Machine-learning-based acquisition of a radio cell timing advance

Abstract

A method performed by a user equipment (UE) served by a source cell is provided. The method includes determining an estimated timing advance (TA) value for a candidate cell, and a confidence level for the estimated TA value, according to a machine learning (ML)-based TA acquisition scheme. The method includes making a determination that a handover condition is fulfilled for the candidate cell, and the confidence level for the estimated TA value is at or above a confidence level threshold, which may be network calibrated. Based on the determination, the method includes applying a configuration of the candidate cell as a target cell for a handover. And the method includes carrying out the handover as a random access channel (RACH)-less handover in which a RACH preamble is skipped, and the estimated TA value is used as a TA value for the target cell to access the target cell.

Description

MACHINE-LEARNING-BASED ACQUISITION OF A RADIO CELL TIMING ADVANCE TECHNOLOGICAL FIELD

[0001] The present disclosure relates generally to telecommunications and, inparticular, to acquisition of a timing advance value for a cell of a radio access network.BACKGROUND

[0002] A telecommunications system can be seen as a facility that enablescommunication sessions between two or more entities such as user terminals, base stations and / or other nodes by providing carriers between the various entities involved in the communications path. A telecommunications system can be provided for example by means of a communication network and one or more compatible communication devices. The communication sessions may comprise, for example, communication of data for carrying communications such as voice, video, electronic mail (email), text message, multimedia and / or content data and so on. Non-limiting examples of services provided comprise two-way or multi-way calls, data communication or multimedia services and access to a data network system, such as the Internet.

[0003] In a wireless telecommunications system, at least a part of a communicationsession between at least two stations occurs over a wireless link. Examples of wireless telecommunications systems comprise public land mobile networks (PLMN), satellite based communication systems and different wireless local networks, for example wireless local area networks (WLAN). Some wireless systems can be divided into cells, and are therefore often referred to as cellular systems.

[0004] A user can access the telecommunications system by means of an appropriatecommunication device or terminal. A communication device of a user may be referred to as user equipment (UE) or user device. A communication device is provided with an appropriate signal receiving and transmitting apparatus for enabling communications, for example enabling access to a communication network or communications directly withother users. The communication device may access a carrier provided by a station, for example a base station of a cell, and transmit and / or receive communications on the carrier.

[0005] The telecommunications system and associated devices typically operate inaccordance with a given standard or specification which sets out what the various entities associated with the communication system are permitted to do and how operations should be achieved. Communication protocols and / or parameters which shall be used for connection of the various entities are also typically defined. One example of atelecommunications system is the Universal Mobile Telecommunications System(UMTS). Other examples of telecommunications systems are Long-Term Evolution (LTE), LTE Advanced and the so-called 5G or New Radio (NR) networks. NR is being standardized by the 3rd Generation Partnership Project (3GPP). BRIEF SUMMARY

[0006] Example implementations of the present disclosure are directed totelecommunications and, in particular, to acquisition of a timing advance value for a cell of a radio access network. The present disclosure includes, without limitation, the following example implementations.

[0007] Some example implementations provide an apparatus implemented by a userequipment (UE) served by a source cell, the apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: determine an estimated timing advance (TA) value for a candidate cell, and a confidence level for the estimated TA value, according to a machine learning (ML)-based TA acquisition scheme; make a determination that a handover condition is fulfilled for the candidate cell, and the confidence level for the estimated TA value is at or above a confidence level threshold; and based on the determination, apply a configuration of the candidate cell as a target cell for a handover; and carry out the handover as a random access channel (RACH)-less handover in which a RACH preamble is skipped, and the estimated TA value is used as a TA value for the target cell to access the target cell.

[0008] Some example implementations provide an apparatus implemented by an userequipment (UE) served by a source cell, the apparatus comprising: means for determining an estimated timing advance (TA) value for a candidate cell, and a confidence level for the estimated TA value, according to a machine learning (ML)-based TA acquisition scheme; means for making a determination that a handover condition is fulfilled for the candidate cell, and the confidence level for the estimated TA value is at or above a confidence level threshold; and based on the determination, means for applying a configuration of the candidate cell as a target cell for a handover; and means for carrying out the handover as a random access channel (RACH)-less handover in which a RACH preamble is skipped, and the estimated TA value is used as a TA value for the target cell to access the target cell.

[0009] Some example implementations provide a method performed by a userequipment (UE) served by a source cell, the method comprising: determining an estimated timing advance (TA) value for a candidate cell, and a confidence level for the estimated TA value, according to a machine learning (ML)-based TA acquisition scheme; making a determination that a handover condition is fulfilled for the candidate cell, and the confidence level for the estimated TA value is at or above a confidence levelthreshold; and based on the determination, applying a configuration of the candidate cellas a target cell for a handover; and carrying out the handover as a random access channel (RACH)-less handover in which a RACH preamble is skipped, and the estimated TA value is used as a TA value for the target cell to access the target cell.

[0010] Some example implementations provide a computer-readable storage mediumthat is non-transitory and has instructions stored therein that, in response to execution by at least one processing circuitry, causes a user equipment (UE) served by a source cell to at least: determine an estimated timing advance (TA) value for a candidate cell, and a confidence level for the estimated TA value, according to a machine learning (ML)-based TA acquisition scheme; make a determination that a handover condition is fulfilled for the candidate cell, and the confidence level for the estimated TA value is at or above a confidence level threshold; and based on the determination, apply a configuration of the candidate cell as a target cell for a handover; and carry out the handover as a randomaccess channel (RACH)-less handover in which a RACH preamble is skipped, and the estimated TA value is used as a TA value for the target cell to access the target cell.

[0011] Some example implementations provide an apparatus implemented by a radioaccess node providing a source cell, the apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: receive an indication from a user equipment (UE) served by the source cell that the UE supports a machine learning (ML)-based timing advance (TA) acquisition scheme; initiate a handover preparation for a candidate cell during which an indication that the handover preparation is for a random access channel (RACH)-less handover is sent to the candidate cell; receive, from the candidate cell, an indication of a confidence level threshold associated with the RACH-less handover; and send a handover command message and an indication of the confidence level threshold to the UE that cause the UE to carry out the RACH-less handover to the candidate cell when a handover condition is fulfilled and a confidence level for an estimated TA value determined by the UE according to the ML-based TA acquisition scheme is at or above the confidence level threshold.

[0012] Some example implementations provide an apparatus implemented by a radioaccess node providing a source cell, the apparatus comprising: means for receiving an indication from an user equipment (UE) served by the source cell that the UE supports a machine learning (ML)-based timing advance (TA) acquisition scheme; means for initiating a handover preparation for a candidate cell during which an indication that the handover preparation is for a random access channel (RACH)-less handover is sent to the candidate cell; means for receiving, from the candidate cell, an indication of a confidence level threshold associated with the RACH-less handover; and means for sending a handover command message and an indication of the confidence level threshold to the UE that cause the UE to carry out the RACH-less handover to the candidate cell when a handover condition is fulfilled and a confidence level for an estimated TA value determined by the UE according to the ML-based TA acquisition scheme is at or above the confidence level threshold.

[0013] Some example implementations provide a method performed by a radioaccess node providing a source cell, the method comprising: receiving an indication from a user equipment (UE) served by the source cell that the UE supports a machine learning (ML)-based timing advance (TA) acquisition scheme; initiating a handover preparation for a candidate cell during which an indication that the handover preparation is for a random access channel (RACH)-less handover is sent to the candidate cell; receiving, from the candidate cell, an indication of a confidence level threshold associated with the RACH-less handover; and sending a handover command message and an indication of the confidence level threshold to the UE that cause the UE to carry out the RACH-less handover to the candidate cell when a handover condition is fulfilled and a confidence level for an estimated TA value determined by the UE according to the ML-based TA acquisition scheme is at or above the confidence level threshold.

[0014] Some example implementations provide a computer-readable storage mediumthat is non-transitory and has instructions stored therein that, in response to execution by at least one processing circuitry, causes a radio access node providing a source cell to at least: receive an indication from a user equipment (UE) served by the source cell that the UE supports a machine learning (ML)-based timing advance (TA) acquisition scheme; initiate a handover preparation for a candidate cell during which an indication that the handover preparation is for a random access channel (RACH)-less handover is sent to the candidate cell; receive, from the candidate cell, an indication of a confidence level threshold associated with the RACH-less handover; and send a handover command message and an indication of the confidence level threshold to the UE that cause the UE to carry out the RACH-less handover to the candidate cell when a handover condition is fulfilled and a confidence level for an estimated TA value determined by the UE according to the ML-based TA acquisition scheme is at or above the confidence level threshold.

[0015] These and other features, aspects, and advantages of the present disclosurewill be apparent from a reading of the following detailed description together with the accompanying figures, which are briefly described below. The present disclosure includesany combination of two, three, four or more features or elements set forth in thisdisclosure, regardless of whether such features or elements are expressly combined orotherwise recited in a specific example implementation described herein. The present disclosure is intended to be read holistically such that any separable features or elementsof the disclosure, in any of its aspects and example implementations, should be viewed ascombinable unless the context of the disclosure clearly dictates otherwise.

[0016] It will therefore be appreciated that this Brief Summary is provided merely forpurposes of summarizing some example implementations so as to provide a basic understanding of some aspects of the disclosure. Accordingly, it will be appreciated that the above described example implementations are merely examples and should not be construed to narrow the scope or spirit of the disclosure in any way. Other example implementations, aspects and advantages will become apparent from the following detailed description taken in conjunction with the accompanying figures which illustrate, by way of example, the principles of some described example implementations. BRIEF DESCRIPTION OF THE FIGURE(S)

[0017] Having thus described example implementations of the disclosure in generalterms, reference will now be made to the accompanying figures, which are not necessarily drawn to scale, and wherein:

[0018] FIG. 1 illustrates a telecommunications system that includes one or morepublic land mobile networks (PLMNs) coupled to one or more external data networks, according to some example implementations of the present disclosure;

[0019] FIG. 2 illustrates a deployment of a PLMN, according to some exampleimplementations;

[0020] FIGS. 3A, 3B and 3C illustrate a signaling chart of a radio access channel(RACH)-less conditional handover (CHO) procedure for a user equipment (UE);

[0021] FIG. 4 illustrates a UE-based timing advance (TA) acquisition scheme forlower-layer triggered mobility (LTM) which is based on downlink (DL) measurement;

[0022] FIGS. 5 and 6 illustrate respective machine learning (ML) model trainingscenarios, according to various example implementations;

[0023] FIG. 7 illustrates a deployment scenario for a ML model to determine anestimated TA value, and a confidence level for the estimated TA value, according to some example implementations;

[0024] FIGS. 8A and 8B illustrate respective training and deployment scenarios for aregression ML model, according to some example implementations;

[0025] FIGS. 9A, 9B and 9C illustrate a signaling chart of a RACH-less CHOprocedure for a UE using a ML-based TA acquisition scheme, according to some example implementations;

[0026] FIGS. 10A, 10B, 10C, 10D and 10E are flowcharts illustrating various steps ina method performed by a UE served by a source cell, according to various example implementations;

[0027] FIGS. 11A, 11B, 11C and 11D are flowcharts illustrating various steps in amethod performed by a radio access node providing a source cell, according to various example implementations; and

[0028] FIG. 12 illustrates an apparatus according to some example implementations.DETAILED DESCRIPTION

[0029] Some implementations of the present disclosure will now be described morefully hereinafter with reference to the accompanying figures, in which some, but not all implementations of the disclosure are shown. Indeed, various implementations of the disclosure may be embodied in many different forms and should not be construed as limited to the implementations set forth herein; rather, these example implementations are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art. Like reference numerals refer to like elements throughout.

[0030] Unless specified otherwise or clear from context, references to first, second orthe like should not be construed to imply a particular order. A feature described as beingabove another feature (unless specified otherwise or clear from context) may instead bebelow, and vice versa; and similarly, features described as being to the left of another feature else may instead be to the right, and vice versa. Also, while reference may be made herein to quantitative measures, values, geometric relationships or the like, unless otherwise stated, any one or more if not all of these may be absolute or approximate to account for acceptable variations that may occur, such as those due to engineering tolerances or the like.

[0031] As used herein, unless specified otherwise or clear from context, the “or” of aset of operands is the “inclusive or” and thereby true if and only if one or more of the operands is true, as opposed to the “exclusive or” which is false when all of the operands are true. Thus, for example, “[A] or [B]” is true if [A] is true, or if [B] is true, or if both [A] and [B] are true. Further, the articles “a” and “an” mean “one or more,” unless specified otherwise or clear from context to be directed to a singular form. Furthermore, it should be understood that unless otherwise specified, the terms “data,” “content,” “digital content,” “information,” and similar terms may be at times used interchangeably. The term “network” may refer to a group of interconnected computers including clients and servers; and within a network, these computers may be interconnected directly or indirectly by various means including via one or more switches, routers, gateways, access points or the like.

[0032] The present disclosure discusses systems and architectures that, while specificterms may be used, are broadly applicable across various technologies. For instance, while the present disclosure may reference technologies from 3GPP such as Global System for Mobile Communications (GSM), UMTS, LTE, LTE Advanced, 5G NR, 5G Advanced, and 6G, the present disclosure is equally relevant to non-3GPP technologies like IEEE 802, Bluetooth, and Bluetooth Low Energy. Example implementations of the present disclosure described herein also mention public land mobile networks (PLMNs) and mobile network operators (MNOs), but example implementations are similarly applicable to standalone non-public networks (SNPNs) and the private entities operating these networks. Furthermore, although some examples and figures focus on radio access networks (RANs) and 3GPP access, example implementations are applicable to any type of network access. This includes not only 5G or 6G 3GPP access but also non-3GPP access, such as wireline access, untrusted non-3GPP access, and trusted non-3GPP access using wireless access gateway function (W-AGF), non-3GPP interworking function (N3IWF), or trusted non-3GPP gateway function (TNGF) to connect to a 5G or 6G core network.

[0033] Further, as used in this application, the term “circuitry” may refer to one ormore or all of the following: (a) hardware-only circuit implementations (such as implementations in only analog and / or digital circuitry); (b) combinations of hardwarecircuits and software, such as (as applicable): (i) a combination of analog and / or digital hardware circuit(s) with software / firmware and (ii) any portions of hardware processor(s) with software (including digital signal processor(s)), software, and memory(ies) that work together to cause an apparatus, such as a mobile phone or server, to perform various functions); or (c) hardware circuit(s) and / or processor(s), such as a microprocessor(s) or a portion of a microprocessor(s), that requires software (e.g., firmware) for operation, but the software may not be present when it is not needed for operation.

[0034] The above definition of circuitry applies to all uses of this term in thisapplication, including in any claims. As a further example, as used in this application, the term circuitry also covers an implementation of merely a hardware circuit or processor (or multiple processors) or portion of a hardware circuit or processor and its (or their) accompanying software and / or firmware. The term circuitry also covers, for example andif applicable to the particular claim element, a baseband integrated circuit or processorintegrated circuit for a mobile device or a similar integrated circuit in server, a cellular network device, or other computing or network device.

[0035] FIG. 1 illustrates a telecommunications system 100 according to variousexample implementations of the present disclosure. The telecommunications system generally includes one or more telecommunications networks. As shown, for example,the system includes one or more PLMNs 102 coupled to one or more other external datanetworks 104 – notably including a wide area network (WAN) such as the Internet. Eachof the PLMNs includes a core network (CN) 106 backbone such as the Evolved PacketCore (EPC) of LTE, the 5G core network (5GC) or the like; and each of the corenetworks and the Internet are coupled to one or more RANs 108, air interfaces or the likethat implement one or more radio access technologies (RATs). As used herein, a “network device” refers to any suitable device at a network side of a telecommunications network. Examples of suitable network devices are described in greater detail below.

[0036] In addition, the system includes one or more radio units that may be varyinglyknown as user equipment (UE) 110, terminal device, terminal equipment, mobile stationor the like. The UE is generally a device configured to communicate with a network device or a further UE in a telecommunications network. The UE may be a portable computer (e.g., laptop, notebook, tablet computer), mobile phone (e.g., cell phone,smartphone), wearable computer (e.g., smartwatch), or the like. In other examples, theUE may be an Internet of things (IoT) device, an industrial IoT (IIoT device), a vehicleequipped with a vehicle-to-everything (V2X) communication technology, or the like. Insome examples, as referenced by 3GPP, the UE may be a narrowband IoT (NB-IoT) device, an enhanced machine-type communication (eMTC) device, a reduced capability (RedCap) device, an ambient IoT device, or the like.

[0037] In operation, these UEs 110 may be configured to connect to one or more ofthe RANs 108 according to their particular radio access technologies to thereby access aparticular CN 106 of a PLMN 102, or to access one or more of the external datanetworks 104 (e.g., the Internet). The external data network may be configured to provideInternet access, operator services, 3rd party services, etc. For example, the InternationalTelecommunication Union (ITU) has classified 5G mobile network services into threecategories: enhanced mobile broadband (eMBB), ultra-reliable and low-latency communications (URLLC), and massive machine type communications (mMTC) or massive internet of things (MIoT).

[0038] Examples of radio access technologies include 3GPP radio accesstechnologies such as GSM, UMTS, LTE, LTE Advanced, 5G NR, 5G Advanced, and 6G. Other examples of radio access technologies include IEEE 802 technologies such as IEEE 802.11 (Wi-Fi), IEEE 802.15 (including 802.15.1 (WPAN / Bluetooth), 802.15.4 (Zigbee) and 802.15.6 (WBAN)), Bluetooth, Bluetooth Low Energy (BLE), ultra wideband (UWB), and the like. Generally, a radio access technology may refer to any 2G, 3G, 4G, 5G, 6G or higher generation mobile communication technology and their different versions, as well as to any other wireless radio access technology that may be arranged to interwork with such a mobile communication technology to provide access tothe CN 106 of a mobile network operator (MNO).

[0039] In various examples, a RAN 108 may be configured as one or moremacrocells, microcells, picocells, femtocells or the like. The RAN may generally includeone or more radio access nodes that are configured to interact with UEs 110. In variousexamples, a radio access node may be referred to as a base station (BS), access point (AP), base transceiver station (BTS), Node B (NB), evolved NB (eNB), macro BS, NB (MNB) or eNB (MeNB), home BS, NB (HNB) or eNB (HeNB), next generation NB(gNB), enhanced gNB (en-gNB), next generation eNB (ng-eNB), or the like. The RAN may include some type of network controlling / governing entity responsible for control of the radio access nodes. The network controlling / governing entity and radio access node may be separate or integrated into a single apparatus. The network controlling / governing entity may include processing circuity configured to carry out various management functions, etc. The processing circuity may be associated with a memory, computer- readable storage medium or database for maintaining information required in the management functions.

[0040] A RAN 108 may be centralized or distributed. In various examples,components of a RAN may be interconnected by Ethernet, Gigabit Ethernet, Asynchronous Transfer Mode (ATM), optical fiber, dark fiber, passive wavelength division multiplexing (WDM), WDM passive optical network (WDM-PON), optical transport network (OTN), time sensitive networking (TSN) and / or any other data link layer network, possibly including radio links. The RAN may be connected to a CN 106 through one or more gateways, network functions or the like.

[0041] As will be appreciated, a PLMN 102 may be deployed in a number ofdifferent manners. FIG. 2 illustrates a deployment 200 of a PLMN, such as a 4G LTE, 5Gor 6G deployment, according to some example implementations. As shown, thedeployment includes a CN 106, and RAN 108 with one or more radio access nodes 202configured to interact with UEs 110. In a 4G LTE deployment, the EPC is the CN, andthe evolved UMTS terrestrial radio access network (E-UTRAN) is the RAN; and the E- UTRAN includes one or more eNBs (radio access nodes) configured to connect UEs to the E-UTRAN to thereby access the EPC. Similarly, in a 5G deployment, the 5GC is the CN 106, and the next generation (NG) radio access network (NG-RAN) is the RAN 108; and the NG-RAN includes one or more gNBs (radio access nodes) configured to connectUEs 110 to the NG-RAN to thereby access the 5GC (at times referred to as the NGC).The term ‘gNB’ in 5G may correspond to the eNB in 4G LTE.

[0042] Some deployments of 4G LTE and 5G in particular are considered standalone(SA) deployments. Other deployments combine 4G LTE and 5G technologies, and are referred to as non-standalone (NSA) deployments. In some deployments, the E-UTRAN includes one or more ng-eNBs that are configured to communicate with the 5GC, andthat may also be configured to communicate with one or more gNBs. Similarly, in another deployment, the NG-RAN may include one or more en-gNBs that are configured to communicate with the EPC, and that may also be configured to communicate with one or more eNBs. In various instances, a single UE 110, a dual-mode or multimode UE, may support multiple (two or more) RANs—thereby being configured to connect to multiple RANs, such as 4G LTE and 5G.

[0043] In some deployments, such as deployment 200, operations of the radio accessnode 202 may be carried out, at least partly, in a central / centralized unit (CU), such as aserver, host or node, operationally coupled to a distributed unit (DU), such as a radio head / node. It is also possible that node operations may be distributed among a plurality of servers, hosts or nodes.

[0044] It should also be understood that the distribution of work between CN 106operations and radio access node 202 operations may vary depending on implementation.Thus, a 5G network architecture may be based on a so-called CU-DU split. One gNB-CU (central node) may control one or more gNB-DUs. The gNB-CU may control a plurality of spatially separated gNB-DUs, acting at least as transmit / receive (Tx / Rx) nodes. In some example implementations, however, the gNB-DUs (also called DU) may include, for example, a radio link control (RLC), medium access control (MAC) layer and aphysical (PHY) layer, whereas the gNB-CU (also called a CU) may include the layersabove the RLC layer, such as a packet data convergence protocol (PDCP) layer, a radio resource control (RRC), and an internet protocol (IP) layer. Other functional splits are also possible. It is considered that a skilled person is familiar with the open systems interconnection (OSI) model and the functionalities within each layer.

[0045] In some example implementations, the server or CU may generate a virtualnetwork through which the server communicates with the radio node. In general, virtual networking may involve a process of combining hardware and software network resources and network functionality into a single, software-based administrative entity, a virtual network. Such virtual network may provide flexible distribution of operations between the server and the radio head / node. In practice, any digital signal processing task may be performed in either the CU or the DU, and the boundary where the responsibility is shifted between the CU and the DU may be selected according to implementation.

[0046] Although only one radio access node 202 is shown in FIG. 2, the deploymentmay include multiple radio access nodes, and at least some of the radio access nodes may be connected to one another by a network interface, such as an Xn interface. Similarly,the radio access nodes may be connected to the CN 106 by a network interface. In 5GNR, the network interface between a gNB and the 5GC is referred to as the NG interface. The NG interface is divided into the NG control plane interface (NG-C), which is anetwork interface between the gNB and access node – access and mobility managementfunction (AMF) in the 5GC, and the NG user plane interface (NG-U), which is the interface between the gNB and user plane function (UPF) in the 5GC. These and other network interfaces may support the exchange of signaling messages between network entities. The signaling messages may be formatted according to an application layer protocol. Examples of application protocols include the NG application protocol (NGAP) for the NG-C interface between the gNB and AMF, and the Xn application protocol (XnAP) for the Xn interface between gNBs.

[0047] For a UE 110 in an RRC connected state, it is generally desirable to keep theUE’s traffic uninterrupted when the UE moves within a cell (at times referred to as aradio cell) or across different cells of one or more radio access nodes 202 (e.g., gNBs). Tocontinuously monitor the UE’s radio link condition toward a serving cell provided by a serving radio access node, the UE may be configured to measure received signal level and quality from the serving cell as well as a list of configured neighboring cells, and report the results to the radio access node periodically and / or whenever a configured reporting event is met. These measurements may then be evaluated at the radio access node, and may result in preparation of one target cell (conventional handover) or multiple candidate (candidate target) cells (conditional handover), and a conventional / conditional handover of the UE from the serving cell provided by the serving radio access node (a “source access node” or more simply a “source node”) to a new cell provided by a target radio access node (a “target access node” or more simply a “target node”).

[0048] In conventional / conditional handover, execution of the ‘handover’ from onecell to another may occur at higher layers, such as RRC (or layer 3, L3). In another type of handover, execution of the ‘handover’ may be moved from the higher layers to lowerlayers. These lower layers may be either PHY (or layer 1, L1) or MAC (or layer 2, L2).This type of handover is currently referred to as L1 / L2-triggered mobility, or lower-layer triggered mobility (LTM), which may reduce latency, overhead and interruption time when compared to L3 handover based mobility.

[0049] In the case of conditional handover (CHO), a UE 110 may be configured witha CHO command containing the candidate cell configuration and one or more CHO conditions for carrying out the handover (HO) to one or more candidate cells. The condition may be based on radio measurements. For example, a condition may be that a measured reference signal received power (RSRP) from the serving cell falls below a threshold RSRP. Each of the candidate cell(s) may have prepared a necessary configuration, such as contention free random access (CFRA) resources for the UE. Such configuration may then be communicated to the UE in the CHO command.

[0050] When UE 110 evaluates the CHO condition and determines that the conditionholds for a specific candidate cell, UE may be configured to apply the CHO command and use the reserved CFRA resources to initiate a random access channel (RACH)procedure towards the candidate cell as a target cell for handover. By the RACHprocedure, the UE may obtain downlink (DL) and uplink (UL) synchronization andRACH access with the candidate node 202B that provides the candidate cell. And duringUL synchronization, the UE may obtain a timing advance (TA) value for the target cell.

[0051] The interruption time of mobility procedures needs to be reduced to make surethe connection up time is enhanced. This is critical for use-cases that have strict end-to- end delay requirements, such as URLLC, future railway mobile communication system(FRMCS), and the like. A RACH-less CHO may reduce the interruption time for mobilityprocedures. In a RACH-less CHO, the UE 110 may perform the RACH proceduretowards the target cell, while still connected to the serving (source) cell so that the TAvalue for the target cell is known at the UE, prior to the UE triggers handover to the targetcell. This approach may reduce the amount of “gap” that occurs at the time of handover,but it still requires the UE to carry out a RACH procedure towards the target cell.

[0052] FIGS. 3A, 3B and 3C illustrate a signaling chart 300 of a RACH-less CHOprocedure involving a UE 110, a source node 202A that is currently serving (connectedto) the UE, and one or more candidate nodes 202B (a target node 202B’ and otherpotential target node(s) shown). Similar to the conventional HO procedure, the UE maybe configured to report measurements of one or more neighboring cells, such as based on one or more reporting events. One example of a suitable reporting event, referred to as an A3 event, occurs when a neighboring cell becomes better than the source node by an offset. As shown in FIG. 3A, at step 301, a reporting event may be triggered at the UE,and the UE may at step 302 send an L3 measurement report indicating relevantmeasurements for one or more candidate cells provided by the candidate node(s). Themeasurement report may indicate, for example, a RSRP for one or more of the candidatecell(s) is 6 dB better than the RSRP for the source node.

[0053] The source node 202A may at step 303 decide to initiate the handoverprocedure as a CHO procedure; and similar to a conventional HO procedure, initiate ahandover preparation in which the source node may at step 304 send a handover requestmessage with a current configuration of the UE 110 towards each of the candidatenode(s) 202B. The source node may also indicate to the candidate node(s) that the handover preparation is for RACH-less CHO.

[0054] Each candidate node 202B may at step 305 perform admission control, suchas to accept or reject the handover request. Each candidate node may use the RACH-less CHO indication to determine if the candidate node supports an early TA acquisition procedure for CHO or conventional (baseline) HO. If a candidate node supports an earlyTA acquisition procedure, the candidate node may indicate / allocate early TA acquisitionresources (preambles for TA acquisition). The candidate node may at step 306 provide ahandover request acknowledgement (ACK) message (to the source node 202A) including a configuration of the early TA acquisition resources, and a candidate cell configuration. The candidate cell configuration may be the same as or similar to the conventional HO procedure, including for example, a cell radio network temporary identifier (C-RNTI), adata radio bearer (DRB) configuration, quality of service (QoS) flow to DRB mapping,UE capability related features enabled by the candidate node, or the like.

[0055] The source node 202A may then at step 307 prepare an RRC reconfigurationbased on the configuration for TA acquisition including the early TA acquisitionresources, and the source node may at step 308 send a RRC reconfiguration (towards theUE 110) with a CHO command message. The CHO command message may indicate early TA acquisition including the early TA acquisition resources, and include thecandidate cell configuration, for each of the candidate node(s) 202B. The CHO command message may also include one or more CHO conditions for one or more candidate cells of the candidate node(s). The UE may then at step 309 send an RRC reconfiguration complete message to the source node.

[0056] As shown in FIG. 3B, at step 310, a reporting event may be triggered at theUE 110, and the UE may at step 311 send an L3 measurement report indicating relevantmeasurements for one or more candidate cells provided by the candidate node(s) 202B.Similar to before, the measurement report may indicate, for example, a RSRP for one ormore of the candidate cell(s) is 3 dB better than the RSRP for the source node.

[0057] The source node 202A at step 312 decides to trigger the TA acquisition of thecandidate cell(s), and the source node at step 313 sends a physical downlink controlchannel (PDCCH) order (using downlink control information (DCI) format 1_0) or otherTA acquisition command to the UE 110. The UE at step 314 triggers early RRCprocessing, and at step 315 sends a RACH preamble to the candidate cell(s) (shown to thetarget node 202B’) to signal the candidate cell(s) to estimate the TA between the UE and the candidate cell(s). And at step 316, the UE cell receives a random access response (RAR) from respective ones of the candidate cell(s) indirectly via the source node 202A, and the RAR from each candidate cell may include an UL grant (physical uplink shared channel (PUSCH) resource), and an indication of the TA value for the candidate cell.

[0058] As shown in FIG. 3C, the UE 110 maintains its connection with the sourcenode, and at step 317 starts evaluating the CHO condition(s) for the candidate cell(s). Ifat least one candidate cell satisfies a corresponding CHO condition, as shown at step 318, the UE may apply the candidate cell configuration for the candidate cell as a target cellfor a handover. The UE may at steps 319, 320, 321 determine the TA value for thecandidate cell as a target cell is valid, use the UL grant and the TA value for the targetcell, and send a RRC reconfiguration complete message to the target node 202B’ / cell toaccess the target node / cell.

[0059] The target node 202B’ may at step 322 send a handover success message tothe source node 202A to inform that the UE 110 has successfully accessed the target cell.In return, the source node may at step 323 stop transmission / reception of user datato / from the UE, start forwarding user data to the target node. The source node may at step324 send a serving node (SN) status transfer message to the target node. Also, the sourcenode may at step 325 send a CHO release preparation message toward the other candidatenodes 202B’ / cells, if any, to cancel CHO for the UE. The target node and CN 106 mayat step 326 carry out a path switch to switch the DL data path towards the target node andestablish an interface instance towards the target node.

[0060] In Release 18, 3GPP has agreed to allow the network to configure the UE 110to perform TA estimation if supported by the UE. A common method is based on DLmeasurement, which may be used if a condition is satisfied in which the timing difference between the source cell and candidate cell is no more than a threshold time (e.g., 260 nanoseconds). When the condition is satisfied, a flag may be indicated to enable the UE to perform UE-based TA acquisition.

[0061] FIG. 4 illustrates a UE-based TA acquisition scheme for LTM which is basedon DL measurement. As shown, a UE 110 may be configured to communicate withmultiple transmission-reception points (TRPs) including a first TRP (TRP1) 402A and asecond TRP (TRP2) 404B. A TRP may be a set of geographically collocated antennas supporting transmission point (TP) and / or reception point (RP) functionality. In various examples, TRPs may include radio access nodes 202, radio access node antennas, remote radio heads (RRHs), radio units (RUs), a remote antenna of a radio access node, or the like.

[0062] In the illustrated example, the UE 110 may be configured to determine (orestimate) the TA of TRP2402B while the UE is served by TRP1402A. In this regard, theTA of TRP1 (TA1) may be known to the UE, and the UE may determine the TA of TRP2 (TA2) as follows: TA2 = TA1 + 2 RTD – 2 Realized (TAE) – OtherEstErrorIn the preceding, RTD refers to relative time difference, TAE refers to timing alignmenterror, and OtherEstError refers to any error that can be caused by the UL / DL reciprocityor estimator implementation / method error.

[0063] The TAE may be the relative difference in time of transmission ofsimultaneous DL signals between TRP1402A and TRP2402B, such as reference signalslike synchronization signal (SS) / physical broadcast channel (PBCH) blocks (SSB),channel state information – reference signals (CSI-RS), or the like. A similar quantity,referred to as cell phase synchronization accuracy, pertains to transmissions from a pair of cells. The cell phase synchronization accuracy for time-division duplex (TDD) may be defined as the maximum absolute deviation in frame start timing between any pair of cells on the same frequency that have overlapping coverage areas.

[0064] Furthermore, a transmit timing error may be defined as the result of a transmittime delay D1 / D2 involved in the transmission of a DL signal, which may be in turn defined as the time delay from the time when a digital signal is generated at baseband to the time when a corresponding radio frequency (RF) signal is transmitted from a transmitantenna. In addition, RTD may be provided by an information element (IE) NR-RTD-Info,which is used by a location server to provide time synchronization information between a reference TRP and a list of neighbor TRPs. A number of these definitions, IEs andrelevant mechanisms are provided in 3GPP specifications, and allow the UE 110 tobecome aware of timing misalignments in transmissions from different TRPs, and take them into account in TA acquisition.

[0065] As described above, in a RACH-less CHO, the UE 110 may perform theRACH procedure towards the target cell, while still connected to the serving (source) cell so that the TA value for the target cell is known at the UE, prior to the UE triggershandover to the target cell. This approach may reduce the amount of “gap” or“interruption time” that occurs at the time of handover, but it still requires the UE to carryout a RACH procedure towards the target cell. A UE-based TA acquisition scheme based on DL measurement may avoid a RACH procedure in LTM, but measuring DL signalsfrom another cell sometimes requires the UE 110 to synchronize to the other cell, andadjust the UE’s RF hardware to the other cell. This very often leads to DL datainterruption at the serving cell itself.

[0066] In view of the foregoing, example implementations of the present disclosureprovide a TA acquisition scheme which is based on artificial intelligence (AI) / machinelearning (ML) – referred to at times as a ML-based TA acquisition scheme. The ML-based TA acquisition scheme of example implementations may avoid RACH preamble transmission (and its power, interference, reception, delay, preamble reservation, etc.), while reducing data interruption relative to a UE-based scheme which is based on DL measurement. The ML-based TA acquisition scheme may enable a more seamlesshandover from one cell to another cell, and with reduced signaling / overhead. The network may also be provided an ability to control and optimize the balance between that overhead reduction and the seamlessness of the handover, as well as use of AI / ML at the UE to achieve improvements in TA acquisition. Some example implementations of the present disclosure are described below, mainly in the context of RACH-less CHO. It should be understood, however, that some example implementations may also be applicable to other types of handover, such as LTM.

[0067] According to some example implementations of the present disclosure, theML-based TA acquisition scheme includes application of at least one input including a TAvalue for the source node 202A to a ML model to determine an estimated TA value for acandidate cell that may be a target cell for a handover of the UE 110, as well as a confidence level for the estimated TA value. Although described as residing at the UE, in various example implementations, the ML model may recite at either the UE or the network (RAN 108).

[0068] The ML-based TA acquisition scheme may also incorporate a confidencelevel threshold that indicates when the UE may use the estimated TA value for the target cell as the TA value for the target cell. In this regard, the UE may be allowed to use the estimated TA value as the TA value for the target cell when the confidence level is at or above the confidence level threshold. The UE may then carry out a RACH-less handover with the target cell in which a RACH preamble is skipped, and the estimated TA value is used as the TA value to access the target cell.

[0069] In some examples, the UE 110 may indicate to the source node 202A that theUE supports (or does not support) the ML-based TA acquisition scheme (using a ML model to determine an estimated TA value for the target cell). The source node may provide control over the ML-based TA acquisition scheme, such as by providing an indication that the UE is allowed (or not allowed) to use the ML-based TA acquisitionscheme. In some examples, the source node or target node 202B’ may provide a selectionof the ML model to be used, among one or more ML models available at the UE. This selection may be based on a number of factors, such as prior success rate for the estimated TA value (in similar contexts, for similar UEs) from each of the available ML model(s).

[0070] The source node 202A or target node 202B’ may also indicate the confidencelevel threshold to be used by the UE 110. In this regard, the RAN 108 may indicate arequirement that the ML model have at least some minimum level of confidence (e.g.,Y%) that its estimated TA value is accurate within some bounds (e.g., X microseconds).The confidence level threshold may be expressed in a number of different manners, such as by an indication of the X, Y values that the network (e.g., target node) may tolerate. Similarly, the UE may express the confidence level for the estimated TA value in a number of different manners, such as by X, Y values; and in some examples, the UE may report the confidence level to the source node.

[0071] In some examples, the network may provide assistance to help to the UE 110to improve its ML model, such as in cases in which the confidence level for the estimated TA value output by the ML model is (continually) below the confidence level threshold.

[0072] Although described primarily in the context of handover, and in particularCHO, the ML-based TA acquisition scheme may be used for maintaining the TA value ofa cell, without performing a DL signal measurement. For example, the ML model in one possibility may be viewed as associated with a terrain of an environment of the UE 110, such as an urban area or rural area, which may indicate different ML models. The ML model may be applicable not only to mobility but for other purposes. It may be possible that the ML-based TA acquisition scheme is carried out multiple times, and when the estimated TA value is useful for another purpose, the prediction may be run in a sequential manner. In this regard, after a ML model is used in the ML-based TA acquisition scheme to determine or maintain a TA value, the ML model may be used for another purpose, and then return for use again in the ML-based TA acquisition scheme.

[0073] As indicated above, the ML-based TA acquisition scheme may includeapplication of at least one input including a TA value for the source cell of the source node 202A. In this regard, the input(s) may indicate a UE context. FIGS.5 and 6illustrate respective ML model training scenarios 500, 600 according to various exampleimplementations. As shown, the ML model may be built using a training dataset including observations of UE context and actual (verified) TA value for one or moretarget cell(s) determined after prior UE handovers in similar UE contexts. And as moreparticularly shown in FIG.6, the ML model may be trained for both the estimated TA value, as well as the confidence level for the estimated TA value.

[0074] The UE context may include the TA value for the source node. In someexamples, the UE context may also include one or more of already estimated TA values for other candidate cell(s) during prior, similar handover events, geographic location of the candidate node(s) that provide the candidate cell(s), distance from the UE to the candidate node(s), RSRP, speed of movement of the UE, or the like.

[0075] The ML model may leverage the UE context (UE context information) for apopulation of UEs. In this regard, for a given (group of) international mobile equipment identity software version (IMEISV) or Apple UEs, UE context information gathered on other (e.g., Apple) UEs may be operatively leveraged to train the ML model. After a handover or RACH-less handover, the UE may also gather feedback on the degree to which the estimated TA value for the new serving cell was accurate (or inaccurate).

[0076] FIG. 7 illustrates a deployment scenario 700 for a ML model to determine anestimated TA value, and a confidence level for the estimated TA value, according to some example implementations. In some more specific examples, a dedicated ML model (e.g., supervised learning neural network based model) may be used to determine the estimated TA value for a target cell, considering at least one input the TA value for the source (serving) cell. The ML model may also provide a confidence level for the estimated TA value. And to prepare the ML model, the ML model may be built using a training dataset including the TA values for the source and target cells over different positions, and ground truth TA values.

[0077] FIGS. 8A and 8B illustrate respective training and deployment scenarios800A, 800B for a regression ML model, according to some example implementations. Inboth scenarios, the regression model may be trained to determine an estimated TA value for a target cell based on at least one input including a TA value for the source cell. In order to also provide the confidence level for the estimated TA value, the regression ML model may be enhanced with a classification type model which includes a softmax layer at the end of its ML architecture. The classification type model may therefore be trained to output a probability value corresponding to the confidence level.

[0078] According to various examples, the confidence level for an estimated TAvalue is defined as the rate of successfully estimated TA value as compared to the actual TA value / network-determined TA value with an allowable error margin in lived condition and statistically. In this regard, the estimated TA value may be compared with the TA value determined by the network. The network may provide feedback to the UE on the network’s estimated TA value which may help the UE correct for an error in its estimated TA value. In addition, the confidence level may viewed as how often the TA value is successfully estimated with an allowable error from the network-determined TA value (determined while the communication link is still working).

[0079] According to various examples, a trigger event may be defined based on theestimated TA value and the confidence level. The trigger event may be defined with a two-dimensional threshold in which one dimension is on the confidence level, and another dimension is on the estimated TA value. When both thresholds are met, the UE may trigger evaluation of CHO condition(s), or more directly trigger handover of the UE.

[0080] Consider an example in which event XX is defined where the TA value of aneighbor cell becomes better than a threshold TA value. In this example, the UE may consider an entering condition for the event to be satisfied when a condition A is fulfilled, and consider a leaving condition for the event to be satisfied when a condition B is fulfilled.

[0081] In some examples, condition A and condition B may be specified asinequalities that are functions of the confidence level as follows: Inequality A (Entering condition) Conf-Level + Offset – Hys > ThreshInequality B (Leaving condition) Conf-Level + Offset + Hys < ThreshIn the inequalities, Conf-Level is the confidence level defined of the estimated TA valuefor the candidate cell, Offset is the offset value defined for the candidate cell, Hys is ahysteresis parameter for the event, and Thresh is a threshold (confidence level) parameterfor the event. One or more of these variables may be defined or otherwise indicated in a report configuration or measurement configuration, or updated via a MAC control element (MAC-CE).

[0082] Additionally or alternatively, condition A and condition B may be specified asinequalities that are functions of the estimated TA value as follows: Inequality A (Entering condition) TA-Level + Offset – Hys > ThreshInequality B (Leaving condition) TA-Level + Offset + Hys < ThreshHere, TA-Level is the estimated TA value for the candidate cell, and Thresh is a threshold(TA value) parameter for the event.

[0083] In some examples, the trigger event may measure the confidence level of theestimated TA value for the candidate (neighbor) cell. If the confidence level of the estimated TA value is high, it shows the estimated TA value is reliable. From theestimated TA value itself, the UE 110 may project the distance of the UE to the candidate(neighbor) node 202B providing the candidate cell. Naturally, the closer the UE to thecandidate node, the higher the conventional RSRP value. This is a knowledge that the UE may derive without performing an actual measurement of RSRP; and as a result, the UE may use the estimated TA value as an alternative to RSRP threshold. In some examples, then, as long as the estimated TA value is reliable (e.g., indicated by a confidence level ator above the confidence level threshold), the estimated TA value may be used as analternative trigger for handover (e.g., CHO condition).

[0084] Some example implementations may also provide a calibration of theconfidence level threshold. In some of these examples, in cases in which the ML model is vendor specific and its related information is not shared with the network, the network may not be able to assess the confidence sensibility with regards to the performance and prediction accuracy. In these cases, the network may carry out the calibration in order to evaluate an appropriate confidence level threshold blindly considering the ML model as a black box.

[0085] The calibration of the confidence level threshold may be implemented in anumber of different manners, such as based on a training dataset (e.g., a labeled dataset) at the network. This may include at least one input including a TA value for the source cell which indicate a UE context (e.g., which may correspond to RSRP source + candidate cells), and an output value of an estimated TA value for a target cell. Thenetwork may provide a sample input to the UE, and receive in return, an output value and a confidence level for the output value. The network may calculate an error between the output value and a corresponding ground truth value for the sample input, and determine a confidence level threshold that indicates a maximum allowed prediction error for the estimated TA value.

[0086] FIGS. 9A, 9B and 9C illustrate a signaling chart 900 of a RACH-less CHOprocedure for a UE using a ML-based TA acquisition scheme, according to some exampleimplementations. As shown, the UE 110 at step 901 informs the source node 202A(source cell) that the UE is capable of at least one of performing a skippable TA andmaintaining the TA according to a ML-based TA acquisition scheme. A reporting eventmay be triggered at the UE at step 902, and the UE may at step 903 send an L3measurement report indicating relevant measurements for one or more candidate cells provided by the candidate node(s). The measurement report may indicate, for example, aRSRP for one or more of the candidate cell(s) is 6 dB better than the RSRP for the sourcenode.

[0087] The source node 202A may at step 904 decide to initiate the handoverprocedure as a CHO procedure; and similar to a conventional HO procedure, initiate ahandover preparation in which the source node may at step 905 send a handover requestmessage with a current configuration of the UE 110 towards each of the candidatenode(s) 202B. The source node may also indicate to the candidate node(s) that the handover preparation is for RACH-less CHO.

[0088] Each candidate node 202B may at step 906 perform admission control, suchas to accept or reject the handover request. Each candidate node may use the RACH-less CHO indication to determine if the candidate node supports an early TA acquisitionprocedure for CHO or conventional (baseline) HO. The candidate node may at step 907provide a handover request ACK message (to the source node 202A) including a candidate cell configuration, similar to before. If a candidate node supports an early TA acquisition procedure, the handover request ACK message message may also include an UL grant (PUSCH) resource), an indication of a confidence level threshold associated with the RACH-less handover, and perhaps also a selection of a ML model for the UE 110.

[0089] The source node 202A may at 908 carry out a calibration of the confidencelevel threshold. As shown, for example, the source node may at step 909 prepare an RRCreconfiguration based on the UL grant, confidence level threshold and / or ML modelselection, and the source node may at step 910 send a RRC reconfiguration (towards theUE 110) with a CHO command message. Similar to before, the CHO command message may include the candidate cell configuration for each of the candidate node(s) 202B, and one or more CHO conditions for one or more candidate cells of the candidate node(s). The CHO command message may also include a sample input, and a request for the UE to process the sample input using the ML model.

[0090] The UE 110 may apply the sample input to the ML model to determine anoutput value and a confidence level for the output value, and the UE may at step 911 sendthe output value and confidence level to the source node 202A.

[0091] Upon receiving the output value and confidence level for the output value, thesource node 202A may at step 912 assess performance of the ML model, such as bycalculating an error between the output value and a corresponding ground truth value for the sample input. The source node may calibrate the confidence level threshold based on this error and confidence level for the output value reported by the UE 110. In this regard, the source node may update the confidence level threshold to correspond to an upper bound of acceptable error between the output value from the ML model and the groundtruth. The source node may then at step 913 send the calibrated / updated confidencelevel threshold to the UE, and the UE may at step 914 send a RRC reconfigurationcomplete message to the source node.

[0092] The UE 110 may at step 915 begin performing TA value estimation withskippable RACH resource using the ML model, according to the ML-based TA acquistionscheme. The UE may at step 916 determine the confidence level for estimated TAvalue(s) for the candidate cell(s), and evaluate the confidence level against the confidence level threshold. As shown at step 917, the UE may determine a trigger event (event XX) for evaluation of the CHO condition(s) is fulfilled based on the estimated TA value and the confidence level. As explained above, this may include a determination that the confidence level is at or above the confidence level threshold.

[0093] As shown in FIG. 9C, the UE 110 may at step 918 send an L3 measurementreport indicating relevant measurements (performed in the background) for one or more candidate cells provided by the candidate node(s) 202B. The UE may also include theconfidence level of the estimated TA value(s) determined from the ML model. The UEmay maintain its connection with the source node, and at steps 919, 920 determine theestimated TA value is valid, and start evaluating the CHO condition(s) for the candidate cell(s).

[0094] If at least one candidate cell satisfies a corresponding CHO condition and theconfidence level is at or above the confidence level threshold, as shown at step 921, the UE may apply the candidate cell configuration for the candidate cell as a target cell for ahandover. The UE may at step 922 execute a RACH-less handover to the target cell of thethe target node 202B’. The UE may at steps 923, 924 use the UL grant and the estimatedTA value for the target cell, and send a RRC reconfiguration complete message to the target node / cell to access the target node / cell.

[0095] The target node 202B’ may at step 925 send a handover success message tothe source node 202A to inform that the UE 110 has successfully accessed the target cell.In return, the source node may at step 926 stop transmission / reception of user datato / from the UE, start forwarding user data to the target node. The source node may at step927 send a SN status transfer message to the target node. Also, the source node may atstep 928 send a CHO release preparation message toward the other candidate nodes / cells, if any, to cancel CHO for the UE. The target node and CN 106 may at step 929carry out a path switch to switch the DL data path towards the target node and establish an interface instance towards the target node.

[0096] In some examples, the source node 202A may provide network assistance toimprove model performance at the UE 110. As explained above, the UE may be allowedto use its estimated TA value for the target node 202B’ when the confidence level for theestimated TA value is at or above the confidence level threshold. But if the confidence level remains under the confidence level threshold for some period of time, the UE may determine that the ML model needs to be replaced or updated. In some of these examples, then, the UE may share one or more capabilities of its ML model with the source node, which may effectively steer the degree to which the source ndoe may provide assistanceto help the UE improve the ML model. Additionally or alternatively, for example, selection of the level of network assistance may account for UE capabilities, such as device battery level, which may impact whether or not the UE is allowed to perform model retraining or further training.

[0097] In some examples in which the UE 110 has the ability to perform local MLmodel training, the UE may request assistance from the source node 202A by requestinga training dataset, such as a labeled dataset (e.g., TA on serving cell, TA on target cell),gathered by the RAN 108 from multiple UEs in various conditions and locations. Uponreceiving the training dataset from the source node, the UE may retrain or further train the ML model to improve the confidence level for its output values, and thereby update the ML model.

[0098] In some examples in which the UE 110 is unable or unwilling to perform localML model training, the UE may be willing to share its ML model with the source node 202A. In some of these examples, UE may send the ML model to the source node which may retrain or further train the ML model to produce an updated ML model. The source node may then send the updated ML model to the UE which may switch from the ML model to the updated ML model.

[0099] In some examples, the UE 110 may be willing to use another ML model (e.g.,a ML model with higher confidence level for its output values). In some of these examples, the UE may request another (trained) ML model from the source node 202A. The UE may receive the other ML model from the source node, and switch from the ML model to the other ML model for determining the estimated TA value. Or the UE may otherwise decide to either use the ML model or the other ML model based on the confidence levels achievable by the respective ML models.

[0100] After the RACH-less handover, the target cell of the target node 202B may bea new serving cell for the UE 110; and in some examples, the UE may maintain the TA value for the new serving cell according to the ML-based TA acquisition scheme, and without DL signal measurement. As described above, the ML model may map inputs toan estimated TA value. The ML model may be trained using a ML algorithm that definesa relationship between the inputs and the output, possibly a multi-dimensional relationship, which one cannot simply obtain without using AI / ML.

[0101] In various examples, the TA value for a cell may be a combined result ofmany input variables, such as distance, obstructing objects, and / or weather which could influence the air density which in turn impact the electromagnetic wave movement in the air. The TA value for a cell is currently simply measured explicitly, but its accuracy may not be long lasting because of changes in the input variables that impact the TA value. Even outside of handover, then, the TA value for the serving cell of the UE may change over time.

[0102] In some example implementations, then, the UE may use the ML-based TAacquisition scheme (with the same or a different ML model) to maintain the TA value for its serving cell. For example, the UE may access the input variables over time (e.g., weather forecast, geogaphic map of the location of the UE, etc.), and apply those variables to a ML model to plot a distribution of the TA value for the serving cell over a certain time period (e.g., 10 minutes), which may be used for UL transmissions to the serving cell. This information may be different from the conventional DL signal measurement, and obtained in other ways such as sensors, broadcast information from other channel(s), or the like.

[0103] FIGS. 10A – 10E are flowcharts illustrating various steps in a method 1000performed by a user equipment (UE) served by a source cell, according to various example implementations. The method includes determining an estimated timing advance (TA) value for a candidate cell, and a confidence level for the estimated TA value, according to a machine learning (ML)-based TA acquisition scheme, as shown at block 1002 of FIG. 10A. The method includes making a determination that a handover condition is fulfilled for the candidate cell, and the confidence level for the estimated TA value is at or above a confidence level threshold, as shown at block 1004. Based on the determination, the method includes applying a configuration of the candidate cell as a target cell for a handover, as shown at block 1006. And the method includes carrying out the handover as a random access channel (RACH)-less handover in which a RACH preamble is skipped, and the estimated TA value is used as a TA value for the target cell to access the target cell, as shown at block 1008.

[0104] In some examples, the method 1000 further includes sending an indication tothe source cell that the UE supports the ML-based TA acquisition scheme, as shown atblock 1010 of FIG. 10B. In some of these examples, the method also includes receivingan indication of the confidence level threshold from the source cell based on the indication that the UE supports the ML-based TA acquisition scheme, as shown at block 1012.

[0105] In some examples, determining the estimated TA value at block 1002 includesapplying at least one input including a TA value for the source cell to a ML model todetermine the estimated TA value, as shown at block 1014 of FIG. 10C.

[0106] In some examples, the ML model is among one or more ML models availableat the UE, and the method 1000 further includes receiving an indication of a selection ofthe ML model from the source cell.

[0107] In some examples, the method 1000 further includes receiving network-assistance information from the source cell, as shown at block 1016 of FIG. 10D. In someof these examples, the mehtod also includes updating the ML model based on the network assistance information, as shown at block 1018.

[0108] In some examples, the network-assistance information received from thesource cell includes a training dataset. In some of these examples, updating the MLmodel at block 1018 includes applying the training dataset to the ML model to retrain orfurther train the ML model to determine the estimated TA value.

[0109] In some examples, the method 1000 further includes sending a request to thesource cell for another ML model trained to determine the estimated TA value. In some of these examples, the network-assistance information received from the source cellincludes the other ML model, and updating the ML model at block 1018 includesswitching from the ML model to the other ML model to determine the estimated TA value.

[0110] In some examples, the method 1000 further includes sending the ML model tothe source cell for further training to produce an updated ML model. In some of these examples, the network-assistance information received from the source cell includes the updated ML model, and updating the ML model at block 1018 includes switching from the ML model to the updated ML model to determine the estimated TA value.

[0111] In some examples, the method 1000 further includes receiving at least onesample input from the source cell, as shown at block 1020 of FIG. 10E. In some of theseexamples, the method includes applying the at least one sample input to the ML model to determine an output value and a confidence level for the output value, as shown at block 1022. The method includes sending the output value and the confidence level for the output value to the source cell for calibration of the confidence level threshold based on a comparison of the output value and a ground truth value for the sample input, and the confidence level for the output value, as shown at block 1024. And the method includes receiving an indication of the confidence level threshold from the source cell based on the calibration, as shown at block 1026.

[0112] In some examples, the determination that the handover condition is fulfilled ismade at block 1004 based on an evaluation of the estimated TA value and the confidencelevel.

[0113] In some examples, the determination that the handover condition is fulfilled ismade at block 1004 based on an evaluation of the handover condition for one or morecandidate cells. In some of these examples, the method 1000 further includes determininga trigger event for the evaluation of the handover condition is fulfilled based on the estimated TA value and the confidence level.

[0114] In some examples, the target cell is a new serving cell for the UE after theRACH-less handover, and the method 1000 further includes maintaining the TA value forthe new serving cell according to the ML-based TA acquisition scheme.

[0115] FIGS. 11A – 11D are flowcharts illustrating various steps in a method 1100performed by a radio access node providing a source cell, according to various example implementations. The method includes receiving an indication from a user equipment (UE) served by the source cell that the UE supports a machine learning (ML)-basedtiming advance (TA) acquisition scheme, as shown at block 1102 of FIG. 11A. Themethod includes initiating a handover preparation for a candidate cell during which an indication that the handover preparation is for a random access channel (RACH)-less handover is sent to the candidate cell, as shown at block 1104. The method includes receiving, from the candidate cell, an indication of a confidence level threshold associated with the RACH-less handover, as shown at block 1106. And the method includes sending a handover command message and an indication of the confidence level threshold to the UE that cause the UE to carry out the RACH-less handover to thecandidate cell when a handover condition is fulfilled and a confidence level for an estimated TA value determined by the UE according to the ML-based TA acquisition scheme is at or above the confidence level threshold, as shown at block 1108.

[0116] In some examples, the ML-based TA acquisition scheme includes anapplication by the UE of at least one input including a TA value for the source cell to a ML model to determine the estimated TA value. In some of these examples, the handovercommand message sent to the UE at block 1108 includes an indication of a selection ofthe ML model.

[0117] In some examples, the method 1100 further includes receiving the indicationof the selection of the ML model from the candidate cell.

[0118] In some examples, the method 1100 further includes sending network-assistance information to the UE for updating the ML model, as shown at block 1110 ofFIG. 11B.

[0119] In some examples, the network-assistance information sent to the UE at block1110 includes a training dataset for application by the UE to the ML model to retrain orfurther train the ML model to determine the estimated TA value.

[0120] In some examples, the method 1100 further includes receiving a request fromthe UE for another ML model trained to determine the estimated TA value. In some of these examples, the network-assistance information sent to the UE includes the other ML model for the UE to switch from the ML model to the other ML model to determine the estimated TA value.

[0121] In some examples, the method 1100 further includes receiving the ML modelfrom the UE, as shown at block 1112 of FIG. 11C. In some of these examples, themethod includes retraining or further training the ML model to produce an updated ML model, as shown at block 1114. Also in some of these examples, the network-assistanceinformation sent to the UE at block 1110 includes the updated ML model for the UE toswitch from the ML model to the updated ML model to determine the estimated TA value.

[0122] In some examples, the handover command message sent to the UE at block1108 further includes at least one sample input for application by the UE to the MLmodel to determine an output value and a confidence level for the output value, as shownin FIG. 11D. In some of these examples, the method 1100 further includes receiving theoutput value and the confidence level for the output value from the UE, as shown at block 1116. And the method includes performing a calibration of the confidence level threshold based on a comparison of the output value and a ground truth value for the sample input, and the confidence level for the output value, as shown at block 1118. The indication of the confidence level threshold is then sent to the UE based on the calibration, as shown at block 1120.

[0123] According to example implementations of the present disclosure, atelecommunications system 100 or PLMN 102, and its components such as a UE 110, CN106, RAN 108, radio access node 202, source node 202A, candidate node 202B, target node 202B’, and / or other potential target node(s), may be implemented by various means. Means for implementing the system and its components may include hardware, firmware, software, or combinations thereof. In some examples, one or more apparatuses may be configured to function as or otherwise implement the system and its components shown and described herein. In examples involving more than one apparatus, the respective apparatuses may be connected to or otherwise in communication with one another in a number of different manners, such as directly or indirectly via a wired or wireless network or the like.

[0124] According to some example implementations, at least some of the method1000 described with respect to FIGS. 10A – 10E may be carried out by an apparatuscomprising means for performing functions corresponding steps of the method. Similarly,at least some of the method 1100 described with respect to FIGS. 11A – 11D may becarried out by an apparatus comprising means for performing functions corresponding steps of the method. Examples of a suitable apparatus may include a gNB (e.g., gNB-DU, gNB-CU), ng-eNB or any suitable apparatus, such as a server, host or node. Other examples of a suitable apparatus may include a user equipment, user device, user terminal or the like.

[0125] FIG. 12 illustrates an apparatus 1200 in which means for performing variousfunctions includes hardware, alone or under direction of one or more computer programs from a computer-readable storage medium or other memory, such as computer memory, according to some example implementations of the present disclosure. Generally, anapparatus of example implementations of the present disclosure may comprise, include or be embodied in one or more fixed or portable electronic devices. Examples of suitable electronic devices include a wearable computer, mobile phone, portable computer, desktop computer, workstation computer, server (server computer) or the like. The apparatus may include one or more of each of a number of components such as, forexample, processing circuitry 1202 connected to computer-readable storage medium orother memory 1204.

[0126] The processing circuitry 1202 may be composed of one or more processorsalone or in combination with one or more computer-readable storage media. The processing circuitry is generally any piece of computer hardware that is capable of processing information such as, for example, data, computer programs and / or other suitable electronic information. The processing circuitry is composed of a collection of electronic circuits some of which may be packaged as an integrated circuit or multiple interconnected integrated circuits (an integrated circuit at times more commonly referred to as a “chip”). The processing circuitry may be configured to execute computer programs, which may be stored onboard the processing circuitry or otherwise stored inthe memory 1204 (of the same or another apparatus).

[0127] The processing circuitry 1202 may be a number of processors, a multi-coreprocessor or some other type of processor, depending on the particular implementation. Further, the processing circuitry may be implemented using a number of heterogeneous processor systems in which a main processor is present with one or more secondary processors on a single chip. As another illustrative example, the processing circuitry may be a symmetric multi-processor system containing multiple processors of the same type. In yet another example, the processing circuitry may be embodied as or otherwise include one or more ASICs, FPGAs or the like. Thus, although the processing circuitry may be capable of executing a computer program to perform one or more functions, the processing circuitry of various examples may be capable of performing one or more functions without the aid of a computer program. In either instance, the processing circuitry may be appropriately programmed to perform functions or operations according to example implementations of the present disclosure.

[0128] The memory 1204 is generally any piece of computer hardware that is capableof storing information such as, for example, data, computer programs, instructions 1206 (e.g., computer-readable program code) and / or other suitable information either on a temporary basis and / or a permanent basis. The memory may include volatile and / or non- volatile memory, and may be fixed or removable. Examples of suitable memory include recording media, random access memory (RAM), read-only memory (ROM), a hard drive, a flash memory, a thumb drive, a removable computer diskette, an optical disk or some combination thereof.

[0129] The memory 1204 is a non-transitory device capable of storing information.One example of a suitable memory is a computer-readable storage medium, which is distinguishable from a computer-readable transmission medium capable of carrying information from one location to another. Examples of suitable computer-readable transmission media comprise electronic carrier signals, telecommunications signals, or some combination thereof. As used herein, the term “non-transitory” is a limitation of the medium itself (i.e., tangible, not a signal) as opposed to a limitation on data storagepersistency (e.g., RAM versus ROM). A computer-readable medium as described hereingenerally refers to a computer-readable storage medium or computer-readable transmission medium. A computer-readable medium is any entity or device capable in which information, such as one or more computer programs or portions thereof, may be stored and carried.

[0130] In addition to the memory 1204 (e.g., computer-readable storage medium), theprocessing circuitry 1202 may also be connected to one or more interfaces for displaying,transmitting and / or receiving information. The interfaces may include a communicationsinterface 1208 and / or one or more user interfaces. The communications interface may beconfigured to transmit and / or receive information, such as to and / or from other apparatus(es), network(s) or the like. The communications interface may be configured to transmit and / or receive information by physical (wired) and / or wireless communications links. Examples of suitable communication interfaces include a network interface controller (NIC), wireless NIC (WNIC) or the like.

[0131] The user interfaces may include a display 1210 and / or one or more user inputinterfaces 1212. The display may be configured to present or otherwise displayinformation to a user, suitable examples of which include a liquid crystal display (LCD), light-emitting diode (LED) display, organic LED (OLED) display, active-matrix OLED (AMOLED) or the like. The user input interfaces may be wired or wireless, and may be configured to receive information from a user into the apparatus, such as for processing, storage and / or display. Suitable examples of user input interfaces include a microphone, image or video capture device, keyboard or keypad, joystick, touch-sensitive surface (separate from or integrated into a touchscreen), biometric sensor or the like. The user interfaces may further include one or more interfaces for communicating with peripherals such as printers, scanners or the like.

[0132] Execution of the instructions 1206 by the processing circuitry 1202, or storageof the instructions in the memory 1204, supports combinations of operations for implementing example implementations of the present disclosure. In this manner, anapparatus 1200 may comprise at least one processing circuitry and at least one memorycoupled to the at least one processing circuitry, where the at least one processing circuitry is configured to execute instructions stored in the at least one memory. It will also be understood that one or more functions, and combinations of functions, may be implemented by special purpose hardware-based computer systems and / or processing circuitry which perform the specified functions, or combinations of special purpose hardware and program code instructions.

[0133] Some example implementations of the present disclosure may also be carriedout in the form of a computer process defined by one or more computer programs or portions thereof. Example implementations of the present disclosure may be carried out by executing at least one portion of a computer program comprising instructions. The computer program may be in source code form, object code form, or in some intermediate form. The computer program may be stored in a computer-readable medium that is readable by a computer, processing circuitry or other suitable apparatus. As indicated above, for example, the computer program may be stored in a memory, such as a computer-readable storage medium. Additionally or alternatively, for example, the computer program may be stored in a computer-readable transmission medium. The coding of software for carrying out example implementations of the present disclosure is well within the scope of a person of ordinary skill in the art.

[0134] As will be appreciated, any suitable instructions may be loaded onto acomputer, a processing circuitry or other programmable apparatus from a memory or a computer-readable medium (e.g., computer-readable storage medium, computer-readable transmission medium) to produce a particular machine, such that the particular machine becomes a means for implementing the functions specified herein. The instructions may also be stored in a computer-readable medium that can direct a computer, a processing circuitry or other programmable apparatus to function in a particular manner to thereby generate a particular machine or particular article of manufacture. In some examples, the instructions stored in the computer-readable medium may produce an article of manufacture, where the article of manufacture becomes a means for implementing functions described herein. The instructions may be retrieved from a computer-readable medium and loaded into a computer, processing circuitry or other programmable apparatus to configure the computer, processing circuitry or other programmable apparatus to execute operations to be performed on or by the computer, processing circuitry or other programmable apparatus.

[0135] Retrieval, loading and execution of instructions comprising program codeinstructions may be performed sequentially such that one instruction is retrieved, loaded and executed at a time. In some example implementations, retrieval, loading and / or execution may be performed in parallel such that multiple instructions are retrieved, loaded, and / or executed together. Execution of the program code instructions may produce a computer-implemented process such that the instructions executed by the computer, processing circuitry or other programmable apparatus provide operations for implementing functions described herein.

[0136] As explained above and reiterated below, the present disclosure includes,without limitation, the following example implementations.

[0137] Clause 1. An apparatus implemented by a user equipment (UE) served by asource cell, the apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: determine an estimated timing advance (TA) value for a candidate cell, and a confidence level for the estimated TA value, according to a machine learning (ML)-based TA acquisition scheme;make a determination that a handover condition is fulfilled for the candidate cell, and the confidence level for the estimated TA value is at or above a confidence level threshold;and based on the determination, apply a configuration of the candidate cell as a target cellfor a handover; and carry out the handover as a random access channel (RACH)-less handover in which a RACH preamble is skipped, and the estimated TA value is used as a TA value for the target cell to access the target cell.

[0138] Clause 2. The apparatus of clause 1, wherein the at least one processingcircuitry is configured to execute the instructions to cause the apparatus to further at least: send an indication to the source cell that the UE supports the ML-based TA acquisition scheme; and receive an indication of the confidence level threshold from the source cell based on the indication that the UE supports the ML-based TA acquisition scheme.

[0139] Clause 3. The apparatus of clause 1 or clause 2, wherein the apparatus causedto determine the estimated TA value includes the apparatus caused to apply at least oneinput including a TA value for the source cell to a ML model to determine the estimated TA value.

[0140] Clause 4. The apparatus of clause 3, wherein the ML model is among one ormore ML models available at the UE, and the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further receive an indication of a selection of the ML model from the source cell.

[0141] Clause 5. The apparatus of clause 3 or clause 4, wherein the at least oneprocessing circuitry is configured to execute the instructions to cause the apparatus to further at least: receive network-assistance information from the source cell; and update the ML model based on the network assistance information.

[0142] Clause 6. The apparatus of clause 5, wherein the network-assistanceinformation received from the source cell includes a training dataset, and the apparatuscaused to update the ML model includes the apparatus caused to apply the training dataset to the ML model to retrain or further train the ML model to determine the estimated TA value.

[0143] Clause 7. The apparatus of clause 5 or clause 6, wherein the at least oneprocessing circuitry is configured to execute the instructions to cause the apparatus tofurther send a request to the source cell for another ML model trained to determine theestimated TA value, and wherein the network-assistance information received from thesource cell includes the other ML model, and the apparatus caused to update the MLmodel includes the apparatus caused to switch from the ML model to the other ML model to determine the estimated TA value.

[0144] Clause 8. The apparatus of any of clauses 5 to 7, wherein the at least oneprocessing circuitry is configured to execute the instructions to cause the apparatus tofurther send the ML model to the source cell for retraining or further training to producean updated ML model, and wherein the network-assistance information received from thesource cell includes the updated ML model, and the apparatus caused to update the MLmodel includes the apparatus caused to switch from the ML model to the updated ML model to determine the estimated TA value.

[0145] Clause 9. The apparatus of any of clauses 3 to 8, wherein the at least oneprocessing circuitry is configured to execute the instructions to cause the apparatus to further at least: receive at least one sample input from the source cell; apply the at least one sample input to the ML model to determine an output value and a confidence level for the output value; send the output value and the confidence level for the output valueto the source cell for calibration of the confidence level threshold based on a comparisonof the output value and a ground truth value for the sample input, and the confidence level for the output value; and receive an indication of the confidence level thresholdfrom the source cell based on the calibration.

[0146] Clause 10. The apparatus of any of clauses 1 to 9, wherein the determinationthat the handover condition is fulfilled is made based on an evaluation of the estimated TA value and the confidence level.

[0147] Clause 11. The apparatus of any of clauses 1 to 10, wherein the determinationthat the handover condition is fulfilled is made based on an evaluation of the handovercondition for one or more candidate cells, and wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further determine a trigger event for the evaluation of the handover condition is fulfilled based on the estimated TA value and the confidence level.

[0148] Clause 12. The apparatus of any of clauses 1 to 11, wherein the target cell is anew serving cell for the UE after the RACH-less handover, and the at least oneprocessing circuitry is configured to execute the instructions to cause the apparatus to further maintain the TA value for the new serving cell according to the ML-based TA acquisition scheme.

[0149] Clause 13. An apparatus implemented by an user equipment (UE) served by asource cell, the apparatus comprising: means for determining an estimated timing advance (TA) value for a candidate cell, and a confidence level for the estimated TA value, according to a machine learning (ML)-based TA acquisition scheme; means for making a determination that a handover condition is fulfilled for the candidate cell, and the confidence level for the estimated TA value is at or above a confidence levelthreshold; and based on the determination, means for applying a configuration of thecandidate cell as a target cell for a handover; and means for carrying out the handover as a random access channel (RACH)-less handover in which a RACH preamble is skipped, and the estimated TA value is used as a TA value for the target cell to access the target cell.

[0150] Clause 14. The apparatus of clause 13, wherein the apparatus furthercomprises: means for sending an indication to the source cell that the UE supports the ML-based TA acquisition scheme; and means for receiving an indication of the confidence level threshold from the source cell based on the indication that the UEsupports the ML-based TA acquisition scheme.

[0151] Clause 15. The apparatus of clause 13 or clause 14, wherein the means fordetermining the estimated TA value includes means for applying at least one input including a TA value for the source cell to a ML model to determine the estimated TA value.

[0152] Clause 16. The apparatus of clause 15, wherein the ML model is among oneor more ML models available at the UE, and the apparatus further comprises means for receiving an indication of a selection of the ML model from the source cell.

[0153] Clause 17. The apparatus of clause 15 or clause 16, wherein the apparatusfurther comprises: means for receiving network-assistance information from the source cell; and means for updating the ML model based on the network assistance information.

[0154] Clause 18. The apparatus of clause 17, wherein the network-assistanceinformation received from the source cell includes a training dataset, and the means forupdating the ML model includes means for applying the training dataset to the ML model to retrain or further train the ML model to determine the estimated TA value.

[0155] Clause 19. The apparatus of clause 17 or clause 18, wherein the apparatusfurther comprises means for sending a request to the source cell for another ML modeltrained to determine the estimated TA value, and wherein the network-assistanceinformation received from the source cell includes the other ML model, and the meansfor updating the ML model includes means for switching from the ML model to the other ML model to determine the estimated TA value.

[0156] Clause 20. The apparatus of any of clauses 17 to 19, wherein the apparatusfurther comprises means for sending the ML model to the source cell for retraining orfurther training to produce an updated ML model, and wherein the network-assistanceinformation received from the source cell includes the updated ML model, and the meansfor updating the ML model includes means for switching from the ML model to the updated ML model to determine the estimated TA value.

[0157] Clause 21. The apparatus of any of clauses 15 to 20, wherein the apparatusfurther comprises: means for receiving at least one sample input from the source cell; means for applying the at least one sample input to the ML model to determine an output value and a confidence level for the output value; means for sending the output value andthe confidence level for the output value to the source cell for calibration of theconfidence level threshold based on a comparison of the output value and a ground truth value for the sample input, and the confidence level for the output value; and means forreceiving an indication of the confidence level threshold from the source cell based on thecalibration.

[0158] Clause 22. The apparatus of any of clauses 13 to 21, wherein thedetermination that the handover condition is fulfilled is made based on an evaluation of the estimated TA value and the confidence level.

[0159] Clause 23. The apparatus of any of clauses 13 to 22, wherein thedetermination that the handover condition is fulfilled is made based on an evaluation ofthe handover condition for one or more candidate cells, and wherein the apparatus further comprises means for determining a trigger event for the evaluation of the handover condition is fulfilled based on the estimated TA value and the confidence level.

[0160] Clause 24. The apparatus of any of clauses 13 to 23, wherein the target cell isa new serving cell for the UE after the RACH-less handover, and the apparatus further comprises means for maintaining the TA value for the new serving cell according to the ML-based TA acquisition scheme.

[0161] Clause 25. A method performed by a user equipment (UE) served by a sourcecell, the method comprising: determining an estimated timing advance (TA) value for a candidate cell, and a confidence level for the estimated TA value, according to a machine learning (ML)-based TA acquisition scheme; making a determination that a handover condition is fulfilled for the candidate cell, and the confidence level for the estimated TA value is at or above a confidence level threshold; and based on the determination,applying a configuration of the candidate cell as a target cell for a handover; and carryingout the handover as a random access channel (RACH)-less handover in which a RACH preamble is skipped, and the estimated TA value is used as a TA value for the target cell to access the target cell.

[0162] Clause 26. The method of clause 25, wherein the method further comprises:sending an indication to the source cell that the UE supports the ML-based TA acquisition scheme; and receiving an indication of the confidence level threshold from the source cell based on the indication that the UE supports the ML-based TA acquisition scheme.

[0163] Clause 27. The method of clause 25 or clause 26, wherein determining theestimated TA value includes applying at least one input including a TA value for the source cell to a ML model to determine the estimated TA value.

[0164] Clause 28. The method of clause 27, wherein the ML model is among one ormore ML models available at the UE, and the method further comprises receiving an indication of a selection of the ML model from the source cell.

[0165] Clause 29. The method of clause 27 or clause 28, wherein the method furthercomprises: receiving network-assistance information from the source cell; and updating the ML model based on the network assistance information.

[0166] Clause 30. The method of clause 29, wherein the network-assistanceinformation received from the source cell includes a training dataset, and updating theML model includes applying the training dataset to the ML model to retrain or further train the ML model to determine the estimated TA value.

[0167] Clause 31. The method of clause 29 or clause 30, wherein the method furthercomprises sending a request to the source cell for another ML model trained to determinethe estimated TA value, and wherein the network-assistance information received fromthe source cell includes the other ML model, and updating the ML model includesswitching from the ML model to the other ML model to determine the estimated TA value.

[0168] Clause 32. The method of any of clauses 29 to 31, wherein the method furthercomprises sending the ML model to the source cell for retraining or further training toproduce an updated ML model, and wherein the network-assistance information receivedfrom the source cell includes the updated ML model, and updating the ML modelincludes switching from the ML model to the updated ML model to determine the estimated TA value.

[0169] Clause 33. The method of any of clauses 27 to 32, wherein the method furthercomprises: receiving at least one sample input from the source cell; applying the at least one sample input to the ML model to determine an output value and a confidence level for the output value; sending the output value and the confidence level for the outputvalue to the source cell for calibration of the confidence level threshold based on acomparison of the output value and a ground truth value for the sample input, and the confidence level for the output value; and receiving an indication of the confidence levelthreshold from the source cell based on the calibration.

[0170] Clause 34. The method of any of clauses 25 to 33, wherein the determinationthat the handover condition is fulfilled is made based on an evaluation of the estimated TA value and the confidence level.

[0171] Clause 35. The method of any of clauses 25 to 34, wherein the determinationthat the handover condition is fulfilled is made based on an evaluation of the handovercondition for one or more candidate cells, and wherein the method further comprises determining a trigger event for the evaluation of the handover condition is fulfilled based on the estimated TA value and the confidence level.

[0172] Clause 36. The method of any of clauses 25 to 35, wherein the target cell is anew serving cell for the UE after the RACH-less handover, and the method further comprises maintaining the TA value for the new serving cell according to the ML-based TA acquisition scheme.

[0173] Clause 37. A computer-readable storage medium that is non-transitory and hasinstructions stored therein that, in response to execution by at least one processing circuitry, causes a user equipment (UE) served by a source cell to at least: determine anestimated timing advance (TA) value for a candidate cell, and a confidence level for theestimated TA value, according to a machine learning (ML)-based TA acquisition scheme; make a determination that a handover condition is fulfilled for the candidate cell, and the confidence level for the estimated TA value is at or above a confidence level threshold;and based on the determination, apply a configuration of the candidate cell as a target cellfor a handover; and carry out the handover as a random access channel (RACH)-less handover in which a RACH preamble is skipped, and the estimated TA value is used as a TA value for the target cell to access the target cell.

[0174] Clause 38. The computer-readable storage medium of clause 37, wherein thecomputer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the UE to further atleast: send an indication to the source cell that the UE supports the ML-based TAacquisition scheme; and receive an indication of the confidence level threshold from thesource cell based on the indication that the UE supports the ML-based TA acquisition scheme.

[0175] Clause 39. The computer-readable storage medium of clause 37 or clause 38,wherein the UE caused to determine the estimated TA value includes the UE caused toapply at least one input including a TA value for the source cell to a ML model todetermine the estimated TA value.

[0176] Clause 40. The computer-readable storage medium of clause 39, wherein theML model is among one or more ML models available at the UE, and the computer- readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the UE to further receive an indication of a selection of the ML model from the source cell.

[0177] Clause 41. The computer-readable storage medium of clause 39 or clause 40,wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the UE to further at least: receive network-assistance information from the source cell; and update the ML model based on the network assistance information.

[0178] Clause 42. The computer-readable storage medium of clause 41, wherein thenetwork-assistance information received from the source cell includes a training dataset,and the UE caused to update the ML model includes the UE caused to apply the training dataset to the ML model to retrain or further train the ML model to determine the estimated TA value.

[0179] Clause 43. The computer-readable storage medium of clause 41 or clause 42,wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the UE tofurther send a request to the source cell for another ML model trained to determine theestimated TA value, and wherein the network-assistance information received from thesource cell includes the other ML model, and the UE caused to update the ML modelincludes the UE caused to switch from the ML model to the other ML model to determine the estimated TA value.

[0180] Clause 44. The computer-readable storage medium of any of clauses 41 to 43,wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the UE tofurther send the ML model to the source cell for retraining or further training to producean updated ML model, and wherein the network-assistance information received from thesource cell includes the updated ML model, and the UE caused to update the ML modelincludes the UE caused to switch from the ML model to the updated ML model to determine the estimated TA value.

[0181] Clause 45. The computer-readable storage medium of any of clauses 39 to 44,wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the UE to further at least: receive at least one sample input from the source cell; apply the at least one sample input to the ML model to determine an output value and a confidence levelfor the output value; send the output value and the confidence level for the output valueto the source cell for calibration of the confidence level threshold based on a comparisonof the output value and a ground truth value for the sample input, and the confidence level for the output value; and receive an indication of the confidence level thresholdfrom the source cell based on the calibration.

[0182] Clause 46. The computer-readable storage medium of any of clauses 37 to 45,wherein the determination that the handover condition is fulfilled is made based on an evaluation of the estimated TA value and the confidence level.

[0183] Clause 47. The computer-readable storage medium of any of clauses 37 to 46,wherein the determination that the handover condition is fulfilled is made based on anevaluation of the handover condition for one or more candidate cells, and wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the UE to further determine a trigger event for the evaluation of the handover condition is fulfilled based on the estimated TA value and the confidence level.

[0184] Clause 48. The computer-readable storage medium of any of clauses 37 to 47,wherein the target cell is a new serving cell for the UE after the RACH-less handover, and the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the UE to further maintain the TA value for the new serving cell according to the ML-based TA acquisition scheme.

[0185] Clause 49. An apparatus comprising means for performing the method of anyof clauses 25 to 36.

[0186] Clause 50. A computer-readable medium comprising instructions that, inresponse to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 25 to 36.

[0187] Clause 51. A computer-readable storage medium comprising instructions that,in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 25 to 36.

[0188] Clause 52. A computer program comprising instructions that, in response toexecution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 25 to 36.

[0189] Clause 53. An apparatus implemented by a radio access node providing asource cell, the apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: receive anindication from a user equipment (UE) served by the source cell that the UE supports amachine learning (ML)-based timing advance (TA) acquisition scheme; initiate a handover preparation for a candidate cell during which an indication that the handover preparation is for a random access channel (RACH)-less handover is sent to the candidate cell; receive, from the candidate cell, an indication of a confidence level threshold associated with the RACH-less handover; and send a handover command message and an indication of the confidence level threshold to the UE that cause the UE to carry out the RACH-less handover to the candidate cell when a handover condition is fulfilled and a confidence level for an estimated TA value determined by the UE according to the ML- based TA acquisition scheme is at or above the confidence level threshold.

[0190] Clause 54. The apparatus of clause 53, wherein the ML-based TA acquisitionscheme includes an application by the UE of at least one input including a TA value for the source cell to a ML model to determine the estimated TA value, and wherein the handover command message sent to the UE includes an indication of a selection of the ML model.

[0191] Clause 55. The apparatus of clause 54, wherein the at least one processingcircuitry is configured to execute the instructions to cause the apparatus to further receive the indication of the selection of the ML model from the candidate cell.

[0192] Clause 56. The apparatus of clause 54 or clause 55, wherein the at least oneprocessing circuitry is configured to execute the instructions to cause the apparatus to further send network-assistance information to the UE for updating the ML model.

[0193] Clause 57. The apparatus of clause 56, wherein the network-assistanceinformation sent to the UE includes a training dataset for application by the UE to the ML model to retrain or further train the ML model to determine the estimated TA value.

[0194] Clause 58. The apparatus of clause 56 or clause 57, wherein the at least oneprocessing circuitry is configured to execute the instructions to cause the apparatus to further receive a request from the UE for another ML model trained to determine the estimated TA value, and wherein the network-assistance information sent to the UE includes the other ML model for the UE to switch from the ML model to the other ML model to determine the estimated TA value.

[0195] Clause 59. The apparatus of any of clauses 56 to 58, wherein the at least oneprocessing circuitry is configured to execute the instructions to cause the apparatus to further at least: receive the ML model from the UE; and retrain or further train the ML model to produce an updated ML model, and wherein the network-assistance information sent to the UE includes the updated ML model for the UE to switch from the ML model to the updated ML model to determine the estimated TA value.

[0196] Clause 60. The apparatus of any of clauses 54 to 59, wherein the handovercommand message sent to the UE further includes at least one sample input for application by the UE to the ML model to determine an output value and a confidence level for the output value, and wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further at least: receive the output value and the confidence level for the output value from the UE; and perform a calibration of the confidence level threshold based on a comparison of the output value and a ground truth value for the sample input, and the confidence level for the output value, and wherein the indication of the confidence level threshold is sent to the UE based on the calibration.

[0197] Clause 61. An apparatus implemented by a radio access node providing asource cell, the apparatus comprising: means for receiving an indication from an userequipment (UE) served by the source cell that the UE supports a machine learning (ML)- based timing advance (TA) acquisition scheme; means for initiating a handover preparation for a candidate cell during which an indication that the handover preparation is for a random access channel (RACH)-less handover is sent to the candidate cell; means for receiving, from the candidate cell, an indication of a confidence level threshold associated with the RACH-less handover; and means for sending a handover command message and an indication of the confidence level threshold to the UE that cause the UEto carry out the RACH-less handover to the candidate cell when a handover condition is fulfilled and a confidence level for an estimated TA value determined by the UE according to the ML-based TA acquisition scheme is at or above the confidence level threshold.

[0198] Clause 62. The apparatus of clause 61, wherein the ML-based TA acquisitionscheme includes an application by the UE of at least one input including a TA value for the source cell to a ML model to determine the estimated TA value, and wherein the handover command message sent to the UE includes an indication of a selection of the ML model.

[0199] Clause 63. The apparatus of clause 62, wherein the apparatus furthercomprises means for receiving the indication of the selection of the ML model from the candidate cell.

[0200] Clause 64. The apparatus of clause 62 or clause 63, wherein the apparatusfurther comprises means for sending network-assistance information to the UE for updating the ML model.

[0201] Clause 65. The apparatus of clause 64, wherein the network-assistanceinformation sent to the UE includes a training dataset for application by the UE to the ML model to retrain or further train the ML model to determine the estimated TA value.

[0202] Clause 66. The apparatus of clause 64 or clause 65, wherein the apparatusfurther comprises means for receiving a request from the UE for another ML model trained to determine the estimated TA value, and wherein the network-assistance information sent to the UE includes the other ML model for the UE to switch from the ML model to the other ML model to determine the estimated TA value.

[0203] Clause 67. The apparatus of any of clauses 64 to 66, wherein the apparatusfurther comprises: means for receiving the ML model from the UE; and means for retraining or further training the ML model to produce an updated ML model, and wherein the network-assistance information sent to the UE includes the updated ML model for the UE to switch from the ML model to the updated ML model to determine the estimated TA value.

[0204] Clause 68. The apparatus of any of clauses 62 to 67, wherein the handovercommand message sent to the UE further includes at least one sample input forapplication by the UE to the ML model to determine an output value and a confidence level for the output value, and wherein the apparatus further comprises: means for receiving the output value and the confidence level for the output value from the UE; and means for performing a calibration of the confidence level threshold based on a comparison of the output value and a ground truth value for the sample input, and the confidence level for the output value, and wherein the indication of the confidence level threshold is sent to the UE based on the calibration.

[0205] Clause 69. A method performed by a radio access node providing a sourcecell, the method comprising: receiving an indication from a user equipment (UE) servedby the source cell that the UE supports a machine learning (ML)-based timing advance (TA) acquisition scheme; initiating a handover preparation for a candidate cell during which an indication that the handover preparation is for a random access channel (RACH)-less handover is sent to the candidate cell; receiving, from the candidate cell, an indication of a confidence level threshold associated with the RACH-less handover; and sending a handover command message and an indication of the confidence level threshold to the UE that cause the UE to carry out the RACH-less handover to the candidate cell when a handover condition is fulfilled and a confidence level for an estimated TA value determined by the UE according to the ML-based TA acquisition scheme is at or above the confidence level threshold.

[0206] Clause 70. The method of clause 69, wherein the ML-based TA acquisitionscheme includes an application by the UE of at least one input including a TA value for the source cell to a ML model to determine the estimated TA value, and wherein the handover command message sent to the UE includes an indication of a selection of the ML model.

[0207] Clause 71. The method of clause 70, wherein the method further comprisesreceiving the indication of the selection of the ML model from the candidate cell.

[0208] Clause 72. The method of clause 70 or clause 71, wherein the method furthercomprises sending network-assistance information to the UE for updating the ML model.

[0209] Clause 73. The method of clause 72, wherein the network-assistanceinformation sent to the UE includes a training dataset for application by the UE to the ML model to retrain or further train the ML model to determine the estimated TA value.

[0210] Clause 74. The method of clause 72 or clause 73, wherein the method furthercomprises receiving a request from the UE for another ML model trained to determine the estimated TA value, and wherein the network-assistance information sent to the UEincludes the other ML model for the UE to switch from the ML model to the other MLmodel to determine the estimated TA value.

[0211] Clause 75. The method of any of clauses 72 to 74, wherein the method furthercomprises: receiving the ML model from the UE; and retraining or further training the ML model to produce an updated ML model, and wherein the network-assistance information sent to the UE includes the updated ML model for the UE to switch from the ML model to the updated ML model to determine the estimated TA value.

[0212] Clause 76. The method of any of clauses 70 to 75, wherein the handovercommand message sent to the UE further includes at least one sample input for application by the UE to the ML model to determine an output value and a confidence level for the output value, and wherein the method further comprises: receiving the output value and the confidence level for the output value from the UE; and performing a calibration of the confidence level threshold based on a comparison of the output value and a ground truth value for the sample input, and the confidence level for the output value, and wherein the indication of the confidence level threshold is sent to the UE based on the calibration.

[0213] Clause 77. A computer-readable storage medium that is non-transitory and hasinstructions stored therein that, in response to execution by at least one processing circuitry, causes a radio access node providing a source cell to at least: receive anindication from a user equipment (UE) served by the source cell that the UE supports amachine learning (ML)-based timing advance (TA) acquisition scheme; initiate a handover preparation for a candidate cell during which an indication that the handover preparation is for a random access channel (RACH)-less handover is sent to the candidate cell; receive, from the candidate cell, an indication of a confidence level threshold associated with the RACH-less handover; and send a handover command message and an indication of the confidence level threshold to the UE that cause the UE to carry out the RACH-less handover to the candidate cell when a handover condition is fulfilled and aconfidence level for an estimated TA value determined by the UE according to the ML- based TA acquisition scheme is at or above the confidence level threshold.

[0214] Clause 78. The computer-readable storage medium of clause 77, wherein theML-based TA acquisition scheme includes an application by the UE of at least one input including a TA value for the source cell to a ML model to determine the estimated TA value, and wherein the handover command message sent to the UE includes an indication of a selection of the ML model.

[0215] Clause 79. The computer-readable storage medium of clause 78, wherein thecomputer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the radio access node to further receive the indication of the selection of the ML model from the candidate cell.

[0216] Clause 80. The computer-readable storage medium of clause 78 or clause 79,wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the radio access node to further send network-assistance information to the UE for updating the ML model.

[0217] Clause 81. The computer-readable storage medium of clause 80, wherein thenetwork-assistance information sent to the UE includes a training dataset for application by the UE to the ML model to retrain or further train the ML model to determine the estimated TA value.

[0218] Clause 82. The computer-readable storage medium of clause 80 or clause 81,wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the radio access node to further receive a request from the UE for another ML model trained to determine the estimated TA value, and wherein the network-assistance information sent to the UE includes the other ML model for the UE to switch from the ML model to the other ML model to determine the estimated TA value.

[0219] Clause 83. The computer-readable storage medium of any of clauses 80 to 82,wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the radioaccess node to further at least: receive the ML model from the UE; and retrain or further train the ML model to produce an updated ML model, and wherein the network- assistance information sent to the UE includes the updated ML model for the UE to switch from the ML model to the updated ML model to determine the estimated TA value.

[0220] Clause 84. The computer-readable storage medium of any of clauses 78 to 83,wherein the handover command message sent to the UE further includes at least one sample input for application by the UE to the ML model to determine an output value and a confidence level for the output value, and wherein the computer-readable storage medium has further instructions stored therein that, in response to execution by the at least one processing circuitry, causes the radio access node to further at least: receive the output value and the confidence level for the output value from the UE; and perform a calibration of the confidence level threshold based on a comparison of the output value and a ground truth value for the sample input, and the confidence level for the output value, and wherein the indication of the confidence level threshold is sent to the UE based on the calibration.

[0221] Clause 85. An apparatus comprising means for performing the method of anyof clauses 69 to 76.

[0222] Clause 86. A computer-readable medium comprising instructions that, inresponse to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 69 to 76.

[0223] Clause 87. A computer-readable storage medium comprising instructions that,in response to execution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 69 to 76.

[0224] Clause 88. A computer program comprising instructions that, in response toexecution by at least one processing circuitry, causes an apparatus to perform the method of any of clauses 69 to 76.

[0225] Many modifications and other implementations of the disclosure set forthherein will come to mind to one skilled in the art to which the disclosure pertains having the benefit of the teachings presented in the foregoing description and the associated figures. Therefore, it is to be understood that the disclosure is not to be limited to thespecific implementations disclosed and that modifications and other implementations are intended to be included within the scope of the appended claims. Moreover, although the foregoing description and the associated figures describe example implementations in the context of certain example combinations of elements and / or functions, it should be appreciated that different combinations of elements and / or functions may be provided by alternative implementations without departing from the scope of the appended claims. In this regard, for example, different combinations of elements and / or functions than thoseexplicitly described above are also contemplated as may be set forth in some of theappended claims. Although specific terms are employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation.

Claims

WHAT IS CLAIMED IS:

1. An apparatus implemented by a user equipment (UE) served by a sourcecell, the apparatus comprising: at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: determine an estimated timing advance (TA) value for a candidate cell, and a confidence level for the estimated TA value, according to a machine learning (ML)-based TA acquisition scheme; make a determination that a handover condition is fulfilled for the candidate cell, and the confidence level for the estimated TA value is at or above a confidence level threshold; and based on the determination, apply a configuration of the candidate cell as a target cell for a handover; andcarry out the handover as a random access channel (RACH)-less handover in which a RACH preamble is skipped, and the estimated TA value is used as a TA value for the target cell to access the target cell.

2. The apparatus of claim 1, wherein the at least one processing circuitry isconfigured to execute the instructions to cause the apparatus to further at least: send an indication to the source cell that the UE supports the ML-based TA acquisition scheme; and receive an indication of the confidence level threshold from the source cell basedon the indication that the UE supports the ML-based TA acquisition scheme.

3. The apparatus of claim 1, wherein the apparatus caused to determine theestimated TA value includes the apparatus caused to apply at least one input including aTA value for the source cell to a ML model to determine the estimated TA value.

4. The apparatus of claim 3, wherein the ML model is among one or moreML models available at the UE, and the at least one processing circuitry is configured toexecute the instructions to cause the apparatus to further receive an indication of a selection of the ML model from the source cell.

5. The apparatus of claim 3, wherein the at least one processing circuitry isconfigured to execute the instructions to cause the apparatus to further at least: receive network-assistance information from the source cell; and update the ML model based on the network assistance information.

6. The apparatus of claim 5, wherein the network-assistance informationreceived from the source cell includes a training dataset, and the apparatus caused to update the ML model includes the apparatus caused to apply the training dataset to the ML model to retrain or further train the ML model to determine the estimated TA value.

7. The apparatus of claim 5, wherein the at least one processing circuitry isconfigured to execute the instructions to cause the apparatus to further send a request to the source cell for another ML model trained to determine the estimated TA value, and wherein the network-assistance information received from the source cell includes the other ML model, and the apparatus caused to update the ML model includes the apparatus caused to switch from the ML model to the other ML model to determine theestimated TA value.

8. The apparatus of claim 5, wherein the at least one processing circuitry isconfigured to execute the instructions to cause the apparatus to further send the ML model to the source cell for retraining or further training to produce an updated ML model, and wherein the network-assistance information received from the source cell includes the updated ML model, and the apparatus caused to update the ML model includes the apparatus caused to switch from the ML model to the updated ML model to determine the estimated TA value.

9. The apparatus of claim 3, wherein the at least one processing circuitry isconfigured to execute the instructions to cause the apparatus to further at least: receive at least one sample input from the source cell; apply the at least one sample input to the ML model to determine an output value and a confidence level for the output value; send the output value and the confidence level for the output value to the source cell for calibration of the confidence level threshold based on a comparison of the output value and a ground truth value for the sample input, and the confidence level for the output value; and receive an indication of the confidence level threshold from the source cell based on the calibration.

10. The apparatus of claim 1, wherein the determination that the handovercondition is fulfilled is made based on an evaluation of the estimated TA value and theconfidence level.

11. The apparatus of claim 1, wherein the determination that the handovercondition is fulfilled is made based on an evaluation of the handover condition for one ormore candidate cells, and wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further determine a trigger event for the evaluation of the handover condition is fulfilled based on the estimated TA value and the confidence level.

12. The apparatus of claim 1, wherein the target cell is a new serving cell forthe UE after the RACH-less handover, and the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further maintain the TA value for the new serving cell according to the ML-based TA acquisition scheme.

13. An apparatus implemented by a radio access node providing a source cell,the apparatus comprising:at least one memory configured to store instructions; and at least one processing circuitry configured to access the at least one memory, and execute the instructions to cause the apparatus to at least: receive an indication from a user equipment (UE) served by the source cell thatthe UE supports a machine learning (ML)-based timing advance (TA) acquisition scheme; initiate a handover preparation for a candidate cell during which an indication that the handover preparation is for a random access channel (RACH)-less handover is sent to the candidate cell; receive, from the candidate cell, an indication of a confidence level threshold associated with the RACH-less handover; and send a handover command message and an indication of the confidence level threshold to the UE that cause the UE to carry out the RACH-less handover to the candidate cell when a handover condition is fulfilled and a confidence level for an estimated TA value determined by the UE according to the ML-based TA acquisition scheme is at or above the confidence level threshold.

14. The apparatus of claim 13, wherein the ML-based TA acquisition schemeincludes an application by the UE of at least one input including a TA value for the source cell to a ML model to determine the estimated TA value, and wherein the handover command message sent to the UE includes an indication of a selection of the ML model.

15. The apparatus of claim 14, wherein the at least one processing circuitry isconfigured to execute the instructions to cause the apparatus to further receive the indication of the selection of the ML model from the candidate cell.

16. The apparatus of claim 14, wherein the at least one processing circuitry isconfigured to execute the instructions to cause the apparatus to further send network- assistance information to the UE for updating the ML model.

17. The apparatus of claim 16, wherein the network-assistance informationsent to the UE includes a training dataset for application by the UE to the ML model to retrain or further train the ML model to determine the estimated TA value.

18. The apparatus of claim 16, wherein the at least one processing circuitry isconfigured to execute the instructions to cause the apparatus to further receive a request from the UE for another ML model trained to determine the estimated TA value, and wherein the network-assistance information sent to the UE includes the other ML model for the UE to switch from the ML model to the other ML model to determine the estimated TA value.

19. The apparatus of claim 16, wherein the at least one processing circuitry isconfigured to execute the instructions to cause the apparatus to further at least: receive the ML model from the UE; and retrain or further train the ML model to produce an updated ML model, and wherein the network-assistance information sent to the UE includes the updated ML model for the UE to switch from the ML model to the updated ML model to determine the estimated TA value.

20. The apparatus of claim 14, wherein the handover command message sentto the UE further includes at least one sample input for application by the UE to the ML model to determine an output value and a confidence level for the output value, and wherein the at least one processing circuitry is configured to execute the instructions to cause the apparatus to further at least: receive the output value and the confidence level for the output value from the UE; and perform a calibration of the confidence level threshold based on a comparison of the output value and a ground truth value for the sample input, and the confidence level for the output value, and wherein the indication of the confidence level threshold is sent to the UE based on the calibration.

Images