing the example actions, various examples of how UE 110 can monitor the conditions to be met will be described below.

[0063] In the first example, the condition is met when the maximum value of all monitoring metrics for the configured AI feature / model is greater than the threshold. In the second example, the condition is met when the minimum value of all monitoring metrics for the configured AI feature / model is less than the threshold. In the third example, the condition is met when the maximum value of monitoring metrics for a subset of the configured AI feature / model is greater than the threshold. In the fourth example, the condition is met when the minimum value of a subset of the configured AI feature / model is less than the threshold. In the fifth example, the condition is met when the maximum value of all monitoring metrics for the configured AI feature / model is greater than the threshold from which these metrics were last reported. In the sixth example, the condition is met when the minimum value of all monitoring metrics for the configured AI feature / model is less than the threshold from which these metrics were last reported. In the seventh example, the condition is met when the maximum value of monitoring metrics for a subset of the configured AI feature / model is greater than the threshold from which these metrics were last reported. In the eighth example, the condition is met when the minimum value of a subset of the configured AI feature / model is less than the threshold from which these metrics were last reported.

[0064] In step 630, when UE 110 determines that the conditions are met, UE 110 reports a set of updated AI features / models in a UAI message in step 640. The reported AI features / models can be a subset of the AI ​​features / models in the list configured in step 620. In the hybrid approach, timers are not used because the network has already set conditions for when to trigger the additional condition report.

[0065] Example In a first embodiment, a method includes: processing a list of supported artificial intelligence (AI) features or AI models based on signals received from a network; determining one or more AI features or AI models from the list that are suitable for current operation; generating a message indicating that the one or more AI features or AI models are suitable for the current operation for transmission to the network; and starting a timer when the message has been transmitted to the network.

[0066] In a second embodiment, according to the method of the first embodiment, the method further includes: determining a list of updates to AI features or AI models that the network has transmitted while the timer is running; stopping the timer; and applying the list of updates to AI features or AI models.

[0067] In a third embodiment, according to the method described in the first embodiment, the method further includes: determining that the timer has expired; and applying the one or more AI features or AI models applicable to the current operation.

[0068] In the fourth embodiment, according to the method of the first embodiment, the list is received in an RRCReconfiguration message.

[0069] In a fifth embodiment, according to the method of the first embodiment, the processing circuitry determines one or more AI features or AI models suitable for the current operation from the list based on the operating environment or internal state.

[0070] In the sixth embodiment, according to the method of the first embodiment 1, the message includes a bitmap corresponding to each AI feature or AI model in the list, wherein a first value of the bitmap indicates that the corresponding AI feature or AI model is suitable for the current operation.

[0071] In the seventh embodiment, according to the method of the first embodiment, the message includes an index of AI features and AI models in the list that are not applicable to the current operation.

[0072] In the eighth embodiment, according to the method of the first embodiment, the message includes an index of AI features and AI models applicable to the current operation from the list.

[0073] In the ninth embodiment, according to the method of the first embodiment, the message further includes an indication that one or more AI features and AI models in the list are preferred.

[0074] In the tenth embodiment, according to the method of the ninth embodiment, the one or more AI features or AI models applied are preferred AI features and AI models from the list.

[0075] In the eleventh embodiment, according to the method of the first embodiment, the message further includes an indication that one or more AI features and AI models not in the list are preferred.

[0076] In the twelfth embodiment, according to the method of the eleventh embodiment, the method further includes applying one or more AI features and AI models that are not in the list and are preferred.

[0077] In a thirteenth embodiment, a processor is configured to perform any one of the methods described according to the first to twelfth embodiments.

[0078] In a fourteenth embodiment, a user equipment (UE) includes: a transceiver configured to communicate with a network; and a processor communicatively coupled to the transceiver and configured to perform any one of the methods described in the first to twelfth embodiments.

[0079] In a fifteenth embodiment, a method includes: determining one or more artificial intelligence (AI) features or AI models applicable to a current operation; generating a message indicating the one or more AI features or AI models applicable to the current operation for transmission to a network; and starting a timer when the message has been transmitted to the network.

[0080] In a sixteenth embodiment, according to the method of the fifteenth embodiment, the method further includes: determining one or more updated AI features or AI models applicable to the updated current operation while the timer is running; and generating a message indicating the one or more updated AI features or AI models applicable to the updated current operation when the timer expires, for transmission to the network.

[0081] In the seventeenth embodiment, according to the method of the fifteenth embodiment, the processing circuitry determines one or more AI features or AI models suitable for the current operation based on the operating environment or internal state.

[0082] In the eighteenth embodiment, according to the method of the fifteenth embodiment, the message includes a UEAssistanceInformation (UAI) message.

[0083] In the nineteenth embodiment, according to the method of the fifteenth embodiment, the message includes an identifier (ID) for each AI feature or AI model among the AI ​​features or AI models applicable to the current operation.

[0084] In the twentieth embodiment, according to the method of the nineteenth embodiment, the ID includes an AI feature ID, an AI model ID, a functional ID, or an index of a feature group.

[0085] In the twenty-first embodiment, a processor is configured to perform any one of the methods described according to the fifteenth to the twenty-first embodiments.

[0086] In the twenty-second embodiment, a user equipment (UE) includes: a transceiver configured to communicate with a network; and a processor communicatively coupled to the transceiver and configured to perform any one of the methods described according to the fifteenth to twenty-first embodiments.

[0087] In a twenty-third embodiment, a method includes: processing a message based on a signal received from a network, the message including a list of supported artificial intelligence (AI) features or AI models and the network, as well as one or more conditions; determining that the one or more conditions are met; and generating a message indicating one or more AI features or AI models on the list and applicable to current operation for transmission to the network.

[0088] In the twenty-fourth embodiment, according to the method of the twenty-third embodiment, wherein satisfying one or more conditions includes one of the following: (i) the maximum value of all monitoring metrics of the AI ​​feature or AI model on the list is greater than a first threshold; (ii) the minimum value of all monitoring metrics of the AI ​​feature or AI model on the list is less than a second threshold; (iii) the maximum value of monitoring metrics of a subset of AI features or AI models on the list is greater than a third threshold; (iv) the minimum value of monitoring metrics of a subset of AI features or AI models on the list is less than a fourth threshold; (v) the maximum value of all monitoring metrics of the AI ​​feature or AI model on the list is greater than a fifth threshold from the last reported monitoring metric; (vi) the minimum value of all monitoring metrics of the AI ​​feature or AI model on the list is less than a sixth threshold from the last reported monitoring metric; (vii) the maximum value of monitoring metrics of a subset of AI features or AI models on the list is greater than a seventh threshold from the last reported monitoring metric; or (viii) the minimum value of a subset of AI features or AI models on the list is less than an eighth threshold from the last reported monitoring metric.

[0089] In the twenty-fifth embodiment, according to the method of the twenty-third embodiment, the list and the one or more conditions are received in an RRCReconfiguration message.

[0090] In the twenty-sixth embodiment, according to the method of the twenty-third embodiment, the message indicating one or more AI features or AI models on the list and applicable to the current operation includes a UEAssistanceInformation (UAI) message.

[0091] In the twenty-seventh embodiment, a processor is configured to perform any one of the methods described according to the twenty-third to twenty-sixth embodiments.

[0092] In the twenty-eighth embodiment, a user equipment (UE) includes: a transceiver configured to communicate with a network; and a processor communicatively coupled to the transceiver and configured to perform any one of the methods described according to the twenty-third to twenty-sixth embodiments.

[0093] Those skilled in the art will understand that the example embodiments described above can be implemented with any suitable software or hardware configuration or combination thereof. Example hardware platforms for implementing the example embodiments may include, for example, Intel x86-based platforms with compatible operating systems, Windows OS, Mac platforms and MAC OS, and mobile devices with operating systems such as iOS, Android, etc. Example embodiments of the methods described above may be embodied as programs containing lines of code stored on a non-transitory computer-readable storage medium, which, at compile time, can be executed on a processor or microprocessor.

[0094] Although this application describes various embodiments that have different features in various combinations, those skilled in the art will understand that any feature of one embodiment can be combined with features of other embodiments in any way that is not expressly excluded or that is not functionally or logically inconsistent with the operation of the device or the specified function of the disclosed embodiment.

[0095] As is widely recognized, the use of personally identifiable information should comply with privacy policies and practices that are generally accepted to meet or exceed industry or governmental requirements for protecting user privacy. Specifically, personally identifiable information data should be managed and processed to minimize the risk of unintentional or unauthorized access or use, and the nature of authorized use should be clearly explained to users.

[0096] It will be apparent to those skilled in the art that various modifications can be made to this disclosure without departing from its spirit or scope. Therefore, this disclosure is intended to cover modifications and variations thereof, provided they fall within the scope of the appended claims and their equivalents.

Claims

1. An apparatus comprising a processing circuit configured to: A list of supported artificial intelligence (AI) features or AI models that are processed based on signals received from the network; From the list, identify one or more AI features or AI models that are suitable for the current operation; Generate a message indicating the applicability of the one or more AI features or AI models to the current operation for transmission to the network; and A timer is started when the message has been sent to the network.

2. The apparatus of claim 1, wherein the processing circuit is further configured to: While the timer is running, determine the list of updated AI features or AI models that the network has transmitted; Stop the timer; and The updated list applies AI features or AI models.

3. The apparatus of claim 1, wherein the processing circuit is further configured to: It is determined that the timer has expired; and The application is one or more AI features or AI models applicable to the current operation.

4. The apparatus of claim 1, wherein the processing circuitry determines, from the list, one or more AI features or AI models suitable for the current operation based on the operating environment or internal state.

5. The apparatus of claim 1, wherein the message includes a bitmap corresponding to each of the AI ​​features or AI models in the list, wherein a first value of the bitmap indicates that the corresponding AI feature or AI model is suitable for the current operation.

6. The apparatus of claim 1, wherein the message includes an index of AI features and AI models in the list that are not applicable to the current operation.

7. The apparatus of claim 1, wherein the message includes an index of AI features and AI models applicable to the current operation from the list.

8. The apparatus of claim 1, wherein the message further includes an indication that one or more of the AI ​​features and AI models in the list are preferred.

9. The apparatus of claim 8, wherein the one or more AI features or AI models applied are preferred from the list of AI features and AI models.

10. The apparatus of claim 1, wherein the message further includes an indication that one or more AI features and AI models not in the list are preferred.

11. The apparatus of claim 10, wherein the processing circuit is further configured to: The application is one or more of the AI ​​features and AI models that are not in the list and are preferred.

12. An apparatus comprising a processing circuit configured to: Identify one or more artificial intelligence (AI) features or AI models applicable to the current operation; Generate a message indicating the one or more AI features or AI models applicable to the current operation for transmission over the network; and A timer is started when the message has been sent to the network.

13. The apparatus of claim 12, wherein the processing circuit is further configured to: While the timer is running, determine one or more updated AI features or AI models applicable to the updated current operation; and When the timer expires, a message is generated indicating one or more AI features or AI models of the update that are applicable to the current operation of the update, for transmission to the network.

14. The apparatus of claim 12, wherein the processing circuitry determines one or more AI features or AI models suitable for the current operation based on the operating environment or internal state.

15. The apparatus of claim 12, wherein the message includes an identifier (ID) for each of the AI ​​features or AI models applicable to the current operation.

16. The apparatus of claim 15, wherein the ID includes an AI feature ID, an AI model ID, a functional ID, or an index of a feature group.

17. An apparatus comprising a processing circuit configured to: The message is processed based on signals received from the network, and the message includes a list of supported artificial intelligence (AI) features or AI models and the network, as well as one or more conditions. Determine that one or more of the conditions are met; and Generate a message indicating one or more AI features or AI models on the list and applicable to the current operation for sending to the network.

18. The apparatus of claim 17, wherein satisfying one or more of the conditions includes one of the following: (i) the maximum value of all monitoring metrics of the AI ​​feature or AI model on the list is greater than a first threshold; (ii) the minimum value of all monitoring metrics of the AI ​​feature or AI model on the list is less than a second threshold; (iii) the maximum value of monitoring metrics of a subset of AI features or AI models on the list is greater than a third threshold; (iv) the minimum value of monitoring metrics of a subset of AI features or AI models on the list is less than a fourth threshold; (v) the maximum value of all monitoring metrics of the AI ​​feature or AI model on the list is greater than a fifth threshold from the last reported monitoring metric; (vi) the minimum value of all monitoring metrics of the AI ​​feature or AI model on the list is less than a sixth threshold from the last reported monitoring metric; (vii) the maximum value of monitoring metrics of a subset of AI features or AI models on the list is greater than a seventh threshold from the last reported monitoring metric; or (viii) the minimum value of a subset of AI features or AI models on the list is less than an eighth threshold from the last reported monitoring metric.

19. The apparatus of claim 17, wherein the list and the one or more conditions are received in an RRCReconfiguration message.

20. The apparatus of claim 17, wherein the message indicating one or more AI features or AI models on the list and applicable to the current operation includes a UEAssistanceInformation (UAI) message.

Images