ML model performance monitoring host switching

Abstract

A method for performance monitoring of artificial intelligence (AI) or machine learning (ML) (AI / ML) models includes using, by a UE, a first AI / ML model. The UE receives a configuration for starting monitoring of performance of the used first model, wherein the request includes performance monitoring criteria associated with the used model, a triggering criterion for indicating a switch of triggering of AI / ML model performance monitoring, wherein the UE continues to use the first AI / ML model, and a triggering criterion for indicating a switch of triggering of AI / ML model performance monitoring from the first device to the UE. The UE determines whether to start at least a portion of the configured AI / ML model performance monitoring, and determines whether at least one triggering criterion is satisfied, and transmits a request to the first device to switch at least a portion of the model performance monitoring while continuing to use the determined AI / ML model.

Description

Technical Field

[0001] The various example implementations generally relate to wireless networks, and more specifically, to host switching for machine learning (ML) performance monitoring. Background Technology

[0002] Machine learning (ML) model or function performance monitoring can be performed at the user equipment (UE) or network. Performing model or function performance monitoring on the UE side can be useful when the UE needs to perform actions based on model inference or function output, and the ML model or function performs a UE-centric function (e.g., beam management).

[0003] However, there may be situations where UE-side monitoring is less reliable than network-side monitoring. Summary of the Invention

[0004] In one aspect of this disclosure, a method for performance monitoring of an artificial intelligence (AI) or machine learning (ML) AI / ML model, performed by a user equipment (UE) connected to a radio access network (RAN), the method comprising the UE using a first AI / ML model, receiving from a network node a configuration for initiating performance monitoring of the first AI / ML model being used, wherein the request includes: performance monitoring criteria associated with the model being used, trigger criteria for indicating a switch of AI / ML model performance monitoring from the UE to a first device (where the UE continues to use the first AI / ML model), and trigger criteria for indicating a switch of AI / ML model performance monitoring from the first device to the UE, the UE determining whether to initiate at least a portion of the configured AI / ML model performance monitoring, the UE determining whether at least one trigger criterion is met, and in response to at least one trigger criterion being met, the UE transmitting a request to the first device to switch at least a portion of the configured AI / ML model performance monitoring from the UE to the first device while continuing to use the determined AI / ML model.

[0005] In one aspect of this method, upon receiving confirmation from the network node, the UE suspends the execution of model or function performance monitoring.

[0006] In one aspect of the method, the method further includes the UE continuing to use the model associated with the model or functional performance monitoring.

[0007] In one aspect of this method, the UE switches to the fallback function upon receiving a negative acknowledgment from the network node.

[0008] In one aspect of this method, the AI / ML model is an ML AI / ML function, and the performance monitoring of the AI / ML model is the performance monitoring of the AI / ML function.

[0009] In one aspect of this approach, the triggering criteria for interrupting AI / ML model performance monitoring include the UE's lack of resources for performing monitoring and lifecycle management (LCM) decisions.

[0010] In one aspect of this method, the triggering criteria for interrupting AI / ML model performance monitoring include the lack of options for model switching by the UE.

[0011] In one aspect of this method, the triggering criteria for interrupting AI / ML model performance monitoring include the AI / ML model performance for beam selection being less accurate than the non-ML model performance for beam selection.

[0012] In one aspect of this method, the triggering criteria for interrupting AI / ML model performance monitoring include downlink channel conditions falling below a threshold.

[0013] In one aspect of this method, the triggering criteria for interrupting AI / ML model performance monitoring include the fact that the performance of the ML model used to estimate the location differs from that of the Global Navigation Satellite System (GNSS) location.

[0014] In one aspect of this method, the triggering criteria for interrupting AI / ML model performance monitoring include the UE's inability to activate the AI / ML model.

[0015] In one aspect of the method, the method also includes the UE determining whether to restart the performance monitoring of the configured AI / ML model.

[0016] In one aspect of the method, the method also includes transmitting a readiness message from the UE to the network node, the readiness message instructing the UE to resume the configured AI / ML model performance monitoring capability.

[0017] In one aspect of this method, upon receiving confirmation from the network node, the UE restarts the performance monitoring of the configured AI / ML model.

[0018] In one aspect of this method, AI / ML model performance monitoring is interrupted by network nodes.

[0019] In one aspect of this disclosure, a UE includes at least one processor and at least one memory storing instructions that, when executed by the at least one processor, cause the UE to perform at least any of the aforementioned methods.

[0020] In one aspect of this disclosure, a processor-readable medium stores instructions that, when executed by at least one processor of the device, cause the device to perform at least any of the aforementioned methods.

[0021] The independent claims provide the subject matter for several aspects. Additional aspects are defined in the dependent claims. Attached Figure Description

[0022] Some exemplary embodiments will now be described with reference to the accompanying drawings.

[0023] Figure 1 This is a diagram of an example embodiment of wireless networking between a network system and a user equipment (UE) according to an illustrative aspect of this disclosure; Figure 2 This is a diagram illustrating an example component of a network system according to an aspect of this disclosure; Figure 3 This is a diagram of an example embodiment of signaling and operation between a network (NW) node and a UE for UE-triggered model monitoring handover, according to one aspect of this disclosure; Figure 4 This is a diagram illustrating an example embodiment of signals and operations between an NW node and a UE for NW-triggered model monitoring of handover, based on an aspect of this disclosure; and Figure 5 This is an illustration of an example embodiment of a component of a UE or network device according to one aspect of this disclosure. Detailed Implementation

[0024] In the following description, certain specific details are set forth in order to provide a thorough understanding of the disclosed aspects. However, those skilled in the art will recognize that the aspects can be practiced without one or more of these specific details or using other methods, components, materials, etc. In other instances, well-known structures associated with transmitters, receivers, or transceivers have not been shown or described in detail to avoid unnecessarily obscuring the description of the aspects.

[0025] Throughout this specification, the reference to "an aspect" or "aspect" means that a particular feature, structure, or characteristic described in connection with that aspect is included in at least one aspect. Therefore, the phrases "in an aspect" or "in one respect" appearing in various places throughout this specification do not necessarily refer to the same aspect. Furthermore, a particular feature, structure, or characteristic may be combined in one or more aspects in any suitable manner.

[0026] The embodiments described in this disclosure can be implemented in wireless networking devices, such as, but not limited to, devices utilizing Global Microwave Access Interoperability (WiMAX), Global System for Mobile Communications (GSM, 2G), GSM EDGE Radio Access Network (GERAN), General Packet Radio Service (GRPS), Universal Mobile Telecommunications System based on Basic Wideband Code Division Multiple Access (W-CDMA) (UMTS, 3G), High-Speed ​​Packet Access (HSPA), Long Term Evolution (LTE), Advanced Long Term Evolution, Enhanced LTE (eLTE), 5G New Radio (5GNR), 5G Advanced, 6G (and above), and 802.11ax (Wi-Fi 6), and other wireless networking systems. The term "eLTE" here refers to LTE evolution connected to a 5G core network. LTE is also referred to as Evolved UMTS Terrestrial Radio Access (EUTRA) or Evolved UMTS Terrestrial Radio Access Network (EUTRAN).

[0027] This disclosure may use the term "serving network device" to refer to a network node or network device (or part thereof) serving a UE. As used herein, the terms "transmit to," "receive from," and "cooperate with" (and variations thereof) include communication that may or may not involve communication through one or more intermediate devices or nodes. The term "acquire" (and variations thereof) includes acquiring in a first instance or re-acquiring after a first instance. The term "connection" may mean a physical connection or a logical connection.

[0028] This disclosure uses 5G NR as an example of a wireless network, and may use smartphones and / or extended reality headsets as examples of UEs. It is intended and should be understood that such examples are merely illustrative, and this disclosure applies to other wireless networks and user equipment.

[0029] Figure 1 This is a diagram illustrating an example of wireless networking between network system 100 and user equipment (UE) 150. Network system 100 may include one or more network nodes 120, one or more servers 110, and / or one or more network devices 130 (e.g., test equipment). Network node 120 will be described in more detail below. As used herein, the term "network apparatus" may refer to any component of network system 100, such as server 110, network node 120, network device 130, any of the foregoing components(s), and / or any other components(s) of network system 100. Examples of network apparatuses include, but are not limited to, apparatuses for implementing various aspects of 5G NR. This disclosure describes embodiments related to 5G NR and embodiments relating to aspects defined by the 3rd Generation Partnership Project (3GPP). However, it is contemplated that embodiments related to other wireless network technologies are included within the scope of this disclosure.

[0030] The following description provides further details of examples of network nodes. In a 5G NR network, a gNodeB (also known as a gNB) may include, for example, a node that provides New Radio (NR) user plane and control plane protocol termination for the UE and is connected to the 5G core network (5GC) via an NG interface, such as, according to Section 3.2 of 3GPP TS 38.300 V16.6.0 (2021-06), which is incorporated herein by reference.

[0031] gNB supports various protocol layers, such as Layer 1 (L1) physical layer, Layer 2 (L2) and Layer 3 (L3).

[0032] NR's Layer 2 (L2) is divided into the following sublayers: Media Access Control (MAC), Radio Link Control (RLC), Packet Data Convergence Protocol (PDCP), and Service Data Adaptation Protocol (SDAP), among which, for example: The physical layer provides a transmission channel to the MAC sublayer; The MAC sublayer provides logical channels to the RLC sublayer; The RLC sublayer provides RLC channels to the PDCP sublayer; The PDCP sublayer provides radio bearers to the SDAP sublayer; The SDAP sublayer provides Quality of Service (QoS) flows to 5GC; The control channels include the Broadcast Control Channel (BCCH) and the Physical Control Channel (PCCH).

[0033] Layer 3 (L3) includes, for example, Radio Resource Control (RRC) according to Section 6 of 3GPP TS 38.300 V16.6.0 (2021-06), which is incorporated herein by reference.

[0034] A gNB centralized unit (gNB-CU) comprises, for example, a logical node that hosts, the radio resource control (RRC), service data adaptation protocol (SDAP), and packet data convergence protocol (PDCP) of a gNB, or the RRC and PDCP protocols of an en-gNB, and controls the operation of one or more gNB distributed units (gNB-DUs). The gNB-CU terminates the F1 interface connected to the gNB-DU. The gNB-CU may also be referred to herein as a CU, centralized unit, centralized control unit, or control unit.

[0035] A gNB Distribution Unit (gNB-DU) comprises, for example, a logical node that hosts the Radio Link Control (RLC), Media Access Control (MAC), and Physical (PHY) layers of a gNB or en-gNB, and its operation is partially controlled by the gNB-CU. A gNB-DU supports one or more cells. A cell is supported by only one gNB-DU. The gNB-DU terminates the F1 interface connected to the gNB-CU. The gNB-DU may also be referred to herein as a DU or Distribution Unit.

[0036] As used herein, the term "network node" may refer to any one or any combination of gNB, gNB-CU, or gNB-DU. RAN (Radio Access Network) nodes or network nodes (such as, for example, gNB, gNB-CU, or gNB-DU, or portions thereof) may be implemented using means, for example, having at least one processor and / or at least one memory having processor-readable instructions ("programs") configured to support and / or provision and / or process CU and / or DU related functions and / or features, and / or at least one protocol (sub)layer (e.g., layer 2 and / or layer 3) of the RAN (Radio Access Network). Different functional splits between centralized and distributed units are possible. The following will combine... Figure 5 Examples describing such devices and components.

[0037] The gNB-CU and gNB-DU portions can be co-located or physically separated, for example. The gNB-DU can even be further divided into, for example, two parts, one including processing equipment and the other including an antenna. The centralized unit (CU) can also be referred to as a baseband unit / radio equipment controller / cloud RAN / virtual RAN (BBU / REC / C-RAN / V-RAN), open RAN (O-RAN), or a portion thereof. The distributed unit (DU) can also be referred to as a remote radio head / remote radio unit / radio equipment / radio unit (RRH / RRU / RE / RU), or a portion thereof. In the various exemplary embodiments of this disclosure, a network node supporting at least one of the centralized unit functions or Layer 3 protocols of a radio access network can be, for example, a gNB-CU. Similarly, a network node supporting at least one of the distributed unit functions or Layer 2 protocols of a radio access network can be, for example, a gNB-DU.

[0038] A gNB-CU can support one or more gNB-DUs. A gNB-DU can support one or more cells, and therefore can support serving cells for user equipment (UE), or candidate cells for handover, dual connectivity and / or carrier aggregation and other procedures.

[0039] User equipment (UE) 150 may be or include wireless or mobile devices, devices having a radio interface for interacting with a RAN (Radio Access Network), smartphones, in-vehicle devices, IoT devices, or M2M devices, and other types of user equipment. Such a UE 150 may include: at least one processor; and at least one memory including program code; wherein the at least one memory and the computer program code are configured, together with the at least one processor, to cause the device to perform at least certain operations, such as, for example, an RRC connection to the RAN. Figure 5 Examples of components of the UE are described below. In an embodiment, UE 150 may be configured to generate messages (e.g., including a cell ID) to be transmitted to the RAN via radio (e.g., to reach and communicate with the serving cell). In an embodiment, UE 150 may generate, transmit, and receive RRC messages containing one or more RRC PDUs (Packet Data Units). Those skilled in the art will understand the RRC protocol and other processes that the UE may perform.

[0040] Continue to refer to Figure 1 In an example of a 5G NR network, network system 100 provides one or more cells that define the coverage area of ​​network system 100. As described above, network system 100 may include a gNB of the 5G NR network, or may include any other means configured to control radio communications and manage radio resources within the cells. As used herein, the term "resource" may refer to radio resources such as resource blocks (RBs), physical resource blocks (PRBs), radio frames, subframes, time slots, subbands, frequency regions, subcarriers, beams, etc. In embodiments, network node 120 may be referred to as a base station.

[0041] Figure 1 Examples are provided, and only network system 100 and UE 150 are illustrated. Those skilled in the art will understand that network system 100 includes... Figure 1 Components not shown in the diagram, and it will be understood that other user equipment can communicate with network system 100.

[0042] Figure 2 yes Figure 1 A block diagram of example components of network system 100. A 5G NR network can be described as an example of network system 100, and the aspects described below should also be applicable to other types of network systems. The network system can be configured according to... Figure 1The signals and connections shown operate to enable UE 150 to communicate with network system 100 via radio access network 225. Furthermore, the network system can be divided into user plane components and functions and control plane components and functions, as shown and described herein. Unless otherwise stated, the terms “component,” “function,” and “service” are used interchangeably herein and can refer to instructions executed by and implemented by one or more processors.

[0043] The following describes example functionality of the components. This example functionality is merely illustrative, and it should be understood that additional operations and functions can be performed by the components described herein. Furthermore, connections between components can be virtual connections based on service interfaces, allowing any component to communicate with any other component. In this way, any component can act as a service "producer," providing network functionality services to any other component acting as a service "consumer."

[0044] For example, a core network 210 is described in the control plane of the network system. The core network 210 may include an Authentication Server Function (AUSF) 211, an Access and Mobility Function (AMF) 212, and a Session Management Function (SMF) 213. The core network 210 may also include a Network Slice Selection Function (NSSF) 214, a Network Exposure Function (NEF) 215, a Network Repository Function (NRF) 216, and a Unified Data Management Function (UDM) 217, which may include a Unified Data Repository (UDR) 224.

[0045] Additional components and functions of the core network 210 may include application functions 218, policy control functions (PCF) 219, network data analysis functions (NWDAF) 220, analytical data repository functions (ADRF) 221, management data analysis functions (MDAF) 222, and operation and management functions (OAM) 223.

[0046] The user plane includes UE 150, Radio Access Network (RAN) 225, User Plane Functions (UPF) 226, and Data Network (DN) 227. RAN 225 may include a combination of Figure 1 The RAN 225 describes one or more components, such as one or more network nodes. However, the RAN 225 may not be limited to such components. The UPF 226 provides connectivity for data transmitted over the RAN 225. For example, DN226 identifies services from service providers, internet access, and third-party services.

[0047] AMF 212 handles connectivity and mobility tasks. AUSF 211 receives authentication requests from AMF 212 and interacts with UDM 217 to authenticate and verify network responses to determine successful authentication. SMF 213 performs Packet Data Unit (PDU) session management and manages session context with UPF 226.

[0048] NSSF 214 can select a Network Slice Instance (NSI) and determine the allowed Network Slice Selection Auxiliary Information (NSSAI). This selection and determination are used to set up AMF 212 to provide services to UE 150. NEF 215 ensures third-party access to network services to create private network services. NRF 216 acts as a repository for storing network functions to allow functions to register and discover each other.

[0049] UDM 217 generates authentication vectors for use by AUSF 211 and ADM 212 and provides user identity processing. UDM 217 can connect to UDR 224, which stores data associated with authentication, applications, etc. AF 218 provides application services (e.g., streaming media services) to users. PCF 219 provides policy control functions. For example, PCF 219 can assist in network slicing and mobility management, as well as provide Quality of Service (QoS) and accounting functions.

[0050] NWDAF 220 collects data (e.g., from UE 150 and network systems) to perform network analytics and provide insights into the capabilities that leverage analytics when providing services. ADRF 221 allows consumers to store, retrieve, and remove data and analytics. MDAF 222 provides additional data analytics services for network functions. OAM 223 provides provisioning and management processing capabilities to manage elements in or connected to the network (e.g., UE 150, network nodes, etc.).

[0051] Figure 2 These are merely examples of components of a network system, and variations are contemplated within the scope of this disclosure. In embodiments, the network system may include... Figure 2 Other components not shown. In embodiments, the network system may not include... Figure 2 Each component is shown. In an embodiment, components and connections can be used with... Figure 2 The connections shown are implemented using different connections. These and other embodiments are considered to be within the scope of this disclosure.

[0052] Although further details will be provided below, this document describes a method for a UE to notify a network node of its intention to suspend, interrupt, or stop performance monitoring of an artificial intelligence (AI) / machine learning (ML) model or function, or for the network to signal to the UE to suspend its ML model or function performance management, wherein the function is enabled by an ML model, and wherein ML model or function performance monitoring can be used synonymously with ML performance monitoring. In various embodiments, the network performs model performance monitoring over a period of time, and the UE can request to start model performance monitoring when ready, and restart model performance monitoring when the network allows it to make such a decision. In the various embodiments described herein, ML model or function can refer to AI / ML model or function, and these terms are used interchangeably herein.

[0053] As mentioned above, lifecycle management (LCM) for ML model or function performance monitoring can be performed at the UE or network (e.g., network node). Performing model or function performance monitoring on the UE side may be useful when the UE needs to perform actions based on model inference or function output, and the ML model or function performs a UE-centric function (e.g., beam management).

[0054] However, as mentioned above, there may be situations where the UE is less reliable than the network in terms of ML model or function performance monitoring. For example, if the UE is configured to report monitoring results to the network, it may lack the resources (power, computing resources) for performance monitoring and reporting of monitoring results. In addition to model performance monitoring of active models, the model or function management framework also includes offline performance monitoring of inactive models, which may consume some energy. For example, if the UE's battery is low, expending additional energy on ML model performance monitoring tasks may be impractical. In various embodiments, the UE has the option to completely abandon the use of the ML model or function and switch to fallback functionality, but the UE may have an alternative option to stop performing ML model or function performance monitoring while still continuing to use the ML model or function.

[0055] In various embodiments, the UE may not be able to identify the best model or function to use or switch to (e.g., the best performing model or function), and the network may be more suitable for temporarily performing LCM.

[0056] When model or functional performance monitoring is performed on the UE side, the UE is responsible for LCM decisions. However, the UE notifies the network of its decisions / actions in some form. The network can periodically analyze these decisions and observe discrepancies in the LCM decisions made by the UE. Due to the discrepancy between the performance expected by the network and the results of the UE's ML model or functional performance monitoring, the network may lack confidence in the model or functional performance monitoring on the UE side.

[0057] In various embodiments, for beam management, for example when many UEs are configured to use beam management models and / or functions, network nodes (e.g., gNodeBs) may be better suited to monitor beam management models and / or functions because they can use the outputs of many UEs to perform performance monitoring, resulting in improved accuracy and performance of model or function performance monitoring. Additionally, when a UE is configured with many component carriers (CCs) in a carrier aggregation (CA) configuration, or is configured with a high multiple-input multiple-output (MIMO) rank (e.g., the number of MIMO layers), measuring the complete set of beams transmitted by the gNodeB / gNB for performance monitoring purposes can be burdensome. Beam management enhancements provided by the UE-side AI / ML model enable the UE to measure a subset of the complete set of beams transmitted by the gNodeB for beam selection. Therefore, if the gNB performs monitoring using an alternative mechanism (e.g., throughput or MCS selection performance analysis), the UE will be able to support a larger number of CCs.

[0058] In various embodiments, for positioning enhancement, such as when the UE can use a large number of Global Navigation Satellite System (GNSS) satellites, or can use traditional gNB-based positioning (e.g., Enhanced Cell ID Positioning (E-CID)) and / or LMF-based positioning (e.g., New Radio (NR) Time Difference of Arrival (TDOA)) to calculate an accurate positioning estimate, it may be feasible to monitor the performance of the UE-side positioning model or function. If the UE is unable to estimate an accurate location for performance monitoring of its UE-side positioning model or function, network assistance may be utilized.

[0059] Therefore, in various embodiments, when the UE has been configured to make decisions for ML functions, the network can take over model or function performance monitoring and make monitoring decisions.

[0060] In the various embodiments described below, the UE triggers a change in the entity performing model / function performance monitoring. In other various embodiments, the network (e.g., a network node) triggers a change in the model / function performance monitoring decision-making entity.

[0061] When the UE lacks sufficient resources to perform the additional task of ML model / function performance monitoring, or when the UE is unable to activate the correct model from its available options, the UE can trigger a change or handover. When the network deems the UE's LCM decision incorrect / unreliable and wishes to take over the LCM decision-making process, the network can trigger a change or handover.

[0062] As used herein, communication with the Radio Access Network (RAN) can refer to and means communication with a portion of the RAN (such as with network nodes (e.g., DU and / or CU)) or another portion of the RAN. As used herein, communication with the core network can refer to and means communication with one or more services / applications of the core network (such as AMF or another service of the core network).

[0063] As used herein, the terms “first” and “second” can refer to a first or second instance of a message transmitted / received by a component (e.g., a UE, a device, etc.) or by a first or second component in the described sequence of components. Therefore, these terms are used in a non-limiting manner and can refer to any message, operation, device, component, etc.

[0064] Based on the brief description, Figure 3 This is a diagram illustrating an example embodiment of signaling and operation between a network (NW) node and a UE for UE-triggered model monitoring handover, according to one aspect of this disclosure. In various embodiments, Figure 3 The components described above can correspond to those in the above. Figure 1 and 2 Similar components as described herein. In various embodiments, for example, a network node may include a gNB. It should be understood that the described signals may have associated operations, and the described operations may have associated signals.

[0065] At operation 300, the network node transmits a message to configure the UE to perform model or function performance monitoring, and the UE receives a message to configure itself to perform model or function performance monitoring. In various embodiments, the UE is configured to make LCM decisions for ML functions, including model activation / deactivation, model selection / switching, etc. In various embodiments, the message used to configure the UE to perform model or function performance monitoring includes LCM decision triggering and decision reporting configuration (such as when and how information related to the LCM decision is reported).

[0066] At operation 301, the network node transmits a message to the UE to trigger the stopping (e.g., interrupting) and starting of model or function performance monitoring, and the UE receives a message to trigger the UE to trigger the stopping and starting of model or function performance monitoring. Thus, the UE is configured with triggers for starting and stopping LCM decisions. In various embodiments, these triggers are a set of conditions that the UE must meet to perform LCM, and can be described in further detail below. In various embodiments, triggers for starting and stopping LCM decisions may include performance degradation exceeding a threshold. Although it may not be clear whether the poor performance is due to an incorrect model or performance monitoring and decisions based on that monitoring, the network node can take further action to resolve it.

[0067] At operation 302, the UE makes an LCM decision based on the received configuration and measurements. Depending on the configuration settings, the UE may transmit information to network nodes (e.g., gNB / LMF).

[0068] At operation 303, the UE decides to stop making LCM decisions because one of the triggering conditions (e.g., lack of resources for performing monitoring and LCM decisions or lack of options for model switching) is met.

[0069] At operation 304, the UE transmits a UE model or function performance monitoring stop message to the network node, and the network node receives the UE model or function monitoring stop message. In various embodiments, the UE requests the network to take over the LCM decision. In various embodiments, the UE sends a reason code indicating the reason for the request. The request can be provided as an indication in a MAC CE associated with an index provided in an RRC configuration related to the monitoring configuration (including where monitoring is performed); the request can also be provided as UE assistance information (UAI) containing an index similar to the aforementioned index and a reason code; or the request can be provided as part of an RRC reconfiguration complete message when AI / ML functionality is configured or during handover.

[0070] At operation 305, the network node assesses the likelihood of taking over the LCM decision. In various embodiments, the network may not favor such takeover for some ML models or functions (e.g., if the NW has agreed to perform LCM for too many UEs, and using the ML model or function is not particularly critical to the UE). Furthermore, the UE is able to perform measurements and then transmit simple indications about the model's performance. Otherwise, the UE would expend air interface resources to transmit measurements used solely for monitoring.

[0071] At operation 306, the network node transmits an ACK or NACK to the UE, and the UE receives the ACK or NACK. In various embodiments, if the network cannot take over LCM decision-making and monitoring, the network sends a NACK message to the UE, and the UE is configured to switch to a non-ML function. If the network accepts the UE's request, the network begins ML model or function performance monitoring and prepares to make an LCM decision, and sends an ACK message to the UE.

[0072] If the UE receives a NACK at operation 306, then at operation 307a, the UE stops performance monitoring and making LCM decisions, and switches to a non-ML function. Since model or function performance management is a critical part of LCM, the UE is not allowed to run an ML model or function without model or function performance monitoring. In this case, the UE must deactivate the ML function.

[0073] If the UE receives an ACK at Operation 306, then at Operation 307b, the UE stops monitoring and making LCM decisions, but continues to use the ML function. At Operation 307c, the network node begins network-side model / function performance monitoring.

[0074] After a period of time or based on configured triggering conditions, at operation 308, the UE can determine, based on the network settings configuration, that it is ready to perform performance monitoring again and make LCM decisions. As an example of configured triggering conditions, the UE can evaluate model or function performance and make one or more LCM decisions. If these decisions are the same as those made by NW monitoring, the UE can trigger a procedure to request the restart of LCM monitoring.

[0075] Therefore, at operation 309, the UE transmits a UE model or function management ready message to the network node, and the network node receives the UE model or function management ready message. For example, the UE sends an indication message to inform the network of the UE's intention to begin model monitoring and make LCM decisions. In various embodiments, the UE transmits information about its readiness to perform model performance monitoring (e.g., improvements in its battery status, availability of more models that can be used to better match the model to the input, etc.).

[0076] At operation 310, the network node transmits either an ACK or a NACK to the UE, and the UE receives the ACK or NACK. In other words, the network node acknowledges the UE's request. If a NACK is transmitted at operation 310, the UE continues with operation 307a, and the network node continues with operation 307c.

[0077] If an ACK is transmitted at operation 310, then at operation 311a, the network node interrupts model or function performance monitoring, and at operation 311b, the UE starts ML model or function performance monitoring.

[0078] Figure 3 The operations described are merely illustrative and variations are expected within the scope of this disclosure. In embodiments, the operations may include... Figure 3 Other operations not shown. In embodiments, operations may not include... Figure 3 Each operation is shown. In an embodiment, the operation can be in conjunction with... Figure 3 The different orders shown are implemented. These and other embodiments are considered to be within the scope of this disclosure. Those skilled in the art will understand that although the various example components are described as performing various functions, other components may perform... Figure 3 The functions described in [the document].

[0079] Figure 4This is a diagram illustrating an example embodiment of signal and operation between an NW node and a UE for NW-triggered model monitoring of handover, according to one aspect of this disclosure. In various embodiments, Figure 3 The components described above can correspond to those in the above. Figure 1 and 2 Similar components as described herein. In various embodiments, for example, a network node may include a gNB. It should be understood that the described signals may have associated operations, and the described operations may have associated signals.

[0080] At operation 400, the network node transmits a message to configure the UE to perform model or function performance monitoring, and the UE receives a message to configure itself to perform model or function performance monitoring. In various embodiments, the UE is configured to make LCM decisions for ML functions, including model activation / deactivation, model selection / switching, etc. In various embodiments, the messages used to configure the UE to perform model or function performance monitoring include LCM decision triggering and decision reporting configuration (such as when and how information related to the LCM decision is reported).

[0081] At operation 401, the network node transmits a message to configure the UE with a trigger for restarting model or function performance monitoring, and the UE receives a message to configure the UE with a trigger for restarting model or function performance monitoring after stopping it based on a network indication. Therefore, the UE is configured with triggers for restarting LCM decisions. In various embodiments, these triggers are a set of conditions that the UE must meet to perform LCM, and can be described in further detail below. In various embodiments, triggers for starting and stopping LCM decisions may include performance degradation below expectations. Although it may not be clear whether the poor performance is due to an incorrect model or performance monitoring, the network node can take further action to resolve it.

[0082] At operation 402, the UE makes an LCM decision based on the received configuration and measurements. Depending on the configuration settings, the UE may transmit information to network nodes (e.g., gNB / LMF).

[0083] At operation 403, the network node assesses that the UE's LCM decision does not meet performance requirements, and the network node has low confidence in the UE's LCM decision. In various embodiments, the assessment can be triggered by the quality of LCM decisions made by other UEs under the same conditions. For example, a UE is reporting the performance of a function, and this performance is inconsistent with the performance of other UEs using the same ML function under the same conditions. This could be due to measurement errors in performance monitoring. Such conditions trigger a process that switches the LCM decision-making party from the UE to the NW. In various embodiments, the network node can receive reports from the UE.

[0084] At operation 404, the network node transmits a UE model or function performance monitoring stop message to the network node, and the network node receives the UE model or function monitoring stop message. In various embodiments, the network node sends an indication message to the UE for handover LCM decision. In various embodiments, the network node transmits parameters / codes specifying the reason for the indication. This request can be provided as an indication in a MAC CE associated with an index provided in an RRC configuration related to the monitoring configuration (including where monitoring is performed), or the request can be provided as part of an RRC reconfiguration complete message when AI / ML functionality is configured or during handover.

[0085] At operation 405, the UE transmits an ACK to the network node, and the network node receives the ACK.

[0086] At operation 406a, the UE stops monitoring and making LCM decisions, but continues to use the ML function. At operation 406b, the network node begins network-side model / function performance monitoring.

[0087] After a period of time or based on configured triggering conditions, at operation 407, the UE can determine, based on the network settings configuration, that it is ready to perform performance monitoring and make LCM decisions again. As an example of configured triggering conditions, the UE can evaluate functional performance and make one or more LCM decisions. If these decisions are the same as those made by NW monitoring, the UE can trigger a request to restart the LCM monitoring process.

[0088] Therefore, at operation 408, the UE transmits a UE model or function management ready message to the network node, and the network node receives the UE model or function management ready message. For example, the UE sends an indication message to inform the network of the UE's intention to begin model monitoring and make LCM decisions. In various embodiments, the UE transmits information about its readiness to perform model performance monitoring (e.g., improvements in its battery status, availability of more models that can be used to better match the model to the input, etc.).

[0089] At operation 409, the network node transmits either an ACK or a NACK to the UE, and the UE receives the ACK or NACK. In other words, the network node acknowledges the UE's request. If a NACK is transmitted at operation 409, the UE continues with operation 306a, and the network node continues with operation 306b.

[0090] If an ACK is transmitted at operation 409, then at operation 410a, the network node interrupts model or function performance monitoring, and at operation 410b, the UE starts model or function performance monitoring.

[0091] Figure 4 The operations described are merely illustrative and variations are expected within the scope of this disclosure. In embodiments, the operations may include... Figure 4 Other operations not shown. In embodiments, operations may not include... Figure 4 Each operation is shown. In an embodiment, the operation can be in conjunction with... Figure 4 The different orders shown are implemented. These and other embodiments are considered to be within the scope of this disclosure. Those skilled in the art will understand that although the various example components are described as performing various functions, other components may perform... Figure 4 The functions described in [the document].

[0092] As described above, in various embodiments, triggering can be used to determine the switching of a model or functional performance monitoring entity (e.g., a UE or network node). In various embodiments, the LCM control switching mechanism triggered by the UE and NW can rely on the configuration of a threshold that causes the UE or NW to request a switch of the entity performing monitoring.

[0093] For example, regarding beam management, the UE measures N_SetB_Beams, where N_SetB_Beams < N_SetA_Beams, and SetB is a subset of SetA. The ML model used to implement ML-enabled beam management functionality takes the measurements of the SetB beams as input and outputs the configured number of best beams N_Best_Beams, which considers the entire set of SetA beams, where at least some beams have not been measured.

[0094] In various embodiments, to monitor the performance of the beam management model, the UE may periodically or as configured by the network, measure not only SetB but also the entire SetA. Using methods for optimal beam determination, the UE can determine that beams other than those predicted based on measurements on SetB are continuously determined. Conversely, if the network is responsible for performance monitoring, the network may request the UE to transmit measurements of the SetA beams.

[0095] During a performance monitoring instance, the UE or network determines the optimal beam based on the UE's performance and according to the configured trigger conditions. Although there is a difference between the optimal beam determined by performance monitoring and the optimal beam determined by the beam management method enabled by ML, the performance still matches expectations based on the following first and second trigger conditions.

[0096] In various embodiments, the ML model or functional output can be compared with the non-ML output. For example, the non-ML output may indicate the order of beam quality in the order of beam A, beam B, beam C, beam D, beam E, beam F, beam G, beam H, while the ML output indicates the order of beam quality in the order of beam E, beam F, beam G, beam H, beam A, beam B, beam C, beam D. In various embodiments, the trigger may be required to exist in the ML enable method output N. BestBeams,ML Example threshold N BestBeams_nonML .

[0097] In various embodiments, triggering can be based on the determination of RSRP or RSRQ measurements. For example, monitoring output: beams in the order of RS measurements (e.g., SSB RSRP or RSRQ on SetA).

[0098] In various embodiments, the ML-enabled methods used for beam management may differ from what is expected by conventional algorithmic approaches. For example, given conventional calculations or measurements of the SetA beam, beam predictions generated by ML-enabled methods may not immediately seem reasonable. For instance, training the ML-enabled methods may include implicit information about beam adjacency, and because overly close beams may be difficult for the UE to distinguish, certain combinations of optimal beams may actually be detrimental to UE performance, even though each of these ideal optimal beams would individually exhibit high performance (e.g., if the UE receives DL transmissions from only a single beam at a time).

[0099] Therefore, in various embodiments, additional triggering conditions can be evaluated after the first triggering condition is met, indicating at least a mismatch between the performance monitoring output and the ML enable method output.

[0100] For example, in various embodiments, an additional trigger may include the order of the modulation scheme (MCS) for downlink reception of the UE or uplink reception of the NW, wherein a potential trigger may include MCS > Threshold. MCS The specific threshold may depend on other factors (e.g., the configured MCS table, the average synchronization block (SSB) RSRP and / or RSRQ or channel state information (CSI) report).

[0101] In various embodiments, additional triggers may include downlink channel conditions (e.g., RSRP, RSRQ, channel state information).

[0102] In various embodiments, additional triggers may include the number of low bit error rate (BER) or negative acknowledgment (NACK) signals being generated by the UE or NW, wherein potential triggers may include BER < Threshold. BER(%) or NACK < Threshold NACK (%).

[0103] In various embodiments, additional triggering may include a high number of simultaneous beams scheduled to the UE, wherein potential triggering may include N DLBeams > Threshold NDLBeams .

[0104] Furthermore, any combination of the above criteria can be used to determine performance. In various embodiments, the criteria can also be reversed to trigger performance below a threshold.

[0105] Network nodes (e.g., gNBs) can pre-provide different trigger files to the UE, and based on the gNB's perception of the UE's performance, may be able to quickly switch the UE using MAC CE to change its trigger.

[0106] In various embodiments, the aforementioned triggering conditions can be configured on the UE. The gNB includes this information and can evaluate any of the triggering conditions independently or in combination with each other without reporting the results to the UE. Additionally, the gNB can use measurements transmitted by the UE or signals measured by the gNB to determine the UE's expected performance, including, for example, reports of SSB RSRP or RSRQ and / or measurements of the sounding reference signal (SRS), which is an uplink transmission of a reference signal to the gNB.

[0107] In various embodiments, the general criteria for requesting the cessation of performance monitoring of the UE or NW may include the UE or NW detecting high performance, but the ML model or functional performance monitoring indicating poor performance of the ML enabling method.

[0108] Regarding the monitoring of positioning performance, the trigger may include the UE using a GNSS positioning method to estimate its location. The trigger for switching monitoring back to the gNB may include the following: during the monitoring period, the UE uses a GNSS positioning method to estimate its location, and the UE compares the GNSS estimate with the prediction of an ML-enabled method, and the outputs of the GNSS and ML-enabled methods differ by more than a threshold, which may be provided (e.g., in meters).

[0109] For example, ML model or functional performance monitoring (e.g., based on preset samples of ground reality (GT) markers for a region) indicates that the model's accuracy reaches 50 cm. However, the UE determines, based on the difference between the GNSS and its output, that the accuracy threshold has not been met, and therefore, the ML model or functional performance monitoring is not functioning correctly. Therefore, once the UE indicates to the gNB that its performance monitoring mechanism is not performing accurately, the gNB can take over the monitoring using its own performance monitoring mechanism, which in various embodiments may utilize conventional positioning methods.

[0110] The following describes the operation from the UE's perspective. From this perspective, a method may include: the UE using a first AI / ML model; receiving from a network node a configuration for initiating performance monitoring of the first AI / ML model used, wherein the request includes performance monitoring criteria associated with the model used, triggering criteria for instructing the switching of AI / ML model performance monitoring from the UE to a first device (where the UE continues to use the first AI / ML model), and triggering criteria for instructing the switching of AI / ML model performance monitoring from the first device to the UE; the UE determining whether to initiate at least a portion of the configured AI / ML model performance monitoring; the UE determining whether at least one of the triggering criteria is met; and in response to the satisfaction of at least one of the triggering criteria, the UE transmitting a request to the first device to switch at least a portion of the configured AI / ML model performance monitoring from the UE to the first device while continuing to use the determined AI / ML model.

[0111] Now for reference Figure 5 This diagram illustrates a block diagram of example components of a UE or network device (e.g., RAN or core network). The device includes electronic storage device 510, processor 520, network interface 540, and memory 550. The various components can be communicatively coupled to each other. Processor 520 can be and may include any type of processor, such as a single-core central processing unit (CPU), multi-core CPU, microprocessor, digital signal processor (DSP), system-on-a-chip (SoC), or any other type of processor. Memory 550 can be a volatile type of memory, such as RAM, or a non-volatile type of memory, such as NAND flash memory. Memory 550 includes processor-readable instructions that can be executed by processor 520 to cause the device to perform various operations, including those mentioned herein, such as... Figure 3 The operation described in A-4B.

[0112] Electronic storage device 510 can be and includes any type of electronic storage device for storing data, such as hard disk drives, solid-state drives, optical disks and / or other non-transitory computer-readable media, as well as other types of electronic storage devices. Electronic storage device 510 stores processor-readable instructions for causing or being configured to cause the device to perform its operations, and also stores data associated with such operations, such as data related to the 5G NR standard and other data. Network interface 540 enables wireless networking technologies, such as 5G NR and / or other wireless networking technologies.

[0113] Figure 5The components shown are merely examples, and those skilled in the art will understand that the apparatus includes other components not shown, and may include multiple instances of any of the shown components. These and other embodiments are considered to be within the scope of this disclosure. For example, a transmitter and a receiver may be included as components for transmitting and receiving signals.

[0114] Other embodiments of this disclosure include the following examples.

[0115] Example 1.1. A UE, comprising: Components used by the UE to employ the first AI / ML model; A component for receiving from a network node a configuration for initiating performance monitoring of a first AI / ML model being used, wherein the request includes performance monitoring criteria associated with the model being used, trigger criteria for indicating a switch of AI / ML model performance monitoring from the UE to the first device, wherein the UE continues to use the first AI / ML model, and trigger criteria for indicating a switch of AI / ML model performance monitoring from the first device to the UE. Components used by the UE to determine whether to start at least a portion of the performance monitoring of the configured AI / ML model; A component used by the UE to determine whether at least one triggering criterion is met; and A component for transmitting a request from the UE to the first device in response to at least one trigger criterion being met, to switch at least a portion of the configured AI / ML model performance monitoring from the UE to the first device while continuing to use the determined AI / ML model.

[0116] Example 1.2. According to the UE described in Example 1.1, the UE suspends the execution of model or function performance monitoring upon receiving an acknowledgment from the network node.

[0117] Example 1.3. The UE according to Example 1.2 further includes: a component for the UE to continue using the model associated with the model or functional performance monitoring.

[0118] Example 1.4. The UE according to Example 1.1, wherein the UE switches to the fallback function upon receiving a negative acknowledgment from the network node.

[0119] Example 1.5. Based on the UE described in Example 1.1, where the AI / ML model is an ML AI / ML function, and the AI / ML model performance monitoring is the AI / ML function performance monitoring.

[0120] Example 1.6. According to any one of Examples 1.1 to 1.5, the triggering criteria for interrupting AI / ML model performance monitoring include the UE lacking resources for performing monitoring and lifecycle management (LCM) decisions.

[0121] Example 1.7. A UE according to any one of Examples 1.1 to 1.5, wherein the triggering criteria for interrupting AI / ML model performance monitoring includes the UE lacking an option for model switching.

[0122] Example 1.8. According to any one of Examples 1.1 to 1.5, the triggering criteria for interrupting AI / ML model performance monitoring include the AI / ML model performance for beam selection being less accurate than the non-ML model performance for beam selection.

[0123] Example 1.9. According to any one of Examples 1.1 to 1.5, the triggering criteria for interrupting AI / ML model performance monitoring include downlink channel conditions falling below a threshold.

[0124] Example 1.10. According to any one of Examples 1.1 to 1.5, the triggering criteria for interrupting AI / ML model performance monitoring include the ML model performance used to estimate the location being different from the Global Navigation Satellite System (GNSS) location.

[0125] Example 1.11. According to any one of Examples 1.1 to 1.5, the triggering criteria for interrupting AI / ML model performance monitoring include the UE's inability to activate the AI / ML model.

[0126] Example 1.12. The UE according to any one of Examples 1.1 to 1.11 further includes a component for determining by the UE whether to restart the performance monitoring of the configured AI / ML model.

[0127] Example 1.13. The UE according to Example 1.12 also includes a component for transmitting a readiness message from the UE to a network node, the readiness message instructing the UE to restart the performance monitoring of the configured AI / ML model.

[0128] Example 1.14. The UE described in Example 1.13, wherein upon receiving an acknowledgment from the network node, the UE restarts the configured AI / ML model performance monitoring.

[0129] Example 1.15. The UE described in Example 1.13, where AI / ML model performance monitoring is interrupted by a network node.

[0130] The embodiments and aspects disclosed herein are examples of this disclosure and may be embodied in various forms. For example, although some embodiments herein are described as separate embodiments, each embodiment herein may be combined with one or more other embodiments herein. The specific structural and functional details disclosed herein should not be construed as limiting, but should be regarded as the basis of the claims and a representative basis for teaching those skilled in the art to implement this disclosure in various ways with virtually any suitable detailed structure. Throughout the description of the drawings, the same reference numerals may refer to similar or identical elements.

[0131] The phrases “in one aspect,” “in all aspects,” “in all dimensions,” “in some aspects,” or “in other aspects” may each refer to one or more of the same or different aspects under this disclosure. The phrase “multiple” may refer to two or more.

[0132] In various embodiments, the terms "first message" and "second message," as well as any subsequent messages, may refer to any message transmitted or received in sequence, and are not necessarily limited to any particular message.

[0133] The phrases “in an embodiment,” “in an embodiment,” “in various embodiments,” “in some embodiments,” or “in other embodiments” may each refer to one or more of the same or different embodiments according to this disclosure. The phrase “A or B” means “(A), (B), or (A and B).” The phrase “at least one of A, B, or C” means “(A); (B); (C); (A and B); (A and C); (B and C); or (A, B, and C).”

[0134] Any method, program, algorithm, or code described herein can be translated into or expressed in a programming language or computer program. As used herein, the terms "programming language" and "computer program" each include any language used to specify instructions to a computer, and include (but are not limited to) the following languages ​​and their derivatives: assembly language, Basic, batch files, BCPL, C, C+, C++, Delphi, Fortran, Java, JavaScript, machine code, operating system command languages, Pascal, Perl, PL1, Python, scripting languages, Visual Basic, meta-languages ​​that specify their own programs, and all first-, second-, third-, fourth-, fifth-, and more generation computer languages. Databases and other data schemas, and any other meta-languages, are also included. There is no distinction between languages ​​that are interpreted, compiled, or use compilation and interpretation methods. There is no distinction between compiled and source versions of a program. Therefore, a reference to a program in which a programming language can exist in more than one state (such as source, compiled, object, or linked) is a reference to any and all such states. A reference to a program can encompass the actual instructions and / or the intent of those instructions.

[0135] While various aspects of this disclosure have been shown in the accompanying drawings, this disclosure is not intended to be limited thereto, as its scope is intended to be as broad as is permitted in the art, and this specification should be interpreted accordingly. Therefore, the above description should not be construed as restrictive, but merely as an example of particular aspects. Other modifications within the scope and spirit of the appended claims will be contemplated by those skilled in the art.

Claims

1. A method for performance monitoring of artificial intelligence (AI) or machine learning (ML) AI / ML models, performed by a user equipment (UE) connected to a radio access network (RAN), the method comprising: The UE uses the first AI / ML model; Receive configuration from network node for initiating performance monitoring of the first AI / ML model used, wherein the request includes: performance monitoring criteria associated with the model used, trigger criteria for indicating the switching of AI / ML model performance monitoring from the UE to the first device, wherein the UE continues to use the first AI / ML model, and trigger criteria for indicating the switching of AI / ML model performance monitoring from the first device to the UE; The UE determines whether to initiate at least a portion of the configured AI / ML model performance monitoring; The UE determines whether at least one of the triggering criteria is met; and In response to at least one of the triggering criteria being met, the UE transmits a request to the first device to switch at least a portion of the configured AI / ML model performance monitoring from the UE to the first device while continuing to use the determined AI / ML model.

2. The method according to claim 1, wherein upon receiving confirmation from the network node, the UE suspends the execution of model or function performance monitoring.

3. The method according to claim 2, further comprising: The UE continues to use the model associated with the model or functional performance monitoring.

4. The method according to claim 1, wherein upon receiving a negative acknowledgment from the network node, the UE switches to the fallback function.

5. The method according to claim 1, wherein the AI / ML model is an ML AI / ML function, and the AI / ML model performance monitoring is AI / ML function performance monitoring.

6. The method according to any one of claims 1 to 5, wherein the triggering criterion for interrupting AI / ML model performance monitoring includes the UE lacking resources for performing monitoring and lifecycle management (LCM) decisions.

7. The method according to any one of claims 1 to 5, wherein the triggering criteria for interrupting AI / ML model performance monitoring includes the UE lacking an option for model switching.

8. The method according to any one of claims 1 to 5, wherein the triggering criterion for interrupting AI / ML model performance monitoring includes the AI / ML model performance for beam selection being less accurate than the non-ML model performance for beam selection.

9. The method according to any one of claims 1 to 5, wherein the triggering criterion for interrupting AI / ML model performance monitoring includes downlink channel conditions falling below a threshold.

10. The method according to any one of claims 1 to 5, wherein the triggering criteria for interrupting AI / ML model performance monitoring include the ML model performance used to estimate the location being different from the Global Navigation Satellite System (GNSS) location.

11. The method according to any one of claims 1 to 5, wherein the triggering criteria for interrupting AI / ML model performance monitoring includes the UE's inability to activate the AI / ML model.

12. The method according to any one of claims 1 to 11, further comprising the UE determining whether to restart the configured AI / ML model performance monitoring.

13. The method of claim 12, further comprising transmitting a ready message from the UE to the network node, the ready message indicating that the UE can restart the configured AI / ML model performance monitoring capability.

14. The method of claim 13, wherein upon receiving an acknowledgment from the network node, the UE restarts the configured AI / ML model performance monitoring.

15. The method of claim 13, wherein AI / ML model performance monitoring is interrupted by the network node.

16. A user equipment (UE), comprising: At least one processor; as well as At least one memory storing instructions that, when executed by the at least one processor, cause the device to perform at least the following: The UE uses a first artificial intelligence / machine learning (AI / ML) model; The system receives configuration and triggering criteria from a network node. The configuration is used to begin monitoring the performance of a first AI / ML model being used, and the triggering criteria are used to instruct the system to switch AI / ML model performance monitoring from the first device to the UE. The request includes performance monitoring criteria associated with the model being used and triggering criteria to instruct the system to switch AI / ML model performance monitoring from the UE to the first device, wherein the UE continues to use the first AI / ML model. The UE determines whether to initiate at least a portion of the configured AI / ML model performance monitoring; The UE determines whether at least one of the triggering criteria is met; as well as In response to at least one of the triggering criteria being met, the UE transmits a request to the first device to switch at least a portion of the configured AI / ML model performance monitoring from the UE to the first device while continuing to use the determined AI / ML model.

17. A user equipment (UE), comprising: At least one processor; as well as At least one memory storing instructions that, when executed by the at least one processor, cause the device to perform at least the method as claimed in any one of claims 1 to 14.

18. A processor-readable medium storing instructions that, when executed by at least one processor of the device, cause the device to perform at least the method as claimed in any one of claims 1 to 14.

Images