Management of federated learning

Abstract

Methods, systems, and devices for wireless communication are described. A server (e.g., a network entity, a model repository) may select a user equipment (UE) to participate in a federated learning process for training a prediction model. Based on selecting the UE, the server may determine a set of training parameters for a training configuration of the federated learning process. The server may send an indication of the training configuration to the UE. The server may activate the federated learning process by sending an activation indication to the UE. Each UE may train the predictive model locally according to the training configuration and may report the model parameters to the server. The server may aggregate the reported model parameters into an updated set of model parameters. The server may assign an updated parameter set identifier (PS ID) to the updated set of model parameters, and may notify the UE of the updated PS ID.

Description

[0001] Cross-references

[0002] This patent application claims priority to U.S. patent application No. 17 / 954,824, entitled “MANAGEMENT OF FEDERATED LEARNING,” filed by Kumar et al. on September 28, 2022, which is assigned to the assignee of this application and is expressly incorporated herein by reference. Technical Field

[0003] The following relates to wireless communications, including the management of federated learning. Background Art

[0004] Wireless communication systems are widely deployed to provide various types of communication content, such as voice, video, packet data, messaging, broadcast, etc. 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 communication for a communication device, which may be referred to as a user equipment (UE).

[0005] Some devices in a wireless communication system may implement machine learning techniques. For example, a network entity may utilize a machine learning model based on a data-driven algorithm (e.g., a machine learning algorithm). The UE may perform measurements and communicate data based on the measurements to the network entity, and the network entity may use the data to train and test the machine learning model. Summary of the invention

[0006] The described technology relates to improved methods, systems, devices, and apparatuses for supporting the management of federated learning. For example, the described technology provides that a user equipment (UE) participates in a federated learning process according to a training configuration. A server for the federated learning process (which may include or be an example of a network entity, a model repository (MR), etc.) may select a UE to participate in the federated learning process, and may select or otherwise determine a model structure (MS) and a baseline parameter set (PS) for a prediction model (e.g., a machine learning model) to be trained during one or more training rounds of the federated learning process. Based on selecting the UE, the server may determine a set of training parameters for the training configuration of the federated learning process. The server may send an indication of the training configuration to the UE. In some examples, the server may activate the training round of the federated learning process at all participating UEs at the same time, for example, by sending an activation indication to the UE. Each UE may locally train the prediction model according to the training configuration and based on a data set collected by the UE to obtain model parameters for the prediction model. The UE may report the model parameters to the server.

[0007] Additionally, the present disclosure supports techniques for aggregating the model parameters reported by each UE into a set of updated model parameters at the server. The server may assign a new PS ID to the set of updated parameters and may notify the UE of the new PS ID. After receiving the new PS ID, each UE may send a message to the server indicating that the prediction model is ready to be activated. The server may activate subsequent training rounds of the federated learning process, and each UE may train the prediction model locally based on the new PS ID.

[0008] A method for wireless communication at a network node is described. The method may include: selecting one or more UEs for a training process for a prediction model based on a trigger to activate the training process; and sending an indication of a training configuration for the prediction model to the one or more UEs, the training configuration including a set of training parameters based on the one or more UEs.

[0009] 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 may be executable by the processor to cause the apparatus to: select one or more UEs for a training process for a prediction model based on a trigger to activate the training process; and send an indication of a training configuration for the prediction model to the one or more UEs, the training configuration including a set of training parameters based on the one or more UEs.

[0010] Another apparatus for wireless communication at a network node is described. The apparatus may include: a component for selecting one or more UEs for a training process for a prediction model based on a trigger to activate the training process; and a component for sending an indication of a training configuration for the prediction model to the one or more UEs, the training configuration including a set of training parameters based on the one or more UEs.

[0011] 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 perform the following operations: selecting one or more UEs for a training process for a prediction model based on a trigger to activate the training process; and sending an indication of a training configuration for the prediction model to the one or more UEs, the training configuration including a set of training parameters based on the one or more UEs.

[0012] A method for wireless communication at a server is described. The method may include: selecting one or more UEs for a training process for a prediction model based on a trigger to activate the training process; and sending an indication of a training configuration for the prediction model to the one or more UEs, the training configuration including a set of training parameters based on the one or more UEs.

[0013] An apparatus for wireless communication at a server is described. The apparatus may include a processor, a memory coupled to the processor, and instructions stored in the memory. The instructions may be executable by the processor to cause the apparatus to: select one or more UEs for a training process for a prediction model based on a trigger to activate the training process; and send an indication of a training configuration for the prediction model to the one or more UEs, the training configuration including a set of training parameters based on the one or more UEs.

[0014] Another apparatus for wireless communication at a server is described. The apparatus may include: a component for selecting one or more UEs for a training process for a prediction model based on a trigger to activate the training process; and a component for sending an indication of a training configuration for the prediction model to the one or more UEs, the training configuration including a set of training parameters based on the one or more UEs.

[0015] A non-transitory computer-readable medium storing code for wireless communication at a server is described. The code may include instructions executable by a processor to perform the following operations: selecting one or more UEs for a training process for a prediction model based on a trigger to activate the training process; and sending an indication of a training configuration for the prediction model to the one or more UEs, the training configuration including a set of training parameters based on the one or more UEs.

[0016] A method for wireless communication at a UE is described. The method may include: receiving a first message indicating a training configuration for a training process for a prediction model, the training configuration including a set of training parameters; and sending a second message indicating whether the UE has implemented the training configuration for the training process based on the set of training parameters.

[0017] An apparatus for wireless communication at a UE is described. The apparatus may include a processor, a memory coupled to the processor, and instructions stored in the memory. The instructions may be executable by the processor to cause the apparatus to: receive a first message indicating a training configuration for a training process for a prediction model, the training configuration including a set of training parameters; and send a second message indicating whether the UE has implemented the training configuration for the training process based on the set of training parameters.

[0018] Another apparatus for wireless communication at a UE is described. The apparatus may include: a component for receiving a first message indicating a training configuration for a training process for a prediction model, the training configuration including a set of training parameters; and a component for sending a second message indicating whether the UE has implemented the training configuration for the training process based on the set of training parameters.

[0019] 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 training configuration for a training process for a prediction model, the training configuration including a set of training parameters; and sending a second message indicating whether the UE has implemented the training configuration for the training process based on the set of training parameters.

[0020] A method for wireless communication at a server is described. The method may include: sending a training configuration for a training process associated with a prediction model to a set of UEs, the training configuration including a first set of model parameters associated with a first PS ID; receiving one or more reports from one or more UEs in the set of UEs indicating one or more subsets of model parameters output at the UE from the training process for the prediction model; aggregating the subsets of model parameters into a second set of model parameters; assigning a second PS ID to the second set of model parameters, the second PS ID being different from the first PS ID; and sending an indication of the second PS ID.

[0021] An apparatus for wireless communication at a server is described. The apparatus may include a processor, a memory coupled to the processor, and instructions stored in the memory. The instructions may be executable by the processor to cause the apparatus to: send a training configuration for a training process associated with a prediction model to a set of UEs, the training configuration including a first set of model parameters associated with a first PS ID; receive one or more reports from one or more UEs in the set of UEs indicating one or more subsets of model parameters output from the training process for the prediction model at the UE; aggregate the subsets of model parameters into a second set of model parameters; assign a second PSID to the second set of model parameters, the second PS ID being different from the first PS ID; and send an indication of the second PS ID.

[0022] Another apparatus for wireless communication at a server is described. The apparatus may include: means for sending a training configuration for a training process associated with a prediction model to a set of UEs, the training configuration including a first set of model parameters associated with a first PSID; means for receiving from one or more UEs in the set of UEs one or more reports indicating one or more subsets of model parameters output at the UE from the training process for the prediction model; means for aggregating the subsets of model parameters into a second set of model parameters; means for assigning a second PS ID to the second set of model parameters, the second PS ID being different from the first PS ID; and means for sending an indication of the second PSID.

[0023] A non-transitory computer-readable medium storing code for wireless communication at a server is described. The code may include instructions executable by a processor to perform the following operations: send a training configuration for a training process associated with a prediction model to a set of UEs, the training configuration including a first set of model parameters associated with a first PS ID; receive one or more reports from one or more UEs in the set of UEs indicating one or more subsets of model parameters output from the training process for the prediction model at the UE; aggregate the subsets of model parameters into a second set of model parameters; assign a second PS ID to the second set of model parameters, the second PS ID being different from the first PS ID; and send an indication of the second PS ID.

[0024] A method for wireless communication at a UE is described. The method may include: receiving a training configuration for a training process associated with a prediction model, the training configuration including a first set of model parameters associated with a first PS ID; sending a report indicating a subset of model parameters output from the training process for the prediction model at the UE; and receiving an indication of a second PS ID associated with a second set of model parameters based on sending the report, the second PS ID being different from the first PS ID.

[0025] An apparatus for wireless communication at a UE is described. The apparatus may include a processor, a memory coupled to the processor, and instructions stored in the memory. The instructions may be executable by the processor to cause the apparatus to: receive a training configuration for a training process associated with a prediction model, the training configuration including a first set of model parameters associated with a first PS ID; send a report indicating a subset of model parameters output from the training process for the prediction model at the UE; and receive an indication of a second PS ID associated with a second set of model parameters based on sending the report, the second PS ID being different from the first PS ID.

[0026] Another apparatus for wireless communication at a UE is described. The apparatus may include: a component for receiving a training configuration for a training process associated with a prediction model, the training configuration including a first set of model parameters associated with a first PS ID; a component for sending a report indicating a subset of model parameters output from the training process for the prediction model at the UE; and a component for receiving an indication of a second PS ID associated with a second set of model parameters based on sending the report, the second PS ID being different from the first PS ID.

[0027] 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: receive a training configuration for a training process associated with a prediction model, the training configuration including a first set of model parameters associated with a first PS ID; send a report indicating a subset of model parameters output from the training process for the prediction model at the UE; and receive an indication of a second PS ID associated with a second set of model parameters based on sending the report, the second PS ID being different from the first PS ID. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 An example of a wireless communication system supporting management of federated learning according to one or more aspects of the present disclosure is illustrated.

[0029] Figure 2An example of a wireless communication system supporting management of federated learning according to one or more aspects of the present disclosure is illustrated.

[0030] Figures 3 to 5 An example of a process flow supporting management of federated learning in accordance with one or more aspects of the present disclosure is illustrated.

[0031] Figure 6 and Figure 7 A block diagram of a device supporting management of federated learning according to one or more aspects of the present disclosure is shown.

[0032] Figure 8 A block diagram of a communication manager supporting management of federated learning is shown in accordance with one or more aspects of the present disclosure.

[0033] Fig. 9 A diagram of a system including a device supporting management of federated learning in accordance with one or more aspects of the present disclosure is shown.

[0034] Fig.10 and Fig.11 A block diagram of a device supporting management of federated learning according to one or more aspects of the present disclosure is shown.

[0035] Fig.12 A block diagram of a communication manager supporting management of federated learning is shown in accordance with one or more aspects of the present disclosure.

[0036] Fig.13 A diagram of a system including a device supporting management of federated learning in accordance with one or more aspects of the present disclosure is shown.

[0037] Figures 14 to 18 A flow chart illustrating a method of supporting management of federated learning according to one or more aspects of the present disclosure is shown. DETAILED DESCRIPTION

[0038] The wireless communication system may support a federated learning process to train a prediction model (e.g., a machine learning model). For the federated learning process, multiple clients (such as user equipment (UE)) may adopt the same model structure (MS) and train model parameters for the prediction model locally based on local observations or data. In some cases, a server for the federated learning process may configure the MS and a baseline parameter set (PS), and may distribute the MS and the baseline PS to the clients participating in the federated learning process.

[0039] After performing training for the prediction model, the client may transmit information for updated model parameters to the server, which compiles information from all clients to determine aggregate model parameters. The server may then transmit updated model parameters or an updated prediction model to the client. The federated learning process may include multiple training rounds. The server may receive information for model parameters, update the machine learning model, and send information for an updated version of the machine learning model to the client for another round of the federated learning process.

[0040] The techniques described herein support training configuration for a federated learning process initiated by a network entity or server, which may be an example of a model repository (MR). The network entity or server may select a UE to participate in a training round (e.g., a training process) of a federated learning process. The network entity or server may determine training parameters for the training configuration based on the selected UE. For example, the network entity or server may select the number of epochs (e.g., iterations) to be performed at the UE for a given training round based on the computing power of the UE, the estimated link capacity of the UE, etc. Other training parameters may include a training deadline (e.g., a time at which the UE will stop training a prediction model and report information on model parameters), a training validity region, and the like.

[0041] The network entity or server may indicate the training configuration to the selected UE. In some cases, each UE may determine whether to participate in the federated learning process based on the training configuration (e.g., based on training parameters). For example, the UE may determine whether the UE is able to complete the indicated number of epochs within the indicated training deadline based on the computing power of the UE. The UE may send a message to the network entity or server indicating whether the UE is participating in the federated learning process. Additionally, in some examples, the UE may send a message to the network entity or server indicating that the UE has implemented the training configuration, the prediction model is ready to be activated at the UE, or both. The network entity or server may simultaneously activate the federated learning process (e.g., training rounds of the federated learning process) by sending an activation message to all selected UEs. Based on receiving the activation message, each UE may begin training the prediction model according to the training configuration.

[0042] The UE may report to a network entity or server the model parameters output from the prediction model after completing a training round (e.g., when a training deadline is reached, after a number of epochs are completed, etc.). In some examples, the server may aggregate the reported model parameters into an updated set of model parameters. The server may assign a new PS ID to the updated set of model parameters and may indicate the new PS ID to the UE. In some examples, the server may additionally update one or more training parameters of the training configuration and may deliver the updated training parameters to the UE. Each UE may send a message confirming receipt of the new PS ID and indicating that the prediction model is ready to be activated again.

[0043] Aspects of the present disclosure are first described in the context of wireless communication systems and process flows. Aspects of the present disclosure are further illustrated by apparatus diagrams, system diagrams, and flow diagrams related to the management of federated learning, and are described with reference to these diagrams.

[0044] Figure 1 An example of a wireless communication system 100 supporting management of federated learning 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).

[0045] 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 names. 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), and the UE 115 and the network entity 105 may establish one or more communication links 125 over the coverage area. The coverage area 110 may be an example of a geographic area, and the network entity 105 and the UE 115 may support signal communications over the geographic area according to one or more radio access technologies (RATs).

[0046] 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 communication with various types of devices, such as Figure 1 Communicate with other UEs 115 or network entities 105) as shown.

[0047] 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.

[0048] In some examples, the network entities 105 may communicate with the core network 130, or with each other, or both. For example, the network entities 105 may communicate with the core network 130 via one or more backhaul 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 directly (e.g., directly between the network entities 105) or indirectly (e.g., via the core network 130) via the backhaul communication link 120 (e.g., according to X2, Xn, or other interface protocols). 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), or other examples or various combinations thereof. UE 115 may communicate with core network 130 via communication link 155 .

[0049] 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, standalone) base station architecture, which may be configured to utilize a protocol stack that is physically or logically integrated within a single network entity 105 (e.g., a single RAN node, such as a base station 140).

[0050] 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)).

[0051] The functional split between CU 160, DU 165, and RU 170 is flexible and can 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 can be employed between CU 160 and DU 165, such that CU 160 can support one or more layers of a protocol stack and DU 165 can support one or more different layers of a protocol stack. In some examples, CU 160 can 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 DUs 165 and RUs 170, such that the DUs 165 may support one or more layers of the protocol stack, and the RUs 170 may support one or more different layers of the protocol stack. The DUs 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 may be within the protocol layer (e.g., some functions of the protocol layer may be performed by one of CU 160, DU 165, or RU 170, while other functions of the protocol layer are performed by the other of CU 160, DU 165, or RU 170). CU 160 may be further functionally split into CU control plane (CU-CP) functions and CU user plane (CU-UP) functions. CU 160 may be connected to one or more DUs 165 via midhaul communication links 162 (e.g., F1, F1-c, F1-u), and DU 165 may 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 interfaces (eg, channels) between layers of a protocol stack supported by respective network entities 105 that communicate via those communication links.

[0052] 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., a 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 DU 165 of a coupled IAB donor. 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.

[0053] 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 a connection 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), wherein 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., CU 160 associated with an alternative IAB donor) via an Xn-C interface (which may be an example of a portion of a backhaul link).

[0054] An IAB node 104 may refer to a RAN node that provides IAB functionality (e.g., access for UE 115, wireless self-backhaul capabilities, etc.). DU 165 may act as a distributed scheduling node toward child nodes associated with IAB node 104, and IAB-MT may act as a scheduled node toward a parent node associated with IAB node 104. That is, an IAB donor may be referred to as a parent node that communicates with one or more child nodes (e.g., an IAB donor may relay for transmissions to UEs through one or more other IAB nodes 104). Additionally or alternatively, depending on the relay chain or configuration of the AN, an IAB node 104 may also be referred to as a parent node or child node of other IAB nodes 104. Thus, 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 a DU interface (eg, 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 .

[0055] For example, the IAB node 104 may be referred to as a parent node supporting communications for a child IAB node or as a child IAB 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 transmission to the UE 115 through the IAB node 104, or may directly signal the transmission to the UE 115, or both. The CU 160 of the IAB donor may signal the establishment of a communication link to the IAB node 104 via the F1 interface, and the IAB node 104 may schedule transmission (e.g., transmission relayed from the IAB donor to the UE 115) via the DU 165. That is, data may be relayed to and from the IAB node 104 via signaling via the NR Uu interface to the MT of 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 .

[0056] 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 management of federated learning 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).

[0057] 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 terminology, 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.

[0058] 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.

[0059] 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 defined physical layer structure 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) component carriers and time division duplex (TDD) component carriers. Communication between the network entity 105 and other devices may refer to communication between a device 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 the network entity 105 (e.g., base station 140, CU 160, DU 165, RU 170) of the RAN communicating with another device (e.g., directly or via one or more other network entities 105).

[0060] 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 according to a channel raster 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.

[0061] The communication link 125 shown in the wireless communication system 100 may include downlink transmissions (e.g., forward link transmissions) from the network entity 105 to the UE 115, uplink transmissions (e.g., return link transmissions) from the UE 115 to the network entity 105, or both, as well as other transmission configurations. A carrier may carry downlink communications or uplink communications (e.g., in FDD mode), or may be configured to carry downlink communications and uplink communications (e.g., in TDD mode).

[0062] A carrier may be associated with a particular bandwidth of the RF spectrum, and in some examples, the carrier bandwidth may be referred to as the "system bandwidth" of the carrier or wireless communication system 100. For example, the carrier bandwidth may be one of a set of bandwidths of carriers for a particular radio access technology (e.g., 1.4 megahertz (MHz), 3 MHz, 5 MHz, 10 MHz, 15 MHz, 20 MHz, 40 MHz, or 80 MHz). A device of the wireless communication system 100 (e.g., a network entity 105, a UE 115, or both) may have a hardware configuration that supports communications using a particular carrier bandwidth, or may be capable of being configured to support communications using one of the carrier bandwidths in the set of carrier bandwidths. In some examples, the wireless communication system 100 may include a network entity 105 or a UE 115 that supports concurrent communications using carriers associated with multiple carrier bandwidths. In some examples, each served UE 115 may be configured to operate using a portion (e.g., a subband, a BWP) or all of the carrier bandwidth.

[0063] The signal waveform transmitted via the carrier wave may be composed of 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 techniques, a resource element may refer to a resource of one symbol period (e.g., the duration of one modulation symbol) and one 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 communication rate. Wireless communication resources may refer to a combination of RF spectrum resources, time resources, and spatial resources (e.g., spatial layers, beams), and the use of multiple spatial resources may increase the data rate or data integrity of the communication with the UE 115.

[0064] One or more numerologies for a carrier may be supported, and the numerologies may include subcarrier spacing (Δf) and cyclic prefixes. A carrier may be divided into one or more BWPs with the same or different numerologies. In some examples, a UE 115 may be configured with multiple BWPs. In some examples, a single BWP for a carrier may be active at a given time, and communications of a UE 115 may be constrained to one or more active BWPs.

[0065] 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, where 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).

[0066] 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 multiple 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 multiple symbol periods (e.g., depending on the length of the cyclic prefix added before each symbol period). In some wireless communication systems 100, the time slot may also be divided into multiple mini-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.

[0067] A subframe, a time slot, a mini-time slot, or a 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 transmission 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 (e.g., in bursts of shortened TTIs (sTTIs)) of the wireless communication system 100 may be dynamically selected.

[0068] Physical channels may be multiplexed using carriers for communication according to various techniques. For example, physical control channels and physical data channels may be multiplexed via downlink carriers for signaling 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)) for a physical control channel may be defined by a set of symbol periods and may extend across a system bandwidth or a subset of a system bandwidth of a carrier. 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 for 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 a plurality of UEs 115 , and a UE-specific search space set for transmitting control information to a specific UE 115 .

[0069] The network entity 105 may provide communication coverage via one or more cells (e.g., macro cells, small cells, hot spots, 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 identifier used to distinguish adjacent cells (e.g., a physical cell identifier (PCID), a virtual cell identifier (VCID), or other cell identifier). 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 the 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.

[0070] A macro cell generally covers a relatively large geographic area (e.g., several kilometers in radius) and may allow unrestricted access to UEs 115 that have service subscriptions with a network provider that supports the macro cell. As compared to a macro cell, a small cell may be associated with a lower power network entity 105 (e.g., a lower power base station 140), 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.

[0071] 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.

[0072] 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.

[0073] 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 either synchronous operation or asynchronous operation.

[0074] Some UEs 115, such as MTC or IoT devices, may be low-cost or low-complexity devices and may provide 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 with integrated 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, wildlife monitoring, weather and geographic event monitoring, queue management and tracking, remote security sensing, physical access control, and transaction-based commercial charging.

[0075] Some UEs 115 may be configured to employ a mode of operation 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 power conservation 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.

[0076] 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). The 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 applications or general commercial applications. The terms ultra-reliable, low-latency, and ultra-reliable low-latency are used interchangeably herein.

[0077] In some examples, a 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 various aspects of such D2D communication configured (e.g., scheduled) by the network entity 105. In some examples, one or more UEs 115 of such a group may be outside of 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, network entity 105 may facilitate scheduling of resources for D2D communications. In some other examples, D2D communications may be conducted between UEs 115 without involving network entity 105.

[0078] In some systems, the D2D communication link 135 can be an example of a communication channel (such as a sidelink communication channel) between vehicles (e.g., UE 115). In some examples, the vehicles can communicate using vehicle-to-everything (V2X) communication, vehicle-to-vehicle (V2V) communication, or some combination of these items. The vehicles can signal information related to traffic conditions, signal scheduling, weather, safety, emergency situations, or any other information related to the V2X system. In some examples, vehicles in the V2X system can communicate with roadside infrastructure (such as roadside units), or communicate with the network via one or more network nodes (e.g., network entity 105, base station 140, RU 170) using vehicle-to-network (V2N) communication, or both.

[0079] The core network 130 may provide user authentication, access authorization, tracking, Internet Protocol (IP) connections, 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 for managing access and mobility (e.g., a mobility management entity (MME), an access and mobility management function (AMF)) and at least one user plane entity for routing packets or interconnecting to an external network (e.g., a serving gateway (S-GW)), a packet data network (PDN) gateway (P-GW), or a user plane function (UPF)). The control plane entity may manage non-access stratum (NAS) functions, such as mobility, authentication, and bearer management for 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 via the user plane entity, which may provide IP address allocation and other functions. The user plane entity may be connected to an IP service 150 for 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.

[0080] 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). Typically, the region from 300 MHz to 3 GHz is referred to as the ultra-high frequency (UHF) region or decimeter band because the wavelength range is approximately one decimeter to one meter. 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) than communications using lower frequencies and longer waves in the high frequency (HF) or very high frequency (VHF) portions of the spectrum below 300 MHz.

[0081] The wireless communication system 100 may also operate in a super high frequency (SHF) region (also known as a centimeter band) in the range of 3 GHz to 30 GHz or in an extremely high frequency (EHF) region (e.g., 30 GHz to 300 GHz) (also known as a millimeter band) using a spectrum. In some examples, the wireless communication system 100 may support millimeter wave (mmW) communication between UE 115 and network entity 105 (e.g., base station 140, RU 170), and the EHF antenna of the corresponding device may be smaller and closer than the UHF antenna. In some examples, such technology may facilitate the use of antenna arrays within the device. However, the propagation of EHF transmissions may be affected by greater attenuation and shorter range than SHF or UHF transmissions. The technology disclosed herein may be adopted across transmissions using one or more different frequency regions, and the use of frequency bands specified across these frequency regions may vary by country or regulatory agency.

[0082] The wireless communication system 100 may utilize both licensed RF spectrum bands and unlicensed RF spectrum bands. For example, the wireless communication system 100 may employ license assisted access (LAA), LTE unlicensed (LTE-U) radio access technology, or NR technology using unlicensed bands such as the 5 GHz industrial, scientific, and medical (ISM) band. When operating using unlicensed RF spectrum bands, devices such as network entities 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 in combination with component carriers operating using licensed bands (e.g., LAA). Operations using unlicensed spectrum may include downlink transmissions, uplink transmissions, P2P transmissions, or D2D transmissions, and the like.

[0083] The network entity 105 (e.g., base station 140, RU 170) or UE 115 may be equipped with multiple antennas that may be used to employ techniques such as transmit diversity, receive diversity, multiple-input multiple-output (MIMO) communications, or beamforming. The antennas of the network entity 105 or UE 115 may be located within one or more antenna arrays or antenna panels that 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.

[0084] The network entity 105 may include an antenna array having a set of rows and columns of antenna ports that the network entity 105 may use to support beamforming for communications with the UE 115. Likewise, 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.

[0085] The network entity 105 or UE 115 may use MIMO communication to utilize multipath signal propagation and improve spectral efficiency by sending or receiving multiple signals via different spatial layers. Such technology may be referred to as spatial multiplexing. Multiple signals may be, for example, sent by a transmitting device via different antennas or different combinations of antennas. Similarly, multiple signals may be received by a receiving device via different antennas or different combinations of antennas. Each of the multiple signals may be referred to as a separate spatial stream and may carry information associated with the same data stream (e.g., the same codeword) or different data streams (e.g., different codewords). Different spatial layers may be associated with different antenna ports for channel measurement and reporting. MIMO technologies include single-user MIMO (SU-MIMO), in which multiple spatial layers are sent to the same receiving device; and multi-user MIMO (MU-MIMO), in which multiple spatial layers are sent to multiple devices.

[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 steer 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 orientation 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 the antenna elements may be defined by a set of beamforming weights associated with a particular orientation (e.g., relative to the antenna array of the transmitting device or the receiving device or relative to some other orientation).

[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, a beam direction associated with transmission along a 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 a reference signal (e.g., a cell-specific reference signal (CRS), a channel state information reference signal (CSI-RS)), which may be precoded or non-precoded. UE 115 may provide feedback for beam selection, which may be a precoding matrix indicator (PMI) or codebook-based feedback (e.g., a multi-panel codebook, a linear combination codebook, a 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] The wireless communication system 100 may be a packet-based network that operates according to a layered protocol stack. In the user plane, the communication at the bearer or PDCP layer may be IP-based. The RLC layer may perform packet segmentation and reassembly to communicate via logical channels. The MAC layer may perform priority processing and multiplexing of logical channels to transport channels. The MAC layer may also implement error detection techniques, error correction techniques, or both to support retransmission to improve link efficiency. In the control plane, the RRC layer may provide the establishment, configuration, and maintenance of an RRC connection that supports a radio bearer for user plane data between the UE 115 and the network entity 105 or the core network 130. The PHY layer may map a transport channel to a physical channel.

[0092] 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 a cyclic redundancy check (CRC)), forward error correction (FEC), and retransmission (e.g., automatic repeat request (ARQ)). HARQ may improve throughput at the MAC layer under poor radio conditions (e.g., low signal-to-noise ratio conditions). In some examples, a 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.

[0093] The devices in the wireless communication system 100 may implement machine learning or artificial intelligence techniques. For example, the network entity 105 may apply machine learning or artificial intelligence techniques to a data-driven algorithm that generates a set of outputs including prediction information, for example, based on a set of inputs. The algorithm may be referred to as a prediction model, which may include or may be an example of a machine learning model, an artificial intelligence model, a neural network model, or a combination thereof. The network entity 105 may collect data from other devices (e.g., UE 115, network entity 105, etc.) or both at the network entity 105 to provide as input data to the prediction model. The prediction model may include a model training function that prepares data (e.g., preprocesses, cleans, formats, and transforms input data) and performs training, validation, and testing of the prediction model using the prepared data. Training the prediction model may include providing training input data so that the prediction model "learns" appropriate outputs for the function or goal of the prediction model. The prediction model may be validated to confirm that the appropriate output is generated for a set of known input data.

[0094] In some cases, other devices in the wireless communication system 100 (e.g., network entity 105, UE 115, etc.) may additionally or alternatively utilize the prediction model, for example, at various levels of collaboration with the network entity 105. For example, the network entity 105 may jointly perform machine learning techniques with one or more other devices. Additionally or alternatively, the network entity 105 and the one or more other devices may each be associated with a corresponding prediction model, but may exchange information related to machine learning (e.g., related to the prediction model) with each other. For example, in federated learning, centralized training may be implemented using a single computing device or apparatus (e.g., a server), which may store the machine learning model and associated training data. The server may be an example of a network device (e.g., network entity 105) or a non-network device (e.g., MR). Federated learning may train a machine learning model, such as a global machine learning model, in a distributed manner. Using federated learning, one or more clients (e.g., client devices, such as UE 115 or other devices) may adopt the same machine learning model or machine learning model structure. The machine learning model or machine learning model structure may be associated with a common task associated with or performed by the UE 115. The UE 115 may independently or locally train model parameters of the machine learning model based on corresponding observations or data collection.

[0095] A machine learning model, such as an artificial neural network (ANN), may include a set of interconnected artificial neurons (e.g., a neuron model) and may be a computing device or may represent a method to be performed by a computing device. The connections of the neuron model may be modeled as weights. The machine learning model may provide predictive models, adaptive control, and other applications by training via a data set. The model may be adaptive based on external or internal information processed by the machine learning model. Machine learning may provide nonlinear statistical data models or decision making, and may model complex relationships between input data and output information.

[0096] The machine learning model or neural network may be trained. For example, the machine learning model may be trained based on supervised learning. During training, the machine learning model may be presented with inputs that the model uses to calculate to produce an output. The actual output may be compared with the target output, and the difference may be used to adjust the parameters (such as weights and biases) of the machine learning model to provide an output closer to the target output. Before training, the output may be incorrect or less accurate, and the error or difference between the actual output and the target output may be calculated. The weights of the machine learning model may then be adjusted so that the output is more closely aligned with the target. This way of adjusting the weights may be referred to as back propagation through a neural network. The process may continue until the achievable error rate stops decreasing, or until the error rate has reached a target level. The signal may travel through multiple layers of the neural network from the input at the first layer to the output at the last layer of the neural network, and may traverse these layers multiple times. As an example, the UE 115 may input information to the neural network, and may receive an output. The UE 115 may report information to a network node such as the network entity 105 based on the output.

[0097] The federated learning training process may be associated with a technique for the network to send a model to the UE 115 and / or for the UE 115 to send updated model weights to the network. One or more machine learning models may be trained before performing the training process (e.g., at a server). A server, a centralized unit (CU)-distributed unit (DU) (CU-DU), a model repository, etc. may send a training request to a network manager based on the state of the machine learning model. The training request may be based on a protocol of an operation, administration and maintenance (OAM) entity, a network entity 105 (e.g., including a RAN-based machine learning controller), and / or a network data analysis function (NWDAF). The network manager may be a service management and orchestration (SMO) / OAM, a base station, an NWDAF, etc. The request may be sent to a mobile network. In some cases, training may be performed at the mobile network. In other cases, a configuration may be sent from the mobile network to a training host, which may correspond to the network entity 105 and / or the UE 115. The UE 115 may perform the training process and send corresponding training data to the network.

[0098] For example, in a cellular network, training may correspond to centralized learning or distributed learning. Centralized learning may refer to training at a single network entity 105, which may include offline learning, where the training may be based on pre-collected data from different network entities 105 or UEs 115, such as historical data. Centralized learning may also include online learning, where the training may be based on real-time data generated or collected by the network entity 105 or UE 115. The training host may be collocated with the inference host. Distributed learning may refer to training at multiple network entities 105, UEs 115, or both. For example, distributed learning may be based on different network entities 105 or UEs 115 sending data indicating weights (e.g., updates) to a centralized MR that may aggregate the data. An example of distributed learning may include federated learning, which may be similar to centralized / online training, except that the training includes multiple training hosts (e.g., multiple UEs 115 may perform training by transmitting weights to a network node (e.g., a network entity 105 or MR that aggregates the weights)). For example, a model manager (e.g., a model repository) may aggregate weights for each training host, which may be contrasted with online learning where one node may receive / collect training data.

[0099] Typically, the RAN machine learning architecture may include different network entities corresponding to the CU control plane (CU-CP), the RAN-based machine learning controller, the CU data repository (CU-DR), the CU model repository (CU-MR), etc. The RAN-based machine learning controller may be used to control machine learning-based processes, such as updating machine learning model rules for the CU-DR. The RAN-based machine learning controller may perform RAN machine learning management processes for machine learning management at the RAN. The CU-DR may receive and store data from multiple network entities and / or provide model repository functionality. That is, the CU-DR may store and process data at the RAN. Additionally or alternatively, the CU-MR may be used to provide a model repository. The CU-MR may store machine learning models, maintain the state of machine learning models, perform machine learning weight aggregation based on configurations received from the RAN-based machine learning controller, etc. The CU-CP may be used for RAN control functions, and the CU-user plane (CU-UP) may be used for RAN user plane functions.

[0100] The techniques described herein support the management of a federated learning process. In some cases, participating in a federated learning training round may be referred to as performing a federated learning process or a training process. The federated learning process may be controlled (e.g., initiated, configured, and activated) by a MR (e.g., a server) or a network node (e.g., an OAM, a RAN node, a network entity 115). For example, the OAM or RAN may initiate a federated learning process (e.g., one or more training rounds of a federated learning process) based on performance or regional standards. In another example, the MR may initiate a training process via the OAM or RAN, for example, by sending an indication to the network entity 105 that training (e.g., for a predictive model) is required. Additionally or alternatively, the network node and the MR may coordinate control of the federated learning process. For example, the OAM or RAN may determine that training is required (e.g., for a predictive model) based on a standard and may indicate a training request to the MR, and the MR may initiate a federated learning process based on the request.

[0101] For example, the network entity 105 may configure a federated learning process for a group of UEs 115 selected by the network entity 105. The network entity 105 may configure a prediction model to be trained in the federated learning process by determining weights, a model structure (MS), a baseline parameter set (PS), etc. Additionally, the network entity 105 may select a set of training parameters for a training configuration for one or more training rounds of the federated learning process. The set of training parameters may include an MS identifier (MS ID), a PS identifier (PSID), a training periodicity, a number of epochs E, a training valid area, a training deadline, or any combination thereof. The network entity 105 may indicate the training configuration to the group of UEs 115. Each UE 115 may download an MS and a PS (e.g., based on an MS ID and a PS ID).

[0102] In some examples, the network entity 105 may select one or more training parameters based on one or more UEs 115 in the group of UEs 115. For example, the network entity 105 may select the number of epochs (e.g., iterations) to be performed at the UE 115 for a given training round based on the computing power of the UE 115, the estimated link capacity of the UE 115, etc. In addition, the techniques described herein support the UE 115 to choose to participate in a training round or avoid participating in a training round, which may be based on a flexible number of epochs. The network entity 105 may configure the minimum number of epochs as a training parameter for the training configuration. The UE 115 facilitating the federated learning session may estimate the number of local epochs that the UE 115 is capable of performing (e.g., based on the computing power of the UE 115, the estimated link capacity of the UE 115, etc.), and may compare the number of local epochs with the minimum number of epochs. If the UE 115 is capable of performing the minimum number of epochs (e.g., if the local number of epochs of the UE 115 is greater than the minimum number of epochs), the UE 115 may determine to participate in the training round. Alternatively, if UE 115 cannot support the minimum number of epochs (e.g., if the local number of epochs of UE 115 is less than the minimum number of epochs), UE 115 may determine to avoid participating in the training round. In either case, UE 115 may indicate to network entity 105 whether UE 115 is participating in the training round.

[0103] The network entity 105 may activate the federated learning process at all UEs 115 simultaneously by sending an activation message to the UEs 115. The activation message may trigger each UE 115 to start training the prediction model locally according to the training configuration. After completing the training rounds (e.g., at a configured training deadline or after executing a configured number of epochs), each UE 115 may report the model parameters output from the prediction model to the network entity 105.

[0104] In some cases, each UE 115 may report model parameters to a server (e.g., MR). The reported model parameters may be understood as updated model parameters that are different from the baseline PS. The server may compile the updated model parameters from the UE 115 to generate global or aggregated model parameters. In some examples, this may be referred to as updating the prediction model, where the information reported from the UE 115 is used to train the prediction model at the server, and the updated version of the prediction model is generated using the collected data. The server may assign a PS ID (e.g., a new PS ID that is different from the PS ID associated with the baseline PS) to the set of aggregated model parameters. The server may transmit or forward the aggregated model parameters and the new PS ID to the UE 115, and the UE 115 may update the local prediction model with the updated model parameters. For example, the server may send an indication of the updated prediction model (such as an indication of the corresponding MS ID) to the UE 115.

[0105] Figure 2 An example of a wireless communication system 200 supporting management of federated learning 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 and may include a plurality of UEs 115 (UE 115-a, UE 115-b, UE 115-c, and UE 115-d), a network entity 105-a, and a server 203, which may be reference Figure 1 Examples of corresponding devices. Although described as communication between UE 115, network entity 105-a and server 203, any type or number of devices can implement the technology described herein. In addition, the technology described herein can be implemented by any type or number of devices of any wireless communication system.

[0106] The network entity 105-a, the UE 115, and the server 203 may be examples of devices that implement data-driven machine learning techniques. Each of the network entity 105-a, the UE 115, and the server 203 may use one or more prediction models (e.g., machine learning models, artificial intelligence models, neural network models, etc.) for one or more functions or goals. The server 203 may include an MR or may be an example of an MR. The network entity 105-a may correspond to an OAM, a base station (e.g., a gNB, a RAN-based machine learning controller), an NWDAF, etc.

[0107] The wireless communication system 200 may support a federated learning process to train a prediction model. For the federated learning process, multiple clients (such as UE 115) may adopt the same prediction model structure to locally train model parameters for the prediction model based on observations or data. The network entity 105-a or the server 203 may select or otherwise determine a model configuration for the prediction model. The model configuration may include an MS and a baseline PS. The baseline PS may include one or more weights (e.g., provided by UE 115 during the federated learning process) to be provided as input to the prediction model for evaluation or be associated with the one or more weights. For example, the network entity 105-a or the server 203 may configure the prediction model to be trained in the federated learning process by determining the weights, the MS, and the baseline PS. The network entity 105-a or the server 203 may additionally determine an MS ID associated with the MS and a PS ID associated with the baseline PS. The network entity 105-a or the server 203 may configure the UE 115 for the federated learning process by, for example, indicating the MS ID and the PS ID to the UE 115.

[0108] Each UE 115 may download the MS and PS via the user plane (e.g., based on the MS ID and the PS ID). In some cases, the network entity 105-a or the server 203 may configure a dedicated data radio bearer for the UE 115 to download the MS and PS, and may send an indication of the data radio bearer to the UE 115. After successfully downloading the MS and PS (e.g., via the data radio bearer) and configuring the prediction model at the UE 115, each UE 115 may send a configuration complete message to the network entity 105-a. The configuration complete message may be an example of a radio resource control (RRC) message (e.g., an RRCConfigurationComplete message).

[0109] The UE 115 may train the machine learning model according to a training configuration 215 received from the network entity 105-a or the server 203, wherein the training configuration 215 includes a set of training parameters. In some examples, the set of training parameters may include an MS ID and a baseline PS ID, so that the UE 115 may receive an indication of the MS ID and the baseline PS ID as part of the training configuration 215. Based on the training, the UE 115 may transmit information about updated model parameters to the network entity 105-a, the server 203, or both. The federated learning process may include multiple training rounds, each of which may be dynamically scheduled by the network entity 105-a. The training round may include the UE 115 receiving and implementing the training configuration 215, training the prediction model according to the training configuration 215, and reporting updated model parameters output from the training. The network entity 105-a or the server 203 may compile information from all participating UEs 115 to determine global or aggregated model parameters, and may then transmit updated model parameters or an updated machine learning model to the client. For example, the network entity 105 - a or the server 203 may receive information regarding model parameters, update the machine learning model, and send information regarding an updated version of the machine learning model to the client for another round of the federated learning process.

[0110] In general, the network entity 105-a or the server 203 may determine to initiate a federated learning process (e.g., a training process, a training round for a federated learning process) to train a prediction model based on a trigger (e.g., a trigger condition). The trigger may include or may be an example of a trigger condition such as a communication condition, an operating condition, etc. For example, the network entity 105-a or the server 203 may be triggered to initiate a federated learning process based on satisfying a performance standard or a region-specific standard. In some cases, the server 203 may send an initiation message 205 to the network entity 105-a based on the server 203 determining that the trigger condition is satisfied. The initiation message 205 may indicate that the network entity 105-a is to initiate or activate a federated learning process. Additionally or alternatively, the network entity 105-a may, for example, send a request message 210 indicating a request for model training to the server 203 based on the trigger. That is, the network entity 105-a may determine that the trigger condition for performing the federated learning process is satisfied, and may send the request message 210 to the server 203.

[0111] Initiation of a federated learning process may refer to determining that a prediction model needs to be trained, e.g., a federated learning process or a training round of a federated learning process is to be performed. In some examples, initiation may also include selecting a UE 115 to participate in the federated learning process, configuring a prediction model, determining a training configuration 215, and indicating information about the federated learning process (e.g., prediction model configuration, training configuration 215) to the participating UE 115. Activation of a federated learning process may refer to activating a federated learning process (e.g., a training round of a federated learning process) at a UE 115, e.g., by sending an activation message commanding the UE to start training the prediction model locally according to the training configuration 215. In some cases, the server 203 may initiate and activate a federated learning process, which may be referred to as an MR-controlled federated learning process. In other cases, the network entity 105-a may initiate and activate a federated learning process, which may be referred to as or understood to be an OAM- or RAN-controlled federated learning process (e.g., the federated learning process is controlled by the OAM or RAN via the network entity 105-a).

[0112] Additionally or alternatively, the network entity 105-a and the server 203 may coordinate management of the federated learning process. For example, if the server 203 determines that training is needed based on a trigger, the server 203 may initiate the federated learning process and may send an initiation message 205 to the network entity 105-a commanding the network entity 105-a to start the federated learning process. Based on receiving the initiation message 205, the network entity 105-a may select the UE 115 to participate in the federated learning process, determine the MS and baseline PS for the prediction model, select training parameters for the training configuration 215, and indicate the MS, PS, and training configuration 215 to the UE 115. In some examples, the network entity 105-a may request the training configuration 215, the model configuration, or both from the server 203. The request message 210 may include an example of a model provisioning request. In some cases, the network entity 105-a may send the request message 210 based on receiving the initiation message 205. For example, the network entity 105-a may receive an initiation message 205 indicating that the network entity 105-a is to initiate a federated learning process, and may send a request message 210 to obtain a training configuration 215 for the federated learning process. In response to the request message 210, the server 203 may send an indication of the training configuration 215 to the network entity 105-a, the UE 115, or both. The network entity 105-a may activate the federated learning process after the server 203 sends the training configuration 215.

[0113] In some examples, the network entity 105-a or the server 203 may select a subset of the configured UEs 115 to participate in a federated learning process or a training round of a federated learning process. For example, the network entity 105-a or the server 203 may configure a total number N of UEs 115 for a federated learning process by sending an MS, a baseline PS, and a training configuration 215 to the N UEs 115. For each training round of the federated learning process, the network entity 105-a or the server 203 may select a number K of UEs 115 to participate, where K is less than N. In some cases, the number K of UEs 115 may be different for different training rounds. The network entity 105-a or the server 203 may activate the training round at the K participating UEs 115 by sending an activation message to the K UEs 115 (e.g., and avoiding sending an activation message to any non-participating UEs 115).

[0114] Additionally, the network entity 105-a or the server 203 may select a set of training parameters for the training configuration 215 of the federated learning process. The set of training parameters may include parameters according to which the UE 115 is to perform a training process and send a report indicating updated model parameters. For example, the set of training parameters may include a training periodicity (e.g., a periodicity for training a prediction model), a training granularity, an MR address, a reporting granularity, a reporting periodicity (e.g., a periodicity for sending reports), a reporting configuration, a training valid area, an epoch number E (e.g., a maximum epoch number, a minimum epoch number, or both), a training deadline, or any combination thereof. The set of training parameters may also include an MS ID and a PS ID.

[0115] The training effective area may indicate a geographical area in which the federated learning process will be performed by the UE 115, such as all or part of a cell served by the network entity 105-a. That is, if or when the UE 115 is located within the training effective area, the UE 115 may perform the federated learning process. The number of epochs E and the training deadline may be associated with the completion status of the training round. For example, the training deadline may indicate the time when the UE 115 will stop (e.g., complete) performing the training process, the duration for which the UE 115 will complete the training process, and the like. In some cases, the training deadline may additionally include the time or duration for which the UE 115 will report the updated model parameters. The number of epochs E may indicate the number of iterations (e.g., weight updates) to be performed by the UE 115 during the training process, wherein the iteration includes providing a set of input data to a prediction model to obtain a set of output data. The UE 115 may stop performing the training process after reaching the configured number of epochs.

[0116] The network entity 105-a or the server 203 may indicate the training configuration 215 (e.g., a set of training parameters) to the UE 115. For example, the network entity 105-a or the server 203 may send control signaling (e.g., layer 3 (L3) control signaling), such as an RRC message, to the UE 115 indicating the training configuration 215 and the set of training parameters. In some examples, the control signaling may also include a model configuration (e.g., MS and PS) in addition to the training configuration 215. In some cases, the network entity 105-a or the server 203 may independently configure each UE 115 by sending corresponding control signaling to each UE 115, while in other cases, the network entity 105-a or the server 203 may configure multiple UEs 115 (e.g., K UEs 115) using the same control signal.

[0117] In some cases, changes in the capabilities or conditions of the UE 115 may introduce constraints on the federated learning process. For example, each of the UEs 115 may have different computing capabilities, processing capabilities, power, memory space, link capabilities, etc., so that, for example, some UEs 115 may take an extended amount of time to download the MS and PS or perform the federated learning process compared to other UEs 115. Therefore, the network entity 105-a or the server 203 may receive reports including updated model parameter information from the UE 115 at different times, which may introduce latency and reduce the efficiency of the federated learning process. Additionally or alternatively, some UEs 115 may be unavailable or unable to participate in the federated learning process or a given training round of the federated learning process. For example, limited power or computing power may prevent the UE 115 from successfully completing the configured number of epochs E before the configured training deadline. In some cases, if the network entity 105-a or server 203 receives a report from the UE 115 after a given time window associated with the federated learning process, the network entity 105-a or server 203 may discard or otherwise drop the report, and the signaling used to configure and activate the federated learning process at the UE 115 and the computational effort of the UE 115 may be wasted. In addition, if the UE 115 is in an idle or inactive mode, has a relatively poor connection to the network entity 105-a or server 203, or moves out of the coverage area of ​​the network entity 105-a, the UE 115 may not be available to participate in the federated learning process or a given training round of the federated learning process.

[0118] Thus, the network entity 105-a or the server 203 may select one or more training parameters based on one or more of the UEs 115, e.g., based on computational, connectivity, and availability constraints of the UEs 115. For example, the network entity 105-a or the server 203 may select a set of training parameters based on an estimated link capacity associated with one or more of the UEs 115, a computational capability of one or more of the UEs 115, and the like. In some cases, the network entity 105-a or the server 203 may support a flexible number of epochs for the training configuration 215, such that each UE 115 may determine a corresponding number of local epochs to be performed at the UE 115 (e.g., based on the capabilities of the UE 115). Additionally or alternatively, participation in the federated learning process or the number of training rounds to participate in the federated learning process may be optional based on the constraints of a given UE 115. That is, a UE 115 receiving the training configuration 215 may determine whether to participate in the federated learning process based on the one or more training parameters and the capabilities of the UE 115.

[0119] As a specific example, the network entity 105-a (e.g., or the server 203) may select as a training parameter a minimum number of epochs M that a given UE 115 is to complete for a training round of a federated learning process. The minimum number of epochs M may be selected based on an estimated link capacity or computing power of one or more UEs 115. A UE 115 that receives a training configuration 215 including the minimum number of epochs M may estimate a number of local epochs F that the UE 115 is able to complete based on the computing power or link capacity associated with the UE 115. In some cases, the number of local epochs F may be based on a training deadline, for example, may represent a number of epochs that the UE 115 is able to complete within or before the training deadline. The UE 115 may compare the estimated number of local epochs F with the minimum number of epochs M to determine whether the UE 115 is able to participate in the federated learning process.

[0120] If F is greater than M, the UE 115 may be able to complete the federated learning process according to the training configuration 215 (e.g., may be able to execute at least M epochs within the training deadline and, in some cases, send a report within the training deadline). The UE 115 may choose to participate in the federated learning process based on F being greater than M. Alternatively, if F is less than M, the UE 115 may not be able to complete the federated learning process according to the training configuration 215, and may determine to avoid implementing the training configuration 215 and avoid participating in the federated learning process. In some examples, the UE 115 may send an indication of the decision (e.g., whether the UE 115 will participate in the federated learning process) to the network entity 105-a or the server 203. For example, the UE 115 may send a message (e.g., an RRC message, such as an RRC configuration complete message) to the network entity 105-a or the server 203 indicating that the UE 115 has implemented the training configuration 215 or has avoided implementing the training configuration 215. Additionally or alternatively, UE 115 may report the estimated number of local epochs F to network entity 105-a or server 203. In some cases, network entity 105-a or server 203 may select training parameters for subsequent training rounds based on F. For example, network entity 105-a or server 203 may select a minimum number of epochs for subsequent training rounds that is equal to or less than F so that UE 115 can participate in subsequent training rounds.

[0121] Each UE 115 may configure the training process at the UE 115 according to the training configuration 215 and the MS and the baseline PS. Based on receiving (e.g., implementing) the training configuration 215, and after downloading the MS and the baseline PS, each UE 115 participating in the federated learning process may send a message to the network entity 105-a or the server 203 indicating that the prediction model is ready to be activated at the UE 115. The message may include an RRC configuration complete message or may be an example of an RRC configuration complete message. In some cases, the message may additionally include an indication of whether the UE 115 is participating in the federated learning process, an indication of the corresponding estimated number of local epochs F, or both.

[0122] After receiving a corresponding message from each UE 115 indicating whether the prediction model is ready to be activated at the UE 115, the network entity 105-a or the server 203 can activate the federated learning process (e.g., training rounds) at all participating UEs 115 simultaneously by sending an activation indication to the UE 115. The activation indication may include or may be an example of layer 1 (L1) or layer 2 (L2) control signaling, such as downlink control information (DCI) or a medium access control (MAC) control element (MAC-CE). The UE 115 may start training the prediction model locally based on receiving the activation indication and according to the training configuration 215.

[0123] The UE 115 may complete the federated learning process (e.g., training rounds) based on a training deadline, after achieving a configured number of epochs (e.g., number of epochs E or at least a minimum number of epochs M), or both. After completing the federated learning process, the UE 115 may send a report indicating a set of model parameters (e.g., updated model parameters) obtained based on the trained prediction model to the network entity 105-a or the server 203. The report may include information associated with the set of model parameters (e.g., weights), which the network entity 105-a or the server 203 may use to generate aggregated (e.g., global) model parameters. For example, the network entity 105-a or the server 203 may compile (e.g., aggregate) the updated model parameters received from the UE 115 into a set of aggregated model parameters.

[0124] The network entity 105-a or the server 203 may assign a new PS ID 220 (e.g., a PS ID different from the PS ID associated with the baseline PS) to the set of aggregated model parameters. For example, the server 203 may select the new PS ID 220 for the set of aggregated parameters including the PS ID associated with the baseline PS and the version tag. Alternatively, the server 203 may select a temporary PS ID as the new PS ID 220. The server 203 may transmit or forward the new PS ID 220 to the UE 115. In some cases, the server 203 may send the new PS ID 220 to the network entity 105-a, and the network entity 105-a may forward the new PS ID 220 to the UE 115, for example, via L2 or L3 signaling (e.g., via RRC signaling, MAC-CE, DCI). The UE 115 may download (e.g., via a user plane) or otherwise obtain the set of aggregated model parameters based on the new PS ID 220.

[0125] In some examples, the UE 115 may update the local prediction model with the set of aggregated model parameters. Additionally or alternatively, the aggregated model parameters may be used to train the prediction model at the network entity 105-a or the server 203 to obtain an updated version of the prediction model, and the network entity 105-a or the server 203 may send an indication of the updated prediction model (such as an indication of the corresponding MS ID) to the UE 115. The UE 115 may download (e.g., via a user plane) the updated prediction model based on the corresponding MS ID. The UE 115 and the network entity 105-a or the server 203 may perform another round of the federated learning process using the set of aggregated model parameters and the updated prediction model. In some examples, the UE 115 may continue to use the previously indicated training configuration 215. In other examples, the network entity 105-a or the server 203 may update or otherwise modify one or more training parameters of the training configuration 215, and may send the updated training configuration 215 to the UE 115.

[0126] Figure 3 An example of a process flow 300 for supporting management of federated learning in accordance with one or more aspects of the present disclosure is illustrated. The process flow 300 may implement or be implemented to implement aspects of the wireless communication system 100 or the wireless communication system 200. For example, the process flow 300 may illustrate communications between a UE 115-e, a network entity 105-b, and a server 303 (which may be examples of corresponding devices described herein).

[0127] In the following description of the process flow 300, operations may be performed (e.g., reported or provided) in a different order than shown, or operations performed by the example devices may be performed in a different order or at different times. Additionally, although the process flow 300 is described with reference to the UE 115-e, the network entity 105-c, and the server 303, any type of device or combination of devices may perform the described operations. Some operations may also be omitted from the process flow 300, or other operations may be added to the process flow 300. Furthermore, although some operations or signaling are shown as occurring at different times for discussion purposes, these operations may actually occur simultaneously or otherwise concurrently.

[0128] The process flow 300 illustrates an example of an OAM or RAN controlled training process (e.g., a federated learning process) for a predictive model. The training process may be controlled by the OAM or RAN via a network node (such as the network entity 105-b). For example, although the training process may be initiated by the server 303 or the network entity 105-b, or the server 303 may supply a training configuration or a model configuration for the training process to the network entity 105-b, the network entity 105-b may control the training process via communication with the UE 115-e (and any additional UEs 115 participating in the training process). That is, the network entity 105-b may indicate the training configuration to the UE 115-e, and may activate or deactivate the training process at the UE 115-e. In addition, the network entity 105-b may collect reports (e.g., updated model parameters) from the UE 115-e according to the training configuration. While process flow 300 illustrates a training process with UE 115 - e , it should be understood that additional UEs 115 not depicted may also participate in the training process.

[0129] At 305, server 303 may optionally initiate a training process. For example, server 303 may send an initiation message indicating that a training process is to be activated by network entity 105-b, and network entity 105-b may receive the initiation message. Server 303 may send the initiation message to network entity 105-b based on a trigger (e.g., based on server 303 determining that a trigger condition is satisfied). In some examples, network entity 105-b receiving the initiation message may be a trigger for network entity 105-b to activate the training process.

[0130] At 310, the network entity 105-b may optionally send a request message indicating a request to initiate or activate a training process, and the server 303 may receive the request message. In some examples, the request message may include a model provisioning request indicating a request for a model configuration of a prediction model or an example of the model provisioning request. The model configuration may include a prediction model, an MS associated with the prediction model, a baseline PS of the prediction model, an MS ID, a baseline PS ID, or a combination thereof. Additionally or alternatively, the request message may indicate a request for a training configuration of the training process. In some examples, the server 303 may send an initiation message to the network entity 105-b based on receiving the request message at 310.

[0131] At 315, network entity 105-b may select one or more UEs (including at least UE 115-e) for the training process based on a trigger to activate the training process. In some examples, the trigger to activate the training process may be, for example, receiving an initiation message from server 303 at 305.

[0132] At 320, in some examples, server 303 may send a training configuration including a set of training parameters for the training process, and network entity 105-b may receive the training configuration. In some cases, server 303 may send the training configuration to network entity 105-b based on receiving the request message at 310.

[0133] At 325, if the network entity 105-b does not receive the training configuration from the server 303 (e.g., at 320), the network entity 105-b may select a set of training parameters for the training configuration based on one or more UEs (e.g., at least based on UE 115-e). For example, the network entity 105-b may select an MS ID associated with the prediction model, a baseline PS ID associated with the prediction model, a training valid area, a maximum number of epochs for the training process, a minimum number of epochs for the training process, a training deadline, a set of weights, a training periodicity, a reporting periodicity, a server address (e.g., an MR address), or a combination thereof. In some cases, the network entity 105-b may select one or more of the training parameters based on an estimated link capacity associated with the one or more UEs, a computing capacity associated with the one or more UEs, or a combination thereof. For example, the network entity 105-b may select a minimum number of epochs to be performed by the one or more UEs for the training process based on an estimated link capacity associated with the one or more UEs, a computing capacity associated with the one or more UEs, or both.

[0134] At 330, the network entity 105-b may configure data radio bearers for one or more UEs for downloading MSs and baseline PSs for the training procedure. The network entity 105-b may send an indication of the configured data radio bearers, and the UE 115-e may receive the indication.

[0135] At 335, in some examples, the network entity 105-b may send the MS and the baseline PS, and the UE 115-e may receive the MS and the baseline PS. In some cases, the UE 115-e may download the MS and the baseline PS via the user plane based on the corresponding MS ID and the baseline PS ID. For example, the UE 115-e may download the MS and the baseline PS via the configured data radio bearer indicated at 335.

[0136] At 340 , the network entity 105 - b may send a message (eg, an L3 message such as an RRC message) indicating a training configuration including a set of training parameters, and the UE 115 - e may receive the message.

[0137] At 345, the UE 115-e may optionally estimate a number of local epochs for the training process based on the computing power of the UE 115-e, the link capacity associated with the UE 115-e, or both. In some cases, the UE 115-e may estimate the number of local epochs based on a training configuration that indicates a minimum number of epochs for the UE 115-e.

[0138] At 350, the UE 115-e may compare the estimated number of local epochs to the minimum number of epochs indicated in the training configuration. The UE 115-e may determine whether to participate in the training process and subsequently implement the training configuration based on the comparison. For example, the UE 115-e may determine that the estimated number of local epochs is less than the minimum number of epochs, and therefore may determine to avoid implementing the training configuration (e.g., and avoid participating in the training process). Alternatively, the UE 115-e may determine that the estimated number of local epochs is greater than or equal to the minimum number of epochs, and may determine to implement the training configuration (e.g., and to participate in the training process).

[0139] At 355, the UE 115-e may implement a training configuration. The UE 115-e may configure a training process according to the set of training parameters and the training configuration. For example, the UE 115-e may configure a prediction model according to the MS and MS ID and using a baseline PS (e.g., based on a baseline PS ID). The UE 115-e may determine the number of epochs to use for the training process based on the indicated number of epochs, based on the estimated number of local epochs, or based on both. In some cases, if the UE 115-e determines to implement the training configuration and participate in the training process, for example based on the estimation and the comparison at 350, the UE 115-e may only configure the training process at 355.

[0140] At 360, the UE 115-e may send a message (e.g., a configuration report) indicating whether the UE has implemented the training configuration of the training process, e.g., based on the estimation and the comparison at 350. The message may include an RRC completion message or may be an example of an RRC completion message. In some cases, the message may additionally indicate that the UE 115-e has completed configuration of the training process (e.g., at 355), that the prediction model and the training process are ready to be activated at the UE 115-e, or both.

[0141] At 365, the network entity 105-b may send a message including an activation indication for the training procedure, and the UE 115-e may receive the message. The message including the activation indication may be an example of L1 or L2 signaling, such as MAC-CE or DCI. The activation indication may command the UE 115-e to activate (e.g., start) the training procedure. In some cases, the network entity 105-b may send the activation indication based on receiving the configuration report from the UE 115-e at 360.

[0142] At 370, the UE 115-e may perform a training process using the prediction model to obtain a set of one or more model parameters based on the training configuration. For example, the UE 115-e may locally train the prediction model using the baseline PS as input data. The UE 115-e may determine updates to the parameters of the prediction model based on training the prediction model using the baseline PS. In some cases, the UE 115-e may perform the training process at 370 based on receiving the activation indication at 365.

[0143] At 375, the UE 115-e may send a report indicating a set of model parameters obtained from the training process (e.g., obtained as output from the prediction model after completing the training process), and the network entity 105-b may receive the report. In some cases, the report may include information associated with the set of model parameters, such as weights. In some cases, the UE 115-e may additionally or alternatively send the report to the server 303.

[0144] Figure 4 An example of a process flow 400 for supporting management of federated learning in accordance with one or more aspects of the present disclosure is illustrated. The process flow 400 may implement or be implemented to implement aspects of the wireless communication system 100 or the wireless communication system 200. For example, the process flow 400 may illustrate communications between a UE 115-f, a network entity 105-c, and a server 403 (which may be examples of corresponding devices described herein).

[0145] In the following description of the process flow 400, operations may be performed (e.g., reported or provided) in a different order than shown, or operations performed by the example devices may be performed in a different order or at different times. Additionally, although the process flow 400 is described with reference to the UE 115-f, the network entity 105-c, and the server 403, any type of device or combination of devices may perform the described operations. Some operations may also be omitted from the process flow 400, or other operations may be added to the process flow 400. Furthermore, although some operations or signaling are shown as occurring at different times for discussion purposes, these operations may actually occur simultaneously or otherwise concurrently.

[0146] Process flow 400 illustrates an example of an MR-controlled training process (e.g., a federated learning process) for a prediction model. The training process may be controlled by a server 403, which may include an MR or be an example of an MR. For example, although the training process may be initiated by server 403 or network entity 105-c, server 403 may control the training process via communication with UE 115-f (and any additional UE 115 participating in the training process). In some cases, server 403 may communicate directly with UE 115-f, while in other cases, server 403 may communicate with UE 115-f via network entity 105-c. Although process flow 400 illustrates a training process with UE 115-f, it should be understood that additional UEs 115 not depicted may also participate in the training process.

[0147] At 405, server 403 may optionally initiate a training process. For example, server 403 may send an initiation message indicating that the training process is activated, and network entity 105-c may receive the initiation message. Server 403 may send the initiation message to network entity 105-c based on a trigger (e.g., based on server 403 determining that a trigger condition is satisfied). In some examples, network entity 105-c receiving the initiation message may be a trigger for network entity 105-c to activate the training process.

[0148] At 410, the network entity 105-c may optionally send a request message indicating a request to initiate or activate a training process, and the server 403 may receive the request message. In some cases, the network entity 105-c may send the request message based on a trigger. In some examples, the request message may include a model provisioning request indicating a request for a model configuration of a prediction model or an example of the model provisioning request. The model configuration may include a prediction model, an MS associated with the prediction model, a baseline PS of the prediction model, an MS ID, a baseline PS ID, or a combination thereof. Additionally or alternatively, the request message may indicate a request for a training configuration of the training process. In some examples, the server 403 may send an initiation message to the network entity 105-c based on receiving the request message at 410.

[0149] At 415, server 403 may select one or more UEs (including at least UE 115-f) for the training process based on a trigger to activate the training process. In some examples, the trigger to activate the training process may be, for example, receiving a request message at 405 from network entity 105-c.

[0150] At 420, server 403 may select a set of training parameters for training configuration based on one or more UEs (e.g., at least based on UE 115-f). For example, server 403 may select an MS ID associated with a prediction model, a baseline PS ID associated with a prediction model, a training valid area, a maximum number of epochs for a training process, a minimum number of epochs for a training process, a training deadline, a set of weights, a training periodicity, a reporting periodicity, a server address (e.g., an MR address), or a combination thereof. In some cases, server 403 may select one or more of the training parameters based on an estimated link capacity associated with one or more UEs, a computing power associated with one or more UEs, or a combination thereof. For example, server 403 may select a minimum number of epochs to be performed by one or more UEs for a training process based on an estimated link capacity associated with one or more UEs, a computing power associated with one or more UEs, or both.

[0151] At 425, server 403 may configure data radio bearers for one or more UEs for downloading MSs and baseline PSs for the training process. Server 403 may send an indication of the configured data radio bearers, and UE 115-f may receive the indication.

[0152] At 430, in some examples, server 403 may send the MS and the baseline PS, and UE 115-f may receive the MS and the baseline PS. In some cases, UE 115-f may download the MS and the baseline PS via the user plane based on the corresponding MS ID and the baseline PS ID. For example, UE 115-f may download the MS and the baseline PS via the configured data radio bearer indicated at 425.

[0153] At 435 , the server 403 may send a message (eg, an L3 message such as an RRC message) indicating a training configuration including a set of training parameters, and the UE 115 - f may receive the message.

[0154] At 440, UE 115-f may optionally estimate a number of local epochs for the training process based on a computational capability of UE 115-f, a link capacity associated with UE 115-f, or both. In some cases, UE 115-f may estimate the number of local epochs based on a training configuration that indicates a minimum number of epochs for UE 115-f.

[0155] At 445, UE 115-f may compare the estimated number of local epochs to the minimum number of epochs indicated in the training configuration. UE 115-f may determine whether to participate in the training process and subsequently implement the training configuration based on the comparison. For example, UE 115-f may determine that the estimated number of local epochs is less than the minimum number of epochs, and therefore may determine to avoid implementing the training configuration (e.g., and avoid participating in the training process). Alternatively, UE 115-f may determine that the estimated number of local epochs is greater than or equal to the minimum number of epochs, and may determine to implement the training configuration (e.g., and to participate in the training process).

[0156] At 450, the UE 115-f may implement a training configuration. The UE 115-f may configure a training process based on a set of training parameters and the training configuration. For example, the UE 115-f may configure a prediction model based on the MS and MS ID and using a baseline PS (e.g., based on a baseline PS ID). The UE 115-f may determine the number of epochs to use for the training process based on the indicated number of epochs, based on the estimated number of local epochs, or based on both. In some cases, if the UE 115-f determines to implement the training configuration and participate in the training process, for example based on the estimation and the comparison at 445, the UE 115-f may only configure the training process at 450.

[0157] At 455, the UE 115-f may send a message (e.g., a configuration report) indicating whether the UE has implemented the training configuration of the training process, e.g., based on the estimation and the comparison at 445, and the network entity 105-c may receive the message. The message may include an RRC completion message or may be an example of an RRC completion message. In some cases, the message may additionally indicate that the UE 115-f has completed the configuration of the training process (e.g., at 450), that the prediction model and the training process are ready to be activated at the UE 115-f, or both.

[0158] At 460, the network entity 105-c may send a message including an activation indication for the training procedure, and the UE 115-f may receive the message. The message including the activation indication may be an example of L1 or L2 signaling, such as MAC-CE or DCI. The activation indication may command the UE 115-f to activate (e.g., start) the training procedure. In some cases, the network entity 105-c may send the activation indication based on receiving the configuration report from the UE 115-f at 455.

[0159] At 465, the UE 115-f may perform a training process using the prediction model to obtain a set of one or more model parameters based on the training configuration. For example, the UE 115-f may locally train the prediction model using the baseline PS as input data. The UE 115-f may determine updates to the parameters of the prediction model based on training the prediction model using the baseline PS. In some cases, the UE 115-f may perform the training process at 465 based on receiving the activation indication at 460.

[0160] At 470, the UE 115-f may send a report indicating a set of model parameters obtained from the training process (e.g., obtained as output from the prediction model after completing the training process), and the network entity 105-c, the server 403, or both may receive the report. In some cases, the report may include information associated with the set of model parameters, such as weights.

[0161] Figure 5 An example of a process flow 500 for supporting management of federated learning in accordance with one or more aspects of the present disclosure is illustrated. The process flow 500 may implement or be implemented to implement aspects of the wireless communication system 100 or the wireless communication system 200. For example, the process flow 500 may illustrate communications between a UE 115-g, a network entity 105-d, and a server 503 (which may be examples of corresponding devices described herein). Although the process flow 500 illustrates a training process with a UE 115-g, it should be understood that additional UEs 115 not depicted may also participate in the training process.

[0162] In the following description of the process flow 500, operations may be performed (e.g., reported or provided) in a different order than shown, or operations performed by the example devices may be performed in a different order or at different times. Additionally, although the process flow 500 is described with reference to the UE 115-g, the network entity 105-d, and the server 503, any type of device or combination of devices may perform the described operations. Some operations may also be omitted from the process flow 500, or other operations may be added to the process flow 500. Furthermore, although some operations or signaling are shown as occurring at different times for discussion purposes, these operations may actually occur simultaneously or otherwise concurrently.

[0163] At 505, the server 503 may send a training configuration for a training process associated with the prediction model to a set of UEs including the UE 115-g. The training configuration may include a first set of model parameters associated with the first PS ID. The training configuration may also include a set of training parameters for the training process, such as an MS ID associated with the prediction model, a training validity area, a maximum number of epochs for the training process, a minimum number of epochs for the training process, a training deadline, a weight set, a training periodicity, a reporting periodicity, a server address (e.g., an MR address), or a combination thereof. In some cases, the server 503 may additionally send the training configuration to the network entity 105-d.

[0164] At 510, the set of UEs including UE 115-g may send a report indicating a subset of model parameters obtained from the training process (e.g., obtained as output from the prediction model after completing the training process), and the server 503 may receive the report. In some cases, the report may include information associated with the subset of model parameters, such as weights. For example, the server 503 may receive a respective report from each UE in the set of UEs, where each report includes a respective subset of model parameters obtained by the corresponding UE.

[0165] At 515, the server 503 may aggregate the received subset of model parameters into a second set of model parameters. The second set of model parameters may include an updated set of model parameters for the prediction model or may be an example of an updated set of model parameters for the prediction model, which, for example, will be used in a subsequent round of the training process.

[0166] At 520, server 503 may optionally update the prediction model, the training configuration, or both. For example, server 503 may train the prediction model at server 503 using the second set of model parameters to obtain an updated prediction model, which may be associated with an updated MS ID (e.g., selected by server 503). Additionally or alternatively, server 503 may update or otherwise modify one or more training parameters of the training configuration (e.g., based on the second set of model parameters).

[0167] At 525, server 503 may assign a second PS ID to the second set of model parameters. The second PS ID may be different from the first PS ID. In some examples, the second PS ID may include the first PS ID combined with a version tag. In other examples, the second PS ID may be a temporary PS ID.

[0168] At 530, server 503 may send a message indicating the second PS ID, and the set of UEs including UE 115-g may receive the message. In some examples, UE 115-g may download a second set of model parameters via the user plane based on the second PS ID. In some cases, server 503 may additionally send the second PS ID to network entity 105-d.

[0169] At 535, server 503 may optionally send a message indicating a second set of model parameters for the prediction model, and the set of UEs including UE 115-g may receive the message. UE 115-g may update the prediction model based on (e.g., using) the second PSID, the second set of model parameters, or both. In some cases, if server 503 updates the training configuration (e.g., updates one or more training parameters for the training configuration), server 503 may indicate the updated training configuration to UE 115-g, and UE 115-g may configure the training process according to the updated training configuration.

[0170] At 540, based on receiving the second PS ID and optionally the second set of model parameters, the updated training configuration, or both, UE 115-g may send a message indicating that the prediction model (e.g., and the training process) is ready to be activated at UE 115-g and server 503 may receive the message.

[0171] At 545 , the UE 115 - g may perform a training process for the prediction model using the second set of model parameters and based on the second PS ID.

[0172] Figure 6A block diagram 600 of a device 605 supporting management of federated learning according to one or more aspects of the present disclosure is shown. The device 605 may be an example of aspects of the network entity 105 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).

[0173] Receiver 610 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 communicated to other components of device 605. In some examples, receiver 610 may support obtaining information by receiving signals via one or more antennas. Additionally or alternatively, receiver 610 may support obtaining information by receiving signals via one or more wired (e.g., electrical, optical) interfaces, wireless interfaces, or any combination thereof.

[0174] The transmitter 615 may provide means for outputting (e.g., transmitting, providing, conveying, transmitting) information generated by other components of the device 605. For example, the transmitter 615 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 615 may support outputting information by sending signals via one or more antennas. Additionally or alternatively, the transmitter 615 may support outputting information by sending signals via one or more wired (e.g., electrical, optical fiber) interfaces, wireless interfaces, or any combination thereof. In some examples, the transmitter 615 and the receiver 610 may be co-located in a transceiver, which may include a modem or be coupled to a modem.

[0175] 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 management of federated learning 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.

[0176] 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 DSP, a CPU, an ASIC, an FPGA or other programmable logic device, a microcontroller, discrete gate or transistor logic components, discrete hardware components, or any combination thereof that is configured as or otherwise supports components 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).

[0177] 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 components 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.

[0178] 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 integrate in conjunction with receiver 610, transmitter 615, or both to obtain information, output information, or perform various other operations as described herein.

[0179] According to examples as disclosed herein, the communication manager 620 may support wireless communications at a network node. For example, the communication manager 620 may be configured to or otherwise support components for the following operations: selecting one or more user equipments (UEs) for a training process based on a trigger to activate a training process for a prediction model. The communication manager 620 may be configured to or otherwise support components for the following operations: sending an indication of a training configuration for a prediction model to one or more UEs, the training configuration including a set of training parameters based on one or more UEs.

[0180] Additionally or alternatively, according to examples as disclosed herein, the communication manager 620 may support wireless communication at a server. For example, the communication manager 620 may be configured to or otherwise support components for the following operations: selecting one or more user equipments (UEs) for a training process based on a trigger to activate a training process for a prediction model. The communication manager 620 may be configured to or otherwise support components for the following operations: sending an indication of a training configuration for a prediction model to one or more UEs, the training configuration including a set of training parameters based on one or more UEs.

[0181] Additionally or alternatively, according to examples as disclosed herein, the communication manager 620 may support wireless communication at a server. For example, the communication manager 620 may be configured to or otherwise support components for the following operations: sending a training configuration for a training process associated with a prediction model to a set of user equipment (UE), the training configuration including a first set of model parameters associated with a first PS ID. The communication manager 620 may be configured to or otherwise support components for the following operations: receiving one or more reports indicating one or more subsets of model parameters output from the training process for the prediction model at the UE from one or more UEs in the set of UEs. The communication manager 620 may be configured to or otherwise support components for the following operations: aggregating subsets of model parameters into a second set of model parameters. The communication manager 620 may be configured to or otherwise support components for the following operations: assigning a second PS ID to a second set of model parameters, the second PS ID being different from the first PS ID. The communication manager 620 may be configured to or otherwise support components for the following operations: sending an indication of the second PS ID.

[0182] 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) can support techniques for efficient management of a federated learning process. For example, the device 605 can select training parameters for a federated learning process based on the capabilities of the participating UEs, which can reduce processing, reduce power consumption, and provide more efficient utilization of communication resources.

[0183] Figure 7A block diagram 700 of a device 705 supporting management of federated learning 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 the network entity 105 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).

[0184] Receiver 710 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 communicated to other components of device 705. In some examples, receiver 710 may support obtaining information by receiving signals via one or more antennas. Additionally or alternatively, receiver 710 may support obtaining information by receiving signals via one or more wired (e.g., electrical, optical) interfaces, wireless interfaces, or any combination thereof.

[0185] The transmitter 715 may provide means for outputting (e.g., sending, providing, conveying, transmitting) information generated by other components of the device 705. For example, the transmitter 715 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 715 may support outputting information by sending signals via one or more antennas. Additionally or alternatively, the transmitter 715 may support outputting information by sending signals via one or more wired (e.g., electrical, optical fiber) interfaces, wireless interfaces, or any combination thereof. In some examples, the transmitter 715 and the receiver 710 may be co-located in a transceiver, which may include a modem or be coupled to a modem.

[0186] The device 705 or its various components may be examples of components for performing various aspects of the management of federated learning as described herein. For example, the communication manager 720 may include a UE selection component 725, a training configuration transmitter 730, a report receiver 735, a model parameter component 740, a PS ID component 745, or any combination thereof. The communication manager 720 may be an example of various aspects of the communication manager 620 as described herein. In some examples, the communication manager 720 or its various components may be configured to use or otherwise cooperate with the receiver 710, the transmitter 715, or both to perform various operations (e.g., receive, obtain, monitor, output, send). For example, the communication manager 720 may receive information from the receiver 710, transmit information to the transmitter 715, or integrate with the receiver 710, the transmitter 715, or both to obtain information, output information, or perform various other operations as described herein.

[0187] According to examples as disclosed herein, the communication manager 720 may support wireless communications at a network node. The UE selection component 725 may be configured to or otherwise support components for the following operations: selecting one or more user equipment (UE) for a training process based on a trigger to activate a training process for a prediction model. The training configuration transmitter 730 may be configured to or otherwise support components for the following operations: sending an indication of a training configuration for a prediction model to one or more UEs, the training configuration including a set of training parameters based on one or more UEs.

[0188] Additionally or alternatively, according to examples as disclosed herein, the communication manager 720 may support wireless communication at a server. The UE selection component 725 may be configured to or otherwise support components for the following operations: selecting one or more user equipment (UE) for a training process based on a trigger to activate a training process for a prediction model. The training configuration transmitter 730 may be configured to or otherwise support components for the following operations: sending an indication of a training configuration for a prediction model to one or more UEs, the training configuration comprising a set of training parameters based on one or more UEs.

[0189] Additionally or alternatively, according to examples as disclosed herein, the communication manager 720 may support wireless communication at a server. The training configuration transmitter 730 may be configured to or otherwise support components for the following operations: send a training configuration for a training process associated with a prediction model to a set of user equipment (UE), the training configuration including a first set of model parameters associated with a first PS ID. The report receiver 735 may be configured to or otherwise support components for the following operations: receive one or more reports indicating one or more subsets of model parameters output from the training process for the prediction model at the UE from one or more UEs in the set of UEs. The model parameter component 740 may be configured to or otherwise support components for the following operations: aggregate the subsets of model parameters into a second set of model parameters. The PS ID component 745 may be configured to or otherwise support components for the following operations: assign a second PS ID to the second set of model parameters, the second PS ID being different from the first PS ID. The PS ID component 745 may be configured to or otherwise support components for the following operations: send an indication of the second PS ID.

[0190] Figure 8 A block diagram 800 of a communication manager 820 supporting management of federated learning in accordance with one or more aspects of the present disclosure is shown. The communication manager 820 may 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 may be examples of components for performing various aspects of management of federated learning as described herein. For example, the communication manager 820 may include a UE selection component 825, a training configuration transmitter 830, a report receiver 835, a model parameter component 840, a PS ID component 845, a training request component 850, an activation component 855, a data radio bearer component 860, a training parameter selection component 865, a configuration implementation component 870, a training activation component 875, or any combination thereof. Each of these components may communicate with each other directly or indirectly (e.g., via one or more buses), which communication may include communication within a protocol layer of a protocol stack, communication associated with a logical channel of a protocol stack (e.g., between protocol layers of a protocol stack, within a device, component, or virtualized component associated with the network entity 105, between devices, components, or virtualized components associated with the network entity 105), or any combination thereof.

[0191] According to examples as disclosed herein, the communication manager 820 may support wireless communications at a network node. The UE selection component 825 may be configured to or otherwise support components for the following operations: selecting one or more user equipment (UE) for a training process based on a trigger to activate a training process for a prediction model. The training configuration transmitter 830 may be configured to or otherwise support components for the following operations: sending an indication of a training configuration for a prediction model to one or more UEs, the training configuration including a set of training parameters based on the one or more UEs.

[0192] In some examples, training request component 850 can be configured or otherwise support components for sending a first message to a server indicating a request to activate a training process. In some examples, activation component 855 can be configured or otherwise support components for receiving a second message from a server indicating activation of a training process.

[0193] In some examples, activation component 855 may be configured or otherwise support means for receiving a first message from a server indicating that a network node is to activate a training process, wherein the first message includes a trigger.

[0194] In some examples, data radio bearer component 860 may be configured or otherwise support means for sending an indication to one or more UEs of a data radio bearer configured for downloading a model structure and a baseline parameter set associated with a prediction model.

[0195] In some examples, training parameter selection component 865 may be configured as or otherwise support components for selecting a set of training parameters for a training configuration based on estimated link capacity associated with one or more UEs, computing capabilities associated with one or more UEs, or a combination thereof.

[0196] In some examples, the set of training parameters includes a minimum number of epochs for a training process to be performed at a UE in one or more UEs. In some examples, the set of training parameters includes an MS ID associated with the prediction model, a baseline PS ID associated with the prediction model, a training valid area, a maximum number of epochs for the training process, a minimum number of epochs for the training process, a training deadline, a weight set, a periodicity, a server address, or a combination thereof.

[0197] In some examples, the configuration implementation component 870 may be configured to or otherwise support components for the following operations: receiving a message from at least one of the one or more UEs indicating that the at least one UE has implemented a training configuration. In some examples, the configuration implementation component 870 may be configured to or otherwise support components for the following operations: receiving a message from at least one of the one or more UEs indicating that the at least one UE has avoided implementing a training configuration.

[0198] In some examples, activation component 855 may be configured or otherwise support components for receiving a message from at least one of the one or more UEs indicating that the prediction model is ready to be activated at the at least one UE. In some examples, activation component 855 may be configured or otherwise support components for sending a message to the one or more UEs including an indication to activate a training process at the one or more UEs.

[0199] Additionally or alternatively, according to examples as disclosed herein, the communication manager 820 may support wireless communication at a server. In some examples, the UE selection component 825 may be configured to or otherwise support components for the following operations: selecting one or more user equipment (UE) for a training process based on a trigger to activate a training process for a prediction model. In some examples, the training configuration transmitter 830 may be configured to or otherwise support components for the following operations: sending an indication of a training configuration for a prediction model to one or more UEs, the training configuration including a set of training parameters based on one or more UEs.

[0200] In some examples, training request component 850 may be configured or otherwise support means for receiving a message from a network node indicating a request to activate a training process, wherein the indication of the training configuration is sent based on receiving the message.

[0201] In some examples, training activation component 875 may be configured or otherwise support means for sending a message to a network node indicating that the network node is to activate a training process, wherein the message includes a trigger.

[0202] In some examples, data radio bearer component 860 may be configured or otherwise support means for sending an indication to one or more UEs of a data radio bearer configured for downloading a model structure and a baseline parameter set associated with a prediction model.

[0203] In some examples, the training parameter selection component 865 can be configured to or otherwise support components for selecting a set of training parameters for a training configuration based on an estimated link capacity associated with one or more UEs, a computing capability associated with one or more UEs, or a combination thereof. In some examples, the set of training parameters includes a minimum number of epochs for a training process to be performed at a UE in the one or more UEs.

[0204] In some examples, the set of training parameters includes an MS ID associated with the prediction model, a baseline PS ID associated with the prediction model, a training validity area, a maximum number of epochs for the training process, a minimum number of epochs for the training process, a training deadline, a weight set, a periodicity, a server address, or a combination thereof.

[0205] Additionally or alternatively, according to examples as disclosed herein, the communication manager 820 may support wireless communications at a server. In some examples, the training configuration transmitter 830 may be configured to or otherwise support components for the following operations: sending a training configuration for a training process associated with a prediction model to a set of user equipment (UEs), the training configuration including a first set of model parameters associated with a first PS ID. The report receiver 835 may be configured to or otherwise support components for the following operations: receiving one or more reports indicating one or more subsets of model parameters output from the training process for the prediction model at the UE from one or more UEs in the set of UEs. The model parameter component 840 may be configured to or otherwise support components for the following operations: aggregating a subset of model parameters into a second set of model parameters. The PS ID component 845 may be configured to or otherwise support components for the following operations: assigning a second PS ID to the second set of model parameters, the second PS ID being different from the first PS ID. In some examples, PS ID component 845 may be configured or otherwise support components for sending an indication of the second PS ID.

[0206] In some examples, activation component 855 may be configured or otherwise support means for receiving, based on sending the indication, a message from at least one UE in the set of UEs indicating that the prediction model is ready to be activated at the UE.

[0207] In some examples, the model parameter component 840 may be configured or otherwise support components for the following operations: sending a message indicating a second set of model parameters for the prediction model to the set of UEs, the second set of model parameters including a set of model parameter updates for the prediction model. In some examples, to support sending the indication, the PSID component 845 may be configured or otherwise support components for the following operations: sending the indication to the network node, to the set of UEs, or to both. In some examples, the second PS ID includes a temporary PS ID or a combination of the first PS ID and a version tag.

[0208] Fig. 9 A diagram of a system 900 including a device 905 that supports management of federated learning 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 network entity 105 as described herein, or include components of these devices. The device 905 may communicate with one or more network entities 105, one or more UEs 115, or any combination thereof, which communication may include communication through one or more wired interfaces, through one or more wireless interfaces, or any combination thereof. The device 905 may include components that support output and acquisition of communications, such as a communication manager 920, a transceiver 910, an antenna 915, a memory 925, a code 930, and a processor 935. 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 940).

[0209] The transceiver 910 may support bidirectional communication via a wired link, a wireless link, or both as described herein. In some examples, the transceiver 910 may include a wired transceiver and may communicate bidirectionally with another wired transceiver. Additionally or alternatively, in some examples, the transceiver 910 may include a wireless transceiver and may communicate bidirectionally with another wireless transceiver. In some examples, the device 905 may include one or more antennas 915, which may be capable of (e.g., concurrently) sending or receiving wireless transmissions. The transceiver 910 may also include a modem for modulating a signal, providing a modulated signal for transmission (e.g., through one or more antennas 915, through a wired transmitter), receiving a modulated signal (e.g., from one or more antennas 915, from a wired receiver), and demodulating a signal. In some implementations, the transceiver 910 may include one or more interfaces, such as one or more interfaces coupled to one or more antennas 915 configured to support various receiving or obtaining operations, or one or more interfaces coupled to one or more antennas 915 configured to support various sending or outputting operations, or a combination thereof. In some implementations, the transceiver 910 may include or be configured to be coupled to one or more processors or memory components, which may be operable to: perform or support operations based on received or obtained information or signals; or generate information or other signals for transmission or other output; or any combination thereof. In some implementations, the transceiver 910, or the transceiver 910 and one or more antennas 915, or the transceiver 910 and one or more antennas 915 and one or more processors or memory components (e.g., processor 935, or memory 925, or both) may be included in a chip or chip assembly installed in the device 905. In some examples, the transceiver may be operable to support communications via one or more communication links (eg, communication link 125, backhaul communication link 120, midhaul communication link 162, fronthaul communication link 168).

[0210] The memory 925 may include RAM and ROM. The memory 925 may store computer-readable, computer-executable code 930 including instructions that, when executed by the processor 935, cause the device 905 to perform various functions described herein. The code 930 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 930 may not be directly executable by the processor 935, but may (e.g., when compiled and executed) cause the computer to perform the functions described herein. In some cases, the memory 925 may include a BIOS, etc., which may control basic hardware or software operations, such as interaction with peripheral components or devices.

[0211] The processor 935 may include an intelligent hardware device (e.g., a general-purpose processor, a DSP, an ASIC, a CPU, an FPGA, a microcontroller, a programmable logic device, a discrete gate or transistor logic component, a discrete hardware component, or any combination thereof). In some cases, the processor 935 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 935. The processor 935 may be configured to execute computer-readable instructions stored in a memory (e.g., a memory 925) to enable the device 905 to perform various functions (e.g., functions or tasks supporting the management of federated learning). For example, the device 905 or a component of the device 905 may include a processor 935 and a memory 925 coupled to the processor 935, and the processor 935 and the memory 925 are configured to perform various functions described herein. The processor 935 may be an example of a cloud computing platform (e.g., one or more physical nodes and supporting software such as an operating system, a virtual machine, or a container instance), which may host functions (e.g., by executing code 930) to perform the functions of the device 905. The processor 935 may be any one or more suitable processors capable of executing scripts or instructions of one or more software programs stored in the device 905 (such as in the memory 925). In some specific implementations, the processor 935 may be a component of a processing system. A processing system may generally refer to a system or a series of machines or components that receive inputs and process these inputs to produce a set of outputs (which may be passed to, for example, other systems or components of the device 905). For example, the processing system of the device 905 may refer to a system including various other components or subcomponents of the device 905, such as the processor 935, or the transceiver 910, or the communication manager 920, or other components or combinations of components of the device 905. The processing system of the device 905 may be docked with other components of the device 905, and may process information (such as inputs or signals) received from other components, or output information to other components. For example, a chip or modem of the device 905 may include a processing system and one or more interfaces, which are used to output information or obtain information or both. The one or more interfaces may be implemented as or otherwise include a first interface configured to output information and a second interface configured to obtain information, or the same interface configured to output information and obtain information, etc. In some implementations, the one or more interfaces may refer to an interface between a processing system of a chip or modem and a transmitter, so that the device 905 can send information output from the chip or modem. Additionally or alternatively, in some implementations, the one or more interfaces may refer to an interface between a processing system of a chip or modem and a receiver, so that the device 905 can obtain information or signal input, and the information can be delivered to the processing system.One of ordinary skill in the art will readily recognize that the first interface may also obtain information or signal input, and the second interface may also output information or signal output.

[0212] In some examples, bus 940 may support communications of a protocol layer of a protocol stack (e.g., within the protocol layer). In some examples, bus 940 may support communications associated with a logical channel of a protocol stack (e.g., between protocol layers of the protocol stack), which may include communications performed within a component of device 905 or between different components of device 905 that may be co-located or located in different locations (e.g., where device 905 may refer to a system, where one or more of communication manager 920, transceiver 910, memory 925, code 930, and processor 935 may be located in one of the different components or divided between the different components).

[0213] In some examples, the communication manager 920 may manage aspects of communications with the core network 130 (e.g., via one or more wired or wireless backhaul links). For example, the communication manager 920 may manage the delivery of data communications for client devices such as one or more UEs 115. In some examples, the communication manager 920 may manage communications with other network entities 105 and may include a controller or scheduler for controlling communications with the UE 115 in coordination with the other network entities 105. In some examples, the communication manager 920 may support an X2 interface within an LTE / LTE-A wireless communication network technology to provide communications between network entities 105.

[0214] According to examples as disclosed herein, the communication manager 920 may support wireless communications at a network node. For example, the communication manager 920 may be configured to or otherwise support components for the following operations: selecting one or more user equipments (UEs) for a training process based on a trigger to activate a training process for a prediction model. The communication manager 920 may be configured to or otherwise support components for the following operations: sending an indication of a training configuration for a prediction model to one or more UEs, the training configuration including a set of training parameters based on one or more UEs.

[0215] Additionally or alternatively, according to examples as disclosed herein, the communication manager 920 may support wireless communication at a server. For example, the communication manager 920 may be configured to or otherwise support components for the following operations: selecting one or more user equipments (UEs) for a training process based on a trigger to activate a training process for a prediction model. The communication manager 920 may be configured to or otherwise support components for the following operations: sending an indication of a training configuration for a prediction model to one or more UEs, the training configuration comprising a set of training parameters based on one or more UEs.

[0216] Additionally or alternatively, according to examples as disclosed herein, the communication manager 920 may support wireless communication at a server. For example, the communication manager 920 may be configured to or otherwise support components for the following operations: sending a training configuration for a training process associated with a prediction model to a set of user equipment (UE), the training configuration including a first set of model parameters associated with a first PS ID. The communication manager 920 may be configured to or otherwise support components for the following operations: receiving one or more reports indicating one or more subsets of model parameters output from the training process for the prediction model at the UE from one or more UEs in the set of UEs. The communication manager 920 may be configured to or otherwise support components for the following operations: aggregating subsets of model parameters into a second set of model parameters. The communication manager 920 may be configured to or otherwise support components for the following operations: assigning a second PS ID to a second set of model parameters, the second PS ID being different from the first PS ID. The communication manager 920 may be configured to or otherwise support components for the following operations: sending an indication of the second PS ID.

[0217] By including or configuring the communication manager 920 according to the examples described herein, the device 905 can support techniques for efficient management of the federated learning process. For example, the device 905 can select training parameters for the federated learning process based on the capabilities of the participating UEs, which can reduce processing, reduce power consumption, and provide more efficient use of communication resources.

[0218] In some examples, the communication manager 920 may be configured to perform various operations (e.g., receive, obtain, monitor, output, transmit) using or otherwise cooperating with the transceiver 910, one or more antennas 915 (e.g., where applicable), 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 transceiver 910, the processor 935, the memory 925, the code 930, or any combination thereof. For example, the code 930 may include instructions that are executable by the processor 935 to cause the device 905 to perform various aspects of the management of federated learning as described herein, or the processor 935 and the memory 925 may be otherwise configured to perform or support such operations.

[0219] Fig.10A block diagram 1000 of a device 1005 supporting management of federated learning according to one or more aspects of the present disclosure is shown. The device 1005 may be an example of aspects of the UE 115 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).

[0220] The receiver 1010 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 the management of federated learning). The information may be communicated to other components of the device 1005. The receiver 1010 may utilize a single antenna or a collection of multiple antennas.

[0221] The transmitter 1015 may provide means for transmitting signals generated by other components of the device 1005. For example, the transmitter 1015 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 the management of federated learning). In some examples, the transmitter 1015 may be co-located with the receiver 1010 in a transceiver unit. The transmitter 1015 may utilize a single antenna or a collection of multiple antennas.

[0222] 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 management of federated learning 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.

[0223] 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 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 components, discrete hardware components, or any combination thereof that is configured as 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).

[0224] 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 components 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.

[0225] In some examples, communication manager 1020 may be configured to perform various operations (e.g., receive, obtain, monitor, output, transmit) 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.

[0226] According to examples as disclosed herein, the communication manager 1020 may support wireless communication at the UE. For example, the communication manager 1020 may be configured to or otherwise support components for the following operations: receiving a first message indicating a training configuration for a training process for a prediction model, the training configuration including a set of training parameters. The communication manager 1020 may be configured to or otherwise support components for the following operations: sending a second message indicating whether the UE has implemented the training configuration of the training process based on the set of training parameters.

[0227] Additionally or alternatively, according to examples as disclosed herein, the communication manager 1020 may support wireless communications at a UE. For example, the communication manager 1020 may be configured to or otherwise support components for the following operations: receiving a training configuration of a training process associated with a prediction model, the training configuration including a first set of model parameters associated with a first PS ID. The communication manager 1020 may be configured to or otherwise support components for the following operations: sending a report indicating a subset of model parameters output from the training process for the prediction model at the UE. The communication manager 1020 may be configured to or otherwise support components for the following operations: receiving an indication of a second PS ID associated with a second set of model parameters based on sending the report, the second PS ID being different from the first PS ID.

[0228] 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 to the receiver 1010, the transmitter 1015, the communication manager 1020, or a combination thereof) can support techniques for efficient management of the federated learning process. For example, the device 1005 can determine whether to participate in the FL process based on the capabilities of the device 1005, which can reduce processing, reduce power consumption, and improve efficient utilization of communication resources.

[0229] Fig.11 A block diagram 1100 of a device 1105 supporting management of federated learning 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 UE 115 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).

[0230] The receiver 1110 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 the management of federated learning). The information may be communicated to other components of the device 1105. The receiver 1110 may utilize a single antenna or a collection of multiple antennas.

[0231] The transmitter 1115 may provide means for transmitting signals generated by other components of the device 1105. For example, the transmitter 1115 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 the management of federated learning). In some examples, the transmitter 1115 may be co-located with the receiver 1110 in a transceiver unit. The transmitter 1115 may utilize a single antenna or a collection of multiple antennas.

[0232] Device 1105 or its various components may be examples of components for performing various aspects of management of federated learning as described herein. For example, communication manager 1120 may include training configuration receiver 1125, configuration implementation component 1130, report transmitter 1135, PS ID component 1140, 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 use or otherwise cooperate with receiver 1110, transmitter 1115, or both to perform various operations (e.g., receive, obtain, monitor, output, send). 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.

[0233] According to examples as disclosed herein, the communication manager 1120 may support wireless communication at the UE. The training configuration receiver 1125 may be configured to or otherwise support components for the following operations: receiving a first message indicating a training configuration for a training process for a prediction model, the training configuration including a set of training parameters. The configuration implementation component 1130 may be configured to or otherwise support components for the following operations: sending a second message indicating whether the UE has implemented the training configuration of the training process based on the set of training parameters.

[0234] Additionally or alternatively, according to examples as disclosed herein, the communication manager 1120 may support wireless communications at the UE. The training configuration receiver 1125 may be configured to or otherwise support components for the following operations: receiving a training configuration for a training process associated with a prediction model, the training configuration including a first set of model parameters associated with a first PS ID. The report transmitter 1135 may be configured to or otherwise support components for the following operations: sending a report indicating a subset of model parameters output from the training process for the prediction model at the UE. The PS ID component 1140 may be configured to or otherwise support components for the following operations: receiving an indication of a second PS ID associated with a second set of model parameters based on sending the report, the second PS ID being different from the first PS ID.

[0235] Fig.12A block diagram 1200 of a communication manager 1220 supporting management of federated learning 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 management of federated learning as described herein. For example, the communication manager 1220 may include a training configuration receiver 1225, a configuration implementation component 1230, a report sender 1235, a PS ID component 1240, a data radio bearer component 1245, an activation component 1250, a model parameter component 1255, a training process component 1260, or any combination thereof. Each of these components can communicate with each other directly or indirectly (e.g., via one or more buses).

[0236] According to examples as disclosed herein, the communication manager 1220 may support wireless communication at the UE. The training configuration receiver 1225 may be configured to or otherwise support components for the following operations: receiving a first message indicating a training configuration for a training process for a prediction model, the training configuration including a set of training parameters. The configuration implementation component 1230 may be configured to or otherwise support components for the following operations: sending a second message indicating whether the UE has implemented the training configuration of the training process based on the set of training parameters.

[0237] In some examples, data radio bearer component 1245 may be configured or otherwise support components for receiving an indication of a data radio bearer configured for downloading a model structure and a baseline parameter set associated with a prediction model. In some examples, data radio bearer component 1245 may be configured or otherwise support components for downloading the model structure and the baseline parameter set via a data radio bearer.

[0238] In some examples, the set of training parameters includes a minimum number of epochs for the training process, and the configuration implementation component 1230 may be configured to or otherwise support components for the following operations: estimating the number of local epochs for the training process based on the computing power of the UE or the link capacity associated with the UE. In some examples, the set of training parameters includes a minimum number of epochs for the training process, and the configuration implementation component 1230 may be configured to or otherwise support components for the following operations: comparing the estimated number of local epochs to the minimum number of epochs, wherein the second message is sent based on the comparison.

[0239] In some examples, the second message instructs the UE to avoid implementing the training configuration based on the estimated number of local epochs being less than the minimum number of epochs. In some examples, the second message instructs the UE to implement the training configuration based on the estimated number of local epochs being equal to or greater than the minimum number of epochs.

[0240] In some examples, the training process component 1260 may be configured as or otherwise support components for performing a training process for a prediction model according to a training configuration and based on the estimated number of local epochs. In some examples, the report sender 1235 may be configured as or otherwise support components for sending a report indicating a set of model parameters for a prediction model based on performing the training process.

[0241] In some examples, the second message also includes an indication that the prediction model is ready to be activated at the UE, and the activation component 1250 may be configured as or otherwise support components for the following operations: receiving a third message including an indication to activate the training process based on sending the second message, wherein performing the training process is based on receiving the third message.

[0242] In some examples, configuration implementation component 1230 may be configured to or otherwise support components for configuring a training process according to a set of training parameters, wherein the second message also includes an indication that configuration of the training process is complete.

[0243] In some examples, the set of training parameters includes an MS ID associated with the prediction model, a baseline PS ID associated with the prediction model, a training validity area, a maximum number of epochs for the training process, a minimum number of epochs for the training process, a training deadline, a weight set, a periodicity, a server address, or a combination thereof.

[0244] In some examples, the first message is received from a network node or a server.

[0245] Additionally or alternatively, the communication manager 1220 may support wireless communications at the UE according to examples as disclosed herein. In some examples, the training configuration receiver 1225 may be configured to or otherwise support components for the following operations: receiving a training configuration for a training process associated with a prediction model, the training configuration including a first set of model parameters associated with a first PS ID. The report transmitter 1235 may be configured to or otherwise support components for the following operations: sending a report indicating a subset of model parameters output from the training process for the prediction model at the UE. The PS ID component 1240 may be configured to or otherwise support components for the following operations: receiving an indication of a second PS ID associated with a second set of model parameters based on sending the report, the second PS ID being different from the first PS ID.

[0246] In some examples, activation component 1250 may be configured or otherwise support means for sending a message indicating that the prediction model is ready to be activated at the UE based on receiving the indication.

[0247] In some examples, model parameter component 1255 may be configured or otherwise support means for receiving a message indicating a second set of model parameters for a prediction model, the second set of model parameters comprising a set of model parameter updates for the prediction model.

[0248] In some examples, training process component 1260 can be configured or otherwise support means for performing a training process for a prediction model using a second set of model parameters and based on a second PS ID.

[0249] In some examples, to support receiving the indication, PS ID component 1240 can be configured or otherwise support means for receiving the indication from the server. In some examples, the second PS ID includes a temporary PS ID or a combination of the first PS ID and a version tag.

[0250] Fig.13A diagram of a system 1300 including a device 1305 that supports management of federated learning 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 UE 115 as described herein, or include components of these devices. The device 1305 may communicate (e.g., wirelessly) with one or more network entities 105, one or more UEs 115, or any combination thereof. 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, an input / output (I / O) controller 1310, a transceiver 1315, an antenna 1325, a memory 1330, a code 1335, and a processor 1340. These components may be in electronic communication or otherwise coupled (e.g., operatively, communicatively, functionally, electronically, electrically) via one or more buses (e.g., a bus 1345).

[0251] I / O controller 1310 can manage input and output signals for device 1305. I / O controller 1310 can also manage peripheral devices that are not integrated into device 1305. In some cases, I / O controller 1310 can represent a physical connection or port to an external peripheral device. In some cases, I / O controller 1310 can utilize an operating system such as or another known operating system. Additionally or alternatively, I / O controller 1310 may represent or interact with a modem, keyboard, mouse, touch screen, or similar device. In some cases, I / O controller 1310 may be implemented as part of a processor such as processor 1340. In some cases, a user may interact with device 1305 via I / O controller 1310 or via hardware components controlled by I / O controller 1310.

[0252] In some cases, the device 1305 may include a single antenna 1325. However, in some other cases, the device 1305 may have more than one antenna 1325, which may be capable of concurrently transmitting or receiving multiple wireless transmissions. The transceiver 1315 may communicate bidirectionally via one or more antennas 1325, a wired link, or a wireless link as described herein. For example, the transceiver 1315 may represent a wireless transceiver and may communicate bidirectionally with another wireless transceiver. The transceiver 1315 may also include a modem for modulating packets, providing the modulated packets to one or more antennas 1325 for transmission, and demodulating packets received from one or more antennas 1325. The transceiver 1315, or the transceiver 1315 and one or more antennas 1325 may be examples of the transmitter 1015, the transmitter 1115, the receiver 1010, the receiver 1110, or any combination thereof or any component thereof as described herein.

[0253] Memory 1330 may include random access memory (RAM) and read-only memory (ROM). Memory 1330 may store computer-readable, computer-executable code 1335 including instructions that, when executed by processor 1340, cause device 1305 to perform various functions described herein. Code 1335 may be stored in a non-transitory computer-readable medium such as system memory or another type of memory. In some cases, code 1335 may not be directly executable by processor 1340, but may (e.g., when compiled and executed) cause a computer to perform the functions described herein. In some cases, memory 1330 may also contain, among other things, a basic I / O system (BIOS) that may control basic hardware or software operations, such as interaction with peripheral components or devices.

[0254] Processor 1340 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, processor 1340 may be configured to operate a memory array using a memory controller. In some other cases, the memory controller may be integrated into processor 1340. Processor 1340 may be configured to execute computer-readable instructions stored in a memory (e.g., memory 1330) to enable device 1305 to perform various functions (e.g., functions or tasks that support the management of federated learning). For example, device 1305 or a component of device 1305 may include processor 1340 and memory 1330 coupled to or coupled to processor 1340, and processor 1340 and memory 1330 are configured to perform various functions described herein.

[0255] According to examples as disclosed herein, the communication manager 1320 may support wireless communication at the UE. For example, the communication manager 1320 may be configured to or otherwise support components for the following operations: receiving a first message indicating a training configuration for a training process for a prediction model, the training configuration including a set of training parameters. The communication manager 1320 may be configured to or otherwise support components for the following operations: sending a second message indicating whether the UE has implemented the training configuration of the training process based on the set of training parameters.

[0256] Additionally or alternatively, according to examples as disclosed herein, the communication manager 1320 may support wireless communications at the UE. For example, the communication manager 1320 may be configured to or otherwise support components for the following operations: receiving a training configuration of a training process associated with a prediction model, the training configuration including a first set of model parameters associated with a first PS ID. The communication manager 1320 may be configured to or otherwise support components for the following operations: sending a report indicating a subset of model parameters output from the training process for the prediction model at the UE. The communication manager 1320 may be configured to or otherwise support components for the following operations: receiving an indication of a second PS ID associated with a second set of model parameters based on sending the report, the second PS ID being different from the first PS ID.

[0257] By including or configuring the communication manager 1320 according to the examples described herein, the device 1305 can support techniques for efficient management of the federated learning process. For example, the device 1305 can determine whether to participate in the FL process based on the capabilities of the device 1305, which can reduce latency, reduce signaling overhead, and improve coordination between devices.

[0258] In some examples, the communication manager 1320 may be configured to perform various operations (e.g., receive, monitor, transmit) using or otherwise cooperating with the transceiver 1315, one or more antennas 1325, 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 1340, the memory 1330, the code 1335, or any combination thereof. For example, the code 1335 may include instructions that are executable by the processor 1340 to cause the device 1305 to perform various aspects of the management of federated learning as described herein, or the processor 1340 and the memory 1330 may be otherwise configured to perform or support such operations.

[0259] Fig.14A flowchart illustrating a method 1400 for supporting management of federated learning according to one or more aspects of the present disclosure is shown. The operations of the method 1400 may be implemented by a network entity or a component thereof as described herein. For example, the operations of the method 1400 may be implemented by a network entity or a component thereof as described in reference to Figures 1 to 9 The network entity performs. In some examples, the network entity may execute an instruction set to control the functional elements of the network entity to perform the described functions. Additionally or alternatively, the network entity may use dedicated hardware to perform various aspects of the described functions.

[0260] At 1405, the method may include: selecting one or more user equipment (UE) for a training process based on a trigger to activate a training process for a prediction model. 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 UE selection component 825 is executed.

[0261] At 1410, the method may include: sending an indication of a training configuration for a prediction model to one or more UEs, the training configuration including a set of training parameters based on the one or more UEs. 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 training configuration transmitter 830 is executed.

[0262] Fig.15 A flowchart illustrating a method 1500 for supporting management of federated learning according to one or more aspects of the present disclosure is shown. The operations of the method 1500 may be implemented by a network entity or a component thereof as described herein. For example, the operations of the method 1500 may be implemented by a network entity or a component thereof as described in reference to Figures 1 to 9 The network entity performs. In some examples, the network entity may execute an instruction set to control the functional elements of the network entity to perform the described functions. Additionally or alternatively, the network entity may use dedicated hardware to perform various aspects of the described functions.

[0263] At 1505, the method may include: selecting one or more user equipment (UE) for a training process based on a trigger to activate a training process for a prediction model. 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 UE selection component 825 is executed.

[0264] At 1510, the method may include: sending an indication of a training configuration for a prediction model to one or more UEs, the training configuration including a set of training parameters based on the one or more UEs. 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 training configuration transmitter 830 is executed.

[0265] Fig.16 A flowchart illustrating a method 1600 for supporting management of federated learning 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 5 and Figures 10 to 13 The described UE 115 may be performed. 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.

[0266] At 1605, the method may include: receiving a first message indicating a training configuration for a training process for a prediction model, the training configuration including a set of training parameters. The operations of 1605 may be performed according to examples disclosed herein. In some examples, aspects of the operations of 1605 may be performed as described in reference Fig.12 The training configuration receiver 1225 is executed.

[0267] At 1610, the method may include sending a second message indicating whether the UE has implemented a training configuration of a training process based on a set of training parameters. 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 Fig.12 The configuration is implemented by component 1230.

[0268] Fig.17 A flowchart illustrating a method 1700 for supporting management of federated learning according to one or more aspects of the present disclosure is shown. The operations of the method 1700 may be implemented by a network entity or a component thereof as described herein. For example, the operations of the method 1700 may be implemented by a network entity or a component thereof as described in reference to Figures 1 to 9 The network entity performs. In some examples, the network entity may execute an instruction set to control the functional elements of the network entity to perform the described functions. Additionally or alternatively, the network entity may use dedicated hardware to perform various aspects of the described functions.

[0269] At 1705, the method may include: sending a training configuration for a training process associated with a prediction model to a set of user equipment (UE), the training configuration including a first set of model parameters associated with a first PS ID. 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 Figure 8 The training configuration transmitter 830 is executed.

[0270] At 1710, the method may include: receiving one or more reports from one or more UEs in the set of UEs indicating one or more subsets of model parameters output at the UE from a training process for a prediction model. 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 Figure 8 The described report receiver 835 is executed.

[0271] At 1715, the method may include aggregating the subset of model parameters into a second set of model parameters. The operations of 1715 may be performed according to examples disclosed herein. In some examples, aspects of the operations of 1715 may be performed as described in reference Figure 8 The model parameter component 840 is used to perform the above operation.

[0272] At 1720, the method may include assigning a second PS ID to the second set of model parameters, the second PS ID being different from the first PS ID. The operations of 1720 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1720 may be performed as described in reference to Figure 8 The PS ID component 845 is used to perform the above.

[0273] At 1725, the method may include sending an indication of the second PS ID. The operations of 1725 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1725 may be performed as described in reference to Figure 8 The PS ID component 845 is used to perform the above.

[0274] Fig.18 A flowchart illustrating a method 1800 for supporting management of federated learning according to one or more aspects of the present disclosure is shown. The operations of the method 1800 may be implemented by a UE or a component thereof as described herein. For example, the operations of the method 1800 may be implemented by a UE or a component thereof as described in reference to Figures 1 to 5 and Figures 10 to 13 The described UE 115 may be performed. 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.

[0275] At 1805, the method may include receiving a training configuration for a training process associated with a prediction model, the training configuration including a first set of model parameters associated with a first PS ID. The operations of 1805 may be performed according to examples disclosed herein. In some examples, aspects of the operations of 1805 may be performed as described in reference to Fig.12 The training configuration receiver 1225 is executed.

[0276] At 1810, the method may include sending a report indicating a subset of model parameters output from a training process for a prediction model at a UE. 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 sender 1235 executes.

[0277] At 1815, the method may include receiving an indication of a second PS ID associated with a second set of model parameters based on sending the report, the second PS ID being different from the first PS ID. The operations of 1815 may be performed according to examples as disclosed herein. In some examples, aspects of the operations of 1815 may be performed as described in reference to Fig.12 The PS ID component 1240 is used to perform the above operation.

[0278] The following provides an overview of various aspects of the disclosure:

[0279] Aspect 1: A method for wireless communication at a network node, the method comprising: selecting one or more UEs for a training process based at least in part on a trigger for activating a training process for a prediction model; and sending an indication of a training configuration for the prediction model to the one or more UEs, the training configuration comprising a set of training parameters based at least in part on the one or more UEs.

[0280] Aspect 2: The method according to aspect 1 further comprises: sending a first message indicating a request to activate the training process to a server; and receiving a second message indicating activation of the training process from the server.

[0281] Aspect 3: The method according to aspect 1, further comprising: receiving a first message from a server indicating that the network node is to activate the training process, wherein the first message includes the trigger.

[0282] Aspect 4: According to the method according to any one of aspects 1 to 3, the method also includes: sending an indication of a data radio bearer configured for downloading a model structure and a baseline parameter set associated with the prediction model to the one or more UEs.

[0283] Aspect 5: According to the method described in any one of Aspects 1 to 4, the method also includes: selecting the set of training parameters for the training configuration based at least in part on an estimated link capacity associated with the one or more UEs, a computing capability associated with the one or more UEs, or a combination thereof.

[0284] Aspect 6: The method according to aspect 5, wherein the set of training parameters includes a minimum number of epochs of the training process to be performed at a UE among the one or more UEs.

[0285] Aspect 7: A method according to any one of Aspects 1 to 6, wherein the set of training parameters includes an MS ID associated with the prediction model, a baseline PS ID associated with the prediction model, a training valid area, a maximum number of epochs of the training process, a minimum number of epochs of the training process, a training deadline, a weight set, periodicity, a server address, or a combination thereof.

[0286] Aspect 8: According to the method according to any one of aspects 1 to 7, the method further includes: receiving a message from at least one UE among the one or more UEs indicating that the at least one UE has implemented the training configuration.

[0287] Aspect 9: According to the method according to any one of aspects 1 to 7, the method also includes: receiving a message from at least one UE of the one or more UEs indicating that the at least one UE has avoided implementing the training configuration.

[0288] Aspect 10: According to any one of aspects 1 to 8, the method further includes: receiving a message from at least one UE of the one or more UEs indicating that the prediction model is ready to be activated at the at least one UE.

[0289] Aspect 11: According to the method according to any one of aspects 1 to 10, the method further includes: sending a message including an indication of activating the training process at the one or more UEs to the one or more UEs.

[0290] Aspect 12: A method for performing wireless communications at a server, the method comprising: selecting one or more UEs for a training process based at least in part on a trigger for activating a training process for a prediction model; and sending an indication of a training configuration for the prediction model to the one or more UEs, the training configuration comprising a set of training parameters based at least in part on the one or more UEs.

[0291] Aspect 13: The method according to aspect 12, the method further comprising: receiving a message from a network node indicating a request to activate the training process, wherein the indication of the training configuration is sent at least in part based on receiving the message.

[0292] Aspect 14: The method according to any one of aspects 12 to 13, the method further comprising: sending a message to a network node indicating that the network node is to activate the training process, wherein the message includes the trigger.

[0293] Aspect 15: According to the method of any one of Aspects 12 to 14, the method further comprises: sending an indication of a data radio bearer configured for downloading a model structure and a baseline parameter set associated with the prediction model to the one or more UEs.

[0294] Aspect 16: According to any one of Aspects 12 to 15, the method also includes: selecting the set of training parameters for the training configuration based at least in part on an estimated link capacity associated with the one or more UEs, a computing capability associated with the one or more UEs, or a combination thereof.

[0295] Aspect 17: The method according to aspect 16, wherein the set of training parameters includes a minimum number of epochs of the training process to be performed at a UE among the one or more UEs.

[0296] Aspect 18: A method according to any one of Aspects 12 to 17, wherein the set of training parameters includes an MS ID associated with the prediction model, a baseline PS ID associated with the prediction model, a training valid area, a maximum number of epochs of the training process, a minimum number of epochs of the training process, a training deadline, a weight set, periodicity, a server address, or a combination thereof.

[0297] Aspect 19: A method for performing wireless communications at a UE, the method comprising: receiving a first message indicating a training configuration for a training process for a prediction model, the training configuration comprising a set of training parameters; and sending a second message indicating whether the UE has implemented the training configuration for the training process at least in part based on the set of training parameters.

[0298] Aspect 20: The method according to Aspect 19 further comprises: receiving an indication of a data radio bearer configured for downloading a model structure and a baseline parameter set associated with the prediction model; and downloading the model structure and the baseline parameter set via the data radio bearer.

[0299] Aspect 21: A method according to any one of Aspects 19 to 20, wherein the set of training parameters includes a minimum number of epochs for the training process, and the method further includes: estimating the number of local epochs for the training process based at least in part on the computing power of the UE or the link capacity associated with the UE; and comparing the estimated number of local epochs with the minimum number of epochs, wherein the second message is sent at least in part based on the comparison.

[0300] Aspect 22: The method according to aspect 21, wherein the second message instructs the UE to avoid implementing the training configuration based at least in part on the estimated number of local epochs being less than the minimum number of epochs.

[0301] Aspect 23: The method according to aspect 21, wherein the second message instructs the UE to implement the training configuration based at least in part on the estimated number of local epochs being equal to or greater than the minimum number of epochs.

[0302] Aspect 24: According to the method of Aspect 23, the method also includes: performing the training process for the prediction model according to the training configuration and at least partially based on the estimated number of local epochs; and sending a report indicating a set of model parameters for the prediction model based at least partially on performing the training process.

[0303] Aspect 25: A method according to Aspect 24, wherein the second message also includes an indication that the prediction model is ready to be activated at the UE, and the method further includes: receiving a third message including an indication to activate the training process based at least in part on sending the second message, wherein performing the training process is based at least in part on receiving the third message.

[0304] Aspect 26: According to the method of any one of Aspects 19 to 25, the method further comprises: configuring the training process according to the set of training parameters, wherein the second message further comprises an indication that the configuration of the training process is completed.

[0305] Aspect 27: A method according to any one of Aspects 19 to 26, wherein the set of training parameters includes an MS ID associated with the prediction model, a baseline PS ID associated with the prediction model, a training valid area, a maximum number of epochs of the training process, a minimum number of epochs of the training process, a training deadline, a weight set, periodicity, a server address, or a combination thereof.

[0306] Aspect 28: The method according to any one of aspects 19 to 27, wherein the first message is received from a network node or a server.

[0307] Aspect 29: A method for performing wireless communications at a server, the method comprising: sending a training configuration for a training process associated with a prediction model to a set of UEs, the training configuration comprising a first set of model parameters associated with a first PS ID; receiving one or more reports from one or more UEs in the set of UEs indicating one or more subsets of model parameters output at the UE from the training process for the prediction model; aggregating the subsets of model parameters into a second set of model parameters; assigning a second PS ID to the second set of model parameters, the second PS ID being different from the first PS ID; and sending an indication of the second PS ID.

[0308] Aspect 30: The method according to aspect 29 further comprises: receiving a message from at least one UE in the set of UEs indicating that the prediction model is ready to be activated at the UE, based at least in part on sending the indication.

[0309] Aspect 31: According to the method described in any one of Aspects 29 to 30, the method further includes: sending a message indicating the second set of model parameters used for the prediction model to the set of UEs, wherein the second set of model parameters includes an updated set of model parameters used for the prediction model.

[0310] Aspect 32: The method according to any one of aspects 29 to 31, wherein sending the indication comprises: sending the indication to a network node, to the set of UEs, or both.

[0311] Aspect 33: The method according to any one of aspects 29 to 32, wherein the second PS ID comprises a temporary PS ID or a combination of the first PS ID and a version tag.

[0312] Aspect 34: A method for performing wireless communications at a UE, the method comprising: receiving a training configuration for a training process associated with a prediction model, the training configuration comprising a first set of model parameters associated with a first PS ID; sending a report indicating a subset of model parameters output from the training process for the prediction model at the UE; and receiving an indication of a second PS ID associated with a second set of model parameters based at least in part on sending the report, the second PS ID being different from the first PS ID.

[0313] Aspect 35: The method according to aspect 34 further comprises: sending a message indicating that the prediction model is ready to be activated at the UE based at least in part on receiving the indication.

[0314] Aspect 36: According to the method described in any one of Aspects 34 to 35, the method also includes: receiving a message indicating the second set of model parameters for the prediction model, the second set of model parameters including an updated set of model parameters for the prediction model.

[0315] Aspect 37: The method according to aspect 36, the method also includes: performing the training process for the prediction model using the second set of model parameters and based at least in part on the second PS ID.

[0316] Aspect 38: The method according to any one of aspects 34 to 37, wherein receiving the indication comprises: receiving the indication from a server.

[0317] Aspect 39: The method according to any one of aspects 34 to 38, wherein the second PS ID comprises a temporary PS ID or a combination of the first PS ID and a version tag.

[0318] Aspect 40: An apparatus for performing wireless communications at a network node, the apparatus 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.

[0319] Aspect 41: An apparatus for wireless communication at a network node, the apparatus comprising at least one component for performing a method according to any one of aspects 1 to 11.

[0320] Aspect 42: 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 1 to 11.

[0321] Aspect 43: A device for wireless communication at a server, the device comprising: a processor; a memory coupled to the processor; and instructions stored in the memory and executable by the processor to cause the device to perform a method according to any one of Aspects 12 to 18.

[0322] Aspect 44: An apparatus for wireless communication at a server, the apparatus comprising at least one component for performing the method according to any one of aspects 12 to 18.

[0323] Aspect 45: A non-transitory computer-readable medium storing code for wireless communication at a server, the code comprising instructions executable by a processor to perform the method according to any one of aspects 12 to 18.

[0324] Aspect 46: An apparatus for performing wireless communications at a UE, the apparatus 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 19 to 28.

[0325] Aspect 47: An apparatus for wireless communication at a UE, the apparatus comprising at least one component for performing a method according to any one of aspects 19 to 28.

[0326] Aspect 48: A non-transitory computer-readable medium storing 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 19 to 28.

[0327] Aspect 49: A device for wireless communication at a server, the device comprising: a processor; a memory coupled to the processor; and instructions stored in the memory and executable by the processor to cause the device to perform a method according to any one of Aspects 29 to 33.

[0328] Aspect 50: An apparatus for wireless communication at a server, the apparatus comprising at least one component for performing a method according to any one of aspects 29 to 33.

[0329] Aspect 51: A non-transitory computer-readable medium storing code for wireless communication at a server, the code comprising instructions executable by a processor to perform the method according to any one of aspects 29 to 33.

[0330] Aspect 52: An apparatus for performing wireless communications at a UE, the apparatus 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 34 to 39.

[0331] Aspect 53: An apparatus for wireless communication at a UE, the apparatus comprising at least one component for performing a method according to any one of aspects 34 to 39.

[0332] Aspect 54: A non-transitory computer-readable medium storing code for wireless communication at a UE, the code comprising instructions executable by a processor to perform a method according to any one of aspects 34 to 39.

[0333] It should be noted that the methods described herein describe possible implementations, and that the 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.

[0334] 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.

[0335] The information and signals described herein may be represented using any of a variety of different technologies 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.

[0336] The various illustrative blocks and components described in conjunction with the disclosure herein may be implemented or executed using a general purpose processor, DSP, ASIC, CPU, FPGA or other programmable logic device, discrete gate or transistor logic components, 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, a plurality of microprocessors, one or more microprocessors combined with a DSP core, or any other such configuration).

[0337] The functions described herein may be implemented using hardware, software executed by a processor, firmware, or any combination thereof. If 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 the one or more instructions or codes. Other examples and specific implementations are within the scope of this 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 any combination of these items. Features that implement the functions may also be physically located at different locations, including being distributed so that parts of the functions are implemented at different physical locations.

[0338] Computer-readable media include both non-transient computer storage media and communication media, including any medium that facilitates the transfer of a computer program from one location to another.Non-transient storage media can be any available medium that can be accessed by a general or special computer.By way of example and not limitation, non-transient computer-readable media may include RAM, ROM, electrically erasable programmable ROM (EEPROM), flash memory, compact disc (CD) ROM or other optical disc storage, disk storage or other magnetic storage device, or can be used to carry or store desired program code components and any other non-transient medium that can be accessed by a general or special computer or a general or special processor in the form of an instruction or data structure.Disks and optical discs as used herein include CDs, laser optical discs, optical optical discs, digital versatile discs (DVDs), floppy disks and blue-ray discs.Disks can reproduce data magnetically, and optical discs can reproduce data optically using lasers.The combination of the above items is also included in the scope of computer-readable media.

[0339] As used herein (including in the claims), "or" used in a list of items (e.g., a list of items with 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 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."

[0340] The term "determining" encompasses a variety of actions, and thus, "determining" may include calculating, computing, processing, deriving, investigating, looking up (such as via looking up in a table, database or other data structure), ascertaining, etc. Additionally, "determining" may include receiving (e.g., receiving information), accessing (e.g., accessing data stored in a memory), etc. Additionally, "determining" may include parsing, retrieving, selecting, choosing, establishing, and other such similar actions.

[0341] 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 the 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 reference label, or other subsequent reference labels.

[0342] 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 "having advantages over other examples." The specific embodiments include specific details to provide an understanding of the described technology. However, these technologies may be practiced without these specific details. In some instances, known structures and devices are shown in block diagram form to avoid obscuring the concepts of the described examples.

[0343] 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. A method for wireless communication at a network node, the method comprising: selecting one or more user equipment (UE) for the training process based at least in part on a trigger to activate a training process for a prediction model; as well as An indication of a training configuration for the prediction model is sent to the one or more UEs, the training configuration comprising a set of training parameters based at least in part on the one or more UEs.

2. The method according to claim 1, further comprising: An indication of a data radio bearer configured for downloading a model structure and a baseline parameter set associated with the prediction model is sent to the one or more UEs.

3. The method according to claim 1, further comprising: The set of training parameters for the training configuration is selected based at least in part on an estimated link capacity associated with the one or more UEs, a computing capability associated with the one or more UEs, or a combination thereof.

4. The method of claim 3, wherein the set of training parameters comprises a minimum number of epochs for the training process to be performed at a UE in the one or more UEs.

5. The method of claim 1 , wherein the set of training parameters comprises a model structure identifier associated with the prediction model, a baseline parameter set identifier associated with the prediction model, a training validity region, a maximum number of epochs for the training process, a minimum number of epochs for the training process, a training deadline, a weight set, a periodicity, a server address, or a combination thereof.

6. The method according to claim 1, further comprising: A message is received from at least one UE of the one or more UEs indicating that the at least one UE has implemented the training configuration or indicating that the at least one UE has refrained from implementing the training configuration.

7. The method according to claim 1, further comprising: A message is received from at least one of the one or more UEs indicating that the prediction model is ready to be activated at the at least one UE.

8. The method according to claim 1, further comprising: A message is sent to the one or more UEs including an indication to activate the training procedure at the one or more UEs.

9. A method for wireless communication at a server, the method comprising: selecting one or more user equipment (UE) for the training process based at least in part on a trigger to activate a training process for a prediction model; as well as An indication of a training configuration for the prediction model is sent to the one or more UEs, the training configuration comprising a set of training parameters based at least in part on the one or more UEs.

10. The method according to claim 9, further comprising: An indication of a data radio bearer configured for downloading a model structure and a baseline parameter set associated with the prediction model is sent to the one or more UEs.

11. The method according to claim 9, further comprising: The set of training parameters for the training configuration is selected based at least in part on an estimated link capacity associated with the one or more UEs, a computing capability associated with the one or more UEs, or a combination thereof, wherein the set of training parameters includes a minimum number of epochs for the training process to be performed at a UE among the one or more UEs.

12. The method of claim 9, wherein the set of training parameters comprises a model structure identifier associated with the prediction model, a baseline parameter set identifier associated with the prediction model, a training validity region, a maximum number of epochs for the training process, a minimum number of epochs for the training process, a training deadline, a weight set, a periodicity, a server address, or a combination thereof.

13. A method for wireless communication at a user equipment (UE), the method comprising: receiving a first message indicating a training configuration for a training process for a prediction model, the training configuration comprising a set of training parameters; as well as A second message is sent indicating whether the UE has implemented the training configuration for the training process based at least in part on the set of training parameters.

14. The method according to claim 13, further comprising: receiving an indication of a data radio bearer configured for downloading a model structure and a baseline parameter set associated with the prediction model; as well as The model structure and the baseline parameter set are downloaded via the data radio bearer.

15. The method of claim 13, wherein the set of training parameters includes a minimum number of epochs for the training process, the method further comprising: estimating a number of local epochs for the training process based at least in part on a computational capability of the UE or a link capacity associated with the UE; as well as The estimated local number of epochs is compared to the minimum number of epochs, wherein the second message is sent based at least in part on the comparison.

16. The method of claim 15, wherein the second message instructs the UE to avoid implementing the training configuration based at least in part on the estimated number of local epochs being less than the minimum number of epochs.

17. The method of claim 15, wherein the second message instructs the UE to implement the training configuration based at least in part on the estimated number of local epochs being equal to or greater than the minimum number of epochs.

18. The method according to claim 17, further comprising: performing the training process for the prediction model according to the training configuration and based at least in part on the estimated number of local epochs; as well as A report is sent indicating a set of model parameters for the predictive model based at least in part on performing the training process.

19. The method of claim 18, wherein the second message further includes an indication that the prediction model is ready to be activated at the UE, the method further comprising: Receiving a third message including an indication to activate the training process is based at least in part on sending the second message, wherein performing the training process is based at least in part on receiving the third message.

20. The method according to claim 13, further comprising: The training process is configured according to the set of training parameters, wherein the second message also includes an indication that the configuration of the training process is complete.

21. The method of claim 13, wherein the set of training parameters comprises a model structure identifier associated with the prediction model, a baseline parameter set identifier associated with the prediction model, a training validity region, a maximum number of epochs for the training process, a minimum number of epochs for the training process, a training deadline, a weight set, a periodicity, a server address, or a combination thereof.

22. A method for wireless communication at a server, the method comprising: sending a training configuration for a training process associated with a prediction model to a set of user equipment (UE), the training configuration comprising a first set of model parameters associated with a first parameter set identifier; receiving, from one or more UEs of the set of UEs, one or more reports indicating one or more subsets of model parameters output at the UEs from the training process for the prediction model; aggregating the subset of model parameters into a second set of model parameters; assigning a second parameter set identifier to the second set of model parameters, the second parameter set identifier being different from the first parameter set identifier; as well as An indication of the second parameter set identifier is sent.

23. The method according to claim 22, further comprising: Based at least in part on sending the indication, receiving a message from at least one UE in the set of UEs indicating that the prediction model is ready to be activated at the UE.

24. The method according to claim 22, further comprising: A message is sent to the set of UEs indicating the second set of model parameters for the prediction model, the second set of model parameters comprising an updated set of model parameters for the prediction model.

25. The method of claim 22, wherein the second parameter set identifier comprises a temporary parameter set identifier or a combination of the first parameter set identifier and a version tag.

26. A method for wireless communication at a user equipment (UE), the method comprising: receiving a training configuration for a training process associated with a predictive model, the training configuration comprising a first set of model parameters associated with a first parameter set identifier; sending a report indicating a subset of model parameters output from the training process for the prediction model at the UE; as well as Based at least in part on sending the report, receiving an indication of a second parameter set identifier associated with a second set of model parameters, the second parameter set identifier being different than the first parameter set identifier.

27. The method according to claim 26, further comprising: A message is sent indicating that the prediction model is ready to be activated at the UE based at least in part on receiving the indication.

28. The method according to claim 26, further comprising: A message is received indicating the second set of model parameters for the prediction model, the second set of model parameters comprising an updated set of model parameters for the prediction model.

29. The method according to claim 28, further comprising: The training process for the predictive model is performed using the second set of model parameters and based at least in part on the second parameter set identifier.

30. The method of claim 26, wherein the second parameter set identifier comprises a temporary parameter set identifier or a combination of the first parameter set identifier and a version tag.

Images