Conditional artificial intelligence, machine learning model and parameter set configuration

Abstract

Methods, systems, and devices for wireless communication are described. A user equipment (UE) can receive, from a network entity, a message indicating a set of machine learning models, a set of parameter sets, or both, and one or more usage conditions associated with the machine learning models and the parameter sets. Based on usage conditions being satisfied, the UE can select a machine learning model, a parameter set, or both for generating a machine learning inference. For example, the UE can select the machine learning model or the parameter set based on priority, whether sufficient input data is provided, or based on other usage conditions. The UE can generate the machine learning inference using the selected machine learning model or the selected parameter set, and the UE can send a report to the network entity indicating an output of the machine learning inference.

Description

[0001] Cross-references

[0002] This patent application claims priority to U.S. patent application No. 17 / 841,339, entitled “CONDITIONAL ARTIFICIAL INTELLIGENCE, MACHINE LEARNING MODEL, AND PARAMETER SET CONFIGURATIONS,” filed by Kumar et al. on June 15, 2022, which is assigned to the assignee of this application and is expressly incorporated herein by reference. Technical Field

[0003] The following article relates to wireless communications, including conditional artificial intelligence (AI), machine learning models, and parameter set configuration. Background Art

[0004] Wireless communication systems are widely deployed to provide various types of communication content, such as voice, video, packet data, messaging, broadcasting, and the like. These systems may be able to support communication with multiple users by sharing available system resources (e.g., time, frequency, and power). Examples of such multiple access systems include fourth generation (4G) systems (such as long term evolution (LTE) systems, advanced LTE (LTE-A) systems, or LTE-A Pro systems) and fifth generation (5G) systems (which may be referred to as new radio (NR) systems). These systems may employ techniques such as code division multiple access (CDMA), time division multiple access (TDMA), frequency division multiple access (FDMA), orthogonal FDMA (OFDMA), or discrete Fourier transform spread orthogonal frequency division multiplexing (DFT-S-OFDM). A wireless multiple access communication system may include one or more base stations, each of which supports wireless communications for communication devices, which may be referred to as user equipment (UE).

[0005] In some wireless communication systems, user equipment (UE) can use machine learning models to make predictions based on learned data. However, in some cases, the input data to the machine learning model may be lost or corrupted, which may lead to errors in the predictions. Summary of the invention

[0006] The described technology relates to improved methods, systems, devices, and apparatuses that support conditional artificial intelligence (AI), machine learning models, and parameter set configurations. For example, the described technology provides conditional machine learning models and parameter set configurations, where for AI and machine learning inference and training, a network entity can configure a model identifier (ID) and / or parameter set together with a plurality of machine learning models, a plurality of parameter sets, or conditions for use of both. In some examples, lost input data (e.g., from a network entity) can be one of the conditions for use. In addition, the network entity can configure other conditions for use. Examples of machine learning model and parameter set use conditions for inference and training may include the availability of model input parameters, regional scope, validity timer of the model, validity timer of the parameter set, configuration of the network entity and user equipment (UE), key performance indicators (KPIs) of the model, physical environment (e.g., indoor, outdoor, and other physical settings), or any combination thereof. Using machine learning model configurations, parameter set configurations, and use conditions in this manner can manage lost or corrupted input data of the machine learning model, and thus reduce errors in the output of the machine learning model.

[0007] In some examples, one or more machine learning models and corresponding parameter sets can be trained so that a given machine learning model, a given parameter set, or both can be unaffected by lost or corrupted input data. One or more machine learning models, one or more parameter sets, or both can generate machine learning inferences using input data from the UE and the network entity, or input data from the UE alone. Machine learning inference can be the process of running real-time data points through the trained machine learning model and parameter set to calculate the output. Information about one or more machine learning models, one or more parameter sets, or both can be provided to the network entity, including how to train the machine learning model and the corresponding parameter set, the priority of each machine learning model, each parameter set, or both, and other information.

[0008] A network entity may send a message to a UE indicating one or more machine learning models, one or more parameter sets, or both, and one or more usage conditions that the UE may use to select one of the machine learning models, one of the parameter sets, or both to generate machine learning inferences based on what input data is available, missing, or corrupted. The UE may select a machine learning model, a parameter set, or both based on whether a particular usage condition is satisfied, the priority of the machine learning model and the parameter set, and other factors. In some examples, the UE may switch the machine learning model, the parameter set, or both that the UE uses to generate machine learning inferences and send an indication of the model and / or parameter set switch to the network entity. In some examples, the UE may generate machine learning inferences using the selected machine learning model, the selected parameter set, or both, and the UE may send a report to the network entity indicating the output of the inference, the machine learning model, the parameter set, or both that was selected to generate interference, or both.

[0009] A method for wireless communication at a UE is described. The method may include receiving a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the set of multiple machine learning models, one of the set of multiple parameter sets, or both for performing machine learning inference; selecting a first machine learning model in the set of multiple machine learning models, a first parameter set in the set of multiple parameter sets, or both for performing machine learning inference based on a usage condition in the one or more usage conditions being satisfied; and sending a report indicating machine learning inference generated by the first machine learning model or by using the first parameter set.

[0010] A device for wireless communication at a UE is described. The device may include a processor, a memory coupled to the processor, and instructions stored in the memory. The instructions may be executed by the processor to cause the device to receive a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the sets of multiple machine learning models, one of the sets of multiple parameter sets, or both for performing machine learning inference; selecting a first machine learning model in the set of multiple machine learning models, a first parameter set in the set of multiple parameter sets, or both for performing machine learning inference based on satisfaction of a usage condition in one or more usage conditions; and sending a report indicating machine learning inference generated by the first machine learning model or by using the first parameter set.

[0011] Another apparatus for wireless communication at a UE is described. The apparatus may include a component for receiving a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the set of multiple machine learning models, one of the set of multiple parameter sets, or both for performing machine learning inference; a component for selecting a first machine learning model in the set of multiple machine learning models, a first parameter set in the set of multiple parameter sets, or both for performing machine learning inference based on a usage condition in the one or more usage conditions being satisfied; and a component for sending a report indicating machine learning inferences generated by the first machine learning model or by using the first parameter set.

[0012] A non-transitory computer-readable medium storing code for wireless communication at a UE is described. The code may include instructions executable by a processor to perform the following operations: receiving a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the sets of multiple machine learning models, one of the sets of multiple parameter sets, or both for performing machine learning inference; selecting a first machine learning model in the set of multiple machine learning models, a first parameter set in the set of multiple parameter sets, or both for performing machine learning inference based on a usage condition in the one or more usage conditions being satisfied; and sending a report indicating machine learning inference generated by the first machine learning model or by using the first parameter set.

[0013] Some examples of the methods, apparatus, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for: receiving a first message indicating a set of multiple priorities corresponding to a set of multiple machine learning models, a set of multiple parameter sets, or both; and selecting a first machine learning model, a first parameter set, or both for performing machine learning inference based on the first machine learning model, the first parameter set, or both being associated with a highest priority in the set of multiple priorities.

[0014] Some examples of the methods, apparatus, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for receiving a first message instructing the network node to change the priority of one or more of a set of multiple machine learning models, one or more of a set of multiple parameter sets, or both based on one or more performance indicators reported by the UE, wherein the first machine learning model, the first parameter set, or both may be selected based on the first message.

[0015] Some examples of the methods, apparatus, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for: selecting a first machine learning model, a first parameter set, or both for performing machine learning inferences based on a usage condition indicating that the first machine learning model, a first parameter set, or both are selected when the UE may be able to successfully receive input data from the network node; and generating machine learning inferences using the input data received from the network node by the first machine learning model or by using the first parameter set.

[0016] Some examples of the methods, apparatus, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for: selecting a first machine learning model, a first parameter set, or both for performing machine learning inference based on a usage condition indicating that the first machine learning model, the first parameter set, or both are selected when the UE fails to receive input data for a second machine learning model in a set of multiple machine learning models, a second parameter set in a set of multiple parameter sets, or both from a network node; and sending a report indicating the machine learning inference and indicating the first machine learning model, the first parameter set, or both.

[0017] Some examples of the methods, apparatus, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for generating machine learning inferences using input data measured by the UE by the first machine learning model or by using the first parameter set based on a usage condition indicating that the first machine learning model, the first parameter set, or both are selected when the UE measures the input data.

[0018] Some examples of the methods, apparatus, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for sending a first feedback message indicating one or more statistics associated with switching between selecting a set of multiple machine learning models, a set of multiple parameter sets, or both.

[0019] Some examples of the methods, apparatus, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for receiving a second feedback message indicating one or more statistics associated with switching between selecting a set of multiple machine learning models, a set of multiple parameter sets, or both.

[0020] Some examples of the methods, devices, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for the following operations: receiving a first message or a second message indicating one or more usage conditions for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both, wherein the one or more usage conditions include at least one of the following: instructions for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both when input data may be lost, or priorities corresponding to each machine learning model in a set of multiple machine learning models, each parameter set in a set of multiple parameter sets, or both, for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both.

[0021] In some examples of the methods, apparatus, and non-transitory computer-readable media described herein, instructions for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both include availability of input data, usage scenarios, regional scope, model validity timer, parameter set validity timer, configuration of the UE, configuration of the network node, performance threshold, or any combination thereof.

[0022] In some examples of the methods, apparatus, and non-transitory computer-readable media described herein, receiving a first message may include operations, features, components, or instructions for the following operations: receiving a first message or a second message indicating configuration information of a set of multiple machine learning models, a set of multiple parameter sets, or both, wherein the configuration information indicates a set of alternative machine learning models in the set of multiple machine learning models, a set of alternative parameter sets in the set of multiple parameter sets, one or more usage conditions, a priority corresponding to each machine learning model, a priority corresponding to each parameter set, or any combination thereof.

[0023] A method for wireless communication at a network node is described. The method may include sending a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the set of multiple machine learning models, one of the set of multiple parameter sets, or both for performing machine learning inferences; and receiving a report indicating machine learning inferences generated by a first machine learning model or by using a first parameter set, the first machine learning model, the first parameter set being selected by a UE based on a usage condition in the one or more usage conditions being satisfied.

[0024] An apparatus for wireless communication at a network node is described. The apparatus may include a processor, a memory coupled to the processor, and instructions stored in the memory. The instructions are executable by the processor to cause the apparatus to send a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the sets of multiple machine learning models, one of the sets of multiple parameter sets, or both for performing machine learning inferences; and receive a report indicating machine learning inferences generated by a first machine learning model or by using a first parameter set, the first machine learning model, the first parameter set being selected by a UE based on a usage condition in the one or more usage conditions being satisfied.

[0025] Another apparatus for wireless communication at a network node is described. The apparatus may include means for sending a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the set of multiple machine learning models, one of the set of multiple parameter sets, or both for performing machine learning inferences; and means for receiving a report indicating machine learning inferences generated by a first machine learning model or by using a first parameter set, the first machine learning model, the first parameter set being selected by a UE based on a usage condition in the one or more usage conditions being satisfied.

[0026] A non-transitory computer-readable medium storing code for wireless communication at a network node is described. The code may include instructions executable by a processor to: send a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the sets of multiple machine learning models, one of the sets of multiple parameter sets, or both for performing machine learning inferences; and receive a report indicating machine learning inferences generated by a first machine learning model or by using a first parameter set, the first machine learning model, the first parameter set being selected by a UE based on a usage condition in the one or more usage conditions being satisfied.

[0027] Some examples of the methods, apparatus, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for sending a first message indicating a set of multiple priorities corresponding to a set of multiple machine learning models, a set of multiple parameter sets, or both.

[0028] Some examples of the methods, apparatus, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for sending a first message that instructs the network node to change the priority of one or more of a set of multiple machine learning models, one or more of a set of multiple parameter sets, or both based on one or more performance indicators reported by the UE, wherein the first machine learning model, the first parameter set, or both may be selected based on the first message.

[0029] Some examples of the methods, apparatus, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for receiving a report indicating a machine learning inference and instructing the UE to select a first machine learning model, a first set of parameters, or both to generate the machine learning inference.

[0030] Some examples of the methods, apparatus, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for sending a feedback message indicating one or more statistics associated with switching between a set of multiple machine learning models, a set of multiple parameter sets, or both.

[0031] Some examples of the methods, devices, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for the following operations: sending a first message or a second message indicating one or more usage conditions for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both, wherein the one or more usage conditions include at least one of the following: instructions for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both when input data may be lost, or a priority corresponding to each machine learning model in a set of multiple machine learning models, each parameter set in a set of multiple parameter sets, for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both.

[0032] In some examples of the methods, apparatus, and non-transitory computer-readable media described herein, instructions for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both include availability of input data, usage scenarios, regional scope, model validity timer, parameter set validity timer, configuration of the UE, configuration of the network node, performance threshold, or any combination thereof.

[0033] In some examples of the methods, apparatus, and non-transitory computer-readable media described herein, sending a first message may include operations, features, components, or instructions for the following operations: sending a first message or a second message indicating configuration information of a set of multiple machine learning models, a set of multiple parameter sets, or both, wherein the configuration information indicates a set of alternative machine learning models in the set of multiple machine learning models, a set of alternative parameter sets in the set of multiple parameter sets, one or more usage conditions, a priority corresponding to each machine learning model, a priority corresponding to each parameter set, or any combination thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 An example of a wireless communication system supporting conditional artificial intelligence (AI), machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is illustrated.

[0035] Figure 2 An example of a wireless communication system supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is illustrated.

[0036] Figure 3 Examples of machine learning models and parameter set configurations that support conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure are illustrated.

[0037] Figure 4 and Figure 5 An example of a process flow supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is illustrated.

[0038] Figure 6 and Figure 7 A block diagram of a device supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown.

[0039] Figure 8 A block diagram of a communication manager supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown.

[0040] Fig. 9 A diagram of a system including devices supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown.

[0041] Fig.10 and Fig.11 A block diagram of a device supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown.

[0042] Fig.12 A block diagram of a communication manager supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown.

[0043] Fig.13 A diagram of a system including devices supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown.

[0044] Figures 14 to 19 A flowchart illustrating a method for supporting conditional AI, machine learning models, and parameter set configuration according to one or more aspects of the present disclosure is shown. DETAILED DESCRIPTION

[0045] For example, when operating in a cellular network, some artificial intelligence (AI) and machine learning models may use a defined set of input and output parameters. In some examples, network entities and user equipment (UE) may use machine learning models in wireless communication systems (e.g., fifth generation (5G) systems) to make decisions based on learned data, which can improve the efficiency, latency, and reliability of wireless applications. Offline data may be used to train some machine learning models, in which case the machine learning models may use different data processing methods to manage missing input values ​​or damaged input values ​​of the model (e.g., deleting missing input values, manually adding missing input values). However, when a UE or network entity (e.g., a network node) uses a machine learning model in a cellular network scenario, the unavailability of input parameters may result in incorrect predictions and high error rates in the output of the machine learning model. In a wireless communication system, if the corresponding input data is corrupted, if the network entity, UE, or both are overloaded and cannot report the input data to each other in a timely manner, or a combination thereof, the input value may not be available for the machine learning model. Therefore, the machine learning model may lack input parameters for any given scenario, which may result in errors in the model output and degradation of overall system performance.

[0046] The techniques described herein may support conditional machine learning model and parameter set configurations that may be used to manage missing or corrupted input data to machine learning models and parameter sets to reduce errors in the output of the machine learning models. The parameter sets may correspond to weights for a machine learning model, or neural network weights for an AI or machine learning model. In some examples, a single machine learning model may be associated with multiple machine learning parameter sets, and usage conditions may be used to select one of the parameter sets based on what data is available, missing, corrupted, or any combination thereof. In some examples, the selected parameter set may be used when running a particular machine learning model to generate machine learning perturbations.

[0047] In some examples, one or more machine learning models and corresponding parameter sets may be trained so that a given machine learning model may not be affected or may be less affected by lost or corrupted input data. One or more machine learning models, one or more parameter sets, or both may generate machine learning inferences using input data from the UE and the network entity, or input data from the UE only. Machine learning inferences may be the process of running real-time data points through the trained machine learning model and parameter set to calculate the output. Information about one or more machine learning models, one or more parameter sets, or both may be provided to the network entity, including how the machine learning model was trained to obtain the associated parameter set, the priority of each machine learning model and / or parameter set, and other information.

[0048] The network entity may configure a model identifier (ID) and / or a parameter set with usage conditions for multiple machine learning models, multiple parameter sets, or both. The network entity may send a message to the UE indicating one or more machine learning models, one or more parameter sets, or both, and one or more usage conditions that the UE may use to select one of the machine learning models, one of the parameter sets, or both to perform machine learning inference based on what input data is available, lost, or corrupted. That is, lost input data (e.g., from the network entity) may be one of the usage conditions. In addition, the network entity may configure other usage conditions. Examples of machine learning model and / or parameter set usage conditions for inference and training may include the availability of model input parameters, regional scope, validity timer of the model, validity timer of the parameter set, configuration of the network entity and the UE, key performance indicators (KPIs) of the model and associated parameter sets, physical environment (e.g., indoor, outdoor, and other physical settings), or any combination thereof. The UE may select a machine learning model and parameter set or both based on the satisfaction of a particular usage condition, the priority of the machine learning model, the priority of the parameter set, and other factors. In some examples, the UE may switch the machine learning model, parameter set, or both that the UE uses to perform machine learning inference and send an indication of the model and / or parameter set switch to the network entity. In some examples, the UE may generate machine learning inference using the selected machine learning model or the selected parameter set, and the UE may send a report to the network entity indicating the output of the inference.

[0049] Various aspects of the subject matter described herein may be implemented to achieve one or more of the following potential improvements, among other things. The techniques employed by the described network devices (e.g., UE, network entity) may enable the UE to select a machine learning model, a parameter set, or both for performing machine learning inference based on usage conditions being satisfied, which may enable more accurate machine learning model outputs. For example, the UE may select a model and parameter set that are trained to be agnostic to specific lost or corrupted input data, which may reduce incorrect predictions and errors in the output of the machine learning model and corresponding parameter set because it may be able to manage lost or corrupted input data. In addition, the described techniques may enable the UE to perform model and / or parameter set switching and report feedback about the model and / or parameter switching to the network entity, which may improve the accuracy and efficiency of the machine learning model and improve communication between the UE and the network entity, as well as other benefits.

[0050] Aspects of the disclosure are first described in the context of wireless communication systems. Aspects of the disclosure are then described in the context of machine learning models and parameter set configurations and process flows. Aspects of the disclosure are further illustrated by and described with reference to apparatus diagrams, system diagrams, and flow diagrams related to conditional AI, machine learning models, and parameter set configurations.

[0051] Figure 1 An example of a wireless communication system 100 supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is illustrated. The wireless communication system 100 may include one or more network entities 105, one or more UEs 115, and a core network 130. In some examples, the wireless communication system 100 may be a long term evolution (LTE) network, an advanced LTE (LTE-A) network, an LTE-A Pro network, a new radio (NR) network, or a network operating according to other systems and radio technologies, including future systems and radio technologies not explicitly mentioned herein.

[0052] The network entities 105 may be dispersed throughout a geographic area to form the wireless communication system 100, and may include devices in different forms or with different capabilities. In various examples, the network entities 105 may be referred to as network elements, mobility elements, radio access network (RAN) nodes, or network equipment, among other nomenclature. In some examples, the network entities 105 and the UE 115 may communicate wirelessly via one or more communication links 125 (e.g., radio frequency (RF) access links). For example, the network entity 105 may support a coverage area 110 (e.g., a geographic coverage area) within which the UE 115 and the network entity 105 may establish one or more communication links 125. The coverage area 110 may be an example of a geographic area within which the network entity 105 and the UE 115 may support signal communications according to one or more radio access technologies (RATs).

[0053] The UEs 115 may be dispersed throughout the coverage area 110 of the wireless communication system 100, and each UE 115 may be stationary or mobile, or stationary and mobile at different times. The UEs 115 may be devices in different forms or with different capabilities. Figure 1 Some example UEs 115 are illustrated in FIG. 1 . The UEs 115 described herein may be capable of supporting communications with various types of devices, such as Figure 1 Other UEs 115 or network entities 105 are shown.

[0054] As described herein, a node of the wireless communication system 100 (which may be referred to as a network node or a wireless node) may be a network entity 105 (e.g., any network entity described herein), a UE 115 (e.g., any UE described herein), a network controller, an apparatus, a device, a computing system, one or more components, or another suitable processing entity configured to perform any of the techniques described herein. For example, the node may be a UE 115. As another example, the node may be a network entity 105. As another example, a first node may be configured to communicate with a second node or a third node. In one aspect of this example, the first node may be a UE 115, the second node may be a network entity 105, and the third node may be a UE 115. In another aspect of this example, the first node may be a UE 115, the second node may be a network entity 105, and the third node may be a network entity 105. In other aspects of this example, the first node, the second node, and the third node may be different relative to these examples. Similarly, references to UE 115, network entity 105, apparatus, device, computing system, etc. may include disclosure of UE 115, network entity 105, apparatus, device, computing system, etc. as nodes. For example, a disclosure that UE 115 is configured to receive information from network entity 105 also discloses that the first node is configured to receive information from a second node.

[0055] In some examples, the network entities 105 may communicate with the core network 130, or communicate with each other, or both. For example, the network entities 105 may communicate with the core network 130 via one or more backhaul communication links 120 (e.g., according to S1, N2, N3 or other interface protocols). In some examples, the network entities 105 may communicate with each other via the backhaul communication link 120 (e.g., according to X2, Xn or other interface protocols) directly (e.g., directly between the network entities 105) or indirectly (e.g., via the core network 130). In some examples, the network entities 105 may communicate with each other via the midhaul communication link 162 (e.g., according to the midhaul interface protocol) or the fronthaul communication link 168 (e.g., according to the fronthaul interface protocol) or any combination thereof. The backhaul communication link 120, the midhaul communication link 162, or the fronthaul communication link 168 may be or include one or more wired links (e.g., electrical links, optical fiber links), one or more wireless links (e.g., radio links, wireless optical links), etc. or various combinations thereof. UE 115 may communicate with core network 130 via communication link 155 .

[0056] One or more of the network entities 105 described herein may include or may be referred to as a base station 140 (e.g., a transceiver base station, a radio base station, an NR base station, an access point, a radio transceiver, a Node B, an evolved Node B (eNB), a next-generation Node B, or a Gigabit Node B (any of which may be referred to as a gNB), a 5G NB, a next-generation eNB (ng-eNB), a Home Node B, a Home Evolved Node B, or other suitable terms). In some examples, the network entity 105 (e.g., a base station 140) may be implemented in an aggregated (e.g., monolithic, stand-alone) base station architecture, which may be configured to utilize a protocol stack physically or logically integrated within a single network entity 105 (e.g., a single RAN node, such as a base station 140).

[0057] In some examples, the network entity 105 may be implemented in a disaggregated architecture (e.g., a disaggregated base station architecture, a disaggregated RAN architecture) that may be configured to utilize a protocol stack that is physically or logically distributed between two or more network entities 105, such as an integrated access backhaul (IAB) network, an open RAN (O-RAN) (e.g., a network configuration sponsored by the O-RAN Alliance), or a virtualized RAN (vRAN) (e.g., a cloud RAN (C-RAN)). For example, the network entity 105 may include one or more of the following: a central unit (CU) 160, a distributed unit (DU) 165, a radio unit (RU) 170, a RAN intelligent controller (RIC) 175 (e.g., a near real-time RIC (near RT RIC), a non-real-time RIC (non-RT RIC)), a service management and orchestration (SMO) 180 system, or any combination thereof. The RU 170 may also be referred to as a radio head, a smart radio head, a remote radio head (RRH), a remote radio unit (RRU), or a transmit receive point (TRP). One or more components of the network entity 105 in the decomposed RAN architecture may be co-located, or one or more components of the network entity 105 may be located in distributed locations (e.g., separate physical locations). In some examples, one or more network entities 105 of the decomposed RAN architecture may be implemented as virtual units (e.g., virtual CU (VCU), virtual DU (VDU), virtual RU (VRU)).

[0058] The functional split between CU 160, DU 165, and RU 170 is flexible and may support different functionalities, depending on which functions (e.g., network layer functions, protocol layer functions, baseband functions, RF functions, and any combination thereof) are performed at CU 160, DU 165, or RU 170. For example, a functional split of a protocol stack may be employed between CU 160 and DU 165 such that CU 160 may support one or more layers of a protocol stack and DU 165 may support one or more different layers of a protocol stack. In some examples, CU 160 may host higher protocol layer (e.g., layer 3 (L3), layer 2 (L2)) functionality and signaling (e.g., radio resource control (RRC), service data adaptation protocol (SDAP), packet data convergence protocol (PDCP)). The CU 160 may be connected to one or more DUs 165 or RUs 170, and the one or more DUs 165 or RUs 170 may host lower protocol layers, such as layer 1 (L1) (e.g., physical (PHY) layer) or L2 (e.g., radio link control (RLC) layer, medium access control (MAC) layer) functionality and signaling, and may each be at least partially controlled by the CU 160. Additionally or alternatively, a functional split of the protocol stack may be employed between the DU 165 and the RU 170, such that the DU 165 may support one or more layers of the protocol stack, and the RU 170 may support one or more different layers of the protocol stack. The DU 165 may support one or more different cells (e.g., via one or more RUs 170). In some cases, the functional split between CU 160 and DU 165 or between DU 165 and RU 170 can be within the protocol layer (e.g., some functions of the protocol layer can be performed by one of CU 160, DU 165, or RU 170, while other functions of the protocol layer are performed by different ones of CU 160, DU 165, or RU 170). CU 160 can be further functionally split into CU control plane (CU-CP) and CU user plane (CU-UP) functions. CU 160 can be connected to one or more DUs 165 via midhaul communication links 162 (e.g., F1, F1-c, F1-u), and DU 165 can be connected to one or more RUs 170 via fronthaul communication links 168 (e.g., open fronthaul (FH) interface). In some examples, midhaul communication link 162 or fronthaul communication link 168 may be implemented based on an interface (e.g., a channel) between layers of a protocol stack supported by respective network entities 105 that communicate via such communication links.

[0059] In some wireless communication systems (e.g., wireless communication system 100), infrastructure and spectrum resources for radio access may support wireless backhaul link capabilities to supplement wired backhaul connections, thereby providing an IAB network architecture (e.g., to core network 130). In some cases, in an IAB network, one or more network entities 105 (e.g., IAB nodes 104) may be partially controlled by each other. One or more IAB nodes 104 may be referred to as donor entities or IAB donors. One or more DUs 165 or one or more RUs 170 may be controlled in part by one or more CUs 160 associated with a donor network entity 105 (e.g., donor base station 140). One or more donor network entities 105 (e.g., IAB donors) may communicate with one or more additional network entities 105 (e.g., IAB nodes 104) via supported access and backhaul links (e.g., backhaul communication links 120). The IAB node 104 may include an IAB mobile terminal (IAB-MT) controlled (e.g., scheduled) by a coupled IAB donor's DU 165. The IAB-MT may include an independent set of antennas for relaying communications with the UE 115, or may share the same antennas of the IAB node 104 (e.g., of the RU 170) for access via the DU 165 of the IAB node 104 (e.g., referred to as a virtual IAB-MT (vIAB-MT)). In some examples, the IAB node 104 may include a DU 165 that supports communication links with additional entities (e.g., IAB node 104, UE 115) within a relay chain or configuration (e.g., downstream) of the access network. In such cases, one or more components of the decomposed RAN architecture (e.g., one or more IAB nodes 104 or components of the IAB node 104) may be configured to operate according to the techniques described herein.

[0060] For example, an access network (AN) or RAN may include an access node (e.g., an IAB donor), communications between an IAB node 104, and one or more UEs 115. The IAB donor may facilitate connectivity between the core network 130 and the AN (e.g., via a wired or wireless connection to the core network 130). That is, the IAB donor may refer to a RAN node having a wired or wireless connection to the core network 130. The IAB donor may include a CU 160 and at least one DU 165 (e.g., and RU 170), in which case the CU 160 may communicate with the core network 130 via an interface (e.g., a backhaul link). The IAB donor and the IAB node 104 may communicate via an F1 interface according to a protocol defining a signaling message (e.g., an F1 AP protocol). Additionally or alternatively, CU 160 may communicate with the core network via an interface (which may be an example of a portion of a backhaul link) and may communicate with other CUs 160 (e.g., CUs 160 associated with alternative IAB donors) via an Xn-C interface (which may be an example of a portion of a backhaul link).

[0061] The IAB node 104 may refer to a RAN node that provides IAB functions (e.g., access for UE 115, wireless self-backhaul capabilities, etc.). The DU 165 may act as a distributed scheduling node toward a child node associated with the IAB node 104, and the IAB-MT may act as a scheduled node toward a parent node associated with the IAB node 104. That is, the IAB donor may be referred to as a parent node that communicates with one or more child nodes (e.g., the IAB donor may relay transmissions to the UE through one or more other IAB nodes 104). Additionally or alternatively, the IAB node 104 may also be referred to as a parent node or child node of other IAB nodes 104, depending on the relay chain or configuration of the AN. Therefore, the IAB-MT entity of the IAB node 104 may provide a Uu interface for the child IAB node 104 to receive signaling from the parent IAB node 104, and the DU interface (e.g., DU 165) may provide a Uu interface for the parent IAB node 104 to signal to the child IAB node 104 or the UE 115.

[0062] For example, the IAB node 104 may be referred to as a parent node supporting child IAB node communications or as a child node associated with an IAB donor, or both. The IAB donor may include a CU 160 having a wired or wireless connection (e.g., backhaul communication link 120) to the core network 130, and may act as a parent node of the IAB node 104. For example, the DU 165 of the IAB donor may relay the transmission to the UE 115 through the IAB node 104, or may directly signal the UE 115 of the transmission, or both. The CU 160 of the IAB donor may signal the IAB node 104 of the communication link establishment via the F1 interface, and the IAB node 104 may schedule the transmission (e.g., the transmission relayed from the IAB donor to the UE 115) through the DU 165. That is, data may be relayed to and from the IAB node 104 via signaling via the NR Uu interface of the MT to the IAB node 104. Communications with the IAB node 104 may be scheduled by the DU 165 of the IAB donor, and communications with the IAB node 104 may be scheduled by the DU 165 of the IAB node 104 .

[0063] Where the techniques described herein are applied in the context of a disaggregated RAN architecture, one or more components of the disaggregated RAN architecture may be configured to support conditional AI, machine learning models, and parameter set configurations as described herein. For example, some operations described as being performed by a UE 115 or a network entity 105 (e.g., a base station 140) may additionally or alternatively be performed by one or more components of the disaggregated RAN architecture (e.g., an IAB node 104, a DU 165, a CU 160, a RU 170, a RIC 175, a SMO 180).

[0064] UE 115 may include or may be referred to as a mobile device, a wireless device, a remote device, a handheld device, or a subscriber device, or some other suitable term, where a "device" may also be referred to as a unit, a station, a terminal, or a client, etc. UE 115 may also include or may be referred to as a personal electronic device, such as a cellular phone, a personal digital assistant (PDA), a tablet computer, a laptop computer, or a personal computer. In some examples, UE 115 may include or may be referred to as a wireless local loop (WLL) station, an Internet of Things (IoT) device, an Internet of Everything (IoE) device, or a machine type communication (MTC) device, etc., which may be implemented in various objects such as appliances or vehicles, meters, etc.

[0065] The UE 115 described herein may be capable of communicating with various types of devices such as other UEs 115 which may sometimes act as relays, as well as network entities 105 and network equipment including macro eNBs or gNBs, small cell eNBs or gNBs, or relay base stations, etc. Figure 1 shown.

[0066] The UE 115 and the network entity 105 may wirelessly communicate with each other via one or more communication links 125 (e.g., access links) using resources associated with one or more carriers. The term "carrier" may refer to a collection of RF spectrum resources having a physical layer structure defined for supporting the communication link 125. For example, a carrier for the communication link 125 may include a portion of an RF spectrum band (e.g., a bandwidth portion (BWP)) that operates according to one or more physical layer channels for a given radio access technology (e.g., LTE, LTE-A, LTE-A Pro, NR). Each physical layer channel may carry acquisition signaling (e.g., synchronization signals, system information), control signaling for coordinating carrier operations, user data, or other signaling. The wireless communication system 100 may support communications with the UE 115 using carrier aggregation or multi-carrier operation. Depending on the carrier aggregation configuration, the UE 115 may be configured to have multiple downlink component carriers and one or more uplink component carriers. Carrier aggregation may be used for both frequency division duplex (FDD) and time division duplex (TDD) component carriers. Communication between the network entity 105 and other devices may refer to communication between these devices and any portion (e.g., entity, sub-entity) of the network entity 105. For example, the terms "send," "receive," or "communicate" when referring to the network entity 105 may refer to any portion of a network entity 105 (e.g., base station 140, CU 160, DU 165, RU 170) of a RAN communicating with another device (e.g., directly or via one or more other network entities 105).

[0067] In some examples, such as in a carrier aggregation configuration, a carrier may also have acquisition signaling or control signaling that coordinates the operation of other carriers. A carrier may be associated with a frequency channel (e.g., an Evolved Universal Mobile Telecommunications System Terrestrial Radio Access (E-UTRA) Absolute RF Channel Number (EARFCN)) and may be identified based on a channel raster used for discovery by UE 115. A carrier may operate in a standalone mode, in which case initial acquisition and connection may be made by UE 115 via the carrier, or a carrier may operate in a non-standalone mode, in which case a different carrier (e.g., of the same or different radio access technology) is used to anchor the connection.

[0068] The signal waveform transmitted via the carrier may include multiple subcarriers (e.g., using multi-carrier modulation (MCM) techniques, such as orthogonal frequency division multiplexing (OFDM) or discrete Fourier transform spread OFDM (DFT-S-OFDM)). In a system using MCM technology, a resource element may refer to a symbol period (e.g., the duration of a modulation symbol) and a resource of a subcarrier, in which case the symbol period and the subcarrier spacing may be inversely related. The number of bits carried by each resource element may depend on the modulation scheme (e.g., the order of the modulation scheme, the coding rate of the modulation scheme, or both), so that a relatively high number of resource elements (e.g., in the transmission duration) and a relatively high order modulation scheme may correspond to a relatively high rate of communication. Wireless communication resources may refer to a combination of RF spectrum resources, time resources, and spatial resources (e.g., spatial layers or beams), and the use of multiple spatial resources may increase the data rate or data integrity used for communication with UE 115.

[0069] The time interval for the network entity 105 or the UE 115 may be expressed as a multiple of a basic time unit, which may be, for example, a sampling period T s =1 / (Δf max ·N f ) seconds, for which Δf max It can represent the supported subcarrier spacing, and N f The supported discrete Fourier transform (DFT) size may be indicated. The time intervals of the communication resources may be organized according to radio frames each having a specified duration (e.g., 10 milliseconds (ms)). Each radio frame may be identified by a system frame number (SFN) (e.g., ranging from 0 to 1023).

[0070] Each frame may include multiple consecutively numbered subframes or time slots, and each subframe or time slot may have the same duration. In some examples, the frame may be divided into subframes (e.g., in the time domain), and each subframe may be further divided into a certain number of time slots. Alternatively, each frame may include a variable number of time slots, and the number of time slots may depend on the subcarrier spacing. Each time slot may include a certain number of symbol periods (e.g., depending on the length of the cyclic prefix appended in front of each symbol period). In some wireless communication systems 100, the time slot may be further divided into a plurality of micro time slots associated with one or more symbols. Excluding the cyclic prefix, each symbol period may be associated with one or more (e.g., N f The duration of a symbol period may depend on the subcarrier spacing or the operating frequency band.

[0071] A subframe, slot, mini-slot, or symbol may be the smallest scheduling unit (e.g., in the time domain) of the wireless communication system 100 and may be referred to as a transmit time interval (TTI). In some examples, the TTI duration (e.g., the number of symbol periods in a TTI) may be variable. Additionally or alternatively, the smallest scheduling unit of the wireless communication system 100 may be dynamically selected (e.g., in a burst of a shortened TTI (sTTI)).

[0072] According to various techniques, physical channels may be multiplexed using carriers for communication. Physical control channels and physical data channels may be multiplexed for signaling via downlink carriers, for example, using one or more of a time division multiplexing (TDM) technique, a frequency division multiplexing (FDM) technique, or a hybrid TDM-FDM technique. A control region (e.g., a control resource set (CORESET)) of a physical control channel may be defined by a set of symbol periods and may extend across a system bandwidth of a carrier or a subset of that system bandwidth. One or more control regions (e.g., CORESETs) may be configured for a set of UEs 115. For example, one or more UEs in UE 115 may monitor or search a control region to obtain control information according to one or more search space sets, and each search space set may include one or more control channel candidates in one or more aggregation levels arranged in a cascaded manner. The aggregation level of a control channel candidate may refer to the amount of control channel resources (e.g., control channel elements (CCEs)) associated with coded information for a control information format having a given payload size. The search space sets may include a common search space set configured for transmitting control information to multiple UEs 115 , and a UE-specific search space set for transmitting control information to a specific UE 115 .

[0073] The network entity 105 may provide communication coverage via one or more cells (e.g., macro cells, small cells, hotspots, or other types of cells, or any combination thereof). The term "cell" may refer to a logical communication entity used to communicate with the network entity 105 (e.g., using a carrier), and may be associated with an ID used to distinguish adjacent cells (e.g., a physical cell ID (PCID), a virtual cell ID (VCID), etc.). In some examples, a cell may also refer to a coverage area 110 or a portion of a coverage area 110 (e.g., a sector) on which a logical communication entity operates. Depending on various factors such as the capabilities of the network entity 105, such cells may range from smaller areas (e.g., structures, subsets of structures) to larger areas. For example, a cell may be or may include a building, a subset of a building, or an external space between or overlapping coverage areas 110, and the like.

[0074] A macro cell generally covers a relatively large geographic area (e.g., several kilometers in radius) and may allow unrestricted access to a UE 115 that has a service subscription with a network provider that supports the macro cell. A small cell may be associated with a lower power network entity 105 (e.g., a lower power base station 140) than a macro cell, and the small cell may operate using the same or different (e.g., licensed, unlicensed) frequency band as the macro cell. A small cell may provide unrestricted access to a UE 115 that has a service subscription with a network provider, or may provide restricted access to a UE 115 associated with a small cell (e.g., a UE 115 in a closed subscriber group (CSG), a UE 115 associated with a user in a home or office). A network entity 105 may support one or more cells, and may also use one or more component carriers to support communications via one or more cells.

[0075] In some examples, a carrier may support multiple cells, and different cells may be configured according to different protocol types (e.g., MTC, narrowband IoT (NB-IoT), enhanced mobile broadband (eMBB)) that may provide access to different types of devices.

[0076] In some examples, the network entities 105 (e.g., base stations 140, RUs 170) may be mobile and thus provide communication coverage for mobile coverage areas 110. In some examples, different coverage areas 110 associated with different technologies may overlap, but the different coverage areas 110 may be supported by the same network entity 105. In some other examples, overlapping coverage areas 110 associated with different technologies may be supported by different network entities 105. The wireless communication system 100 may include, for example, a heterogeneous network in which different types of network entities 105 provide coverage for various coverage areas 110 using the same or different radio access technologies.

[0077] The wireless communication system 100 may support synchronous or asynchronous operation. For synchronous operation, the network entities 105 (e.g., base stations 140) may have similar frame timing, and transmissions from different network entities 105 may be approximately aligned in time. For asynchronous operation, the network entities 105 may have different frame timing, and in some examples, transmissions from different network entities 105 may not be aligned in time. The techniques described herein may be used for synchronous or asynchronous operation.

[0078] Some UEs 115, such as MTC or IoT devices, may be low-cost or low-complexity devices and may allow automated communication between machines (e.g., via machine-to-machine (M2M) communication). M2M communication or MTC may refer to data communication technology that allows devices to communicate with each other or with a network entity 105 (e.g., base station 140) without human intervention. In some examples, M2M communication or MTC may include communication from devices that integrate sensors or meters to measure or capture information and relay such information to a central server or application that uses the information or presents the information to a person interacting with the application. Some UEs 115 may be designed to collect information or implement automated behavior of machines or other devices. Examples of applications for MTC devices include: smart metering, inventory monitoring, water level monitoring, equipment monitoring, health care monitoring, field survival monitoring, weather and geographic event monitoring, fleet management and tracking, remote security sensing, physical access control, and transaction-based commercial charging.

[0079] Some UEs 115 may be configured to employ an operating mode that reduces power consumption, such as half-duplex communication (e.g., a mode that supports unidirectional communication via transmission or reception but does not transmit and receive concurrently). In some examples, half-duplex communication may be performed with a reduced peak rate. Other energy-saving techniques for UEs 115 include entering a power-saving deep sleep mode when not engaged in active communications, operating using limited bandwidth (e.g., according to narrowband communications), or a combination of these techniques. For example, some UEs 115 may be configured to operate using a narrowband protocol type that is associated with a defined portion or range (e.g., a set of subcarriers or resource blocks (RBs)) within a carrier, within a guard band of a carrier, or outside a carrier.

[0080] The wireless communication system 100 may be configured to support ultra-reliable communication or low-latency communication or various combinations thereof. For example, the wireless communication system 100 may be configured to support ultra-reliable low-latency communication (URLLC). UE 115 may be designed to support ultra-reliable or low-latency or critical functions. Ultra-reliable communication may include private communication or group communication, and may be supported by one or more services (such as push-to-talk, video, or data). Support for ultra-reliable, low-latency functions may include prioritization of services, and such services may be used for public safety or general commercial applications. The terms ultra-reliable, low-latency, and ultra-reliable low-latency may be used interchangeably herein.

[0081] In some examples, the UE 115 may be configured to support communication directly with other UEs 115 via a device-to-device (D2D) communication link 135 (e.g., according to a peer-to-peer (P2P), D2D, or sidelink protocol). In some examples, one or more UEs 115 in a group that are performing D2D communication may be within a coverage area 110 of a network entity 105 (e.g., a base station 140, a RU 170), which may support aspects of such D2D communication configured (e.g., scheduled) by the network entity 105. In some examples, one or more UEs 115 in such a group may be outside the coverage area 110 of the network entity 105, or may otherwise be unable or not configured to receive transmissions from the network entity 105. In some examples, a group of UEs 115 communicating via D2D communication may support a one-to-many (1:M) system, in which each UE 115 transmits to each of the other UEs 115 in the group. In some examples, the network entity 105 may facilitate scheduling of resources for D2D communication. In some other examples, D2D communications may be performed between UEs 115 without involving network entity 105 .

[0082] The core network 130 may provide user authentication, access authorization, tracking, Internet Protocol (IP) connectivity, and other access, routing, or mobility functions. The core network 130 may be an evolved packet core (EPC) or a 5G core (5GC), which may include at least one control plane entity (e.g., a mobility management entity (MME), an access and mobility management function (AMF)) for managing access and mobility and at least one user plane entity (e.g., a serving gateway (S-GW), a packet data network (PDN) gateway (P-GW), or a user plane function (UPF)) for routing packets or interconnecting to an external network. The control plane entity may manage non-access stratum (NAS) functions, such as mobility, authentication, and bearer management of UE 115 served by a network entity 105 (e.g., a base station 140) associated with the core network 130. User IP packets may be delivered through the user plane entity, which may provide IP address allocation and other functions. The user plane entity may be connected to the IP service 150 of one or more network operators. IP services 150 may include access to the Internet, an intranet, an IP Multimedia Subsystem (IMS), or packet-switched streaming services.

[0083] The wireless communication system 100 may operate using one or more frequency bands that may be in the range of 300 megahertz (MHz) to 300 gigahertz (GHz). Generally speaking, the region from 300 MHz to 3 GHz is referred to as the ultra-high frequency (UHF) region or decimeter band because the wavelength ranges from about one decimeter to one meter in length. UHF waves may be blocked or redirected by buildings and environmental features (which may be referred to as clusters), but these waves may be sufficient to penetrate structures so that macro cells provide services to UEs 115 located indoors. Communications using UHF waves may be associated with smaller antennas and shorter ranges (e.g., less than 100 kilometers) compared to communications using lower frequencies and longer waves in the high frequency (HF) or very high frequency (VHF) portions of the spectrum below 300 MHz.

[0084] The wireless communication system 100 may utilize licensed and unlicensed RF spectrum bands. For example, the wireless communication system 100 may use unlicensed bands (such as the 5 GHz Industrial, Scientific, and Medical (ISM) band) to employ license assisted access (LAA), LTE unlicensed (LTE-U) radio access technology, or NR technology. When operating using unlicensed RF spectrum bands, devices such as network entity 105 and UE 115 may employ carrier sensing for conflict detection and avoidance. In some examples, operations using unlicensed bands may be based on carrier aggregation configuration (e.g., LAA) in conjunction with component carriers operating using licensed bands. Operations using unlicensed spectrum may include downlink transmissions, uplink transmissions, P2P transmissions, or D2D transmissions, and the like.

[0085] The network entity 105 (e.g., base station 140, RU 170) or UE 115 may be equipped with multiple antennas that can be used to employ technologies such as transmit diversity, receive diversity, multiple-input multiple-output (MIMO) communication, or beamforming. The antennas of the network entity 105 or UE 115 may be located in one or more antenna arrays or antenna panels, which may support MIMO operations or transmit or receive beamforming. For example, one or more base station antennas or antenna arrays may be co-located at an antenna assembly, such as an antenna tower. In some examples, antennas or antenna arrays associated with the network entity 105 may be located at different geographical locations. The network entity 105 may include an antenna array having a collection of multiple rows and columns of antenna ports that the network entity 105 can use to support beamforming for communications with the UE 115. Similarly, the UE 115 may include one or more antenna arrays that may support various MIMO or beamforming operations. Additionally or alternatively, the antenna panel may support RF beamforming for signals sent via the antenna ports.

[0086] Beamforming (which may also be referred to as spatial filtering, directional transmission, or directional reception) is a signal processing technique that may be used at a transmitting device or a receiving device (e.g., a network entity 105, a UE 115) to shape or direct an antenna beam (e.g., a transmit beam, a receive beam) along a spatial path between the transmitting device and the receiving device. Beamforming may be achieved by combining signals conveyed via antenna elements of an antenna array so that some signals propagating along a particular direction relative to the antenna array experience constructive interference, while other signals experience destructive interference. Adjustments to signals conveyed via antenna elements may include the transmitting device or the receiving device applying an amplitude offset, a phase offset, or both to signals carried via antenna elements associated with the device. Adjustments associated with each of these antenna elements may be defined by a set of beamforming weights associated with a particular direction (e.g., relative to the antenna array of the transmitting device or the receiving device or relative to some other direction).

[0087] The network entity 105 or UE 115 may use beam scanning techniques as part of a beamforming operation. For example, the network entity 105 (e.g., base station 140, RU 170) may use multiple antennas or antenna arrays (e.g., antenna panels) to perform beamforming operations for directional communication with UE 115. Some signals (e.g., synchronization signals, reference signals, beam selection signals, or other control signals) may be sent multiple times by the network entity 105 along different directions. For example, the network entity 105 may send signals according to different sets of beamforming weights associated with different transmission directions. Transmissions along different beam directions may be used to identify (e.g., by a transmitting device (such as network entity 105), or by a receiving device (such as UE 115)) beam directions for later transmission or reception by the network entity 105.

[0088] Some signals, such as data signals associated with a particular receiving device, may be sent by a transmitting device (e.g., transmitting network entity 105, transmitting UE 115) along a single beam direction (e.g., a direction associated with a receiving device (such as receiving network entity 105 or receiving UE 115)). In some examples, the beam direction associated with the transmission along the single beam direction may be determined based on signals sent along one or more beam directions. For example, UE 115 may receive one or more of the signals sent by network entity 105 along different directions, and may report to network entity 105 an indication of the signal received by UE 115 with the highest signal quality or other acceptable signal quality.

[0089] In some examples, transmission by a device (e.g., by network entity 105 or UE 115) may be performed using multiple beam directions, and the device may use a combination of digital precoding or beamforming to generate a combined beam for transmission (e.g., from network entity 105 to UE 115). UE 115 may report feedback indicating precoding weights for one or more beam directions, and the feedback may correspond to a configured set of beams across a system bandwidth or one or more subbands. Network entity 105 may send reference signals (e.g., cell-specific reference signals (CRS), channel state information reference signals (CSI-RS)), which may or may not be precoded. UE 115 may provide feedback for beam selection, which may be a precoding matrix indicator (PMI) or codebook-based feedback (e.g., multi-panel codebook, linear combination codebook, port selection codebook). Although these techniques are described with reference to signals sent along one or more directions by a network entity 105 (e.g., base station 140, RU 170), UE 115 may use similar techniques to send signals multiple times along different directions (e.g., to identify a beam direction for subsequent transmission or reception by UE 115), or to send signals along a single direction (e.g., to send data to a receiving device).

[0090] A receiving device (e.g., UE 115) may perform receiving operations according to multiple receiving configurations (e.g., directional listening) when receiving various signals (such as synchronization signals, reference signals, beam selection signals, or other control signals) from a receiving device (e.g., network entity 105). For example, the receiving device may perform reception according to multiple receiving directions by: receiving via different antenna subarrays, processing the received signals according to different antenna subarrays, receiving according to different receiving beamforming weight sets (e.g., different directional listening weight sets) applied to signals received at multiple antenna elements of an antenna array, or processing the received signals according to different receiving beamforming weight sets applied to signals received at multiple antenna elements of an antenna array, any of which may be referred to as "listening" according to different receiving configurations or receiving directions. In some examples, the receiving device may use a single receiving configuration to receive along a single beam direction (e.g., when receiving a data signal). A single receive configuration may be aligned along a beam direction determined based on listening according to different receive configuration directions (e.g., a beam direction determined to have the highest signal strength, highest signal-to-noise ratio (SNR), or other acceptable signal quality based on listening according to multiple beam directions).

[0091] UE 115 and network entity 105 may support retransmission of data to increase the likelihood of data being successfully received. Hybrid automatic repeat request (HARQ) feedback is a technique for increasing the likelihood of correctly receiving data via a communication link (e.g., communication link 125, D2D communication link 135). HARQ may include a combination of error detection (e.g., using cyclic redundancy check (CRC)), forward error correction (FEC), and retransmission (e.g., automatic repeat request (ARQ)). HARQ can improve the throughput at the MAC layer under poor radio conditions (e.g., low signal-to-noise ratio conditions). In some examples, the device may support same-slot HARQ feedback, in which case the device may provide HARQ feedback in a specific time slot for data received via a previous symbol in the time slot. In some other examples, the device may provide HARQ feedback in a subsequent time slot or according to some other time interval.

[0092] Some wireless communication systems 100 may support wireless devices such as UE 115 and network entity 105 that use machine learning models or AI models to make decisions based on learned data, which can improve the efficiency, latency, and reliability of wireless applications. Some machine learning models and AI models may use a defined set of input and output parameters, where one or more input values ​​(e.g., corresponding to the input parameters) may be used to predict output values ​​(e.g., corresponding to the output parameters). Some machine learning models may be trained using offline data. During offline training, the machine learning model may adapt different data processing methods to manage lost or corrupted input values.

[0093] In some examples, as a result of machine learning or artificial intelligence model training, multiple parameter sets representing weights of the machine learning or artificial intelligence model may be generated for different conditions. Example conditions for generating different parameter sets for a given model may include availability of model input parameters, regional scope, configuration of network entity 105 and UE 115, KPIs of the model, physical environment (e.g., indoors, outdoors, and other physical settings), or any combination thereof.

[0094] In some examples, the machine learning model may operate in a cellular network or wireless communication scenario. Once the machine learning model is trained, the machine learning model may perform machine learning inference, during which the trained model is used to draw conclusions from new input data. For example, during an inference procedure, the machine learning model may receive new input data and infer new outputs from the new input data based on the data that the model was trained on (e.g., what the machine learning model has learned). Therefore, for a machine learning model operating in a wireless communication system, the unavailability of input values ​​during machine learning inference may lead to incorrect predictions and increased errors in the output of the machine learning model.

[0095] In the wireless communication system 100, the unavailability of the input value may be based on an untimely delivery of data from the network entity 105 or the UE 115. For example, the RAN of the wireless communication system 100 may be overloaded and unable to report data to the UE 115 in a timely manner, wherein machine learning inferences may be performed by a machine learning model and parameter set at the UE 115, and the UE 115 may use a subset of the data reported from the network entity 105 for inference (e.g., input data reported from the network entity 105 to the UE 115). Alternatively, the UE 115 may be overloaded and unable to report data to the network entity 105 in a timely manner, wherein machine learning inferences may be performed by a machine learning model and parameter set at the network entity 105, and the network entity 105 may use a subset of the data reported from the UE 115 for inference (e.g., input data reported from the UE 115 to the network entity 105). In some examples, the input data itself may be corrupted such that the UE 115 or network entity 105 may be unable to decode the reported input data (eg, the input data may be corrupted due to passing through a poor channel).

[0096] The wireless communication system 100 may support conditional machine learning model and parameter set configurations, which may be used to manage lost or corrupted input data to a machine learning model to reduce errors in the output of the machine learning model. In some examples, one or more machine learning models and parameter sets may be trained so that a given machine learning model, a given parameter set, or both may not be affected by lost or corrupted input data. One or more machine learning models, one or more parameter sets, or both may use input data from a UE 115 and a network entity 105, or input data from only a UE 115 to generate machine learning inferences. Machine learning inferences may be the process of running real-time data points through a trained machine learning model to calculate an output. Information about one or more machine learning models, one or more parameter sets, or both may be provided to the network entity 105, including how the machine learning models are trained, the priority of each machine learning model, the priority of each parameter set, and other information.

[0097] The network entity 105 may configure the model ID with usage conditions for multiple machine learning models, multiple parameter sets, or both. The network entity 105 sends a message to the UE 115 indicating one or more machine learning models, one or more parameter sets, or both, and the usage conditions that the UE 115 may use to select one of the machine learning models, one of the parameter sets, or both to perform machine learning inference based on what input data is available, lost, or corrupted. Examples of machine learning model and / or parameter set usage conditions for inference and training may include availability of model input parameters, regional scope, validity timer of the model, validity timer of the parameter set, configuration of the network entity and the UE 115, KPI of the model, physical environment (e.g., indoor, outdoor, and other physical settings), or any combination thereof. The UE 115 may select a machine learning model, parameter set, or both based on whether a particular usage condition is satisfied, the priority of the machine learning model and / or parameter set, and other factors. In some examples, the UE 115 may switch the machine learning model and / or parameter set it uses to perform machine learning inference and send an indication of the model and / or parameter set switch to the network entity 105. In some examples, UE 115 may generate machine learning inferences using the selected machine learning model, the selected parameter set, or both, and UE 115 may send a report to network entity 105 indicating the output of the inferences.

[0098] Figure 2 An example of a wireless communication system 200 that supports conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is illustrated. In some examples, the wireless communication system 200 may implement aspects of the wireless communication system 100, or may be implemented by aspects of the wireless communication system 100. For example, the wireless communication system 200 may include a UE 115-a and a network entity 105-a (e.g., a network node), which may be examples of corresponding devices as described herein. In addition, the wireless communication system 200 may include a model designer 205. In some cases, the UE 115-a may use the machine learning model 210, the parameter set 245, or both, and the UE 115-a may select the machine learning model and the parameter set or both from a set of machine learning models and / or parameter sets based on one or more usage conditions.

[0099] As described herein, a parameter set may correspond to weights for a machine learning model, or neural network weights for an AI or machine learning model. In some examples, a single machine learning model may be associated with multiple machine learning parameter sets, and a usage condition may be used to select one of the parameter sets based on what data is available, missing, corrupted, or any combination thereof. In some examples, the selected parameter set may be used when running a particular machine learning model to generate machine learning interference. Thus, UE 115-a may be configured to conditionally change the parameter set while keeping the machine learning model the same.

[0100] The UE 115-a and the network entity 105-a may communicate via respective communication links, and the UE 115-a and the network entity 105-a may each communicate with the model designer 205. The model designer 205 may configure the UE 115-a (e.g., cloud-based software) with a program for training one or more machine learning models and corresponding parameter sets (including the machine learning model 210 and the parameter set 245). In some examples, during model training, the model designer 205 may train the machine learning model 210 and the parameter set 245 so that the machine learning model 210 and the parameter set 245 may be agnostic to scenarios when the input data is unavailable or compromised. For example, when training the machine learning model 210 and the parameter set 245, the model designer 205 may consider a set of required input parameters and a set of optional input parameters (e.g., for which unavailability may have little impact on the performance of the machine learning model 210 and the parameter set 245, and availability may improve the performance of the machine learning model 210).

[0101] After training the machine learning model 210 and the parameter set 245, the model designer 205 may send an indication of a model registration 215 to the network entity 105-a. In some examples, the model registration 215 may indicate a model descriptor corresponding to the machine learning model 210, the parameter set 245, or both, which may include information about whether the machine learning model 210, the parameter set 245, or both are agnostic to missing or corrupted input data, and information about which missing or corrupted input data the machine learning model 210, the parameter set 245, or both are agnostic. Additionally, the model registration 215 may include information about a data processing method that may be used to manage missing inference input data.

[0102] In some examples, during model registration 215, model designer 205 may register alternative machine learning models, alternative parameter sets, or both (e.g., machine learning models and / or parameter sets that are alternatives to machine learning model 210 and parameter set 245, which may be referred to as baseline machine learning models and baseline parameter sets) and corresponding priorities of the alternative machine learning models and / or parameter sets. The alternative machine learning models, alternative parameter sets, or both may suppress the use of input data (e.g., including measurements and parameters) from network entity 105. In some cases, network entity 105-a may change the priority of the alternative machine learning model, alternative parameter set, or both based on a KPI determined by network entity 105-a or a KPI reported by UE 115-a. In some examples, model registration 215 may include a list of missing input data that is unknown to machine learning model 210, parameter set 245, or both (which are trained by model designer 205). For example, model registration 215 may include a list of input parameters that lack a significant impact on the performance of machine learning model 210, parameter set 245, or both.

[0103] In response to receiving an indication of model registration 215 from the model designer 205, the network entity 105-a may perform model configuration to configure the machine learning model 210, the parameter set 245, or both and one or more alternative machine learning models, one or more alternative parameter sets, or both for use by the UE 115-a. The network entity 105-a may use conditional model and parameter set configurations, wherein for each configured machine learning model and parameter set, the network entity 105-a may additionally configure one or more usage conditions and priorities. In some examples, the network entity 105-a may define a configuration based on the availability of input inference data in the conditional model and parameter set configuration. In addition, the network entity 105-a may define a performance-based configuration, a scenario-based configuration (e.g., a configuration based on the RAT, the physical environment of the wireless communication system 200, antenna port characteristics, and other scenarios), or any combination thereof.

[0104] In some examples, the usage conditions for selecting a machine learning model, a parameter set, or both may include instructions for selecting a specific one of the machine learning models and / or parameter sets when some input data is missing, a priority corresponding to each machine learning model, each parameter set, or both for selecting a specific one of the machine learning models, a specific one of the machine learning models, or both, or a combination thereof. Additionally, the instructions for selecting a machine learning model, a parameter set, or both may include availability of input data (e.g., from UE 115-a, network entity 105-a, or both), usage scenarios, regional scope, model validity timer, parameter set validity timer, configuration of UE 115-a, configuration of network entity 105-a, performance threshold, or any combination thereof.

[0105] In some examples, the network entity 105-a may configure the machine learning model 210, the parameter set 245, or both and one or more alternative machine learning models, one or more alternative parameter sets, or both that the UE 115-a may use, depending on which input data is missing for a given machine learning model. For example, if the machine learning inference uses input data from the network entity 105-a, the network entity 105-a may configure the machine learning model 210, the parameter set 245, or both to use input data from the UE 115-a and the network entity 105-a. If both the network entity 105-a and the UE 115-a provide the required input data for the machine learning model 210, the parameter set 245, or both, the UE 115-a may use the machine learning model 210, the parameter set 245, or both. However, if the network entity 105-a fails to provide the UE 115-a with the input data required by the machine learning model 210, the parameter set 245, or both, thereby making the machine learning model 210, the parameter set 245, or both unavailable, the UE 115-a may use an alternative machine learning model and / or parameter set configured by the network entity 105-a that may use only the input data from the UE 115-a. Figure 3 Detailed description.

[0106] UE 115-a may receive a message 220 from network entity 105-a indicating a set of machine learning models, a set of parameter sets, or both (e.g., including machine learning model 210, parameter set 245, or both and one or more alternative machine learning models, one or more alternative parameter sets, or both configured by network entity 105-a) and indicating usage conditions for selecting one of the machine learning models and / or parameter sets for performing machine learning inference. UE 115-a may load a set of machine learning models and / or parameter sets for inference, and UE 115-a may select a machine learning model, parameter set, or both for machine learning inference based on corresponding usage conditions being satisfied. For example, if one or both of machine learning model 210, parameter set 245, or both has the highest priority, UE 115-a may select the machine learning model, parameter set, or both for which all input inference data is available and trusted, or based on the availability of input data. In some examples, the UE 115-a may select the machine learning model 210, the parameter set 245, or both based on a usage condition indicating that the machine learning model 210, the parameter set 245, or both are selected when the UE 115-a is able to successfully receive and decode input data from the network entity 105-a. Alternatively, the UE 115-a may select the machine learning model 210, the parameter set 245, or both based on a usage condition indicating that the machine learning model 210, the parameter set 245, or both are selected when the UE 115-a fails to receive input data for an alternative machine learning model from the network entity 105-a. If the usage conditions of multiple machine learning models and / or parameter sets are met (e.g., the machine learning model 210, the parameter set 245, or both and one or more alternative machine learning models, one or more alternative machine learning models, or both), the UE 115-a may select the machine learning model 210, the parameter set 245, or both based on the priority of these models and / or parameter sets (e.g., in the case where the machine learning model 210, the parameter set 245, or both may have the highest priority).

[0107] In some examples, UE 115-a may perform model and / or parameter set switching for different inference occasions. For example, UE 115-a may be configured to perform a first machine learning inference during a first inference occasion. UE 115-a may select a machine learning model 210, a parameter set 245, or both to generate a first machine learning inference based on input data received and successfully decoded by UE 115-a from network entity 105-a, where data reported by the network entity and input data measured by UE 115-a may be used in machine learning model 210, parameter set 245, or both. UE 115-a may generate machine learning inferences using the selected machine learning model 210, parameter set 245, or both. In some examples, UE 115-a may be configured to perform a second machine learning inference during a second inference occasion. However, UE 115-a may fail to receive input data from network entity 105-b for the second inference occasion, and therefore, UE 115-a may select an alternative machine learning model, an alternative parameter set, or both to generate machine learning inferences. Because UE 115-a switches from using machine learning model 210, parameter set 245, or both in the first inference opportunity to using an alternative machine learning model, an alternative parameter set, or both in the second inference opportunity, UE 115-a may send a switch indication 225 to network entity 105-a indicating the switch (e.g., by indicating the ID of the machine learning model and / or parameter set). That is, UE 115-a may indicate which model, parameter set, or both UE 115-a used to obtain a particular inference output in the switch indication 225 sent to network entity 105-a.

[0108] In some cases, the UE 115-a may generate a machine learning inference output using the machine learning model 210, the parameter set 245, or both, and the UE 115-a may send a report 230 indicating the generated machine learning inference output to the network entity 105-a. In addition, the UE 115-a may transmit feedback 235 indicating some statistics regarding model and / or parameter set switching to the model designer 205. For example, the feedback 235 may include information such as the rate at which the UE 115-a performs model and / or parameter set switching, the reason for the model and / or parameter set switching (e.g., indicating that the UE 115-a performs the model and / or parameter set switching based on the absence of the parameter), and any other system KPIs obtained by the UE 115-a or the network entity 105-a based on the model and / or parameter set switching. In addition, the UE 115-a may receive feedback 240 from the network entity 105-a, which may include information about the model and / or parameter set switching from the perspective of the network entity 105-a (e.g., based on the KPI determined by the network entity 105-a). The UE 115-a may then send feedback 235 to the model designer 205, where the feedback 235 may include the feedback generated from the UE 115-a and the feedback 240 from the network entity 105-a.

[0109] The described techniques may be used for a machine learning model, a parameter set, or both operating at a UE 115-a, which may use some input data at the UE 115-a from the network entity 105-a or some input data at the network entity 105-a from the UE 115-a. For example, the UE 115-a may use the described techniques for a beam management inference procedure. The UE 115-a may expect to receive some information from the network entity 105-b about the directionality, beam shape, and other characteristics of a transmit or receive beam associated with the network entity 105-a. The network entity 105-a may provide input data including beam parameters to the UE 115-a, and the UE 115-a may use the input data in the machine learning model 210, the parameter set 245, or both (e.g., which may use the input data from the network entity 105-a) to generate an inference output that may correspond to a beam prediction or beam selection from a set of possible beams that the UE 115-a may use to transmit or receive.

[0110] Additionally or alternatively, when the network entity 105-a is configured for performance-based usage conditions for selecting a machine learning model, a parameter set, or both, the UE 115-a may use the described techniques to perform machine learning inferences. For example, the latency of performing the inference may be based on the power capabilities of the UE 115-a. If the UE 115-a is operating in a low power state, the UE 115-a may not be able to run complex machine learning models and parameter sets and obtain inference outputs in a timely manner. Therefore, the overall performance of the inference procedure may be reduced. In order to improve the overall performance of the inference procedure when the UE 115-a is operating in a low power mode, the UE 115-a may use model and / or parameter set switching to select a different machine learning model, a different parameter set, or both for obtaining an inference output that may be less complex and therefore use less time and less power to run. Therefore, the UE 115-a may switch which machine learning model, parameter set, or both it uses to obtain the inference output based on the power mode in which the UE 115-a is operating.

[0111] Additionally or alternatively, the UE 115-a may use the machine learning model 210, the parameter set 245, or both to perform RF fingerprinting (RFFP). The RFFP may be used to identify a transmitting device to a passive receiving device, and may be used for device authentication and access control to reduce the vulnerability of the 5G wireless communication system. In UE-based RFFP, a neural network may be trained for the RFFP and made available for inference at the UE 115-a. The neural network may use channel observations derived from a downlink positioning reference signal (PRS) as input to determine the location of the UE 115-a. The channel observations may come from one TRP (e.g., a single-cell RFFP) or from multiple TRPs (e.g., a multi-cell RFFP). In such cases of UE-based RFFP, the machine learning model 210, the parameter set 245, or both may suppress the use of input data from the network entity 105.

[0112] In some cases, the RFFP may be network entity based, where the network entity 105 may utilize uplink channel observations from one or more TRPs to derive the location of the UE 115. In such cases of network entity based RFFP, the machine learning model 210, the parameter set 245, or both may refrain from using input data from the UE 115. Alternatively, for UE-assisted RFFP, the neural network may be partitioned in a manner that a portion of the machine learning model 210, the parameter set 245, or both may be executed at the UE 115 and the remainder of the machine learning model and / or parameter set may be executed at the neural network. Thus, the machine learning model 210, the parameter set 245, or both may use input data or at least some input parameters sent from the UE 115 to the network entity 105. In some cases, for network-assisted RFFP, the network entity 105 may observe the uplink channel from one or more TRPs and send derived channel features of the UE 115. The UE 115 may execute the neural network and use the network reported features to derive its final location. Thus, the machine learning model 210, the parameter set 245, or both may use input data or at least some input parameters sent from the network entity 105 to the UE 115. In some examples, different machine learning models and / or parameter sets associated with different techniques may have different accuracy, and the machine learning models and / or parameter sets may be required even with additional signaling for reporting the input data.

[0113] By selecting a machine learning model, parameter set, or both from a plurality of configured machine learning models, parameter sets, or both to generate an inferred output, communication between a UE 115-a and a network entity 105-a may be improved. For example, a UE 115-a may select a machine learning model and parameter set that is trained to be agnostic to particular lost or corrupted input data, which may reduce incorrect predictions and errors in the output of the machine learning model and parameter set because it may be able to manage lost or corrupted input data. In addition, the described techniques may enable a UE 115-a to perform model and / or parameter set switching and report feedback about the model and / or parameter set switching to the network entity 105-a, which may improve the accuracy and efficiency of the machine learning model and improve communication between the UE and the network entity, as well as other benefits.

[0114] Figure 3An example of a machine learning model and parameter set configuration 300 supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is illustrated. In some examples, the machine learning model and parameter set configuration 300 may implement aspects of the wireless communication systems 100 and 200, or may be implemented by aspects of the wireless communication systems 100 and 200. For example, the machine learning model and parameter set configuration 300 may include a UE 115-b and a network entity 105-b (e.g., a network node), which may be examples of corresponding devices as described herein. In some examples, the network entity 105-b may configure a machine learning model 305, a parameter set 330, or both for use by the UE 115-b, wherein each machine learning model 305 and each parameter set 330 may use input data from the UE 115-b, the network entity 105-b, or a combination thereof.

[0115] As described herein, the network entity 105-b may receive, from the model designer, model registration information for a set of machine learning models 305 and a set of parameter sets 330 trained by the model designer. Upon receiving the model registration information (e.g., configuration information for the set of machine learning models 305 and / or the set of parameter sets 330), the network entity 105-b may use conditional model and parameter set configuration techniques to configure the machine learning models 305, the parameter sets 330, or both, priorities of the machine learning models 305 and the parameter sets 330, and conditions for use of the machine learning models 305 and the parameter sets 330. For example, the network entity 105-b may configure the machine learning model 305-a, the parameter set 330-a, or both that may use data 310 measured by the UE 115-b and data 315 reported from the network entity 105-b to perform machine learning inference. Alternatively, network entity 105-b may configure machine learning model 305-b, parameter set 330-b, or both that may perform machine learning inferences using only data 310 measured and reported by UE 115-b. In some examples, network entity 105-b may configure machine learning model 305-c, parameter set 330-c, or both that may use data 310, data 315, or a combination thereof according to legacy procedures.

[0116] For example, UE 115-b may report data 310 and network entity 105-b may report data 315 for machine learning model 305-a, parameter set 330-a, or both. Machine learning model 305-a and parameter set 330-a may use data 310 and data 315 to generate output x1, which may be input into neural network 320 to generate inference output 325. In some examples, neural network 320 may include a neural network such as f(x1, ..., x N )=I1x1+...+I N xN Functions such as I1+...+I N = 1. In some examples, if the data 315 reported from the network entity 105-b is not available in a timely manner, the UE 115-b may provide the data 310 to the machine learning model 305-b, the parameter set 330-b, or both. The machine learning model 305-b, the parameter set 330-b, or both may use the data 310 to generate the output x K , which may be input into the neural network 320 to generate the inferred output 325. Alternatively, if the UE 115-b is unable to collect and generate data 310 for the machine learning model 305-b, the parameter set 330-b, or both in a timely manner, the machine learning model 305-c, the parameter set 330-c, or both may generate the output x using the data 315 reported from the network entity 105-b. N , which may be input into the neural network 320 to generate the inferred output 325. Thus, the UE 115-b may use a different machine learning model 305 and a different parameter set 330 based on which of the machine learning model 305, the parameter set 330, or both is being used by the UE 115-b, based on what input data is available to the machine learning model 305, the parameter set 330, or both, and based on other usage conditions, as described herein.

[0117] In some examples, UE 115-b may select machine learning model 305, parameter set 330, or both based on usage conditions configured by network entity 105-b for each machine learning model 305, each parameter set 330, or both. The usage conditions may include the availability of input data, which may include a list containing acceptable missing input data (e.g., input data for which machine learning model 305 and parameter set 330 may not be known) and a list containing unacceptable missing input data (e.g., input data for which no machine learning model 305 or parameter set 330 may not be known). For example, information 315 reported from network entity 105-b may be on the list of acceptable missing input data because machine learning model 305-b, parameter set 330-b, or both may operate using only data 310 from UE 115-b, whereas data 310 may be on the list of unacceptable missing input data because machine learning model 305-a and parameter set 330-a and machine learning model 305-b and parameter set 330-b may use data 310.

[0118] Additionally or alternatively, the usage conditions may indicate a usage scenario associated with the selection of the machine learning model and parameter set by UE 115-b. For example, the usage scenario may refer to whether UE 115-b is operating indoors or outdoors, or the usage scenario may refer to the altitude of UE 115-b (e.g., if UE 115-b is in a non-terrestrial network). In some examples, the usage conditions may include an area scope, which may include a cell, a public land mobile network (PLMN), a radio network controller (RNC), a carrier frequency list, etc. Additionally or alternatively, the usage conditions may include a validity timer for the machine learning model 305, the parameter set 330, or both, which may represent a duration of time during which use of a given machine learning model 305, the corresponding parameter set 330, or both may be considered valid.

[0119] In some examples, the usage conditions may include configuration information for the UE 115-b, the network entity 105-b, or both, which may affect which machine learning model 305, which parameter set 330, or both the UE 115-b selects. For example, the configuration information may include the number and spatial layout of antenna ports, the shape of the transmit and receive beams, the mobility state of the UE 115-b (e.g., low, medium, or high mobility), Doppler spread and subcarrier spacing, and other configuration information corresponding to the UE 115-b, the network entity 105-b, or both. Additionally or alternatively, the usage conditions may be performance-based. For example, the usage conditions may include a threshold of acceptable performance or a hysteresis of acceptable performance corresponding to the output of the machine learning model 305, the parameter set 330, or both. Each usage condition may be applied by the UE 115-b individually or in any combination to select which machine learning model 305 and parameter set 330 in a set of machine learning models and / or parameter sets selected by the UE 115-b.

[0120] Figure 4 An example of a process flow 400 supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is illustrated. In some examples, the process flow 400 may implement aspects of wireless communication systems 100 and 200, or may be implemented by aspects of wireless communication systems 100 and 200. For example, a UE (which may be an example of a UE 115 as described herein) may use the process flow 400 to select a machine learning model, a parameter set, or both based on one or more usage conditions corresponding to the machine learning model, the parameter set, or both.

[0121] As described herein, a model designer may train a set of machine learning models and / or parameter sets. In some examples, the machine learning models and / or parameter sets may be trained such that they may be agnostic to lost or corrupted input data. For example, a machine learning model may use input data from a UE and a network entity, while a different machine learning model may use input data from the UE. The model designer may send an indication of model registration to a network entity, which the network entity may use to configure multiple machine learning models, multiple parameter sets, or both for the UE. In addition to the machine learning models and / or parameter sets, the network entity may configure priorities for the machine learning models, parameter sets, or both, and usage conditions for selecting the machine learning models, parameter sets, or both. The network entity may send a message to the UE indicating the configured machine learning models and / or parameter sets, priorities, and usage conditions.

[0122] At 405, the UE may begin an inference occasion during which the UE will select a machine learning model, a parameter set, or both to perform machine learning inference using some input data. During the inference procedure, the machine learning model and parameter set may receive new input data and predict new outputs from the new input data based on what they were trained on (e.g., what the machine learning model and parameter set have learned).

[0123] At 410, the UE may determine whether the UE has received and successfully decoded the input data reported from the network entity. That is, the UE may evaluate a usage condition corresponding to the availability of input data that may be measured by the UE or reported by the network entity (e.g., a list of acceptable lost input data, a list of unacceptable lost input data). In some examples, the process at 410 may be based on any other usage condition, such as a usage scenario, an area scope, a configuration of the UE or the network entity, or any other usage condition.

[0124] At 415, if the UE has received and successfully decoded input data from the network entity, the UE may select a machine learning model, parameter set, or both with the highest priority for which all required input data from the UE and the network entity is available and successfully decoded. That is, the UE may select a baseline machine learning model, baseline parameter set, or both that uses input data from the UE and the network entity to generate an inferred output. The UE may perform machine learning inference using input data measured by the UE and reported by the network entity. In some examples, if the UE performs a model and / or parameter set switch and the machine learning model and / or parameter set used for machine learning difference is different from that selected by the UE, the UE may send a message to the network entity indicating an ID associated with the machine learning model, parameter set, or both used for machine learning inference. The UE may send a message indicating an ID associated with the machine learning model, parameter set, or both at the beginning of each inference opportunity, or the UE may send the message as part of a routine performance measurement.

[0125] Alternatively, at 420, if the UE fails to receive input data from the network entity, or if the UE receives input data from the network entity but fails to successfully decode the input data, the UE may select an alternative machine learning model, an alternative machine learning model, or both with the highest priority, for which all required input data is available at the UE. That is, the UE may select an alternative machine learning model, an alternative parameter set, or both that uses only input data measured by the UE. The UE may perform machine learning inference using input data measured by the UE. In some examples, if the UE performs model and / or parameter set switching and the machine learning model and parameter set used for machine learning difference are different from those selected by the UE, the UE may send a message to the network entity indicating an ID associated with the machine learning model, parameter set, or both used for machine learning inference.

[0126] Figure 5An example of a process flow 500 that supports conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is illustrated. The process flow 500 may implement aspects of the wireless communication systems 100 and 200, or may be implemented by aspects of the wireless communication systems 100 and 200. For example, the process flow 500 may illustrate operations between a UE 115-c and a network entity 105-c (e.g., a network node), which may be examples of corresponding devices described herein. In the following description of the process flow 500, operations between the UE 115-c and the network entity 105-c may be sent in an order different from the example order shown, or operations performed by the UE 115-c and the network entity 105-c may be performed in a different order or at a different time. Some operations may also be omitted from the process flow 500, and other operations may be added to the process flow 500.

[0127] At 505, the UE 115-c may receive a first message from the network entity 105-c indicating a set of machine learning models, a set of parameter sets, or both, and indicating one or more usage conditions for selecting one of the machine learning models, one of the parameter sets, or both for performing machine learning inference. In some examples, the usage conditions may include availability of input data, usage scenarios, and regional scope, among other usage conditions.

[0128] At 510, the UE 115-c may select a first machine learning model, a first set of parameters, or both for performing machine learning inference based on a usage condition in one or more usage conditions being satisfied. For example, if the UE 115-c receives input data from the network entity 105-c and the UE 115-c measures its own input data, the UE 115-c may select a first machine learning model, a first set of parameters, or both that may use input data from both the UE 115-c and the network entity 105-c. Alternatively, if the UE 115-c fails to receive input data from the network entity 105-c, or if the UE 115-c receives input data from the network entity 105-c but fails to successfully decode the input data, the UE 115-c may select a first machine learning model, a first set of parameters, or both that may use input data from only the UE 115-c.

[0129] At 515, UE 115-c may switch the first machine learning model, the first parameter set, or both to a second machine learning model, the second parameter set, or both, for example, based on satisfying different usage conditions for different inference occasions.

[0130] At 520, the UE 115-c may use the first machine learning model, the first set of parameters, or both to generate machine learning inferences using input data received from the UE 115-c, input data measured by the UE 115-c, or both. Thus, the UE 115-c may use the first machine learning model, the first set of parameters, or both to generate machine learning inferences based on the first machine learning model, the first set of parameters, or both satisfying a usage condition that the first machine learning model, the first set of parameters, or both may use input data from the UE 115-c and the network entity 105-c.

[0131] At 525, UE 115-c may send a report to network entity 105-c indicating machine learning inferences generated using the first machine learning model or the first parameter set, an ID of the first machine learning model, the first parameter set, or both, an indication of why the first machine learning model, the first parameter set, or both were selected (e.g., if the first machine learning model, the first parameter set, or both were not known to the specific date that UE 115-a was lost when selecting between a set of machine learning models, a set of parameter sets, or both), or any combination thereof.

[0132] At 530, the UE 115-c may receive a feedback message from the network entity 105-c indicating one or more statistics associated with the UE 115-c switching between selecting a first machine learning model, a first parameter set, or both and a second machine learning model, a second parameter set, or both. For example, the feedback message may indicate a rate at which the UE 115-c performs model and / or parameter set switching and a reason for the model and / or parameter set switching.

[0133] Figure 6 A block diagram 600 of a device 605 supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown. The device 605 may be an example of aspects of a UE 115 as described herein. The device 605 may include a receiver 610, a transmitter 615, and a communication manager 620. The device 605 may also include a processor. Each of these components may communicate with each other (e.g., via one or more buses).

[0134] The receiver 610 may provide means for receiving information (such as packets, user data, control information, or any combination thereof) associated with various information channels (e.g., control channels, data channels, information channels related to conditional AI, machine learning models, and parameter set configurations). The information may be passed to other components of the device 605. The receiver 610 may utilize a single antenna or a collection of multiple antennas.

[0135] Transmitter 615 may provide means for transmitting signals generated by other components of device 605. For example, transmitter 615 may transmit information (such as packets, user data, control information, or any combination thereof) associated with various information channels (e.g., control channels, data channels, information channels related to conditional AI, machine learning models, and parameter set configurations). In some examples, transmitter 615 may be co-located with receiver 610 in a transceiver module. Transmitter 615 may utilize a single antenna or a collection of multiple antennas.

[0136] The communication manager 620, the receiver 610, the transmitter 615, or various combinations thereof or various components thereof may be examples of means for performing various aspects of conditional AI, machine learning models, and parameter set configurations as described herein. For example, the communication manager 620, the receiver 610, the transmitter 615, or various combinations thereof or components thereof may support methods for performing one or more of the functions described herein.

[0137] In some examples, the communication manager 620, the receiver 610, the transmitter 615, or various combinations or components thereof may be implemented in hardware (e.g., in a communication management circuit). The hardware may include a processor, a digital signal processor (DSP), a central processing unit (CPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, a microcontroller, discrete gate or transistor logic, discrete hardware components, or any combination thereof that is configured to or otherwise supports components for performing the functions described in the present disclosure. In some examples, the processor and a memory coupled to the processor may be configured to perform one or more of the functions described herein (e.g., by executing instructions stored in the memory by the processor).

[0138] Additionally or alternatively, in some examples, the communication manager 620, the receiver 610, the transmitter 615, or various combinations or components thereof may be implemented in code executed by a processor (e.g., as communication management software or firmware). If implemented in code executed by a processor, the functionality of the communication manager 620, the receiver 610, the transmitter 615, or various combinations or components thereof may be performed by a general purpose processor (e.g., configured as or otherwise supporting means for performing the functions described in the present disclosure), a DSP, a CPU, an ASIC, an FPGA, a microcontroller, or any combination of these or other programmable logic devices.

[0139] In some examples, communication manager 620 can be configured to perform various operations (e.g., receive, obtain, monitor, output, send) using or otherwise cooperating with receiver 610, transmitter 615, or both. For example, communication manager 620 can receive information from receiver 610, transmit information to transmitter 615, or be integrated in conjunction with receiver 610, transmitter 615, or both to obtain information, output information, or perform various other operations as described herein.

[0140] According to examples as disclosed herein, the communication manager 620 may support wireless communications at a UE. For example, the communication manager 620 may be configured to or otherwise support a component for receiving a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both and indicating one or more usage conditions for selecting one of the sets of multiple machine learning models, one of the sets of multiple parameter sets, or both for performing machine learning inference. The communication manager 620 may be configured to or otherwise support a component for selecting a first machine learning model, a first parameter set, or both for performing machine learning inference based on satisfaction of a usage condition in one or more usage conditions. The communication manager 620 may be configured to or otherwise support a component for sending a report indicating machine learning inferences generated by the first machine learning model or by using the first parameter set.

[0141] By including or configuring the communication manager 620 according to examples as described herein, the device 605 (e.g., a processor controlling or otherwise coupled to the receiver 610, the transmitter 615, the communication manager 620, or a combination thereof) may support techniques for selecting a machine learning model, a parameter set, or both based on usage conditions, which techniques may reduce errors and mispredictions in the machine learning model outputs and improve machine learning inference results.

[0142] Figure 7 A block diagram 700 of a device 705 supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown. The device 705 may be an example of aspects of the device 605 or UE 115 as described herein. The device 705 may include a receiver 710, a transmitter 715, and a communication manager 720. The device 705 may also include a processor. Each of these components may communicate with each other (e.g., via one or more buses).

[0143] The receiver 710 may provide means for receiving information (such as packets, user data, control information, or any combination thereof) associated with various information channels (e.g., control channels, data channels, information channels related to conditional AI, machine learning models, and parameter set configurations). The information may be passed to other components of the device 705. The receiver 710 may utilize a single antenna or a collection of multiple antennas.

[0144] The transmitter 715 may provide means for transmitting signals generated by other components of the device 705. For example, the transmitter 715 may transmit information (such as packets, user data, control information, or any combination thereof) associated with various information channels (e.g., control channels, data channels, information channels related to conditional AI, machine learning models, and parameter set configurations). In some examples, the transmitter 715 may be co-located with the receiver 710 in a transceiver module. The transmitter 715 may utilize a single antenna or a collection of multiple antennas.

[0145] Device 705 or its various components may be examples of components for performing various aspects of conditional AI, machine learning models, and parameter set configurations as described herein. For example, communication manager 720 may include message receiving component 725, selection component 730, report sending component 735, or any combination thereof. Communication manager 720 may be an example of various aspects of communication manager 620 as described herein. In some examples, communication manager 720 or its various components may be configured to perform various operations (e.g., receive, obtain, monitor, output, send) using or otherwise in conjunction with receiver 710, transmitter 715, or both. For example, communication manager 720 may receive information from receiver 710, transmit information to transmitter 715, or integrate with receiver 710, transmitter 715, or both to obtain information, output information, or perform various other operations as described herein.

[0146] According to examples as disclosed herein, the communication manager 720 may support wireless communications at a UE. The message receiving component 725 may be configured to or otherwise support a component for receiving a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the sets of multiple machine learning models, one of the sets of multiple parameter sets, or both for performing machine learning inference. The model selection component 730 may be configured to or otherwise support a component for selecting a first machine learning model in a set of multiple machine learning models, a first parameter set in a set of multiple parameter sets, or both for performing machine learning inference based on satisfaction of a usage condition in one or more usage conditions. The report sending component 735 may be configured to or otherwise support a component for sending a report indicating a machine learning inference generated by the first machine learning model or by using the first parameter set.

[0147] Figure 8 A block diagram 800 of a communication manager 820 supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown. The communication manager 820 can be an example of aspects of the communication manager 620, the communication manager 720, or both as described herein. The communication manager 820 or its various components can be examples of components for performing various aspects of conditional AI, machine learning models, and parameter set configurations as described herein. For example, the communication manager 820 may include a message receiving component 825, a selection component 830, a report sending component 835, a priority component 840, an input data component 845, an inference generation component 850, a feedback component 855, a usage condition component 860, or any combination thereof. Each of these components can communicate directly or indirectly with each other (e.g., via one or more buses).

[0148] According to examples as disclosed herein, the communication manager 820 may support wireless communications at a UE. The message receiving component 825 may be configured to or otherwise support a component for receiving a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the sets of multiple machine learning models, one of the sets of multiple parameter sets, or both for performing machine learning inference. The model selection component 830 may be configured to or otherwise support a component for selecting a first machine learning model in a set of multiple machine learning models, a first parameter set in a set of multiple parameter sets, or both for performing machine learning inference based on satisfaction of a usage condition in one or more usage conditions. The report sending component 835 may be configured to or otherwise support a component for sending a report indicating a machine learning inference generated by the first machine learning model or by using the first parameter set.

[0149] In some examples, the prioritization component 840 can be configured or otherwise support means for receiving a first message indicating a set of multiple priorities corresponding to a set of multiple machine learning models, a set of multiple parameter sets, or both. In some examples, the selection component 830 can be configured or otherwise support means for selecting a first machine learning model, a first parameter set, or both for performing machine learning inference based on the first machine learning model, the first parameter set, or both being associated with a highest priority in the set of multiple priorities.

[0150] In some examples, the priority component 840 may be configured to or otherwise support a component for receiving a first message that instructs the network node to change the priority of one or more of a set of multiple machine learning models, one or more of a set of multiple parameter sets, or both based on one or more performance indicators reported by the UE, wherein the first machine learning model, the first parameter set, or both are selected based on the first message.

[0151] In some examples, input data component 845 can be configured or otherwise support components for selecting a first machine learning model, a first parameter set, or both for performing machine learning inferences based on a usage condition indicating that the first machine learning model, the first parameter set, or both are selected when the UE is able to successfully receive input data from the network node. In some examples, inference generation component 850 can be configured or otherwise support components for generating machine learning inferences using input data received from the network node by the first machine learning model or by using the first parameter set.

[0152] In some examples, the input data component 845 may be configured as or otherwise support a component for selecting a first machine learning model, a first parameter set, or both for performing machine learning inference based on a usage condition indicating that the first machine learning model, the first parameter set, or both are selected when the UE fails to receive input data for a second machine learning model in a set of multiple machine learning models, a second parameter set in a set of multiple parameter sets, or both from a network node. In some examples, the report sending component 835 may be configured as or otherwise support a component for sending a report indicating machine learning inference and indicating the first machine learning model, the first parameter set, or both.

[0153] In some examples, the inference generation component 850 may be configured to or otherwise support components for generating machine learning inferences using input data measured by the UE by a first machine learning model or by using a first parameter set based on a use condition indicating selection of a first machine learning model, a first parameter set, or both when the UE measures the input data.

[0154] In some examples, feedback component 855 may be configured to or otherwise support components for sending a first feedback message indicating one or more statistics associated with switching between selecting a set of multiple machine learning models, a set of multiple parameter sets, or both.

[0155] In some examples, feedback component 855 may be configured to or otherwise support means for receiving a second feedback message indicating one or more statistics associated with switching between selecting a set of multiple machine learning models, a set of multiple parameter sets, or both.

[0156] In some examples, the usage condition component 860 may be configured to or otherwise support a component for receiving a first message or a second message indicating one or more usage conditions for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both, wherein the one or more usage conditions include at least one of the following: instructions for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both when input data is missing, or priorities corresponding to each machine learning model in the set of multiple machine learning models, each parameter set in the set of multiple parameter sets, or both, for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both.

[0157] In some examples, instructions for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both include availability of input data, usage scenarios, regional scope, model validity timer, parameter set validity timer, configuration of the UE, configuration of the network node, performance threshold, or any combination thereof.

[0158] In some examples, to support receiving a first message, the message receiving component 825 may be configured as or otherwise support a component for receiving a first message or a second message indicating configuration information for a set of multiple machine learning models, a set of multiple parameter sets, or both, wherein the configuration information indicates a set of alternative machine learning models in the set of multiple machine learning models, a set of alternative parameter sets in the set of multiple parameter sets, one or more usage conditions, a priority corresponding to each machine learning model, a priority corresponding to each machine learning model, or any combination thereof.

[0159] Fig. 9 A diagram 900 of a system including a device 905 supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown. The device 905 may be an example of a device 605, a device 705, or a UE 115 as described herein, or include components thereof. The device 905 may communicate (e.g., wirelessly) with one or more network entities 105, one or more UEs 115, or any combination thereof. The device 905 may include components for two-way voice and data communications, including components for sending and receiving communications, such as a communication manager 920, an input / output (I / O) controller 910, a transceiver 915, an antenna 925, a memory 930, a code 935, and a processor 940. These components may be electronically communicated or otherwise coupled (e.g., operatively, communicatively, functionally, electronically, electrically) via one or more buses (e.g., a bus 945).

[0160] I / O controller 910 can manage input signals and output signals of device 905. I / O controller 910 can also manage peripheral devices that are not integrated into device 905. In some cases, I / O controller 910 can represent a physical connection or port to an external peripheral device. In some cases, I / O controller 910 can utilize an operating system, such as or another known operating system. Additionally or alternatively, I / O controller 910 may represent or interact with a modem, keyboard, mouse, touch screen, or similar device. In some cases, I / O controller 910 may be implemented as part of a processor (such as processor 940). In some cases, a user may interact with device 905 via I / O controller 910 or via hardware components controlled by I / O controller 910.

[0161] In some cases, the device 905 may include a single antenna 925. However, in some other cases, the device 905 may have more than one antenna 925, which may be capable of concurrently transmitting or receiving multiple wireless transmissions. The transceiver 915 may communicate bidirectionally via one or more antennas 925, a wired or wireless link, as described herein. For example, the transceiver 915 may represent a wireless transceiver and may communicate bidirectionally with another wireless transceiver. The transceiver 915 may also include a modem for: modulating packets; providing the modulated packets to one or more antennas 925 for transmission; and demodulating packets received from one or more antennas 925. The transceiver 915 or the transceiver 915 and one or more antennas 925 may be examples of transmitters 615, transmitters 715, receivers 610, receivers 710, or any combination thereof or components thereof as described herein.

[0162] The memory 930 may include random access memory (RAM) and read-only memory (ROM). The memory 930 may store computer-readable, computer-executable code 935 including instructions that, when executed by the processor 940, cause the device 905 to perform various functions described herein. The code 935 may be stored in a non-transitory computer-readable medium, such as a system memory or another type of memory. In some cases, the code 935 may not be directly executed by the processor 940, but may (e.g., when compiled and executed) cause the computer to perform the functions described herein. In some cases, the memory 930 may also contain, among other things, a basic I / O system (BIOS), which may control basic hardware or software operations, such as interaction with peripheral components or devices.

[0163] The processor 940 may include an intelligent hardware device (e.g., a general-purpose processor, a DSP, a CPU, a microcontroller, an ASIC, an FPGA, a programmable logic device, a discrete gate or transistor logic component, a discrete hardware component, or any combination thereof). In some cases, the processor 940 may be configured to operate a memory array using a memory controller. In some other cases, the memory controller may be integrated into the processor 940. The processor 940 may be configured to execute computer-readable instructions stored in a memory (e.g., a memory 930) to enable the device 905 to perform various functions (e.g., functions or tasks that support conditional AI, machine learning models, and parameter set configurations). For example, the device 905 or a component of the device 905 may include a processor 940 and a memory 930 coupled to or coupled to the processor 940, and the processor 940 and the memory 930 are configured to perform the various functions described herein.

[0164] According to examples as disclosed herein, the communication manager 920 may support wireless communication at a UE. For example, the communication manager 920 may be configured to or otherwise support a component for receiving a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the sets of multiple machine learning models, one of the sets of multiple parameter sets, or both for performing machine learning inference. The communication manager 920 may be configured to or otherwise support a component for selecting a first machine learning model in a set of multiple machine learning models, a first parameter set in a set of multiple parameter sets for performing machine learning inference based on satisfaction of a usage condition in one or more usage conditions. The communication manager 920 may be configured to or otherwise support a component for sending a report indicating machine learning inferences generated by the first machine learning model or by using the first parameter set.

[0165] By including or configuring the communication manager 920 according to examples as described herein, the device 905 may support techniques for selecting machine learning models, parameter sets, or both based on usage conditions that may reduce errors and mispredictions in machine learning model outputs and improve machine learning inference results.

[0166] In some examples, the communication manager 920 may be configured to perform various operations (e.g., receive, monitor, transmit) using or otherwise cooperating with the transceiver 915, one or more antennas 925, or any combination thereof. Although the communication manager 920 is illustrated as a separate component, in some examples, one or more functions described with reference to the communication manager 920 may be supported or performed by the processor 940, the memory 930, the code 935, or any combination thereof. For example, the code 935 may include instructions that may be executed by the processor 940 to cause the device 905 to perform various aspects of the conditional AI, machine learning models, and parameter set configurations as described herein, or the processor 940 and the memory 930 may be otherwise configured to perform or support such operations.

[0167] Fig.10 A block diagram 1000 of a device 1005 supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown. The device 1005 may be an example of aspects of a network node as described herein. The device 1005 may include a receiver 1010, a transmitter 1015, and a communication manager 1020. The device 1005 may also include a processor. Each of these components may communicate with each other (e.g., via one or more buses).

[0168] The receiver 1010 may provide means for obtaining (e.g., receiving, determining, identifying) information associated with various channels (e.g., control channels, data channels, information channels, channels associated with a protocol stack), such as user data, control information, or any combination thereof (e.g., I / Q samples, symbols, packets, protocol data units, service data units). The information may be passed to other components of the device 1005. In some examples, the receiver 1010 may support obtaining information by receiving signals via one or more antennas. Additionally or alternatively, the receiver 1010 may support obtaining information by receiving signals via one or more wired (e.g., electrical, optical) interfaces, wireless interfaces, or any combination thereof.

[0169] The transmitter 1015 may provide means for outputting (e.g., transmitting, providing, conveying, transmitting) information generated by other components of the device 1005. For example, the transmitter 1015 may output information associated with various channels (e.g., control channels, data channels, information channels, channels associated with a protocol stack) such as user data, control information, or any combination thereof (e.g., I / Q samples, symbols, packets, protocol data units, service data units). In some examples, the transmitter 1015 may support outputting information by sending signals via one or more antennas. Additionally or alternatively, the transmitter 1015 may support outputting information by sending signals via one or more wired (e.g., electrical, optical) interfaces, wireless interfaces, or any combination thereof. In some examples, the transmitter 1015 and the receiver 1010 may be co-located in a transceiver, which may include a modem or be coupled to a modem.

[0170] The communication manager 1020, the receiver 1010, the transmitter 1015, or various combinations thereof or various components thereof may be examples of means for performing various aspects of conditional AI, machine learning models, and parameter set configurations as described herein. For example, the communication manager 1020, the receiver 1010, the transmitter 1015, or various combinations thereof or components thereof may support methods for performing one or more of the functions described herein.

[0171] In some examples, the communication manager 1020, the receiver 1010, the transmitter 1015, or various combinations or components thereof may be implemented in hardware (e.g., in a communication management circuit). The hardware may include a processor, a DSP, a CPU, an ASIC, an FPGA or other programmable logic device, a microcontroller, discrete gate or transistor logic, discrete hardware components, or any combination thereof configured as or otherwise supporting means for performing the functions described in the present disclosure. In some examples, a processor and a memory coupled to the processor may be configured to perform one or more of the functions described herein (e.g., by executing instructions stored in the memory by the processor).

[0172] Additionally or alternatively, in some examples, the communication manager 1020, the receiver 1010, the transmitter 1015, or various combinations or components thereof may be implemented in code executed by a processor (e.g., as communication management software or firmware). If implemented in code executed by a processor, the functionality of the communication manager 1020, the receiver 1010, the transmitter 1015, or various combinations or components thereof may be performed by a general purpose processor (e.g., configured as or otherwise supporting means for performing the functions described in the present disclosure), a DSP, a CPU, an ASIC, an FPGA, a microcontroller, or any combination of these or other programmable logic devices.

[0173] In some examples, communication manager 1020 may be configured to perform various operations (e.g., receive, obtain, monitor, output, send) using or otherwise cooperating with receiver 1010, transmitter 1015, or both. For example, communication manager 1020 may receive information from receiver 1010, transmit information to transmitter 1015, or be integrated in conjunction with receiver 1010, transmitter 1015, or both to obtain information, output information, or perform various other operations as described herein.

[0174] The communication manager 1020 may support wireless communications at a network node according to examples as disclosed herein. For example, the communication manager 1020 may be configured to or otherwise support a component for sending a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the sets of multiple machine learning models, one of the sets of multiple parameter sets, or both for performing machine learning inference. The communication manager 1020 may be configured to or otherwise support a component for receiving a report indicating machine learning inferences generated by a first machine learning model or by using a first parameter set, the first machine learning model, the first parameter set being selected by the UE based on a usage condition in one or more usage conditions being satisfied.

[0175] By including or configuring the communication manager 1020 according to the examples described herein, the device 1005 (e.g., a processor controlling or otherwise coupled with the receiver 1010, the transmitter 1015, the communication manager 1020, or a combination thereof) may support techniques for selecting a machine learning model, a parameter set, or both based on usage conditions, which techniques may reduce errors and mispredictions in the machine learning model outputs and improve machine learning inference results.

[0176] Fig.11A block diagram 1100 of a device 1105 supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown. The device 1105 may be an example of aspects of the device 1005 or network node 105 as described herein. The device 1105 may include a receiver 1110, a transmitter 1115, and a communication manager 1120. The device 1105 may also include a processor. Each of these components may communicate with each other (e.g., via one or more buses).

[0177] The receiver 1110 may provide means for obtaining (e.g., receiving, determining, identifying) information associated with various channels (e.g., control channels, data channels, information channels, channels associated with a protocol stack), such as user data, control information, or any combination thereof (e.g., I / Q samples, symbols, packets, protocol data units, service data units). The information may be passed to other components of the device 1105. In some examples, the receiver 1110 may support obtaining information by receiving signals via one or more antennas. Additionally or alternatively, the receiver 1110 may support obtaining information by receiving signals via one or more wired (e.g., electrical, optical) interfaces, wireless interfaces, or any combination thereof.

[0178] The transmitter 1115 may provide means for outputting (e.g., transmitting, providing, conveying, transmitting) information generated by other components of the device 1105. For example, the transmitter 1115 may output information associated with various channels (e.g., control channels, data channels, information channels, channels associated with a protocol stack) such as user data, control information, or any combination thereof (e.g., I / Q samples, symbols, packets, protocol data units, service data units). In some examples, the transmitter 1115 may support outputting information by sending signals via one or more antennas. Additionally or alternatively, the transmitter 1115 may support outputting information by sending signals via one or more wired (e.g., electrical, optical) interfaces, wireless interfaces, or any combination thereof. In some examples, the transmitter 1115 and the receiver 1110 may be co-located in a transceiver, which may include a modem or be coupled to a modem.

[0179] Device 1105 or its various components may be examples of components for performing various aspects of conditional AI, machine learning models, and parameter set configurations as described herein. For example, communication manager 1120 may include message sending component 1125, report receiving component 1130, or any combination thereof. Communication manager 1120 may be an example of various aspects of communication manager 1020 as described herein. In some examples, communication manager 1120 or its various components may be configured to perform various operations (e.g., receive, obtain, monitor, output, send) using or otherwise in conjunction with receiver 1110, transmitter 1115, or both. For example, communication manager 1120 may receive information from receiver 1110, transmit information to transmitter 1115, or integrate with receiver 1110, transmitter 1115, or both to obtain information, output information, or perform various other operations as described herein.

[0180] The communication manager 1120 may support wireless communications at a network node according to examples as disclosed herein. The message sending component 1125 may be configured to or otherwise support means for sending a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the sets of multiple machine learning models, one of the sets of multiple parameter sets, or both for performing machine learning inferences. The report receiving component 1130 may be configured to or otherwise support means for receiving a report indicating machine learning inferences generated by a first machine learning model or by using a first parameter set, the first machine learning model, the first parameter set being selected by the UE based on a usage condition in the one or more usage conditions being satisfied.

[0181] Fig.12 A block diagram 1200 of a communication manager 1220 supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown. The communication manager 1220 can be an example of aspects of the communication manager 1020, the communication manager 1120, or both as described herein. The communication manager 1220 or its various components can be examples of components for performing various aspects of conditional AI, machine learning models, and parameter set configurations as described herein. For example, the communication manager 1220 may include a message sending component 1225, a report receiving component 1230, a model and parameter set priority component 1235, a feedback sending component 1240, a model and parameter set component 1245, a configuration information sending component 1250, or any combination thereof. Each of these components can communicate with each other directly or indirectly (e.g., via one or more buses).

[0182] The communication manager 1220 may support wireless communications at a network node according to examples as disclosed herein. The message sending component 1225 may be configured to or otherwise support means for sending a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the sets of multiple machine learning models, one of the sets of multiple parameter sets, or both for performing machine learning inferences. The report receiving component 1230 may be configured to or otherwise support means for receiving a report indicating machine learning inferences generated by a first machine learning model or by using a first parameter set, the first machine learning model, the first parameter set being selected by the UE based on a usage condition in the one or more usage conditions being satisfied.

[0183] In some examples, model and parameter set priority component 1235 may be configured as or otherwise support a component for sending a first message indicating a set of multiple priorities corresponding to a set of multiple machine learning models, a set of multiple parameter sets, or both.

[0184] In some examples, the model and parameter set priority component 1235 may be configured as or otherwise support a component for sending a first message instructing the network node to change the priority of one or more of a set of multiple machine learning models, one or more of a set of multiple parameter sets, or both based on one or more performance indicators reported by the UE, wherein the first machine learning model, the first parameter set, or both are selected based on the first message.

[0185] In some examples, report receiving component 1230 may be configured to or otherwise support components for receiving a report indicating machine learning inferences and indicating that the UE selected a first machine learning model, a first set of parameters, or both to generate the machine learning inferences.

[0186] In some examples, feedback sending component 1240 may be configured to or otherwise support components for sending a feedback message indicating one or more statistics associated with switching between a set of multiple machine learning models, a set of multiple parameter sets, or both.

[0187] In some examples, the model and parameter set component 1245 may be configured as or otherwise support a component for sending a first message or a second message indicating one or more usage conditions for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both, wherein the one or more usage conditions include at least one of the following: instructions for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets when input data is missing, or a priority corresponding to each machine learning model in the set of multiple machine learning models, each parameter set in the set of multiple parameter sets, for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both.

[0188] In some examples, instructions for selecting one of a set of multiple machine learning models, one of a set of multiple parameter sets, or both include availability of input data, usage scenarios, regional scope, model validity timer, parameter set validity timer, configuration of the UE, configuration of the network node, performance threshold, or any combination thereof.

[0189] In some examples, to support sending a first message, the configuration information sending component 1250 may be configured as or otherwise support a component for sending a first message or a second message, wherein the first message or the second message indicates configuration information for a set of multiple machine learning models, a set of multiple parameter sets, or both, wherein the configuration information indicates a set of alternative machine learning models in the set of multiple machine learning models, a set of alternative machine learning models in the set of multiple parameter sets, or both, one or more usage conditions, a priority corresponding to each machine learning model, a priority corresponding to each parameter set, or any combination thereof.

[0190] Fig.13 A diagram 1300 of a system including a device 1305 supporting conditional AI, machine learning models, and parameter set configurations according to one or more aspects of the present disclosure is shown. The device 1305 may be an example of a device 1005, a device 1105, or a network node as described herein or a component including these devices. The device 1305 may include components for two-way voice and data communications, including components for sending and receiving communications, such as a communication manager 1320, a transceiver 1310, an antenna 1315, a memory 1325, a code 1330, and a processor 1335. These components may be in electronic communication or otherwise coupled (e.g., operatively, communicatively, functionally, electronically, electrically) via one or more buses (e.g., bus 1340).

[0191] The transceiver 1310 may support bidirectional communication via a wired link, a wireless link, or both as described herein. In some examples, the transceiver 1310 may include a wired transceiver and may communicate bidirectionally with another wired transceiver. Additionally or alternatively, in some examples, the transceiver 1310 may include a wireless transceiver and may communicate bidirectionally with another wireless transceiver. In some examples, the device 1305 may include one or more antennas 1315, which may be capable of sending or receiving wireless transmissions (e.g., concurrently). The transceiver 1310 may also include a modem for modulating a signal, thereby providing a modulated signal for transmission (e.g., through one or more antennas 1315, through a wired transmitter), for receiving a modulated signal (e.g., from one or more antennas 1315, from a wired receiver), and for demodulating a signal. The transceiver 1310 or transceiver 1310 and one or more antennas 1315 or wired interfaces (where applicable) may be examples of transmitter 1015, transmitter 1115, receiver 1010, receiver 1110, or any combination thereof or components thereof as described herein. In some examples, the transceiver may be operable to support communication via one or more communication links (e.g., communication link 125, backhaul communication link 120, midhaul communication link 162, fronthaul communication link 168).

[0192] Memory 1325 may include RAM and ROM. Memory 1325 may store computer-readable, computer-executable code 1330 including instructions that, when executed by processor 1335, cause device 1305 to perform various functions described herein. Code 1330 may be stored in a non-transitory computer-readable medium, such as system memory or another type of memory. In some cases, code 1330 may not be directly executable by processor 1335, but may (e.g., when compiled and executed) cause a computer to perform the functions described herein. In some cases, in addition to this, memory 1325 may also include BIOS, which may control basic hardware or software operations, such as interaction with peripheral components or devices.

[0193] Processor 1335 may include an intelligent hardware device (e.g., a general-purpose processor, DSP, ASIC, CPU, FPGA, microcontroller, programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof). In some cases, processor 1335 may be configured to operate a memory array using a memory controller. In some other cases, the memory controller may be integrated into processor 1335. Processor 1335 may be configured to execute computer-readable instructions stored in a memory (e.g., memory 1325) to enable device 1305 to perform various functions (e.g., functions or tasks that support conditional AI, machine learning models, and parameter set configurations). For example, device 1305 or a component of device 1305 may include processor 1335 and memory 1325 coupled to processor 1335, the processor 1335 and memory 1325 being configured to perform the various functions described herein. Processor 1335 can be an example of a cloud computing platform (e.g., one or more physical nodes and supporting software such as an operating system, virtual machine, or container instance) that can host functionality for performing functions of device 1305 (e.g., by executing code 1330).

[0194] In some examples, bus 1340 may support communications of protocol layers (e.g., within a protocol layer) of a protocol stack. In some examples, bus 1340 may support communications associated with logical channels of a protocol stack (e.g., between protocol layers of a protocol stack), which may include communications performed within components of device 1305, or communications performed between different components of device 1305 that may be co-located or may be located in different locations (e.g., where device 1305 may refer to a system in which one or more of communication manager 1320, transceiver 1310, memory 1325, code 1330, and processor 1335 may be located in one of the different components or divided between the different components).

[0195] In some examples, communication manager 1320 may manage aspects of communications with core network 130 (e.g., via one or more wired or wireless backhaul links). For example, communication manager 1320 may manage delivery of data communications for client devices such as one or more UEs 115. In some examples, communication manager 1320 may manage communications with other network entities 105 and may include a controller or scheduler for controlling communications with UE 115 in collaboration with other network entities 105. In some examples, communication manager 1320 may support an X2 interface within an LTE / LTE-A wireless communication network technology to provide communications between network entities 105.

[0196] The communication manager 1320 may support wireless communications at a network node according to examples as disclosed herein. For example, the communication manager 1320 may be configured to or otherwise support a component for sending a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the sets of multiple machine learning models, one of the sets of multiple parameter sets, or both for performing machine learning inferences. The communication manager 1320 may be configured to or otherwise support a component for receiving a report indicating machine learning inferences generated by a first machine learning model or by using a first parameter set, the first machine learning model, the first parameter set being selected by the UE based on a usage condition in one or more usage conditions being satisfied.

[0197] By including or configuring the communication manager 1320 according to examples as described herein, the device 1305 may support techniques for selecting machine learning models, parameter sets, or both based on usage conditions that may reduce errors and mispredictions in machine learning model outputs and improve machine learning inference results.

[0198] In some examples, the communication manager 1320 may be configured to perform various operations (e.g., receive, obtain, monitor, output, transmit) using or otherwise cooperating with the transceiver 1310, one or more antennas 1315 (e.g., where applicable), or any combination thereof. Although the communication manager 1320 is illustrated as a separate component, in some examples, one or more functions described with reference to the communication manager 1320 may be supported or performed by the processor 1335, the memory 1325, the code 1330, the transceiver 1310, or any combination thereof. For example, the code 1330 may include instructions that may be executed by the processor 1335 to cause the device 1305 to perform various aspects of the conditional AI, machine learning models, and parameter set configurations as described herein, or the processor 1335 and the memory 1325 may be otherwise configured to perform or support such operations.

[0199] Fig.14 A flowchart illustrating a method 1400 for supporting conditional AI, machine learning models, and parameter set configuration according to one or more aspects of the present disclosure is shown. The operations of the method 1400 may be implemented by a UE or a component thereof as described herein. For example, the operations of the method 1400 may be implemented by a UE or a component thereof as described in reference to Figures 1 to 9 The UE 115 described herein may be executed. In some examples, the UE may execute an instruction set to control the functional elements of the UE to perform the described functions. Additionally or alternatively, the UE may use dedicated hardware to perform various aspects of the described functions.

[0200] At 1405, the method may include receiving a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the set of multiple machine learning models, one of the set of multiple parameter sets, or both for performing machine learning inference. The operations of 1405 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1405 may be performed as described in reference to Figure 8 The message receiving component 825 is executed.

[0201] At 1410, the method may include selecting a first machine learning model in a set of multiple machine learning models, a first parameter set in a set of multiple parameter sets, or both for performing machine learning inference based on the satisfaction of a usage condition in one or more usage conditions. The operations of 1410 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1410 may be performed as described in reference to Figure 8 The selection component 830 is executed.

[0202] At 1415, the method may include sending a report indicating machine learning inferences generated by the first machine learning model or by using the first parameter set. The operations of 1415 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1415 may be performed as described in reference to Figure 8 The report sending component 835 is used to execute.

[0203] Fig.15 A flowchart illustrating a method 1500 for supporting conditional AI, machine learning models, and parameter set configuration according to one or more aspects of the present disclosure is shown. The operations of the method 1500 may be implemented by a UE or a component thereof as described herein. For example, the operations of the method 1500 may be implemented by a UE or a component thereof as described in reference to Figures 1 to 9 The UE 115 described herein may be executed. In some examples, the UE may execute an instruction set to control the functional elements of the UE to perform the described functions. Additionally or alternatively, the UE may use dedicated hardware to perform various aspects of the described functions.

[0204] At 1505, the method may include receiving a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the set of multiple machine learning models, one of the set of multiple parameter sets, or both for performing machine learning inference. The operations of 1505 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1505 may be performed as described in reference to Figure 8 The message receiving component 825 is executed.

[0205] At 1510, the method may include selecting a first machine learning model, a first parameter set, or both for performing machine learning inference based on one or more usage conditions indicating that the first machine learning model, a usage condition of the first machine learning model, or both are selected when the UE is able to successfully receive input data from the network node. The operations of 1510 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1510 may be performed as described in reference to Figure 8 The selection component 830 is executed.

[0206] At 1515, the method may include generating machine learning inferences using input data received from the network node by a first machine learning model or by using a first parameter set. The operations of 1515 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1515 may be performed as described in reference to Figure 8 The inference generation component 850 is used to perform.

[0207] At 1520, the method may include sending a report indicating machine learning inferences generated by the first machine learning model or by using the first parameter set. The operations of 1520 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1520 may be performed as described in reference to Figure 8 The report sending component 835 is used to execute.

[0208] Fig.16 A flowchart illustrating a method 1600 for supporting conditional AI, machine learning models, and parameter set configuration according to one or more aspects of the present disclosure is shown. The operations of the method 1600 may be implemented by a UE or a component thereof as described herein. For example, the operations of the method 1600 may be implemented by a UE or a component thereof as described in reference to Figures 1 to 9 The UE 115 described herein may be executed. In some examples, the UE may execute an instruction set to control the functional elements of the UE to perform the described functions. Additionally or alternatively, the UE may use dedicated hardware to perform various aspects of the described functions.

[0209] At 1605, the method may include receiving a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the set of multiple machine learning models, one of the set of multiple parameter sets, or both for performing machine learning inference. The operations of 1605 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1605 may be performed as described in reference to Figure 8 The message receiving component 825 is executed.

[0210] At 1610, the method may include selecting the first machine learning model, the first parameter set, or both for performing machine learning inference based on a usage condition indicative of selecting the first machine learning model, the first parameter set, or both when the UE fails to receive input data for a second machine learning model in a set of multiple machine learning models, a second parameter set in a set of multiple parameter sets, or both from a network node. The operations of 1610 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1610 may be performed as described in reference to Figure 8 The selection component 830 is executed.

[0211] At 1615, the method may include sending a report indicating machine learning inferences generated by the first machine learning model or by using the parameter set, and indicating the first machine learning model, the first parameter set, or both. The operations of 1615 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1615 may be performed as described in reference to Figure 8 The report sending component 835 is used to execute.

[0212] Fig.17 A flowchart illustrating a method 1700 for supporting conditional AI, machine learning models, and parameter set configuration according to one or more aspects of the present disclosure is shown. The operations of the method 1700 may be implemented by a network node or a component thereof as described herein. For example, the operations of the method 1700 may be implemented by a network node or a component thereof as described in reference to Figures 1 to 5 and Figures 10 to 13 In some examples, the network node may execute an instruction set to control the functional elements of the network node to perform the functions. Additionally or alternatively, the network node may use dedicated hardware to perform various aspects of the described functions.

[0213] At 1705, the method may include sending a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the set of multiple machine learning models, one of the set of multiple parameter sets, or both for performing machine learning inference. The operations of 1705 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1705 may be performed as described in reference to Fig.12 The message sending component 1225 is used to execute.

[0214] At 1710, the method may include receiving a report indicating machine learning inferences generated by a first machine learning model or by using a first parameter set, the first machine learning model, the first parameter set being selected by the UE based on a usage condition in one or more usage conditions being satisfied. The operations of 1710 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1710 may be performed as described in reference to Fig.12 The report receiving component 1230 is used to execute.

[0215] Fig.18 A flowchart illustrating a method 1800 for supporting conditional AI, machine learning models, and parameter set configuration according to one or more aspects of the present disclosure is shown. The operations of the method 1800 may be implemented by a network node or a component thereof as described herein. For example, the operations of the method 1800 may be implemented by a network node or a component thereof as described in reference to Figures 1 to 5 and Figures 10 to 13 In some examples, the network node may execute an instruction set to control the functional elements of the network node to perform the functions. Additionally or alternatively, the network node may use dedicated hardware to perform various aspects of the described functions.

[0216] At 1805, the method may include sending a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, indicating one or more usage conditions for selecting one of the set of multiple machine learning models, one of the set of multiple parameter sets, or both for performing machine learning inference, and indicating a set of multiple priorities corresponding to the set of multiple machine learning models, the set of multiple parameter sets, or both. The operations of 1805 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1805 may be performed as described in reference to Fig.12 The message sending component 1225 is used to execute.

[0217] At 1810, the method may include receiving a report indicating machine learning inferences generated by a first machine learning model or by using a first parameter set, the first machine learning model, the first parameter set being selected by the UE based on a usage condition in one or more usage conditions being satisfied. The operations of 1810 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1810 may be performed as described in reference to Fig.12 The report receiving component 1230 is used to execute.

[0218] Fig.19 A flowchart illustrating a method 1900 for supporting conditional AI, machine learning models, and parameter set configuration according to one or more aspects of the present disclosure is shown. The operations of the method 1900 may be implemented by a network node or a component thereof as described herein. For example, the operations of the method 1900 may be implemented by a network node or a component thereof as described in reference to Figures 1 to 5 and Figures 10 to 13 In some examples, the network node may execute an instruction set to control the functional elements of the network node to perform the functions. Additionally or alternatively, the network node may use dedicated hardware to perform various aspects of the described functions.

[0219] At 1905, the method may include sending a first message indicating a set of multiple machine learning models, a set of multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the set of multiple machine learning models, one of the set of multiple parameter sets, or both for performing machine learning inference. The operations of 1905 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1905 may be performed as described in reference to Fig.12 The message sending component 1225 is used to execute.

[0220] At 1910, the method may include receiving a report indicating machine learning inferences generated by a first machine learning model or by using a first parameter set, the first machine learning model, the first parameter set being selected by the UE based on a usage condition in one or more usage conditions being satisfied. The operations of 1910 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1910 may be performed as described in reference to Fig.12 The report receiving component 1230 is used to execute.

[0221] At 1915, the method may include sending a feedback message indicating one or more statistics associated with switching between a set of multiple machine learning models, a set of multiple parameter sets, or both. The operations of 1915 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1915 may be performed as described in reference to Fig.12 The feedback sending component 1240 is used to execute.

[0222] The following provides an overview of various aspects of the disclosure:

[0223] Aspect 1: A method for wireless communication at a UE, comprising: receiving a first message indicating multiple machine learning models, multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both for performing machine learning inference; selecting a first machine learning model among the multiple machine learning models, a first parameter set among the multiple parameter sets, or both for performing the machine learning inference based at least in part on satisfaction of a usage condition among the one or more usage conditions; and sending a report indicating the machine learning inference generated by the first machine learning model or by using the first parameter set.

[0224] Aspect 2: The method according to Aspect 1 also includes: receiving the first message indicating multiple priorities corresponding to the multiple machine learning models, the multiple parameter sets, or both; and selecting the first machine learning model, the first parameter set, or both for performing the machine learning inference based at least in part on the first machine learning model, the first parameter set, or both being associated with the highest priority among the multiple priorities.

[0225] Aspect 3: The method according to Aspect 2 also includes: receiving the first message, wherein the first message indicates that the network node changes the priority of one or more of the multiple machine learning models, one or more of the multiple parameter sets, or both based at least in part on one or more performance indicators reported by the UE, wherein the first machine learning model, the first parameter set, or both are selected at least in part based on the first message.

[0226] Aspect 4: The method according to any one of Aspects 1 to 3 also includes: selecting the first machine learning model, the first parameter set, or both for performing the machine learning inference based at least in part on the usage condition indicating that the first machine learning model, the first parameter set, or both are selected when the UE is able to successfully receive input data from the network node; and generating the machine learning inference by the first machine learning model or by using the first parameter set using the input data received from the network node.

[0227] Aspect 5: The method according to any one of Aspects 1 to 4 also includes: selecting the first machine learning model, the first parameter set, or both for performing the machine learning inference based at least in part on the usage condition indicating that the first machine learning model, the first parameter set, or both are selected when the UE fails to receive input data for a second machine learning model among the multiple machine learning models, a second parameter set among the multiple parameter sets, or both from a network node; and sending the report indicating the machine learning inference and indicating the first machine learning model, the first parameter set, or both.

[0228] Aspect 6: The method according to any one of Aspects 1 to 5 also includes: generating the machine learning inference using the input data measured by the UE by the first machine learning model or by using the first parameter set, at least in part based on the usage condition indicating that the first machine learning model, the first parameter set, or both are selected when the UE measures the input data.

[0229] Aspect 7: The method according to any one of Aspects 1 to 6 further includes: sending a first feedback message, wherein the first feedback message indicates one or more statistics associated with switching between selecting the multiple machine learning models, the multiple parameter sets, or both.

[0230] Aspect 8: The method according to any one of Aspects 1 to 7 further includes: receiving a second feedback message, wherein the second feedback message indicates one or more statistics associated with switching between selecting the multiple machine learning models, the multiple parameter sets, or both.

[0231] Aspect 9: The method according to any one of Aspects 1 to 8 further includes: receiving the first message or the second message, the first message or the second message indicating the one or more usage conditions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both, wherein the one or more usage conditions include at least one of the following: instructions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both when input data is missing, or a priority corresponding to each of the multiple machine learning models, each of the multiple parameter sets, or both, for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both.

[0232] Aspect 10: A method according to Aspect 9, wherein the instructions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both include the availability of input data, usage scenarios, area scope, model validity timer, parameter set validity timer, configuration of the UE, configuration of the network node, performance threshold, or any combination thereof.

[0233] Aspect 11: A method according to any one of Aspects 1 to 10, wherein receiving the first message includes: receiving the first message or the second message, the first message or the second message indicating configuration information of the multiple machine learning models, the multiple parameter sets, or both, wherein the configuration information indicates a set of alternative machine learning models among the multiple machine learning models, a set of alternative parameter sets among the multiple parameter sets, the one or more usage conditions, a priority corresponding to each machine learning model, a priority corresponding to each parameter set, or any combination thereof.

[0234] Aspect 12: A method for wireless communication at a network node, comprising: sending a first message indicating multiple machine learning models, multiple parameter sets, or both, and indicating one or more usage conditions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both for performing machine learning inference; and receiving a report indicating the machine learning inference generated by the first machine learning model or by using a first parameter set, the first machine learning model, the first parameter set being selected by a UE at least in part based on satisfaction of a usage condition among the one or more usage conditions.

[0235] Aspect 13: The method according to Aspect 12 also includes: sending the first message indicating multiple priorities corresponding to the multiple machine learning models, the multiple parameter sets, or both.

[0236] Aspect 14: The method according to Aspect 13 also includes: sending the first message, wherein the first message instructs the network node to change the priority of one or more of the multiple machine learning models, one or more of the multiple parameter sets, or both based at least in part on one or more performance indicators reported by the UE, wherein the first machine learning model, the first parameter set, or both are selected at least in part based on the first message.

[0237] Aspect 15: The method according to any one of Aspects 12 to 14 further includes: receiving the report indicating the machine learning inference and instructing the UE to select the first machine learning model, the first parameter set, or both to generate the machine learning inference.

[0238] Aspect 16: The method according to any one of Aspects 12 to 15 further includes: sending a feedback message indicating one or more statistics associated with switching between the multiple machine learning models, the multiple parameter sets, or both.

[0239] Aspect 17: The method according to any one of Aspects 12 to 16 further includes: sending the first message or the second message, wherein the first message or the second message indicates the one or more usage conditions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both, wherein the one or more usage conditions include at least one of the following: instructions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both when input data is missing, or a priority corresponding to each of the multiple machine learning models, each parameter set in the multiple parameter sets, for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both.

[0240] Aspect 18: A method according to Aspect 17, wherein the instructions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both include the availability of input data, usage scenarios, area scope, model validity timer, parameter set validity timer, configuration of the UE, configuration of the network node, performance threshold, or any combination thereof.

[0241] Aspect 19: A method according to any one of Aspects 12 to 18, wherein sending the first message includes: sending the first message or the second message, the first message or the second message indicating configuration information of the multiple machine learning models, the multiple parameter sets, or both, wherein the configuration information indicates a set of alternative machine learning models among the multiple machine learning models, a set of alternative parameter sets among the multiple parameter sets, the one or more usage conditions, a priority corresponding to each machine learning model, a priority corresponding to each parameter set, or any combination thereof.

[0242] Aspect 20: An apparatus for performing wireless communication at a UE, comprising: a processor; a memory coupled to the processor; and instructions stored in the memory and executable by the processor to cause the apparatus to perform a method according to any one of Aspects 1 to 11.

[0243] Aspect 21: An apparatus for wireless communication at a UE, comprising at least one component for performing the method according to any one of aspects 1 to 11.

[0244] Aspect 22: A non-transitory computer-readable medium storing a code for wireless communication at a UE, the code comprising instructions executable by a processor to perform the method according to any one of aspects 1 to 11.

[0245] Aspect 23: An apparatus for wireless communication at a network node, comprising: a processor; a memory coupled to the processor; and instructions stored in the memory and executable by the processor to cause the apparatus to perform a method according to any one of Aspects 12 to 19.

[0246] Aspect 24: An apparatus for wireless communication at a network node, comprising at least one means for performing a method according to any one of aspects 12 to 19.

[0247] Aspect 25: A non-transitory computer-readable medium storing code for wireless communication at a network node, the code comprising instructions executable by a processor to perform the method according to any one of aspects 12 to 19.

[0248] It should be noted that the methods described herein describe possible implementations, and that the various operations and steps may be rearranged or otherwise modified and that other implementations are possible. Furthermore, aspects from two or more methods may be combined.

[0249] Although aspects of LTE, LTE-A, LTE-A Pro, or NR systems may be described for example purposes, and LTE, LTE-A, LTE-A Pro, or NR terminology may be used in much of the description, the techniques described herein may also be applicable to networks other than LTE, LTE-A, LTE-A Pro, or NR networks. For example, the described techniques may be applicable to various other wireless communication systems, such as Ultra Mobile Broadband (UMB), Institute of Electrical and Electronics Engineers (IEEE) 802.11 (Wi-Fi), IEEE 802.16 (WiMAX), IEEE 802.20, Flash-OFDM, and other systems and radio technologies not explicitly mentioned herein.

[0250] The information and signals described herein may be represented using any of a variety of different techniques and methods. For example, data, instructions, commands, information, signals, bits, symbols, and chips mentioned throughout the specification may be represented by voltage, current, electromagnetic waves, magnetic fields or magnetic particles, optical fields or optical particles, or any combination thereof.

[0251] The various illustrative blocks and components described in conjunction with the disclosure herein may be implemented or performed using a general purpose processor, DSP, ASIC, CPU, FPGA or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor may be a microprocessor, but in an alternative, the processor may be any processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices (e.g., a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration).

[0252] The functions described herein may be implemented using hardware, software executed by a processor, firmware, or any combination thereof. When implemented using software executed by a processor, the functions may be stored as one or more instructions or codes of a computer-readable medium, or sent using one or more instructions or codes of a computer-readable medium. Other examples and specific implementations are within the scope of the present disclosure and the appended claims. For example, due to the nature of software, the functions described herein may be implemented using software executed by a processor, hardware, firmware, hard wiring, or a combination of any of these items. Features that implement the functions may also be physically located at different locations, including being distributed so that the various parts of the functions are implemented at different physical locations.

[0253] Computer-readable medium includes both non-transient computer storage medium and communication medium, and it includes any medium that facilitates computer program to be transmitted from one position to another position.Non-transient storage medium can be any available medium that can be accessed by general or special-purpose computer.By way of example and not limitation, non-transient computer-readable medium can include RAM, ROM, electrically erasable programmable ROM (EEPROM), flash memory, compact disc (CD) ROM or other optical disc storage device, disk storage device or other magnetic storage device, or can be used for carrying or storing desired program code components and any other non-transient medium that can be accessed by general or special-purpose computer or general or special-purpose processor in the form of instruction or data structure.Moreover, any connection is appropriately referred to as computer-readable medium.For example, if software is sent from website, server or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL) or wireless technology such as infrared, radio and microwave, then coaxial cable, fiber optic cable, twisted pair, DSL or wireless technology such as infrared, radio and microwave are included in the definition of computer-readable medium. Disks and optical disks used herein include CDs, laser optical disks, optical optical disks, digital versatile disks (DVDs), floppy disks, and Blu-ray disks. Disks can reproduce data magnetically, while optical disks can reproduce data optically using lasers. Combinations of the above are also included within the scope of computer-readable media.

[0254] As used herein (including in the claims), "or" used in a list of items (e.g., a list of items followed by a phrase such as "at least one of" or "one or more of") indicates an inclusive list, so that, for example, a list of at least one of A, B, or C means A or B or C or AB or AC or BC or ABC (i.e., A and B and C). In addition, as used herein, the phrase "based on" should not be interpreted as a reference to a closed set of conditions. For example, an example step described as "based on condition A" can be based on both condition A and condition B without departing from the scope of the present disclosure. In other words, as used herein, the phrase "based on" should be interpreted in the same manner as the phrase "based at least in part on."

[0255] The terms "determine" or "determining" encompass a variety of actions, and thus, "determining" may include calculating, computing, processing, deriving, investigating, searching (such as via searching in a table, database, or other data structure), ascertaining, and similar actions. Furthermore, "determining" may include receiving (e.g., receiving information), accessing (e.g., accessing data stored in a memory), etc. Additionally, "determining" may include parsing, acquiring, selecting, choosing, establishing, and other such similar actions. Furthermore, as used herein, the phrase "set" should be understood to include the possibility of a set having one member. That is, the phrase "set" should be understood in the same manner as "one or more,"

[0256] In the drawings, similar components or features may have the same reference label. In addition, various components of the same type may be distinguished by following the reference label with a dash and a second label to distinguish between similar components. If only the first reference label is used in the specification, the description may apply to any of the similar components having the same first reference label, regardless of the second or other subsequent reference labels.

[0257] The descriptions set forth herein in conjunction with the accompanying drawings describe example configurations and do not represent all examples that may be implemented or within the scope of the claims. The term "example" as used herein means "used as an example, instance, or illustration," rather than "preferred" or "advantageous over other examples." The specific implementation includes specific details to provide an understanding of the described techniques. However, these techniques may be practiced without these specific details. In some cases, known structures and devices are shown in block diagram form to avoid obscuring the concepts of the described examples.

[0258] The description herein is provided to enable one of ordinary skill in the art to implement or use the present disclosure. Various modifications to the present disclosure will be apparent to one of ordinary skill in the art, and the general principles defined herein may be applied to other variations without departing from the scope of the present disclosure. Thus, the present disclosure is not limited to the examples and designs described herein, but should be granted the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An apparatus for wireless communication at a user equipment (UE), comprising: processor; a memory coupled to the processor; and instructions stored in the memory and executable by the processor to cause the apparatus to: receiving a first message indicating a plurality of machine learning models, a plurality of parameter sets, or both, and indicating one or more usage conditions for selecting one of the plurality of machine learning models, one of the plurality of parameter sets, or both for performing machine learning inference; selecting a first machine learning model from the plurality of machine learning models, a first parameter set from the plurality of parameter sets, or both for use in performing the machine learning inference based at least in part on a usage condition from the one or more usage conditions being satisfied; as well as Sending a report indicating the machine learning inferences generated by the first machine learning model or by using the first set of parameters.

2. The apparatus of claim 1 , wherein the instructions are further executable by the processor to cause the apparatus to: receiving the first message indicating a plurality of priorities corresponding to the plurality of machine learning models, the plurality of parameter sets, or both; and The first machine learning model, the first parameter set, or both are selected for performing the machine learning inference based at least in part on the first machine learning model, the first parameter set, or both being associated with a highest priority among the multiple priorities.

3. The apparatus of claim 2, wherein the instructions are further executable by the processor to cause the apparatus to: Receive the first message, the first message instructing the network node to change the priority of one or more of the multiple machine learning models, one or more of the multiple parameter sets, or both based at least in part on one or more performance indicators reported by the UE, wherein the first machine learning model, the first parameter set, or both are selected at least in part based on the first message.

4. The apparatus of claim 1, wherein the instructions are further executable by the processor to cause the apparatus to: selecting the first machine learning model, the first set of parameters, or both for performing the machine learning inference based at least in part on the usage condition indicating that the first machine learning model, the first set of parameters, or both are selected when the UE is able to successfully receive input data from a network node; and The machine learning inferences are generated by the first machine learning model or by using the first parameter set using the input data received from the network node.

5. The apparatus of claim 1, wherein the instructions are further executable by the processor to cause the apparatus to: selecting the first machine learning model, the first parameter set, or both for performing the machine learning inference based at least in part on the usage condition indicating that the first machine learning model, the first parameter set, or both is selected when the UE fails to receive input data for a second machine learning model in the plurality of machine learning models, a second parameter set in the plurality of parameter sets, or both from a network node; and Sending the report indicating the machine learning inferences and indicating the first machine learning model, the first parameter set, or both.

6. The apparatus of claim 1, wherein the instructions are further executable by the processor to cause the apparatus to: The machine learning inference is generated by the first machine learning model or by using the first parameter set using input data measured by the UE based at least in part on the usage condition indicating selection of the first machine learning model, the first parameter set, or both when the UE measures the input data.

7. The apparatus of claim 1, wherein the instructions are further executable by the processor to cause the apparatus to: Sending a first feedback message indicating one or more statistics associated with switching between selecting the plurality of machine learning models, the plurality of parameter sets, or both.

8. The apparatus of claim 1, wherein the instructions are further executable by the processor to cause the apparatus to: Receive a second feedback message indicating one or more statistics associated with switching between selecting the plurality of machine learning models, the plurality of parameter sets, or both.

9. The apparatus of claim 1, wherein the instructions are further executable by the processor to cause the apparatus to: Receive the first message or the second message, the first message or the second message indicating the one or more usage conditions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both, wherein the one or more usage conditions include at least one of the following: instructions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both when input data is missing, or a priority corresponding to each of the multiple machine learning models, each of the multiple parameter sets, or both, for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both.

10. The apparatus of claim 9, wherein the instructions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both include availability of input data, usage scenarios, area scope, model validity timer, parameter set validity timer, configuration of the UE, configuration of a network node, performance threshold, or any combination thereof.

11. The apparatus of claim 1 , wherein the instructions to receive the first message are executable by the processor to cause the apparatus to: Receive the first message or the second message, wherein the first message or the second message indicates configuration information of the multiple machine learning models, the multiple parameter sets, or both, wherein the configuration information indicates a set of alternative machine learning models among the multiple machine learning models, a set of alternative parameter sets among the multiple parameter sets, the one or more usage conditions, a priority corresponding to each machine learning model, a priority corresponding to each parameter set, or any combination thereof.

12. An apparatus for wireless communication at a network node, comprising: processor; a memory coupled to the processor; and instructions stored in the memory and executable by the processor to cause the apparatus to: sending a first message indicating a plurality of machine learning models, a plurality of parameter sets, or both, and indicating one or more usage conditions for selecting one of the plurality of machine learning models, one of the plurality of parameter sets, or both for performing machine learning inference; as well as Receive a report indicating the machine learning inferences generated by a first machine learning model or by using a first set of parameters, wherein the first machine learning model and the first set of parameters are selected by a user equipment (UE) at least in part based on satisfaction of a usage condition among the one or more usage conditions.

13. The apparatus of claim 12, wherein the instructions are further executable by the processor to cause the apparatus to: Send the first message indicating multiple priorities corresponding to the multiple machine learning models, the multiple parameter sets, or both.

14. The apparatus of claim 13, wherein the instructions are further executable by the processor to cause the apparatus to: sending the first message, the first message instructing the network node to change the priority of one or more of the multiple machine learning models, one or more of the multiple parameter sets, or both based at least in part on one or more performance indicators reported by the UE, wherein the first machine learning model, the first parameter set, or both are selected at least in part based on the first message.

15. The apparatus of claim 12, wherein the instructions are further executable by the processor to cause the apparatus to: Receiving a report indicating the machine learning inference and instructing the UE to select the first machine learning model, the first parameter set, or both to generate the report of the machine learning inference.

16. The apparatus of claim 12, wherein the instructions are further executable by the processor to cause the apparatus to: Sending a feedback message indicating one or more statistics associated with switching between the plurality of machine learning models, the plurality of parameter sets, or both.

17. The apparatus of claim 12, wherein the instructions are further executable by the processor to cause the apparatus to: Send the first message or the second message, wherein the first message or the second message indicates the one or more usage conditions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both, wherein the one or more usage conditions include at least one of the following: instructions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both when input data is missing, or a priority corresponding to each of the multiple machine learning models, each of the multiple parameter sets, or both, for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both.

18. The apparatus of claim 17, wherein the instructions for selecting one of the multiple machine learning models, one of the multiple parameter sets, or both comprise availability of input data, usage scenarios, area scope, model validity timer, parameter set validity timer, configuration of the UE, configuration of a network node, performance threshold, or any combination thereof.

19. The apparatus of claim 12, wherein the instructions to send the first message are executable by the processor to cause the apparatus to: Send the first message or the second message, wherein the first message or the second message indicates configuration information of the multiple machine learning models, the multiple parameter sets, or both, wherein the configuration information indicates a set of alternative machine learning models among the multiple machine learning models, a set of alternative parameter sets among the multiple parameter sets, the one or more usage conditions, a priority corresponding to each machine learning model, a priority corresponding to each parameter set, or any combination thereof.

20. A method for wireless communication at a user equipment (UE), comprising: receiving a first message indicating a plurality of machine learning models, a plurality of parameter sets, or both, and indicating one or more usage conditions for selecting one of the plurality of machine learning models, one of the plurality of parameter sets, or both for performing machine learning inference; selecting a first machine learning model from the plurality of machine learning models, a first parameter set from the plurality of parameter sets, or both for use in performing the machine learning inference based at least in part on a usage condition from the one or more usage conditions being satisfied; as well as Sending a report indicating the machine learning inferences generated by the first machine learning model or by using the first set of parameters.

21. The method according to claim 20, further comprising: receiving the first message indicating a plurality of priorities corresponding to the plurality of machine learning models, the plurality of parameter sets, or both; as well as The first machine learning model, the first parameter set, or both are selected for performing the machine learning inference based at least in part on the first machine learning model, the first parameter set, or both being associated with a highest priority among the multiple priorities.

22. The method according to claim 21, further comprising: Receive the first message, the first message instructing the network node to change the priority of one or more of the multiple machine learning models, one or more of the multiple parameter sets, or both based at least in part on one or more performance indicators reported by the UE, wherein the first machine learning model, the first parameter set, or both are selected at least in part based on the first message.

23. The method of claim 20, further comprising: selecting the first machine learning model, the first set of parameters, or both for performing the machine learning inference based at least in part on the usage condition indicating that the first machine learning model, the first set of parameters, or both are selected when the UE is able to successfully receive input data from a network node; as well as The machine learning inferences are generated by the first machine learning model, the first parameter set, or both using the input data received from the network node.

24. The method of claim 20, further comprising: selecting the first machine learning model, the first parameter set, or both for performing the machine learning inference based at least in part on the usage condition indicating that the first machine learning model, the first parameter set, or both is selected when the UE fails to receive input data for a second machine learning model in the plurality of machine learning models, a second parameter set in the plurality of parameter sets, or both from a network node; as well as Sending the report indicating the machine learning inferences and indicating the first machine learning model, the first parameter set, or both.

25. The method of claim 20, further comprising: The machine learning inference is generated by the first machine learning model, the first parameter set, or both using the input data measured by the UE, at least in part based on the usage condition indicating that the first machine learning model, the first parameter set, or both are selected when the UE measures the input data.

26. A method for wireless communication at a network node, comprising: sending a first message indicating a plurality of machine learning models, a plurality of parameter sets, or both and indicating one or more usage conditions for selecting one of the plurality of machine learning models, one of the plurality of parameter sets, or both for performing machine learning inference; as well as Receive a report indicating the machine learning inferences generated by a first machine learning model, a first parameter set, or both, wherein the first machine learning model, the first parameter set is selected by a user equipment (UE) at least in part based on satisfaction of a usage condition among the one or more usage conditions.

27. The method according to claim 26, further comprising: Send the first message indicating multiple priorities corresponding to the multiple machine learning models, the multiple parameter sets, or both.

28. The method according to claim 27, further comprising: sending the first message, the first message instructing the network node to change the priority of one or more of the multiple machine learning models, one or more of the multiple parameter sets, or both based at least in part on one or more performance indicators reported by the UE, wherein the first machine learning model, the first parameter set, or both are selected at least in part based on the first message.

29. The method of claim 26, further comprising: Receiving a report indicating the machine learning inference and instructing the UE to select the first machine learning model, the first parameter set, or both to generate the report of the machine learning inference.

30. The method of claim 26, further comprising: Sending a feedback message indicating one or more statistics associated with switching between the plurality of machine learning models, the plurality of parameter sets, or both.

Images