Controlling user-plane data collection in a cellular communication system

Abstract

To facilitate user-plane (UP) data collection in a cellular communication network, a user equipment (UE) can transmitting, to a core network (CN), a registration request message including an indication of a capability of the UE with respect to the UP data collection; encapsulate, in a payload container, a UP data collection control message that includes a feature identifier (ID) identifying an artificial intelligence / machine learning (AI / ML)-enabled feature to which the UP data collection pertains; include the payload container in a non-access stratum (NAS) transport message; and transmit, to the CN, the NAS transport message.

Description

PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01PCONTROLLING USER-PLANE DATA COLLECTION IN A CELLULAR COMMUNICATION SYSTEM CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to and the benefit of the filing date of provisional U.S. Patent Application No. 63 / 779,273 entitled “Controlling User-Plane Data Collection in a Cellular Communication System, filed March 27, 2025, and provisional U.S. Patent Application No. 63 / 803,509 entitled “Controlling User-Plane Data Collection in a Cellular Communication System, filed May 9, 2025. The entire content of the provisional applications is hereby expressly incorporated herein by reference.FIELD OF THE DISCLOSURE

[0002] This disclosure relates generally to methods, devices, and articles in wireless communication systems, such as 3GPP communication systems, and in particular to exchanging control messages related to user-plane data collection.BACKGROUND

[0003] This background description is provided for the purpose of generally presenting the context of the disclosure. Work of the presently named inventors, to the extent it is described in this background section, as well as aspects of the description that may not otherwise qualify as prior art at the time of filing, are neither expressly nor impliedly admitted as prior art against the present disclosure.

[0004] The 3rd Generation Partnership Project (3GPP) contemplates convergence between communication system functions and Artificial Intelligence (Al) technology, to improve the intelligence at the fifth-generation core (5GC) and the air interface and support network automation. More particularly, 3 GPP has begun to address such topics as data collection (DC), machine learning (ML) model training, analytics inference, etc. These technical areas require collaborative AI / ML mechanisms for coordinating functionality across such domains as user equipment (UE), radio access network (RAN), 5GC, operations and maintenance (0AM), and application functions (AF).PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0005] To support collection of training data for a UE-side model that generates complete inferences at the UE, based on the specifications related to AI / ML for New Radio (NR) air interface documented in 3 GPP TR 38.843, 3 GPP has agreed to study the potential support of UE data collection to meet the requirements for RAN Al support for air interface operation. Today, the remaining issues include enhancement of UE data collection and policy control in a 5G system.

[0006] 3GPP has identified several options for data collection for UE-side model training: (i) a UE collecting and directly transferring training data to an Over-the-Top (OTT) server, in a transparent or non-transparent manner; (ii) the UE collecting training data and transferring the training data to the core network, which then transfers the training data to the OTT server; or (iii) the UE collecting training data and transferring the training data to an 0 AM, which then transfers the relevant data to the OTT server.

[0007] According to TS 23.288, a network data analytics function (NWDAF) can support data collection for network automation based on the UE application data collection framework described in TS 26.531. The NWDAF can interact with a data collection application function (DCAF) in the trusted domain of the operator network, to collect data from a UE application as an input for analytics generation and ML model training. This framework can apply to the UE collecting and directly transferring training data to an OTT server as discussed above, where the DCAF operates as a data collection network function (DCNF) that connects to the User-Plane Function (UPF) via an N6 interface.

[0008] To comply with local regulations and operator policy, and before collecting AI / ML model training data, the network should obtain, from the UE, user consent for the corresponding AI / ML-enabled features for various purposes, e.g., AI / ML training for a UE-sided model, a RAN-sided model, or a CN-side model. However, according to TS 23.501 and TS 23.288, the user consent for data collection currently is based on the subscription information stored in a Unified Data Management function (UDM). The subscription information indicates whether the user authorizes data collection and usage of the UE data for a particular purpose, where the purpose for data collection is analytics or model training.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0009] The existing mechanisms are designed for collecting UE data managed and stored in the core network, and the NWDAF generates an AI / ML model or analytics requested by consumers (e.g., an AF, a Policy Control Function (PCF), a Session Management Function (SMF), an Access & Mobility Management Function (AMF), or an 0AM) for network-side inferencing. In addition to collecting training data for AI / ML related to the NR air interface, various other 3GPP features can require training data collection. These 3GPP features can correspond to various use cases and can span various domains such as the RAN, the core network, and cloud applications. These AI / ML-enabled features are expected to enhance resource management for the radio interface, radio resource management, RAN mobility, and core network automation.

[0010] However, currently there are no mechanisms for controlling collection of training data, sensing data, and other types of auxiliary data. For example, there is no mechanism a UE can use to establish a user-plane (UP) data connection, release a UP data connection, suspend a UP data connection, resume a UP data collection, etc., when using a network-provided AI / ML model training service. Further, it is unclear how a network and a UE can interact to control the UE operation for collecting and / or transferring training data for each AI / ML-enabled feature, when the UE enables one or more Al / ML-enabled features or when there are different entities collecting and transferring data to the network for the same AI / ML-enabled feature.SUMMARY

[0011] The techniques discussed below address the challenges outlined above and support controlling UP data collection for AI / ML features, sensing, and other services.

[0012] An example embodiment of these techniques is a method for facilitating UP data collection in a cellular communication network, the method implemented in a UE and comprising: transmitting, to a CN, a registration request message including an indication of a capability of the UE with respect to the UP data collection; encapsulating, in a payload container, a UP data collection control message that includes a feature ID identifying an AI / ML-enabled feature to which the UP data collection pertains; including the payload container in a NAS transport message; and transmitting, to the CN, the NAS transport message.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0013] Another example embodiment of these techniques is a method for facilitating UP data collection in a cellular communication network, the method implemented in a CN and comprising: receiving, from a UE, a registration request message including an indication of a capability of the UE with respect to the UP data collection; and encapsulating, in a payload container, a UP data collection control message that includes a feature ID identifying an AI / ML-enabled feature to which the UP data collection pertains; including the payload container in a NAS transport message; and transmitting, to the UE, the NAS transport message.

[0014] Another example embodiment of these techniques is a method for facilitating userplane (UP) data collection in a cellular communication network. The method is implemented in a first network endpoint and comprises negotiating, with a second network endpoint, support of the UP data collection; and communicating, using a non-access stratum (NAS) transport message, information related to managing the UP data collection.

[0015] Another example embodiment of these techniques is a network device comprising processing hardware, the network device configured to implement the method above.

[0016] Another example embodiment of these techniques is a method for facilitating userplane (UP) data collection from a downstream network device. The method is implemented in a core network and comprises negotiating, with the downstream network device, support of the UP data collection; and communicating, using a non-access stratum (NAS) transport message, information related to managing the UP data collection.

[0017] Still another example embodiment of these techniques is a core network of a cellular communication network, the core network comprising processing hardware and configured to implement the method above.BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Fig. 1 is a block diagram of an example wireless communication system that can implement one or more of the techniques of this disclosure for auxiliary data collection to support sensing functions, artificial intelligence (AI) / machine learning (ML) functions, and other user-plane (UP) data functions;PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0019] Fig. 2 is a block diagram of an example protocol stack according to which the UE of Fig. 1 can communicate with the RAN of Fig. 1;

[0020] Fig. 3A is a service-based representation of the CN architecture, which the system of Fig. 1 can implement, and in which multiple network functions (NFs) implement respective instances of a data collection network function (DCNF) to support the corresponding data collection functions;

[0021] Fig. 3B illustrates another example service-based representation of the CN architecture, in which a DCNF operates as a separate, dedicated NF, from which support other NFs can request data collection services;

[0022] Fig. 3C illustrates another example service-based representation of the CN architecture, in which a Network Data Analytics Function (NWDAF) implements a DCNF, so that other NFs in the core network can request data collection services from the DCNF-enhanced NWDAF;

[0023] Fig. 4 is a reference-point based representation of the CN architecture of Fig. 3B, which the system of Fig. 1 can implement;

[0024] Fig. 5 is messaging diagram of an example scenario in which the UE or the CN of Fig.1 initiates auxiliary data collection;

[0025] Fig. 6 is a messaging diagram of an example scenario in which the network initiates a UP data collection management procedure;

[0026] Fig. 7 is a messaging diagram of an example scenario in which the UE initiates a UP data collection management procedure;

[0027] Fig. 8A is a messaging diagram of an example scenario in which a Unified Data Management function (UDM) uses a service call for data set “Policy Data” and subset “UE context policy control data” with a dedicated information element to indicate UE capability with respect to UP data collection;

[0028] Fig. 8B is a messaging diagram of an example scenario in which a UDM uses a service call for data set “Policy Data” and subset “PDU session policy control data” with a dedicatedPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Pinformation element to identify allowed one or more AI / ML-enabled features for UP data collection;

[0029] Fig. 9 is a messaging diagram of an example scenario in which a core network uses Policy and Charging Control (PCC) rules that include UP data collection configuration information with a DCNF address and a port number;

[0030] Fig. 10 is a messaging diagram of an example scenario in which a core network uses N4 rules to enforce a Forwarding Action Rule (FAR) that directs UP data collection traffic;

[0031] Fig. 11 A is a messaging diagram of an example scenario in which network endpoints (implemented in a UE, a RAN node, a component of a CN node, a dedicated CN node, or a set of CN nodes) include a UP data collection control message (UPDC-CM) in a payload container of type UE parameters update transparent container,

[0032] Fig. 1 IB is a messaging diagram of an example scenario in which network endpoints include a UPDC-CM in a payload container of a type dedicated to conveying control information for UP data collection;

[0033] Fig. 12 is a messaging diagram of an example scenario in which network endpoints transmit and receive, respectively, a UPDC-CM including several information elements;

[0034] Fig. 13 A is a messaging diagram illustrating several example UPDC-CMs which network endpoints can exchange to manage UP data collection;

[0035] Fig. 13B illustrates a messaging scheme generally similar to that of Fig. 13 A, but with a generic UPDC-CM including an indication of a respective control function;

[0036] Fig. 14 is a flow diagram of an example method in a CN node for negotiating and initiating UP data collection;

[0037] Fig. 15 is a flow diagram of an example method in a UE for negotiating and initiating UP data collection; and

[0038] Fig. 16 is a flow diagram of an example method for facilitating UP data collection in a cellular communication network, which can be implemented in a network endpoint.DETAILED DESCRIPTION OF THE DRAWINGSPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01POverview

[0039] Generally speaking, a UE and one or more network functions (NT's) operating in core network (CN) support a framework for managing data collection for auxiliary data such as AI / ML training (or simply “training data”), sensing data, interference data, etc. In the examples below, the network devices communicate the auxiliary data over a user plane (UP), and the auxiliary data is referred below to as “user-plane (UP) data. More generally, network devices can exchange auxiliary data over another plane, e.g., a plane dedicated to transporting auxiliary data.

[0040] The network endpoints that communicate UP data can reside in a UE, in a RAN node, or one or more CN nodes.

[0041] The training data the UE transmits to a data control function (DCF) or DCNF (in this disclosure, the terms “DCF” and “DCNF” are used interchangeably) can apply for various 3 GPP system features and can correspond to various use cases related to RAN, CN, and / or cloud functionality. The DCNF can be configured to exchange UE Data Transfer Control Messages (UEDT-CM) with the UE over user plane for managing the UE data transfer for an AI / ML enabled feature; receive standardized data from the UE and store the data (accordingly, the CN obtains the full visibility of standardized UE data contents of AI / ML enabled features); manage event exposure subscription to data consumer; and report UE data to the Application Server (AS), e.g. UE side server or OTT server, over user plane based on event exposure subscription for UE data collection.

[0042] The AI / ML-enabled features can enhance resource management for the radio interface, radio resource management, RAN mobility, core network automation, etc. Some of the specific examples of such features include AI / ML based channel state information (CSI) compression, AI / ML based CSI reference signal (CSLRS) overhead reduction, AI / ML-based CSI prediction, AI / ML based beam management with downlink beam prediction, direct AI / ML positioning, and AI / ML-assisted positioning for positioning accuracy enhancement, supporting AI / ML based network slicing, AI / ML based coverage and capability optimization, AI / ML based Network Energy Saving (NES) for finer granularities, AI / ML enabled radio resource management (RRM) measurement prediction, measurement event prediction, AI / ML enabled quality of service (QoS) sustainability analytics enhancement, QoS and Policy Assistance Analytics enhancement, PacketPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01PData Unit (PDU) session traffic analytics enhancement, movement behavior analytics enhancement, location accuracy analytics enhancement, etc.

[0043] For example, an AI / ML-enabled feature can enhance 3GPP -based sensing services or non-3GPP based sensing services, including sensing for communication and communication for sensing. In this case, the network can collect the sensing data from the UE or a RAN node, for AI / ML-enabled training services the network provides. In the discussion below, the term “network” refers to the core network (or “CN”).

[0044] In at least some of the implementations, the AI / ML-enabled features have respective feature identifiers (IDs). A UE in general supports one or more AI / ML-enabled features and uses a feature ID to associate particular training data with a particular AEML-enabled feature. More generally, a UE and / or the network (e.g., the CN) can use feature IDs when handling UP data collection and policies that control UP data connections for UP data collection. In some cases, for a specific AI / ML-enabled feature, one or more AI / ML models may apply. When only one AI / ML model applies to an AI / M- enabled feature, the Feature ID can identify an AI / ML model as a Model ID (and, conversely, the Model ID can identify the feature in this case).

[0045] The approaches discussed below can apply to a UE, a DCNF (which can be implemented in a legacy NF or can operate as a special-purpose, dedicated NF), or both.

[0046] In some implementations, the DCNF operates within an existing (legacy) NF such as for example a Location Management Function (LMF), a PCF, an AMF, or an SMF. The entity in which the DCNF operates can support service operations for UP data collection directly. For example, an AI / ML-enabled feature can be related to positioning accuracy enhancement, and the LMF can operate as the DCNF to provide the UP data collection service for that feature.

[0047] In other implementations, the DCNF is a dedicated NF configured to support operations related to UP data collection for, or on behalf of, the existing (legacy) NFs such as an LFM, a PCF, an AMF, or an SMF for example. According to this approach, NFs can request a UP data collection service from the dedicated DCNF.

[0048] In yet other implementations, a DCNF operates in an NWDAF with coordination functionality, i.e., with a Data Collection Coordination Function (DCCF). According to thisPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Papproach, NFs operate as consumers and request a data collection service from the NWDAF with the DCCF. Accordingly, the NWDAF with DCCF can enhance operations ofNnwdctf ' DataManagement services, to collect training data from a UE.

[0049] While this disclosure uses 5G networks as an example, one will appreciate that the principles of this disclosure generally apply to 3G networks, 4G networks, 6G networks, and future generations of networks.

[0050] In this disclosure, a UP data collection control message (UPDC-CM) can also be understood as a UE Data Transfer Control Messages (UEDT-CM).

[0051] The techniques of this disclosure can apply to the following example scenario: an Application Function (AF) performing UE-side model training can request data collection from 5G system for an AI / ML enabled feature with UE performing AI / ML model inferencing using UE side AI / ML model. For such an AI / ML-enabled feature, the UE can start AI / ML model inferencing based on a corresponding generalized AI / ML model (e.g. obtained from the application server (AS) supporting Federated learning) or available AI / ML model at the UE.

[0052] To further provide customized AI / ML model for the UE side AI / ML model inferencing, the AS performing UE-side model training can request a data control function (DCF) or data control network function (DCNF) in the core network for reporting related data of the AI / ML-enabled feature.

[0053] The UE would receive data collection request from the DCF to perform data collection and report related inferencing input data, inferencing output data, inferencing results, and corresponding situational information (e g. location, time, etc ), to the DCF and AF / AS. As a result, the UE can then obtain a customized UE side AI / ML model for the AI / ML enabled feature. When depicting UE data collection, transfer, and reporting in this solution, it includes all the above relevant standardized data needed for UE side model training.

[0054] To achieve end-to-end coordination for UE data collection, transfer, and report for a specific AI / ML enabled feature among UE, RAN node, Core network, AF / AS, AI / ML enabled feature ID is introduced with the following assumptions: (i) for a specific AEML-enabled feature, one or more AI / ML models may be applied. As such, when only one AI / ML model isPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Papplied for an AI / ML enabled feature, the AI / ML enabled feature ID can be used to identify an AI / ML model as Model ID, and vise versa; (ii) a UE may support one or more AI / ML-enabled features and each AI / ML-enabled feature may have one or more AI / ML models suited for different situations, e.g. time, location, indoor, outddor, etc.

[0055] To support the full controllability of standardized UE data transfer over UP, a DCF can be introduced in the core network with the following functionalities: (i) to exchange UE Data Transfer Control Messages (UEDT-CM) with the UE over user plane for managing the UE data transfer for an AI / ML enabled feature; (ii) to receive standardized data from the UE and store them; accordingly, 5GC obtains the full visibility of standardized UE data contents of AI / ML enabled features; (iii) to manage event exposure subscription to data consumer; (iv) to report UE data to the Application Server (AS), e.g. UE side server or OTT server, over user plane based on event exposure subscription for UE data collection.

[0056] The techniques discussed below can enhance the Nnef AFsessionWithQoS service, which allows the AF to requests the network for an AF session to provide a specific QoS for UE data collection and transfer. Accordingly, the PCF can generate PCC rules for the SMF to configure UE(s), RAN, PSA UPF, and DCF for managing the UE data transfer for the triggered AI / ML enabled feature, and handling the event exposure subscription for UE data reporting to the AF / AS supporting UE-side model training.Example architecture

[0057] Referring first to Fig. 1, an example wireless communication system 100 can implement one or more of the techniques of this disclosure for managing user consent for UP data collection. The example wireless communication system 100 includes a UE 102, a base station (BS) 104, a base station 106, and a core network (CN) 110, such as a fifth generation (5G) core (5GC). The base stations 104 and 106 can operate in a RAN 105 connected to the CN 110. The CN 110 can also be implemented as a sixth generation (6G) core or another suitable core network.

[0058] The base station 104 covers a cell 124, and the base station 106 covers a cell 126. If the base station 104 is a gNB, the cell 124 is an NR cell. If the base station 124 is an ng-eNB, the cell 124 is an evolved universal terrestrial radio access (E-UTRA) cell. Similarly, if the basePATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Pstation 106 is a gNB, the cell 126 is an NR cell, and if the base station 126 is an ng-eNB, the cell 126 is an E-UTRA cell. The cells 124 and 126 can be in the same Radio Access Network Notification Areas (RNA) or different RNAs. The cells 124 and 126 can partially overlap, so that the UE 102 can select, reselect, or hand over from one of the cells 124 and 126 to the other. In general, the RAN 105 can include any number of base stations, and each of the base stations can cover one, two, three, or any other suitable number of cells. The UE 102 can support at least a 5GNR (or simply, “NR”) air interface to communicate with the base stations 104 and 106. Each of the base stations 104, 106 can connect to the CN 110 via an interface (e.g., SI or NG interface). The base stations 104 and 106 also can be interconnected via an interface (e.g., X2 or Xn interface) for interconnecting NG RAN nodes.

[0059] Several network functions (NFs) that make up the CN 110 are discussed below with reference to Figs. 3A-C. The CN 110 includes a DCNF 150, which in different implementations can operate as a separate NF within the CN 110 or as a component of another NF such as an LMF for example. In some implementations, a data collection application function (DCAF) 152 can operate in the trusted domain of the CN 110. The CN 110 can communicate with an application server (AS) 146, which also can operate in a trusted domain. Further, in some scenarios or implementations, one or more NFs of the CN 110 operate on a CN-side AI / ML model 147.

[0060] While not shown in Fig. 1 to avoid clutter, the CN 110 may include processing hardware, which may include one or more general -purpose processors (e.g., CPUs) and a non-transitory computer-readable memory storing instructions that the one or more general-purpose processors execute. Additionally or alternatively, the processing hardware can include specialpurpose processing units.

[0061] The base station 104 can be equipped with processing hardware that can include one or more general -purpose processors (e.g., CPUs) and a non-transitory computer-readable memory storing instructions that the one or more general-purpose processors execute (not shown).Additionally or alternatively, the processing hardware can include special-purpose processing units.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0062] The UE 102 is equipped with processing hardware 130 that can include one or more general-purpose processors such as CPUs and non-transitory computer-readable memory storing machine-readable instructions executable on the one or more general-purpose processors, and / or special-purpose processing units. The UE 102 also includes a transceiver 132 to communicate with the RAN 105 over a radio interface. Further, the UE 102 includes a memory 134 storing a data collection controller 136 configured to implement one or more of the techniques for managing UP data collection discussed in this disclosure. The memory 134 in some cases further stores a UE-side ML model (not shown).

[0063] The base station 106 is equipped with processing hardware 140 that can include one or more general -purpose processors such as CPUs and non-transitory computer-readable memory storing machine-readable instructions executable on the one or more general-purpose processors, and / or special-purpose processing units. The base station 106 also includes a transceiver 142 to communicate with the UE 102 over a radio interface. The base station 106 further includes a memory 134 storing a data collection controller 146 configured to implement one or more of the techniques for managing UP data collection. The base station 104 may be configured in a similar manner as the base station 106.

[0064] The collection controller 136, the data collection controller 146, and the DCNF 150 define network endpoints for UP data collection in the UE 102, the RAN 105, and the CN 110, respectively. The DCNF 150 can collect UP data from the UE 102 or the RAN 105 to train corresponding models, for example.

[0065] Fig. 2 illustrates, in a simplified manner, an example protocol stack 200 according to which the UE 102 can communicate with an eNB / ng-eNB 230 or a gNB 232 (e.g., one or more of the base stations 104, 106). In the example stack 200, a physical layer (PHY) 202A of EUTRA provides transport channels to the EUTRA MAC sublayer 204A, which in turn provides logical channels to the EUTRA RLC sublayer 206A. The EUTRA RLC sublayer 206A in turn provides RLC channels to a EUTRA PDCP sublayer 208 and, in some cases, to an NR PDCP sublayer 210. Similarly, the NR PHY 202B provides transport channels to the NR MAC sublayer 204B, which in turn provides logical channels to the NR RLC sublayer 206B. The NR RLC sublayer 206B in turn provides data transfer services to the NR PDCP sublayer 210. ThePATENT APPLICATION Attorney Docket No.: 31730 / 309119-01PNR PDCP sublayer 210 in turn can provide data transfer services to Service Data Adaptation Protocol (SDAP) 212 or a radio resource control (RRC) sublayer (not shown in Fig. 2). The UE 102, in some implementations, supports both the EUTRA and the NR stack as shown in Fig. 2, to support handover between EUTRA and NR base stations and / or to support DC over EUTRA and NR interfaces. Further, as illustrated in Fig. 2, the UE 102 can support layering of NR PDCP 210 over EUTRA RFC 206A, and SDAP sublayer 212 over the NR PDCP sublayer 210.

[0066] The EUTRA PDCP sublayer 208 and the NR PDCP sublayer 210 receive packets (e.g., from an Internet Protocol (IP) layer, layered directly or indirectly over the PDCP layer 208 or 210) that can be referred to as service data units (SDUs), and output packets (e. , to the RLC layer 206A or 206B) that can be referred to as protocol data units (PDUs). Except where the difference between SDUs and PDUs is relevant, this disclosure for simplicity refers to both SDUs and PDUs as “packets.”

[0067] On a control plane, the EUTRA PDCP sublayer 208 and the NR PDCP sublayer 210 can provide signaling radio bearers (SRBs) or RRC sublayer (not shown in Fig. 2) to exchange RRC messages or non-access-stratum (NAS) messages, for example. On a user plane, the EUTRA PDCP sublayer 208 and the NR PDCP sublayer 210 can provide data radio bearers (DRBs) to support data exchange. Data exchanged on the NR PDCP sublayer 210 can be SDAP PDUs, Internet Protocol (IP) packets, or Ethernet packets.

[0068] Fig. 3A is a service-based representation 300A of an example CN architecture, which the system of Fig. 1 can implement as the CN 110. In the representation 300A, the overall nonroaming reference architecture of the policy and charging control (PCC) framework for the 5GS includes components illustrated using solid lines, and the other components are illustrated using dashed lines. According to this representation, network functions enable other authorized network functions to access their services. The components that are outside the PCC framework include a Network Slicing Selection Function (NSSF) 302, a Network Repository Function (NRF) 306, a Unified Data Management (UDM) 308, an Edge Application Server Discovery Function (EASDF) 310, a Network Slice Specific Authentication and Authorization Function (NSSAAF) 312, an Authentication Server Function (AUSF) 314, a Service CommunicationPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01PProxy (SCP) 316, and a Network Slice Admission Control Function (NSACF) 318. The non-PCC architecture further includes the UE 102 and the (R)AN 105.

[0069] The PCC framework in the architecture 300 A includes a Unified Data Repository (UDR) 352, a Network Exposure Function (NEF) 354, a network data analytics function (NWDAF) 356, an Application Function (AF) 358, a Policy Control Function (PCF) 360, a Charging Function (CHF) 362, an Access & Mobility Management Function (AMF) 364, a Session Management Function (SMF) 366, and a User Plane Function (UPF) 370. The UPF 370 can access a data network (DN) 330, in which an application server (AS) 311 can operate.

[0070] The AMF 364 is generally configured to manage registration, connection, and mobility of a UE (such as the UE 102) and provide transport for session management (SM) messages between the UE 102A and the SMF 366. The SMF 366 is generally configured to manage sessions, allocate IP addresses for UEs, and provides downlink (DL) notifications. In some implementations, the SMF 366 also includes following functionalities for PIN service: providing per-QoS flow non-3GPP QoS assistance information to the UE (e.g, PEGC), and supporting IP address allocation to UE and Packet Detection Rule (PDR) configuration with packet filter set for PIN to UPF for framed routing based on PIN group information from the UDM 308.

[0071] The UDM 308 is generally configured to handle user identification, access authorization based on subscription data, and subscription management. In some implementations, the UDM 308 supports the functionality of PIN group management handling.

[0072] The UDR 352 is generally configured to store subscription-related information, such as subscription data, policy data, structured data for exposure, and application data. The UPF 370 is generally configured to handle packet routing and forwarding. The NEF 354 is generally configured to expose a network’s capabilities and services to authorized third-party applications.

[0073] The AF 358 in some deployment operates in a trusted domain 311 or outside the trusted domain 311, i.e., in a non-trusted domain. The trusted domain 311 is generally internal to the CN 110 and includes such components as the UDM 308, the UDR 352, the PCF 360, the AMF 364, the SMF 366, and the UPF 370. Generally speaking, an AF operating outside the trusted domain 311 (such as operated by an authorized third-party entity) can access the network functions of the CN 110 only via the NEF 354, whereas an AF operating within the trustedPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Pdomain 311 can access at least some of the network functions of the CN 110 directly, or may access these functions via the NEF 354 in some deployments.

[0074] According to the architecture 300A, several NFs implement respective instances of a DCNF. For example, the LMF 245 implements an LMF DCNF 350A, the PCF 360 implements a PCF DCNF 350B, an SMF 366 implements an SMF DCNF 350C, the AMF 364 implements an AMF DCNF 350D, and the UPF 370 implements a UPF DCNF 350E. Each of the DCNF instances 350A-E can support the corresponding NF directly. As a more specific example, the LMF DCNF 35OA provides UP data collection service for an AI / ML enabled feature for positioning accuracy enhancement.

[0075] Referring to Fig. 3B, a DCNF 35OG according to an architecture 300B operates as a separate, dedicated data collection NF to support service operations for UP data collection on behalf of the legacy NFs (the LMF 345, the PCF 360, the AMF 364, etc.) and special-purpose NFs such as a sensing service network function (SSF) (not shown). Tn this case, the NFs can request UP data collection service from the dedicated DCNF 350G. For example, the DCNF 350G can provide a UP data collection service for the NWDAF 356 operating as a service consumer, for an AI / ML enabled feature configured to enhance NWDAF analytics. As another example, the DCNF 350G can operate as a UP data collection service producer to provide UP data collection service for the LMF 345 operating as a service consumer, for an AI / ML enabled feature configured to enhance positioning accuracy.

[0076] Referring now to Fig. 3C, a DCNF 350G, the NWDAF 356 can implement a DCNF 350H. Here, the other NFs operate as consumers and can request data collection services from the NWDAF 356 equipped with the DCNF 350G. For example, the NWDAF 356 can enhance Nnwdaf Datamanagement services operations to collect training data from the UE 102.

[0077] Fig. 4 is a reference-point based representation 400 of the example 5GS architecture discussed with reference to Fig. 3B. In Fig. 4, the non-roaming reference architecture of the PCC framework for the 5GS is illustrated as blocks and connections with solid lines, and components and connections outside the PCC framework are illustrated using dashed lines. The interfaces are labeled with the corresponding protocols, e.g., N3 between the RAN 105 and the UPF 370, N1 between the AMF 364 and the UE 102, etc.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0078] The DCNF 350G can support respective interfaces with the SMF 366, the PCF 360, the NWDAF 356, and other NFs that may require UP data collection services.

[0079] The communication system shown Figs. 1-4 in may include additional, fewer, and / or alternative devices or functionalities, and may be configured to perform additional, fewer, or alternate actions, including functionalities / actions described herein.

[0080] Generally speaking, similar events in Figs. 5-16 are labeled with similar reference numbers that share two least significant digits, with differences discussed below where appropriate. For example, event 502 is similar to event 603, and event 602 is similar to event 610.Example techniques for controlling UP data collection

[0081] Fig. 5 is messaging diagram of an example scenario 500 in which the UE 102 for example or the CN 110 initiates auxiliary data collection. More particularly, the scenario 500 corresponds to a high-level procedure for data collection management through the exchange of information between a UE and a network to control UP data collection. In this scenario, a UE-side ML model operating in the UE 102, a RAN-side model operating in the RAN 105, or network-side model operating in the CN 110 can require and initiate UP data collection associated with an AI / ML training service, for example.

[0082] The UE 102 performs 502 a registration procedure to negotiate support of UP data collection support with the network. In an example implementation, this procedure can be implemented similar as described in 3GPP TS 23.502, clause 4.2.2.2. The AMF 364 selects an instance of the PCF 360, for the UE 102, and associates the instance with the UE 102.

[0083] When performing 502 the registration procedure, the UE 102 can send, to the CN 110, a Registration Request message including at least one of the following: (i) UE capabilities with respect to UP data collection, i.e., the capability to collect required data and transfer the data over a user plane connection to the network, and (ii) UE capabilities with respect to controlling UP data collection, i.e., is the capability to control a UE-initiated data collection procedure or handle a network-initiated data collection procedure. The UE can 102 send information (i) or (ii) as a part of 5GMM capability or a part of 6GMM capability, for example.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0100] The AMF 364 or the UDM 308 (or the UDR 352) stores the UE capabilities with respect to UP data collection. Additionally, the UE 102 can provide, to the AMF 364, a support indicator for URSP or DCRP (data collection routing policy) transport using control plane, as a part of UE Policy Classmark IE . The network can use the URSP or DCRP to providing UE policies related to UP data collection.

[0084] If the AMF 364 determines that the network supports UP data collection, e.g. for AI / ML model training at the network, the AMF 364 sends a Registration Accept message, including the following information (as a part of 5GS Network Feature Support IE or as part of 6GS Network Feature Support IE), for an AI / ML model training service the network provides to the UE 102: (i) support of UP data collection, and (ii) support of control of the UP data collection. Otherwise, the AMF 364 does not related information related to network support of UP data collection, and in some cases includes the corresponding cause value to indicate lack of supporting with respect to UP data collection. In the event the AMF 364 indicates support of UP data collection only, without indicating support of controlling UP data collection, the UE 102 can transmit a PDU Session Modification / Release request in order to terminate UP data collection for one or more AI / ML enabled feature(s), and the SMF 366 can transmit a PDU Session Modification / Release command to terminate a UP data collection for one or more AI / ML enabled feature(s).

[0085] If the network supports UE policy handling and the triggering condition is met, the PCF or DCNF (Data Collection network function) performs 510 UE Configuration Update procedure to provision UE policies for the UE. The DCNF can performs UE Configuration Update procedure directly or via PCF using Namf_Communication_NlN2MessageTransfer message to provision Session Management UE policies to the UE via AMF, which can refer to TS23.502 clause 4.2.4.3: UE Configuration Update procedure for transparent UE policy delivery (Figure 4.2.4.3-1: UE Configuration Update procedure for transparent UE Policy delivery) when the PCF / DCNF decides to update UE policy based on triggering conditions. The triggering condition at the PCF / DCNF includes at least one of the following: (i) the DCNF determines for enabling UP data collection from the UE for a specific AI / ML enabled feature using AI / MLPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Pmodel training service at the network, (ii) the DCNF receives AF or other network function's request, or (iii) the DCNF determines to update UE policy related to UP data collection.

[0086] The UE 102 and the network use 520 DL NAS Transport messaging and UL NAS Transport messaging to exchange information related to UP data collection management. These techniques are discussed in more detail below.

[0087] With continued reference to Fig. 5, based on the triggering condition of the UE policy evaluation, the UE 102 can perform 550 the following: (i) evaluate UE policies enhanced for UP data collection, (ii) determine to enforce a matched UE policy, and (iii) determine whether to perform a PDU Session Establishment / Modification request procedure for UP data collection for one or more Al / ML enabled features (e.g. , for Al / ML model training service at the network). The triggering condition for UE policy evaluation can be one the following: (i) the UE 102 receiving an upper layer (application) or a lower layer (e.g. RRC / MAC / SDAP) request for establishing or releasing UP data collection, or (ii) the UE 102 receiving an upper layer or lower layer request to update or request information regarding UP data collection of one or more Al / ML enabled features.

[0088] Based on the enforced UE policy enhanced for UP data collection, the UE 102 performs 560 a PDU Session Establishment / Modification request procedure (e.g., generally as described in TS 23.502 clause 4.3.2). The SMF 366 may select the PCF 360 / the DCNF 350 and create PCF 360 / DCNF 350 association for the PDU Session to retrieve and / or update UE policies of the PDU Session. The PCF for the UE 102 and the PCF for the PDU Session may be the same (e.g., the PCF 360) or different.

[0089] As part of the PDU Session Establishment / Modification request procedure, the PCF 360 and the SMF 366 performs the following steps: (i) the PCF 360 can request subscription information at the PDU Session establishment or PDU Session modification during the UE Policy Association Establishment procedure. The UDR 352 can provide policy control subscription profile information during the UE Policy Association Establishment procedure using a Nudr service for Data Set = "Policy Data" and Data Subset = "UE context policy control data.”; (ii) the SMF 366 can obtain PCC rules from the PCF 360, and the PCC rules can contain the UP Data collection configuration information. The SMF 366 accordingly can configure thePATENT APPLICATION Attorney Docket No.: 31730 / 309119-01PUPF 370 using N4 rules; (iii) the N4 rules can include UP data connection information, which can guide the PSA UPF 370 to route the traffic with related AI / ML data toward the corresponding DCNF address and port information or the FQDN (fully qualified domain name) information for a specific AI / ML-enabled feature. The UE 102 then receives, from the network, a PDU Session Establishment Accept message or a PDU Session Modification Command message, which confirms the use of the UP data collection for the PDU Session.

[0090] Based on the UE policy and address information of DCNF for each AI / M-enabled feature, the UE 102 requests 570 one or more secure user=plane connections with the DCNF 350, e.g. using TLS over TCP or QUIC, for UP data collection of one or more AI / ML-enabled features.

[0091] Fig. 6 is a messaging diagram of an example scenario 600 in which the network initiates a UP data collection management procedure. More particularly, Fig. 6 illustrates an example of signaling transport for network -initiated UP Data Collection Management procedure using UPDC-CM messages as discussed with reference to Figs. 13A and 13B. More generally, the scenario 600 can incorporate the techniques discussed below with reference to Figs. 11 A-13B. When the network determines to request UP Data Collection Management from the UE 102, the DCNF 350 / PCF 360 requests the AMF 364 to send a first UPDC-CM (User Plane Control Message) in the DL NAS Transport message. Accordingly, the UE 102 handles the PDU Session, user plane data connection, and sends, to the network , a UL NAS Transport message including a second UPDC-CM in response to the first UPDC-CM received in the DL NAS Transport message.

[0092] Procedures 602 and 610 are similar to the procedures 502 and 510, respectively, discussed above. The DCNF 350 determines 623 to send, to the UE 102 via the AMF 364, a UPDC control message (UPDC CM) for performing UP data collection management for a specific AI / ML enabled feature, identified by an AI / ML enabled feature ID. To this end, the DCNF 350 can generate a correlation ID to enable the correlation between the UPDC CM and the corresponding data to be collected from the UE 102. The DCN 350 sends 624, to the AMF 364, an Namf_Communication_NlN2MessageTransfer message including the UPDC_CM, DCNF info, the correlation ID, and the AI / ML-enabled feature ID. The AMF 364 sends 625, toPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Pthe UE 102, a DL NAS Transport message including the UPDC CM, the DCNF info, the correlation ID, and the AI / ML-enabled feature ID.

[0093] Example UPDC CM messages are discussed with reference to Fig. 13 A and Fig. 13B.

[0094] The DCNF 350 sends 624, and the AMF 364 forwards 625, the correlation ID to enable the correlation between the UPDC CM and the corresponding data to be collected, transferred to, and handled in the DCNF 350 from one entity, e.g. UE, RAN node, Application Server. Further, the DCNF 350 sends 624, and the AMF 364 forwards 625, the feature ID of an AI / ML-enabled feature to associate the collected data with the AI / ML enabled feature. The AI / ML-enabled feature may require data collection from different entities, e.g. a RAN node such as the base station 104, the UE 102, an O&M entity, or a third party application server.

[0095] The DCNF information can include a DCNF address as an end point of a user plane connection in the network for collecting data of the AI / ML enabled feature. For example, the address can be the FQDN address (which can be used for flexibility in managing DNS changes with a separate port numbers configuration), a URL address (which can be used for explicitly specifying the protocol, host, port, and resource path), or an IP address and port number of the DCNF (which can be used without a DNS dependency, but which can be less flexible if the IP changes).

[0096] The UE 102 stores 648 the UPDC information including the UPDC CM, the DCNF info, the correlation ID, and the AI / ML-enabled feature ID for handling UP data collection with the network. Based on the UPDC CM in UPDC information, the UE 102 may trigger 650 a UE policy evaluation.

[0097] For example, the UE 102 can trigger 650 a UE policy evaluation in response to receiving a request from an upper layer for a PDU session, due for example to a UPDC CM including a UP Data Collection Connection Establishment indication (see Fig. 13B) or UP Data Collection Connection Establishment request (see Fig. 13 A).

[0098] Based on the routing selection descriptor (RSD) in the matched UE policy the UE 102 identified 650, the UE 102 may determine 661 to send, to the SMF 366 via the AMF 364, a PDU Session Establishment / Modification Request message (DNN, S-NSSAI) using aPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01PNsmf_PDUSession_CreateSMContext Request (DNN, S-NSSAI) or a Nsmf PDUSession UpdateSMContext Request (DNN, S-NSSAI). ThePDU Session can be specific for the UP data collection in accordance with at least one of the following pieces of information the UE 102 includes in a PDU Session Establishment / Modification Request message: (i) a new (i.e., dedicated, special-purpose) IE with an AI / ML enabled feature ID, (ii) a new (dedicated, special-purpose) PDU session type set to “UP data collection,” or (iii) a specific pair of a DNN and an S-NSSAI configured by the operator for UP data collection only, indicated as a DC-DNN, a DC-SNSSAI (which cannot be used for other purposes such as accessing the Internet).

[0099] With the information indicated in PDU Session Establishment / Modification Request message, e.g. AI / ML enabled feature ID, the PDU session type set to “UP data collection,” and the {DNN, S-NSSAI} tuple, the SMF 366 obtains 663, from UDM 308 / UDR352 for the UE 102, the DCNF instance information and the correlation ID for the AI / ML-enabled feature. More particularly, the SMF 366 can perform 663 the following: (i) send, to the DCNF 350, a new (dedicated, special-purpose) Ndcnf_UPDC_Policy Association service operation message including a UE-ID, a correlation ID, an AI / ML-enabled feature ID according to the DCNF instance information to request an association of policy for the UP data collection (UPDC) with the AI / ML-enabled feature ID and the correlation ID, (ii) obtain, from the DCNF 350, UPDC policy including the QoS parameters and the UP Data collection configuration information (including DCNF address information); perform QoS flow binding for the UP Data collection, based on UPDC policy; and (iv) configure N4 rules with the UP Data collection configuration information to the PSA UPF 360, so as to route the data traffic of the specific AI / ML enabled feature toward the DCNF 350.

[0100] The SMF 366 sends 664, to the UE 102 via the AMF 364, a PDU Session Establishment Accept or a PDU Session Modification Command message using Nsmf_PDUSession_CreateSMContext Response or Nsmf_PDUSession_UpdateSMContext Response. Alternatively, if the UE 102 did not receive 625 the DCNF address information, or if the DCNF address information requires an update, the AMF 364 sends 665 a PDU Session Establishment Accept or a PDU Session Modification Command message including DCNFPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Paddress information, e.g. in a new IE or in ePCO IE. Collectively, events 661, 662, 663, 664, and 655 define a PDU Session Establishment / Modification procedure. The UE 102 stores or updates 666 the DCNF address information the UE 102 received as a part of the UPDC information.

[0101] Based on the UPDC CM the UE 102 received 625, the UE 102 requests 670 to manage secure user plane connections with the DCNF 350. For example, for UPDC CM including a UP Data Collection Connection Establishment indication (see Fig. 13B) or UP Data Collection Connection Establishment request (see Fig. 13 A), the UE 102 requests 670 to establish a secure UP connection (e.g. TLS connection over TCP or QUIC connection over UDP) with the DCNF 350 via the UPF 370 for UP data collection of AI / ML enabled features.

[0102] As another example, for a UPDC CM including a UP Data Collection Connection Release indication (see Fig. 13B) or UP Data Collection Connection Release request (see Fig.13 A), the UE 102 requests 670 to release a secure UP connection (e.g. TLS connection over TCP or QUIC connection over UDP) with the DCNF 350 via the UPF 370 for UP data collection of AI / ML enabled features.

[0103] The UE 102 sends 626, to the AMF 364, a UL NAS Transport message including a UPDC CM message in the payload container, the AI / ML-enabled feature ID, and the correlation ID. The AMF 364 forwards 627 the UPDC_CM message in the payload container, the AI / ML-enabled feature ID, and the correlation ID to the DCNF 350 using a Namf_CommunicationNlMessageNotify message, if the procedure 670 is successful. The UE 102 sends 626 the UPDC_CM message in response to the UPDC_CM message the UE 102 previously received 625.

[0104] For example, the UE 102 sends 626 a UPDC CM including a UP Data Collection Connection Establishment Complete indication (see Fig. 13B) or UP Data Collection Connection Establishment Complete (see Fig. 13 A) to confirm a request for establishing a secure UP connection (e.g. TLS connection over TCP or QUIC connection over UDP) with the DCNF 350 via the UPF 370 for UP data collection of AI / ML-enabled features.

[0105] As another example, the UE 102 sends 626 a UPDC CM including a UP Data Collection Suspend Confirm indication (see Fig. 13B) or a UP Data Collection Suspend ConfirmPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P(see Fig. 13A) to confirm a request for suspending UP data collection with the DCNF 350 via the UPF 370 for UP data collection of AI / ML-enabled features.

[0106] As another example, the UE 102 sends 626 a UPDC CM including a UP Data Collection Resume Confirm indication (see Fig. 13B) or a UP Data Collection Resume Confirm (see Fig. 13A) to confirm a request to resume UP data collection with the DCNF 350 via the UPF 370 for UP data collection of AI / ML-enabled features.

[0107] As yet another example, the UE 102 sends 626 a UPDC CM including a UP Data Collection Update Confirm indication (see Fig. 13B) or a UP Data Collection Update Confirm (see Fig. 13 A) to confirm a request to update a UP data collection task, e.g. updating validity duration, periodicity, averaging window of the data collection, with the DCNF 350 via the UPF 370 for UP data collection of AI / ML-enabled features.

[0108] As yet another example, the UE 102 sends 626 a UPDC CM including a UP Data Collection Connection Release Confirm indication (see Fig. 13B) or a UP Data Collection Connection Release Confirm (see Fig. 13 A) to confirm a release of the secure UP connection (e.g. TLS connection over TCP or QUIC connection over UDP) with the DCNF 350 via the UPF 370 for UP data collection of AIML enabled features.

[0109] As another example, the UE 102 sends 626 a UPDC CM including a UP Data Collection Status Update response indication (see Fig. 13B) or an UP Data Collection Staus Update response (see Fig. 13A) to update information of UP Data Collection Status.

[0110] The UE 102 starts 680 to enforce the network instruction in the UPDC CM, e.g. send data message to the DCNF 350, over the UP data connection that uses TLS over TCP or QUIC for UP data connection, or resume sending data messages to the DCNF 350 via the UPF 370. At a later time, the UE 102 may suspend 690 data collection, suspend the transfer, or stop data collection.

[0111] Fig. 7 is a messaging diagram of an example scenario 700 in which the UE 102 initiates a UP data collection management procedure. When the UE 102 determines to request, from the network (the DCNF 350 / the PCF 358), UP data collection management, the UE 102 sends a first UPDC-CM in the UL NAS Transport message. The UE handles the PDU session,PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Pthe user plane data connection, and sends, to the network and in response to the first UPDC-CM received in the DL NAS Transport message, a UL NAS Transport message including a second UPDC-CM. Figure 7 illustrates example signaling transport for a UE-initiated UP data collection management procedure using UPDC-CM messages as indicated in approach of Fig.13A or approach of Fig. 13B.

[0112] After performing procedures 702 and 710, which are similar to the procedures 602 and 610, respectively, the UE 102 determines 731 to send a UPDC CM for performing UP data collection management for a specific AI / ML-enabled feature, identified by an AI / ML enabled feature ID. The UE 102 sends 732, to the AMF 364, a UL NAS Transport message including the UPDC_CM message and the corresponding AI / ML- enabled feature ID. T he UPDC CM message can be one of the messages discussed with reference to Figs. 13A and 13B.

[0113] The UE 102 includes the AI / ML-enabled feature ID to associate the collected data with the corresponding AI / ML enabled feature. Generally speaking, an AI / ML enabled feature may require data collected from different entities, such as a RAN node, a UE, an O&M, or a third-party application server.

[0114] The AMF 364 then discovers 740 the DCNF 350, via the NRF 306 for example, based on the AI / ML enabled feature ID. The AMF 364 then sends 733, to the DCNF 350, an Ndcnf_Communication_NlN2MessageTransfer message including the UPDC-CM and AI / ML enabled feature ID.

[0115] The DCNF 350 generates a correlation ID for the AI / ML enabled feature and sends 734, to the UE 102 via the AMF 364, a response message including a second UPDC-CM message, the correlation ID, and the DCNF address information. The second UPDC-CM message is optional if the DCNF 350 determines to indicate another UPDC-CM command to the UE 102 for UPDC management.

[0116] The DCNF 350 includes the correlation ID to enable the correlation between the corresponding data to be collected, transferred to, and handled in the DCNF 350 from one entity, e.g. a UE, a RAN node, or an Application Server. The DCNF information can include a DCNF address as an end point of user plane connection in the network for collecting data of the AI / MLPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Penabled feature. Similar to the example above, the address can be the FQDN address, a URL address, or an IP address and port number of the DCNF.

[0117] The AMF 364 then sends 735, to the UE 102, a DL NAS Transport message including the UPDC CM, the DCNF info, the correlation ID, and the AI / ML-enabled feature ID. The UE 102 stores 741 the UPDC information including the UPDC CM, the DCNF information, the correlation ID, and the AI / ML-enabled feature ID for handling UP data collocation with the network. The procedures or messages 760, 766, 726, 727, 780, and 790 are similar to the procedures or messages 660, 666, 626, 627, 680, and 690, discussed above.

[0118] Scenarios 860A, 860B, 960, and 1060 of Figs. 8A-10 discussed next are similar to the PDU Session Establishment / Modification procedure 560 or 660, with the differences specifically considered below. In some implementations, the techniques of Fig. 8A-10 are implemented as a part of the procedure 560 or 660.

[0119] Fig. 8A illustrates an example scenario 860A in which the UDM 308 uses 864A a service call for data set “Policy Data” and subset “UE context policy control data” with a dedicated information element to indicate UE capability with respect to UP data collection. Here, the UDR 352 provides a policy control subscription profile information at the PDU Session establishment, using a Nudr service for Data Set "Policy Data,” and includes at least one of the following enhancements: for Data Subset " UE context policy control data" (enhanced based, for example, on Table 6.2.1.3-1 in TS 23.503), add a new Indication of UE capability of UP data collection which is used to indicate the UE support for transferring UP data collection to the network. The UDR 308 / the UDR 352 thus implement a subscription retrieval procedure 863A rather than the procedure 663 of Fig. 6.

[0120] Fig. 8B illustrates an example scenario 860B in which the UDM 308 uses 864B a service call for data set “Policy Data” and subset “PDU session policy control data” with a dedicated information element to identify allowed one or more AI / ML-enabled features for UP data collection. Here, the UDR 352 provides policy control subscription profile information at PDU session establishment, using a Nudr service for Data Set "Policy Data", and includes at least one of the following enhancements: for Data Subset "PDU Session policy control data" (enhanced based on Table 6.2.1.3-2 in TS23.503): add a new (dedicated, special-purpose) IE ofPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Pallowed AI / ML enabled features for UP data collection. The UDR 308 / the UDR 352 thus implement a subscription retrieval procedure 863B rather than the procedure 663 of Fig. 6.

[0121] Fig. 9 illustrates an example scenario 900 in which the core network uses Policy and Charging Control (PCC) rules that include UP data collection configuration information with a DCNF address and a port number. The PCC rules information in a CN, such as 5GC or a 6GC, can include a UP Data collection configuration information which indicates a list of the DCNF address and a port number with an associated AI / ML-enabled feature (which can be identified by a feature ID, as discussed above) for UP data collection, and which may also include corresponding Protocol ID of secure transport layer mechanism, e.g. TLS over TCP or QUIC. With this information, the SMF 366 can determine 967 how to handle the PDU Session Establishment / Modification Request procedure from the UE 102, or whether to initiate PDU Session Modification procedure with the RAN 105 and the UPF 370 to invoke / revoke UP data collection for the AI / ML enabled feature.

[0122] Fig. 10 illustrates an example scenario 1060 in which a CN uses N4 rules to enforce a Forwarding Action Rule (FAR) that directs UP data collection traffic. Based on the UP Data collection configuration information in the PCC rules, the SMF 366 configures 1068 N4 rules at the PSA UPF 3709, so that when the traffic is detected based on the PDR, the UPF 370 enforces the FAR accordingly: (i) a PDR includes the DCAF address and port number or FQDN of a specific AI / ML enabled feature, which is used to detect the traffic containing training data to be collected and reported by the UE 102; (ii) and a forwarding action rule (FAR) includes the following information. In one implementation, this information in the FAR enhances the attributes currently included in a FAR with new settings: (i) a destination interface set as at coreside and (ii) forwarding policy that indicates to steer traffic toward the target DCNF address and the port number in the CN or the FQDN address for the corresponding AI / ML enabled feature. In another implementation, the information in the FAR adds a new attribute (relative to the existing attributes in a FAR) that indicates in-network forwarding information for the target DCNF address and the port number in the CN.

[0123] Fig. 11 A is a messaging diagram of an example scenario 1100A in which network endpoints (implemented in a UE, a RAN node, a component of a CN node, a dedicated CN node,PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Por a set of CN nodes) include a UP data collection control message (UPDC-CM) in a payload container of type UE parameters update transparent container. This approach is based on enhancing the UL NAS Transport and DL NAS Transport messages for UP data collection management. The UP Data Collection control message (UPDC-CM) 1121 A, 1122A is encapsulated in a payload container which can be identified by an existing payload container type as a UE parameters update transparent container.

[0124] Fig. 1 IB is a messaging diagram of an example scenario 1100B in which network endpoints include a UPDC-CM in a payload container of a type dedicated to conveying control information for UP data collection. This approach is also based on enhancing the UL NAS Transport and DL NAS Transport messages for UP data collection management. The UP Data Collection control message (UPDC-CM) 1121B, 1122B includes a new payload container type for transferring UPDC-CM for UP Data Collection Management.

[0125] As illustrated next in Fig. 12, network endpoints can transmit and receive 1121, respectively, a UPDC-CM including one or more of the following units of information: a UE ID, (e.g., a SUPI), an AI / ML-enabled feature ID, a correlation ID, a validity time period, a validity area, a periodicity of transferring data for collection, a start time of data collection, a duration of data collection or end time of data collection.

[0126] Fig. 13 A is a messaging diagram 1320A illustrating several example UPDC-CMs which network endpoints can exchange to manage UP data collection. The techniques of Figs.13A and 13B are compatible with one or more of the techniques of Figs. 11 A, 1 IB, and 12 discussed above.

[0127] To implement UP data collection management, the UPDC endpoint 1 can transmit 1328A a UP Data Collection Connection Establishment request, which is used to request for establishment of a new UP connection for UP Data Collection between the UE and the DCNF. The UPDC endpoint 2 can respond 1329A with a UP Data Collection Connection Establishment Complete, which is used to acknowledge for the establishment completion of a new UP connection for UP Data Collection between the UE 102 and the DCNF 350.

[0128] The UPDC endpoint 1 can transmit 1328B a UP Data Collection Suspend request, which is used to request for suspending UP Data Collection for the existing UP connectionPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Pbetween the UE 102 and the DCNF 350. The UPDC endpoint 2 can respond 1329B with a UP Data Collection Suspend Confirm, which is used to acknowledge for the confirmation of suspending the UP connection for UP Data Collection between the UE 102 and the DCNF 350.

[0129] The UPDC endpoint 1 can transmit 1328C a UP Data Collection Resume request, which is used to request for resuming UP Data Collection for the existing UP connection between the UE and the DCNF. The UPDC endpoint 2 can respond 1329C with a UP Data Collection Resume Confirm, which is used to acknowledge for the confirmation of resuming UP Data Collection for the existing UP connection between the UE 102 and the DCNF 350.

[0130] The UPDC endpoint 1 can transmit 1328D a UP Data Collection Connection Update request, which is used to request for updating UP connection for UP Data Collection between the UE and the DCNF. The UPDC endpoint 1 can respond 1329D with a UP Data Collection Connection Update Confirm, which is used to acknowledge for confirming the update of UP connection for UP Data Collection between the UE 102 and the DCNF 350.

[0131] The UPDC endpoint 1 can transmit 1328D a UP Data Collection Connection Release request, which is used to request for releasing UP Data Collection for the existing UP connection between the UE and the DCNF. The UPDC endpoint 2 can respond 1329D a UP Data Collection Connection Release Confirm, which is used to acknowledge for confirming the release of UP Data Collection for the existing UP connection between the UE 102 and the DCNF 350.

[0132] The UPDC endpoint 1 can transmit 1328F a UP Data Collection Status Update indication, which is used to request status update for the UP Data Collection for the existing UP connection between the UE 102 and the DCNF 350. The UPDC CM can additionally indicate what the status information is requested for the update, e.g. data volume of data collected, transferred, time window of the collected data to be transferred, earliest timestamp of the collected data to be transferred, last timestamp of the collected data to be transferred, last timestamp of the data being transferred, etc. for UP data collection of AIML enabled features.

[0133] The UPDC endpoint 2 can respond 1329F with UP Data Collection Status Update response indication, which is used to respond status update for the UP Data Collection for the existing UP connection between the UE and the DCNF. The UPDC_CM can include the status information requested for the update, e.g. data volume of data collected, transferred, timePATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Pwindow of the collected data to be transferred, earliest timestamp of the collected data to be transferred, last timestamp of the collected data to be transferred, last timestamp of the data being transferred, etc. for UP data collection of AIML enabled features.

[0134] Each of the messages of Fig. 13A can include one or more of the fields or IES discussed above with reference to Fig. 12.

[0135] Fig. 13B illustrates a messaging scheme generally similar to that of Fig. 13 A, but with a generic UPDC-CM including an indication of the respective control function. In a scenario 1320B, the UPDC endpoint 1 can transmit 1330A, 1330B, 133OC, 1330D, 1330E, or 1330F a generic UPDC-CM with an indication that identifies a message similar to that in transmission 1328A, 1328B, 1328C, 1328D, 1328E, or 1328F, respectively. The UPDC endpoint 2 can transmit 1331 A, 133 IB, 1331C, 133 ID, 133 IE, or 133 IF a generic UPDC-CM with an indication that identifies a message similar to that in transmission 1329A, 1329B, 1329C, 1329D, 1329E, or 1329F, respectively. Similar to Fig. 13 A, each of the messages in a scenario 1320B of Fig. 13B can include one or more of the fields or IEs discussed above with reference to Fig. 12.

[0136] Fig. 14 is a flow diagram of an example method 1400 in a CN node for negotiating and initiating UP data collection. Generally speaking, UE data collection for an AI / M-enabled feature of NR air interface requires the UE and the network supporting AI / ML enabled feature. The techniques of this disclosure enhance the registration procedure for negotiation of UE capability and network support for AVML enabled feature as follows: (i) the UE informs the AMF about its UE capabilities of AI / ML enabled feature by including an AI / ML enabled feature indication in Registration request message, which allows the network to later trigger and configure policies for UE data transfer of AI / ML enabled features accordingly; (ii) the AMF determines network support of AI / ML enabled feature based on UE supports of AI / ML enabled feature indication and whether AI / ML enabled feature is allowed for the UE from the UDM. If supported, the AMF sends N2 Initial Context Setup Request message including network support indication of AI / ML enabled feature and NAS-PDU with Registration Accept message to the RAN node. The Registration Accept message includes Allowed AI / ML enabled Feature indication, which may be as a part of 5GS Network Feature Support IE, is further delivered from the RAN node to the UE. With the network support indication of AI / ML enabled feature, the RAN node can store thePATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Preceived network support indication of AI / ML enabled feature as RAN UE context. When determining whether to trigger a specific AI / ML enabled feature, the NG-RAN can query UE’s capability via RRC messages for the AI / ML enabled feature using the corresponding AI / ML-enabled feature ID before triggering the AI / ML enabled feature.

[0137] At block 1410, the CN node negotiates, with a UE, support of UP data collection. At block 1423, the CN node determines to send, to the UE, a UPDC-CM message. At block 1424, the CN node transmits, to the UE, a DL NAS Transport message including the UPDC-CM message, a feature ID identifying an AI / ML-enabled featured to which the UP data collection pertains, and optionally a correlation ID. At block 1427, the CN node receives, from the UE, a UL NAS Transport message including a response UPDC-CM message and a feature ID (and, optionally, the correlation ID).

[0138] When an AI / ML enabled feature is enabled, the UE can obtain a customized AI / ML model suitable for the moment and situation from the AS. This requires the UE to perform UP data collection and transfer the training data from the UE to the 5G core and then further to the AS. In this scenario, the enabling of the AI / ML feature serves as the trigger for the UE to perform URSP rules evaluation (e.g., based on a received RRC / MAC layer request). To prepare for transferring UE data traffic, the UE then establishes the PDU Session via an Establishment / Modification request procedure based on the RSD for the matched TD.Accordingly, the UE can differentiate the traffic for UE data transfer from regular traffic in the 5GC for different PDU Sessions.

[0139] To establish a PDU Session for UE data transfer, the UE policy of URSP rules can be enhanced as follows. For the TD enhancement, three options are considered. According to option 1, the TD includes existing IP descriptors indicating destination DCF address, port number, and protocol ID for a secure UP connection. This option assumes that the UE is configured with DCF address and port number, e.g. via DL NAS Transport procedure or received application layer information from the AF / AS supporting UE-side model training.

[0140] According to option 2: the TD includes Connection Capabilities indicating a new standardized component value as UP data collection. When the UE needs to transfer UP dataPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Ptraffic, the UE establishes the PDU Session Establishment / Modification request procedure based on RSD for the matched TD indicating Connection capabilities for UP data collection / transfer.

[0141] According to option 3, the TD includes a new component of AI / ML enabled feature ID(s) for data traffic of UP data collection and transfer. This option assumes that the UE is configured with allowed AI / ML enabled feature ID(s) in Registration Procedure, or DL NAS Transport procedure or received application layer information from the AF / AS supporting UE-side model training. When the UE needs to transfer UP data traffic, the UE establishes the PDU Session Establishment / Modification request procedure based on RSD for the matched TD indicating the AI / ML enabled feature ID.

[0142] For the RSD enhancement, the RSD indicates the following PDU Session related information, e.g. dedicated DNN, dedicated S-NSSAI for UP data collection / transfer.

[0143] Fig. 15 is a flow diagram of an example method 1500 in a UE for negotiating and initiating UP data collection. At block 1510, the UE negotiates, with the network support of UP data collection. At block 1511, the UE determines to send, to thesa network a UPDC-CM message. At block 1532, the UE sends, to the network, a UL NAS Transport message including the UPDC-CM and a feature ID identifying an AI / ML-enabled featured to which the UP data collection pertains. At block 1535, the UE receives, from the network, a response UPDC-CM message and a feature ID (and, optionally, the correlation ID).

[0144] Fig. 16 is a flow diagram of an example method 1600 for facilitating UP data collection in a cellular communication network, which can be implemented in a network endpoint such as a UE or a node in the CN. At block 1610, the network endpoint negotiates, with a second network endpoint, support of user-plane data collection. At block 1624, the network endpoint communicates, with the second network endpoint and using a NAS message, information related to managing the UP data collection.

[0145] Finally, the procedure of an AF request for UE Data Collection, Transfer, and Reporting can be summarized as follows. At step 1, the UE performs a PDU Session Establishment procedure (as defined, for example, in step 1-6 in clause 4.3.2.2.1 of TS 23.502; see also the discussion of procedures 560 and 660 above. At step 2, An AF sends an AF request using Nnef_AFsessionWithQoS or Npcf_PolicyAuthorization_Create service operation for thePATENT APPLICATION Attorney Docket No.: 31730 / 309119-01PUE or a group of the UEs that may provide a least one of the following pieces of information for an AI / ML enabled feature: (i) AI / ML enabled feature ID, (ii) QoS parameters for the UE data transfer, (iii) indication of UE data collection, (iv) UE data report configuration information: AS server address and port number for UE data transfer (e.g. from DCF to the AS), Subscription of UE data report with a new Event ID for the DCF including UP data reporting schedule information, e.g. time, frequency.

[0146] At step 3, the PCF gets information of UE capability of AI / ML enabled feature ID and DCF address information from the UDM / UDR. In addition, the PCF checks user consent of data collection with purpose of AI / ML enabled feature for NR air interface, using data-key: SUPI of the UE, and / or the AI / ML enabled feature ID, and proceed with the following steps if user consent indicates allowed. Otherwise, the PCF replies AF with proper cause indicating lacking of user consent for the AI / ML enabled feature for NR air interface.

[0147] At step 4, the SMF performs SM Policy Association Modification procedure to the PCF for the PDU session. The PCF provides PCC rule to the SMF for the impacted PDU session for the UE (or PCC rules for the impacted PDU sessions for a group of UE). The PCC rule includes information obtained from the AF request message, and DCF address information, in which the DCF address information is retrieved from the UDM / UDR if available.

[0148] At step 5, based on received PCC rules, the SMF determines to enable UE data transfer and manage UP data transfer over user plane for the AI / ML enabled feature, and performs QoS flows binding accordingly.

[0149] At step 6, the SMF configures the RAN node with QoS profile including the QoS parameters, and AI / ML enabled feature ID, in N2 container in the Namf_Communication_NlN2MessageTransfer message to the AMF in Step 6a. Then the RAN node receiving N2 PDU Session Request message, in Step 6b, obtains QoS profile from the AMF. Also, the SMF configures the UE with QoS rule, AI / ML enabled feature ID, and Indication of UP data collection, which are included in NAS message of PDU Session Establishment Accept or Modification Command message, and sends the NAS message in N1 container in the Namf_Communication_NlN2MessageTransfer message to the AMF. Then the RAN node receiving N2 PDU Session Request message, in Step 6b, forwards PDU SessionPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01PEstablishment Accept or PDU Session Modification Command message in Step 6c to the UE. Actions at the RAN node: based on the QoS profile, and AI / ML enabled feature ID, the RAN node may query the UE capability of the AI / ML enabled feature ID, and sends RRCReconfiguration message to the UE if determining to enable the AI / ML enabled feature operation (e.g. the AI / ML enabled feature needs inferencing input from the RAN node). Actions at the UE: based on the QoS rule for UP data collection for the AI / ML enabled feature ID, the UE starts the corresponding AI / ML enabled feature operation and data collection of the standardized data of the corresponding AI / ML enabled feature.

[0150] At step 7, the SMF configures the PSA UPF by sending N4 Session Modification Request message including N4 rules with the following information for the uplink direction: (i) a PDR rule, which is used to detect the data traffic flow with destination address set as DCF address and port number, which is associated with the AI / ML enabled feature ID, (ii) a FAR rule includes the DCF address and port number. When the traffic is detected based on the PDR, the PSA UPF enforces Forwarding Action Rule (FAR) that defines how the detected packet is to be forwarded, (iii) QER rules are used to associate the data traffic flow with proper QoS parameters for UP data transfer of the AI / ML enabled feature.

[0151] In addition, the SMF may configure N4 rules for UE data report to the AS with the following information for the uplink direction: (i) a PDR rule, which is used to detect the data traffic flow with source address set as DCF address and port number for UE data report, which is associated with the AI / ML enabled feature ID, (ii) a FAR rule includes the AS address and port number.

[0152] At step 8, the SMF discovers a suitable DCF if the DCF address information is not available in the PCC rules and configures the DCF using a new Ndcf service operation with the following information: (i) AI / ML feature ID; (ii) UP Connection information for data transfer including AS address information, and PSA UPF information. It is noted that if only the AS address information is provided, the DCF requests for a secure connection with the AS. If both PSA UPF information and AS address information are provided, the DCF requests for a secure connection with the AS via PSA UPF. In this case, the PSA UPF requests for a secure connection with the AS; (iii) UP data reporting schedule information, e.g. time, frequency.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0153] At step 9, the RAN node and the UE starts AI / ML enabled feature identified by AI / ML enabled feature ID if not already starting. At step 10, the UE performs UE data collection of the AI / ML enabled feature. At step 11, the UE sends collect data and transfer the data traffic to DCF via PSA UPF. Based on configured N4 rules, when the PSA UPF detects data traffics from the UE based on PDR, it forwards the data traffics to the configured DCF address indicated in FAR.

[0154] At step 12, when the DCF receives data traffics from the UE via PSA UPF, it may perform the following steps based on information, obtained in Step 8, for configured UP data transfer schedule information, e.g. time, frequency. If only AS address information is configured by the SMF, the DCF requests for a secure connection with the AS directly over an new interface between the DCF and the AS. If both PSA UPF information and AS address information are provided, the DCF requests for a secure connection with the AS via PSA UPF. In this case, the PSA UPF requests for a secure connection with the AS. For data traffic received from the DCF, the PSA UPF forwards it to the AS.

[0155] The following list of examples reflects a variety of the embodiments explicitly contemplated by the present disclosure.

[0156] Example 1. A method for facilitating user-plane (UP) data collection in a cellular communication network, the method implemented in a first network endpoint and comprising: negotiating, with a second network endpoint, support of the UP data collection; and communicating, using a non-access stratum (NAS) transport message, information related to managing the UP data collection.

[0157] Example 2. The method of example 1, wherein: the first network endpoint is a user equipment (UE) in communication, via a radio access network (RAN), with a core network (CN) of the cellular communication network.

[0158] Example 3. The method of example 2, wherein the negotiating includes: transmitting, to the CN, a registration request message including an indication of a capability of the UE with respect to the UP data collection.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0159] Example 4. The method of example 3, wherein: the registration request message includes an indication of a capability of the UE with respect to supporting a UE-initiated or a network-initiated UP data collection procedure.

[0160] Example 5. The method of example 3 or 4, further comprising: receiving, from the CN and in response to the registration request message, a registration accept message including at least one of (i) an indication of a capability of the CN with respect to the UP data collection or (ii) a capability of the CN with respect to supporting a UE-initiated or a network-initiated UP data collection procedure.

[0161] Example 6. The method of any of example 2-5, further comprising: receiving, from the CN, one or more UE policies related to the UP data collection.

[0162] Example 7. The method of example 1, wherein: the first network endpoint is a data collection network function (DCNF) implemented in a core network (CN) of the cellular communication network.

[0163] Example 8. The method of example 7, wherein: the DCNF implements the UP data collection as a component of a network function (NF) configured to implement a respective non-data-coll ection function of the CN.

[0164] Example 9. The method of example 8, wherein: the NF is one of (i) a Location Management Function (LMF), a Policy Control Function (PCF), an Access and Mobility Management Function (AMF), or a Session Management Function (SMF), a User-Plane Function (UPF).

[0165] Example 10. The method of example 7, wherein: the DCNF implements the UP data collection as a network network (NF) dedicated to data collection and configured to provide a data collection service to one or more NFs in the core network.

[0166] Example 11. The method of example 7, wherein: the DCNF implements the UP data collection in a Network Data Analytics Function (NWDAF) and configured to provide a data collection service to one or more NFs in the core network.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0167] Example 12. The method of any of examples 7-11, wherein the negotiating includes: receiving, from a UE, a registration request message including an indication of a capability of the UE with respect to the UP data collection.

[0168] Example 13. The method of example 12, wherein: the registration request message includes an indication of a capability of the UE with respect to supporting a UE-initiated or a network-initiated UP data collection procedure.

[0169] Example 14. The method of example 12 or 13, further comprising: transmitting, to the UE and in response to the registration request message, a registration accept message including at least one of (i) an indication of a capability of the CN with respect to the UP data collection or (ii) a capability of the CN with respect to supporting a UE-initiated or a network-initiated UP data collection procedure.

[0170] Example 15. The method of any of the preceding examples, wherein: the communicating of the information related to managing the UP data collection includes encapsulating a UP data collection control message in a payload container, and including the payload container in the NAS transport message.

[0171] Example 16. The method of example 15, wherein: the payload container is of a type UE parameters update transparent container.

[0172] Example 17. The method of example 15, wherein: the payload container is of a type dedicated specifically to conveying UP data collection control messages.

[0173] Example 18. The method of any of the preceding examples, wherein the communicating of the information related to managing the UP data collection includes: transmitting, from the first network endpoint to the second network endpoint, a control message to request an establishing of a new connection for UP data collection between the first network endpoint and the second network endpoint.

[0174] Example 19. The method of example 18, wherein the communicating of the information related to managing the UP data collection further includes: receiving, by the first network endpoint from the second network endpoint, a control message to indicate a completionPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Pof the establishing of the new connection between the first network endpoint and the second network endpoint.

[0175] Example 20. The method of any of examples 1-17, wherein the communicating of the information related to managing the UP data collection includes: transmitting, from the first network endpoint to the second network endpoint, a control message to request suspending an existing connection for UP data collection between the first network endpoint and the second network endpoint.

[0176] Example 21. The method of example 20, wherein the communicating of the information related to managing the UP data collection further includes: receiving, by the first network endpoint from the second network endpoint, a control message to confirm the suspending of the existing connection between the first network endpoint and the second network endpoint.

[0177] Example 22. The method of any of examples 1-17, wherein the communicating of the information related to managing the UP data collection includes: transmitting, from the first network endpoint to the second network endpoint, a control message to request resuming an existing suspended connection for UP data collection between the first network endpoint and the second network endpoint.

[0178] Example 23. The method of example 22, wherein the communicating of the information related to managing the UP data collection further includes: receiving, by the first network endpoint from the second network endpoint, a control message to confirm the resuming of the existing suspended connection between the first network endpoint and the second network endpoint.

[0179] Example 24. The method of any of examples 1-17, wherein the communicating of the information related to managing the UP data collection includes: transmitting, from the first network endpoint to the second network endpoint, a control message to request updating an existing connection for UP data collection between the first network endpoint and the second network endpoint.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0180] Example 25. The method of example 24, wherein the communicating of the information related to managing the UP data collection further includes: receiving, by the first network endpoint from the second network endpoint, a control message to confirm the updating of the existing connection between the first network endpoint and the second network endpoint.

[0181] Example 26. The method of any of examples 1-17, wherein the communicating of the information related to managing the UP data collection includes: transmitting, from the first network endpoint to the second network endpoint, a control message to request releasing an existing connection for UP data collection between the first network endpoint and the second network endpoint.

[0182] Example 27. The method of example 26, wherein the communicating of the information related to managing the UP data collection further includes: receiving, by the first network endpoint from the second network endpoint, a control message to confirm the releasing of the existing connection between the first network endpoint and the second network endpoint.

[0183] Example 28. The method of any of examples 1-17, wherein the communicating of the information related to managing the UP data collection includes: transmitting, from the first network endpoint to the second network endpoint, a control message to request an update of a status of a connection for UP data collection between the first network endpoint and the second network endpoint.

[0184] Example 29. The method of example 28, wherein the communicating of the information related to managing the UP data collection further includes: receiving, by the first network endpoint from the second network endpoint, a control message to provide the update of the status of the connection between the first network endpoint and the second network endpoint.

[0185] Example 30. The method of example 28 or 29, wherein the update of the status includes one or more of: (i) an amount of collected data, (ii) an amount of transferred data, (iii) a time window of the collected data to be transferred, (iv) an earliest timestamp of the collected data to be transferred, or (v) a last timestamp of the collected data to be transferred.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0186] Example 31. The method of any of examples 18-29, wherein the control message is a dedicated control message defined for a corresponding control UP data collection management function.

[0187] Example 32. The method of any of examples 18-29, wherein the control message is a common control message including an indicator to identify a corresponding control UP data collection management function.

[0188] Example 33. The method of any of examples 18-32, wherein the control message includes one or more of: (i) a UE identifier (ID); (ii) an identifier of a feature to which the UP data collection pertains; (iii) a correlation ID, (iv) a validity time period, (v) a validity area; (vl) a periodicity of transferring data for the UP data collection (vii) a start time of the UP data collection; (viii) a duration of the UP data collection; or (ix) an end time of the UP data collection.

[0189] Example 34. The method of any of examples 7-14, further comprising: transmitting, to the second network endpoint, a request to initiate the UP data collection.

[0190] Example 35. The method of example 34, further comprising: generating a correlation identifier (ID) to establish a correlation between a control message and UP data to be collected from the second network endpoint; providing the correlation ID to the second network endpoint.

[0191] Example 36. The method of example 34, further comprising: identifying an AI / ML-enabled feature for the UP data collection; providing a feature ID, corresponding to the identified AI / ME-enabled feature, to the second network endpoint.

[0192] Example 37. The method of example 34, further comprising: generating a DCNF information including an address associated with the UP data collection; providing the DCNF information to the second network endpoint.

[0193] Example 38. The method of any of examples 34-37, further comprising: receiving, from the second network endpoint, a message responsive to the request to initiate the UP data collection.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0194] Example 39. The method of any of examples 2-6, further comprising: transmitting, to the second network endpoint, a request to initiate the UP data collection.

[0195] Example 40. The method of example 39, further comprising: including, in the request to initiate the UP data collection, a feature ID corresponding to AI / ML-enabled feature for the UP data collection.

[0196] Example 41. The method of example 39 or 40, further comprising: receiving, from the second network endpoint and in response to the request to initiate the UP data collection, a message including a correlation ID that establishes a correlation between a control message and UP data to be collected.

[0197] Example 42. The method of any of the preceding examples, further comprising: establishing, with the second network endpoint, a packet data unit (PDU) session for the UP data collection, including communicating a PDU Session Establishment or Modification Request message.

[0198] Example 43. The method of example 42, wherein the PDU Session Establishment or Modification Request message includes an indication that a type of the PDU session is UP data collection.

[0199] Example 44. The method of example 42, wherein the PDU Session Establishment or Modification Request message includes an IE dedicated to conveying a feature ID.

[0200] Example 45. The method of example 42, wherein the PDU Session Establishment or Modification Request message indicates a data network name (DNN) and Single Network Slice Selection Assistance Information (S-NSSAI) allocated specifically for the UP data collection.

[0201] Example 46. The method of any of the preceding examples, wherein: the UP data collections pertain to AI / ML training.

[0202] Example 47. The method of any of examples 1-8C, wherein: the UP data collections pertain to interference information.

[0203] Example 48. The method of any of examples 1-45, wherein: the UP data collections pertain to sensor data collection.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0204] Example 49. A network device comprising processing hardware, the network device configured to implement a method of any of the preceding examples.

[0205] Example 50. A method for facilitating user-plane (UP) data collection from a downstream network device, the method implemented in a core network and comprising: negotiating, with the downstream network device, support of the UP data collection; and communicating, using a non-access stratum (NAS) transport message, information related to managing the UP data collection.

[0206] Example 51. The method of example 50, further comprising: using, at a Unified Data Management function (UDM), a service for a data set “Policy Data” and a data subset “UE Context policy control data,” the data subset including an indication of UE capability with respect to the UP data collection.

[0207] Example 52. The method of example 50, further comprising: using, at a UDM, a service for a data set “Policy Data” and a data subset “PDU Session policy control data,” the data subset including an indication of ALML-enabled features for which the UP data collection is allowed.

[0208] Example 53. The method of any of examples 50-52, further comprising: using, at a Session Management Function (SMF), Policy and Charging Control (PCC) rules that indicate a DCNF address and a port number, for the UP data collection.

[0209] Example 54. The method of any of examples 50-53, further comprising: using, at a User-Plane Function (UPF), a forwarding action rule (FAR) that includes a forwarding policy to steer traffic to a DCNF address and a port number of an AI / ML feature.

[0210] Example 55. The method of any of examples 50-54, further comprising: using, at UPF), an FAR that includes an attribute dedicated to indicating in-network forwarding information for a target DCNF and a port number.

[0211] Example 56. A core network of a cellular communication network, the core network comprising processing hardware and configured to implement a method of any of examples 50-55.Additional considerationsPATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0212] A user device in which the techniques of this disclosure can be implemented (e.g., the UE 102) can be any suitable device capable of wireless communications such as a smartphone, a tablet computer, a laptop computer, a mobile gaming console, a point-of-sale (POS) terminal, a health monitoring device, a drone, a camera, a media-streaming dongle or another personal media device, a wearable device such as a smartwatch, a wireless hotspot, a femtocell, or a broadband router. Further, the user device in some cases may be embedded in an electronic system such as the head unit of a vehicle or an advanced driver assistance system (ADAS). Still further, the user device can operate as an intemet-of-things (loT) device or a mobile-internet device (MID). Depending on the type, the user device can include one or more general-purpose processors, a computer-readable memory, a user interface, one or more network interfaces, one or more sensors, etc.

[0213] Certain embodiments are described in this disclosure as including logic or a number of components or modules. Modules may can be software modules (e.g., code, or machine-readable instructions stored on non-transitory machine-readable medium) or hardware modules. A hardware module is a tangible unit capable of performing certain operations and may be configured or arranged in a certain manner. A hardware module can comprise dedicated circuitry or logic that is permanently configured (e.g., as a special-purpose processor, such as a field programmable gate array (FPGA) or an application-specific integrated circuit (ASIC), a digital signal processor (DSP), etc.) to perform certain operations. A hardware module may also comprise programmable logic or circuitry (e.g., as encompassed within a general -purpose processor or other programmable processor) that is temporarily configured by software to perform certain operations. The decision to implement a hardware module in dedicated and permanently configured circuitry, or in temporarily configured circuitry (e.g., configured by software) may be driven by cost and time considerations.

[0214] When implemented in software, the techniques can be provided as part of the operating system, a library used by multiple applications, a particular software application, etc. The software can be executed by one or more general-purpose processors or one or more specialpurpose processors.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P

[0215] As used herein, the terms “comprises,” “comprising,” “includes,” “including,” “has,” “having” or any other variation thereof, are intended to cover a non-ex elusive inclusion. For example, a process, method, article, or apparatus that comprises a list of elements is not necessarily limited to only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Further, unless expressly stated to the contrary, “or” refers to an inclusive or and not to an exclusive or. For example, a condition A or B is satisfied by any one of the following: A is true (or present) and B is false (or not present), A is false (or not present) and B is true (or present), and both A and B are true (or present).

Claims

PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01PWhat is claimed is:

1. A method for facilitating user-plane (UP) data collection in a cellular communication network, the method implemented in a user equipment (UE) and comprising: transmitting, to a core network (CN), a registration request message including an indication of a capability of the UE with respect to the UP data collection;encapsulating, in a payload container, a UP data collection control message that includes a feature identifier (ID) identifying an artificial intelligence / machine learning (AIZML)-enabled feature to which the UP data collection pertains;including the payload container in a non-access stratum (NAS) transport message; and transmitting, to the CN, the NAS transport message.

2. The method of claim 1, wherein the payload container is one of:(i) a UE parameters update transparent container, or(ii) of a type dedicated specifically to conveying UP data collection control messages.

3. The method of any of the preceding claims, wherein the control message operates to request one of:(i) establishing of a new connection for UP data collection between the UE and the CN, (ii) suspending an existing connection for UP data collection between the UE and the CN, (iii) resuming an existing suspended connection for UP data collection between the UE and the CN,(iv) updating an existing connection for UP data collection between the UE and the CN, (v) releasing an existing connection for UP data collection between the UE and the CN; or(vi) updating a status of a connection for UP data collection between the UE and the CN.

4. The method of claim 3, wherein the updating of the status includes one or more of:(i) an amount of collected data,PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P(ii) an amount of transferred data,(iii) a time window of the collected data to be transferred,(iv) an earliest timestamp of the collected data to be transferred, or(v) a last timestamp of the collected data to be transferred.

5. The method of any of the preceding claims, further comprising:receiving, from the CN, (i) a correlation identifier (ID) to establish a correlation between a control message and UP data to be collected, and (ii) a data collection network function (DCNF) information including an address associated with the UP data collection.

6. The method of any of the preceding claims, further comprising: establishing, with the CN, a packet data unit (PDU) session for the UP data collection, including transmitting a PDU Session Establishment message or a Modification Request message indicating a data network name (DNN) and Single Network Slice Selection Assistance Information (S-NSSAI) allocated specifically for the UP data collection.

7. The method of any of the preceding claims, wherein the registration request message includes an indication of a capability of the UE with respect to supporting a UE-initiated or a network-initiated UP data collection procedure.

8. A method for facilitating user-plane (UP) data collection in a cellular communication network, the method implemented in a core network (CN) and comprising: receiving, from a user equipment (UE), a registration request message including an indication of a capability of the UE with respect to the UP data collection; and encapsulating, in a payload container, a UP data collection control message that includes a feature identifier (ID) identifying an artificial intelligence / machine learning (AI / ML)-enabled feature to which the UP data collection pertains;including the payload container in a non-access stratum (NAS) transport message; and transmitting, to the UE, the NAS transport message.PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01P9. The method of claim 8 implemented in a data collection network function (DCNF), wherein:the DCNF implements the UP data collection as a component of a network function (NF) configured to implement a respective non-data-col lection function of the CN, andthe NF is one of (i) a Location Management Function (LMF), (ii) a Policy Control Function (PCF), (iii) an Access and Mobility Management Function (AMF), a (iv) a Session Management Function (SMF), or a (v) User-Plane Function (UPF).

10. The method of claim 8 or 9, wherein the payload container is one of:(i) a UE parameters update transparent container, or(ii) of a type dedicated specifically to conveying UP data collection control messages.

11. The method of any of claims 8-10, wherein the control message operates to request one of:(i) establishing of a new connection for UP data collection between the UE and the CN, (ii) suspending an existing connection for UP data collection between the UE and the CN, (iii) resuming an existing suspended connection for UP data collection between the UE and the CN,(iv) updating an existing connection for UP data collection between the UE and the CN, (v) releasing an existing connection for UP data collection between the UE and the CN; or(vi) updating a status of a connection for UP data collection between the UE and the CN.

12. The method of any of claims 8-11, further comprising:generating a correlation identifier (ID) to establish a correlation between a control message and UP data to be collected; andproviding the correlation ID to UE.

13. The method of any of claims 8-12, further comprising:PATENT APPLICATION Attorney Docket No.: 31730 / 309119-01Pusing, at a UPF, a forwarding action rule (FAR) that includes a forwarding policy to steer traffic to a DCNF address and a port number of an AI / ML feature.

14. The method of any of claims 8-13, further comprising:using, at a UPF, a FAR that includes an attribute dedicated to indicating in-network forwarding information for a target DCNF and a port number.

15. An apparatus comprising processing hardware and configured to implement a method of any of the preceding claims.

Images