Method for data-driven model signaling for multiple USIM

Abstract

The present disclosure describes a method of data driven AI / ML model signaling for multiple USIM in a wireless mobile communication system including a base station (e.g., a gNB) and a mobile station (e.g., a UE). Multiple model operations are performed sequentially or in parallel across multiple USIMs using independent radio links based on a collaboration mechanism.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to AI / ML based model signaling, wherein techniques are proposed for pre-configuration and signaling of specific information for multi-USIM collaborative model operation with separate radio links. BACKGROUND

[0002] In 3GPP (3rd Generation Partnership Project), one of the selected study items as approved Release 18 package is AI / ML (Artificial Intelligence / Machine Learning), as described in the related documentation (RP-213599) submitted in 3GPP TSG RAN (Technical Specification Group Radio Access Network) Meeting #94e. The formal name of this AI / ML study item is “Study on AI / ML for NR air interface”, and currently RAN WG1 and WG2 are actively working on the specification. The objective of this study item is to identify a common AI / ML framework and areas to use AI / ML based techniques and use cases to gain benefits. According to 3GPP, the main objective of this study item is to study AI / ML framework for air interface by considering performance, complexity, and potential specification impact with target use cases. In particular, AI / ML models, terminology, and description for identifying common and specific characteristics of the framework will be one of the key work scopes. Regarding the AI / ML framework, various aspects are being considered for study, and one of the key items is on the lifecycle management of AI / ML models, which includes mandatory multiple phases of model training, model deployment, model inference, model monitoring, model update, etc. Previously, in 3GPP TR 37.817 of Release 17, named Study on enhancement for Data Collection for NR and EN-DC, UE mobility was also considered as one of the AI / ML use cases, and one of the scenarios for model training / inference is that both functions are located within the RAN node. Subsequently, a new work item “Artificial Intelligence (AI) / Machine Learning (ML) for NG-RAN” was initiated in Release 18 to specify data collection enhancements and signaling support within the existing NG-RAN interfaces and architecture.

[0003] For the above positive standardization work, there is currently no specification defined for signaling methods or network (e.g., gNB) / mobile station (e.g., UE) behavior regarding supporting AI / ML model operation when a UE has multiple USIMs (Universal Subscriber Identity Module). In the case of a single USIM, multiple ML operations are limited for parallel processing such as inference and online training, and in the case of two or more multi-USIMs (MUSIM), it is possible to perform dedicated ML operations with separate radio links. When model performance is degraded on a single USIM-based radio link, model adaptation such as model switching / re-training / fallback is required. As a result, latency and / or signaling overhead will occur. Therefore, MUSIM-based signaling / process needs to be specified for model operation.

[0004] US11463865B1 shows techniques that enable a wireless carrier network to provide subscribers of a competitor wireless carrier network with an opportunity to try wireless telecommunication services provided by the wireless carrier network.

[0005] US2022295343A1 describes systems and processes for high throughput wireless communications, where a UE is configured to receive data indicating a link metric for each available communication link through which the UE is configured to communicate, and the UE is configured to determine whether a network through which the UE is communicating is congested.

[0006] US2020312301A1 explains systems and techniques for model adaptation, where a set of adaptation training data and a set of parameters are received, and the set of adaptation training data can be used to determine the set of adaptation parameters.

[0007] US2021133588A1 shows that machine learning models for classification can adapt to changes in features of input data to provide better classification performance.

[0008] US10990850B1 describes a model adaptation controller for retraining a deployed ML model using samples with ground truth values generated by different ML models, and the retraining process can be performed iteratively to automatically improve and adapt ML models running on edge devices.

[0009] WO2022099425A1 describes a method of adjusting a current set of configuration parameters to minimize a loss function for a machine learning model for a plurality of mixed data elements, where the method results in an adapted machine learning model with improved performance on inference on new target samples. SUMMARY

[0010] Collaborative ML is proposed to support sequential / parallel model operation execution between gNB and UE over multiple USIMs with separate radio links.

[0011] The benefit is reduced latency, reduced per-USIM signaling overhead by joint ML operation.

[0012] The first aspect is a method for multi-USIM, data-driven model signaling for collaborative operation of AI / ML model, comprising the following steps: enabling / disabling collaborative ML (CM) operation by multi-USIM (MUSIM), determining different CM types (e.g., sequential CM and parallel CM) according to the UE capability of supporting DSDS / DSDA of modem / RF chain, performing different CM operation modes assigned to each USIM based on the mapping relationship between CM mode and USIM, determining the main CM operation mode and the supplementary CM operation mode of each USIM.

[0013] In some embodiments of the method according to the first aspect, the method is characterized in that the CM mode is defined and pre-configured for a set of mode categories, wherein the mode category can be a baseline mode or an enhanced mode based on the ML-based life cycle management operation, and then the ML operation is indexed with different CM modes for configuration.

[0014] In some embodiments of the method according to the first aspect, the method is characterized in that the multiple radio links can be collaborative to handle model operation between UE and different gNBs across MUSIM, wherein different model operation modes can be configured.

[0015] In some embodiments of the method according to the first aspect, the method is characterized in that the CM operation is enabled or disabled depending on the MUSIM configuration, wherein an indication message for enabling / disabling MUSIM-based CM is transmitted through L1 / L2 signaling.

[0016] In some embodiments of the method according to the first aspect, the method is characterized in that sequential CM or parallel CM is determined according to the UE Tx / Rx capability of supporting DSDS / DSDA of modem / RF chain, comprising: performing sequential CM, wherein any activated USIM is connected mode, only one CM mode is run at a time, and any associated CM mode in the combination can be sequentially performed by the same USIM or different USIM; performing parallel CM, wherein two or more CM modes are executed in parallel when MUSIM is simultaneously activated as connected mode, so that the combination of CM modes pre-configured with MUSIM can then be performed.

[0017] In some embodiments of the method according to the first aspect, the method is characterized by performing primary CM and supplementary CM for the MUSIM, wherein a CM operation mode is assigned to each USIM, wherein the CM mode is mapped to the MUSIM based on different criteria (e.g., ML support configuration, radio link conditions, etc.), including: assigning a primary CM mode to USIM1 (for mandatory ML operation, e.g., model inference) and a supplementary CM mode to USIM2 (for mandatory / optional ML operation, e.g., model training), pre-configuring a supplementary CM mode associated with the primary CM mode for the CM mode combination, wherein the index value of the CM mode is then indicated by DCI and / or MAC CE and / or RRC signaling.

[0018] In some embodiments of the method according to the first aspect, the method is characterized by setting a priority level for the preferred CM mode combination, including: determining the corresponding CM mode of the MUSIM; indicating the priority level of each CM mode corresponding to the MUSIM; prioritizing the primary ML operation on a dedicated USIM link, wherein the priority level of the CM mode can be dynamically switched or updated; pre-configuring the number of priority levels can be more than two, and each CM mode with priority is associated with the assigned USIM.

[0019] In some embodiments of the method according to the first aspect, the method is characterized by determining the CM mode with priority based on the priority of the CM mode when any of the used radio links is unavailable due to other services, including: pre-configuring the CM mode with high priority (through the primary link) and the CM mode with low priority (through the non-primary link), setting the CM mode priority level for the MUSIM link switching operation.

[0020] In some embodiments of the method according to the first aspect, the method is characterized by configuring the cooperative ML based on DSDA / DSDS between the gNB and the UE, including: generating a candidate combination of CM modes with USIM1 and USIM2 for model operation, receiving the CM mode selection result from the non-primary link, performing the final CM mode combination.

[0021] In some embodiments of the method according to the first aspect, the method is characterized by activating the CM operation with a trigger event, wherein the list of trigger events (e.g., traffic load) for activating the CM is used to activate the CM corresponding to the MUSIM from the non-cooperative ML with single USIM, and the request for the CM can be decided by the network side or by the UE side for implementation-specific use cases.

[0022] In some embodiments of the method according to the first aspect, the method is characterized by utilizing model splitting to activate CM operations, wherein the ML model for the operation between the network and the UE is split and the split model is processed by each MUSIM. In a similar manner, the data set for model training / inference / monitoring can also be split into multiple data set groups to be processed by each USIM link.

[0023] According to a second aspect, the disclosure relates to a device for multi-USIM, data-driven model signaling supporting AI / ML model collaborative operation, the device comprising a wireless transceiver, a processor coupled with a memory, the memory having stored therein computer program instructions configured to implement the steps of the method of the first aspect.

[0024] According to a third aspect, the disclosure relates to a device for multi-USIM, data-driven model signaling supporting AI / ML model collaborative operation implemented by a gNB, the device comprising a wireless transceiver, a processor coupled with a memory, the memory having stored therein computer program instructions configured to implement the steps of the method of the first aspect.

[0025] According to a fourth aspect, the disclosure relates to a user equipment comprising the device according to any one of the embodiments of the second aspect.

[0026] According to a fifth aspect, the disclosure relates to a base station comprising the device according to any one of the embodiments of the third aspect.

[0027] According to a sixth aspect, the disclosure relates to a wireless communication system, wherein the gNB comprises a processor coupled with a memory, the memory having stored therein computer program instructions configured to implement the steps of the method of the first aspect, and wherein the user equipment (UE) comprises a processor coupled with a memory, the memory having stored therein computer program instructions configured to implement the steps of the method of the first aspect.

[0028] According to a fifth aspect, the disclosure relates to a wireless communication system comprising at least one base station according to any one of the embodiments of the disclosure and at least one user equipment for performing the method according to any one of the embodiments of the first aspect.

[0029] According to a sixth aspect, the disclosure relates to a computer program product comprising instructions which, when executed by at least one processor, configure the at least one processor to perform the method according to the first aspect, the at least one processor being configured to perform the method for data exchange according to any of the embodiments of the disclosure. The computer program product can use any programming language, and can be in the form of source code, object code, or any intermediate form such as a partially compiled form, or any other desired form.

[0030] According to a seventh aspect, the disclosure relates to a computer-readable storage medium comprising instructions which, when executed by at least one processor, configure the at least one processor to perform the method according to any of the embodiments of the disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0031] Figure 1 is an exemplary table of collaborative ML modes.

[0032] Figure 2 is an exemplary table of multi-USIM based collaborative ML modes.

[0033] Figure 3 is a flowchart of enabling / disabling collaborative ML for multi-USIM.

[0034] Figure 4 is a flowchart of enabling sequential / parallel collaborative ML for multi-USIM.

[0035] Figure 5 is a flowchart of primary and supplementary collaborative ML for multi-USIM.

[0036] Figure 6 is an exemplary table of setting priority levels of collaborative ML modes.

[0037] Figure 7 is a signaling flow of priority levels of collaborative ML modes.

[0038] Figure 8 is a signaling flow of inter-PLMN collaborative ML.

[0039] Figure 9 is a signaling flow of intra-PLMN collaborative ML.

[0040] Figure 10 is a signaling flow of activating collaborative ML with a trigger event.

[0041] Figure 11 is a signaling flow of activating collaborative ML with model splitting. DETAILED DESCRIPTION

[0042] The following explanation will provide a detailed description of mechanisms for data-driven AI / ML model signaling for Multi-USIM (MUSIM) in a wireless mobile communication system comprising base stations (e.g., gNBs) and mobile stations (e.g., UEs) where multiple model operations are performed sequentially or in parallel across multiple USIMs with separate radio links. AI / ML based techniques are currently being applied to many different applications, and 3GPP has also started its technology study to be applied to multiple use cases based on the observed potential benefits. The AI / ML life cycle can be split into several phases such as data collection / preprocessing, model training, model testing / validation, model deployment / update, model monitoring, model switching / selection, etc., where each phase is equally important to achieve the target performance of any particular model.

[0043] One of the challenging issues in applying AI / ML models to any use case or application is managing the life cycle of the AI / ML model. This is mainly because data / model drift occurs during model deployment / inference, and it causes the performance of the AI / ML model to deteriorate. Fundamentally, data set statistics change after the model is deployed, and the model inference capability is also affected by unseen data as input.

[0044] In a similar aspect, the statistical properties of the data set and the relationship between the input and output of the trained model can change with the occurrence of drift. Model adaptation is then required to support operations such as model switching, retraining, fallback, etc. When a wireless communication network with AI / ML model enabled is deployed, it is then important to consider how to handle the adaptation of the AI / ML model under model training, inference, monitoring, update, etc. There are different types of MUSIM, such as DSDS (dual SIM dual standby) and DSDA (dual SIM dual active). For DSDS, the UE is restricted to connecting to one network at a time, and only one SIM is used for connection at any given time. For example, when the UE is using one SIM for connection (e.g., for a voice call), the other SIM will be idle. For DSDA, the UE can connect to multiple networks, being able to use two SIMs and two radios in order to simultaneously maintain two active data communication sets. When there is a DSDA configuration, mapping to the network can be done concurrently. For example, a voice call using one SIM and data communication (e.g., internet browsing) on the second SIM.

[0045] Techniques are described for pre-configuring and signaling specific information for collaborative model operation with separate radio links based on the type of MUSIM such as DSDS and DSDA. First, to enable collaborative ML (CM) operation by MUSIM, an indication message is transmitted for enabling MUSIM based CM and transmitted using L1 / L2 signaling. For example, a single bit can be used to indicate, bit 0 for enabling CM and bit 1 for disabling CM. The CM type is also determined to be sequential CM or parallel CM depending on the UE transmit (Tx) and receive (Rx) capabilities of the modem / RF chain supporting DSDS / DSDA. For example, if DSDS is configured, sequential CM is used and if DSDA is configured, parallel CM is used. For sequential CM, either active USIM is run as a connected mode at a time and any associated CM mode in the combination can be operated sequentially by the same USIM or different USIMs. For parallel CM, two or more CM modes are operated in parallel when MUSIM is activated as connected mode at the same time. Thus, a combination of CM modes pre-configured with MUSIM can then be performed. A CM operation mode is assigned to each USIM, where the CM mode is mapped to MUSIM based on different criteria such as ML support configuration, radio link conditions, etc. For example, a primary CM mode is assigned to USIM1 (for mandatory ML operation, e.g., model inference) and a supplemental CM mode is assigned to USIM2 (for mandatory / optional ML operation, e.g., model training). If there are more than two USIMs, more than one primary and / or supplemental CM mode can be assigned to these USIMs. Depending on the activated CM type and the primary CM mode, the supplemental CM mode associated with the primary CM mode can be pre-configured in order to reduce model adaptation latency and / or balance signaling overhead across MUSIM radio links. For example, a CM mode combination for MUSIM can be pre-configured in advance for the primary CM mode and the supplemental CM mode, then an index value of the CM mode is indicated (through DCI, MAC CE or RRC signaling).

[0046] The detailed description set forth below in connection with the appended drawings is intended as a description of various configurations and is not intended to represent the only configurations in which the concepts described herein can be practiced. The detailed description includes specific details for the purpose of providing a thorough understanding of various concepts. However, it will be apparent to those skilled in the art that these concepts can be practiced without these specific details. In particular, although terminology from 3GPP 5G NR can be used in this disclosure to exemplify the embodiments herein, this should not be seen as limiting the scope of the application.

[0047] Some embodiments of what is contemplated herein will now be described more fully with reference to the accompanying drawings. Other embodiments, however, are also within the scope of the subject matter disclosed herein, and the disclosed subject matter should not be construed as limited to the embodiments set forth herein; rather, they are provided by way of example to convey the scope of the subject matter to those skilled in the art.

[0048] Generally, all terms used herein are to be interpreted according to their ordinary meaning in the technical field of the technology concerned unless explicitly stated otherwise and / or unless explicitly given a different meaning in the context in which it is used. All references to a / an / the item, apparatus, component, means, step, etc. are to be interpreted openly as referring to at least one instance of whatever is being referred to unless otherwise indicated. The steps of any methods disclosed herein do not have to be performed in the exact order disclosed, unless explicitly stated otherwise or inherently required. Any of the embodiments of the subject matter disclosed herein can be applied to any other embodiments of the subject matter disclosed herein, in any appropriate combination, where appropriate. Likewise, any features of any of the embodiments of the subject matter disclosed herein can be applied to any other embodiments of the subject matter disclosed herein, where appropriate, in any appropriate combination. Other objects, features and advantages of the enclosed embodiments will become apparent from the description provided herein.

[0049] In some embodiments, the more general term“network node” can be used and can correspond to any type of radio network node or any network node that communicates with a UE (directly or via another node) and / or with another network node. Examples of network nodes are NodeB, MeNB, ENB, network nodes belonging to a MCG or SCG, base station (BS), multi-standard radio (MSR) radio node such as a MSR BS, eNodeB, gNodeB, network controller, radio network controller (RNC), base station controller (BSC), relay, donor node controlled relay, base transceiver station (BTS), access point (AP), transmission points, transmission nodes, RRU, RRH, nodes in a distributed antenna system (DAS), core network nodes (e.g., mobile switching center (MSC), mobility management entity (MME), etc.), operation & maintenance (O&M), operation support system (OSS), self-optimizing network (SON), positioning nodes (e.g., evolved serving mobile location center (E-SMLC)), minimization of drive testing (MDT), test equipment (physical node or software), etc.

[0050] In some embodiments, the non-limiting term user equipment (UE) or wireless device can be used, and this term can refer to any type of wireless device communicating with a network node and / or with another UE in a cellular or mobile communication system. Examples of UEs are target devices, device-to-device (D2D) UEs, machine-type UE or UE capable of machine-to-machine (M2M) communication, PDAs, PADs, tablet computers, mobile terminals, smart phones, laptop-embedded equipped (LEE), laptop-mounted equipment (LME), USB dongles, UE category M1, UE category M2, ProSe UE, V2V UE, V2X UE, etc.

[0051] Also, the terms such as base station / gNodeB and UE should be considered as non-limiting and do not imply certain hierarchical relationship between the two; in general, "gNodeB" can be considered as device 1, "UE" can be considered as device 2, and these two devices communicate with each other over some radio channel. And in the following, the transmitter or receiver can be gNodeB (gNB) or UE.

[0052] As those skilled in the art will appreciate, the aspects of the embodiments can be embodied as a system, device, method or program product. Accordingly, the embodiments can take the form of an entirely hardware embodiment, an entirely software embodiment (including firmware, resident software, micro-code, etc.) or an embodiment combining software and hardware aspects.

[0053] For example, the disclosed embodiments can be implemented as hardware circuitry, including custom very-large-scale integration ("VLSI") circuits or gate arrays, off-the-shelf semiconductors such as logic chips, transistors, or other discrete components. The disclosed embodiments can also be implemented in programmable hardware devices such as field programmable gate arrays, programmable array logic, programmable logic devices or the like. As another example, the disclosed embodiments can include one or more physical or logical blocks of executable code, which may, for example, be organized as an object, procedure, or function.

[0054] Furthermore, embodiments can take the form of a program product embodied in one or more computer readable storage devices storing machine-readable code, computer readable code, and / or program code, hereinafter "code". The storage devices can be tangible, non-transitory, and / or non-transmission. The storage devices can not comprise signals. In a certain embodiment, the storage devices only employ signals to access code.

[0055] Any combination of one or more computer-readable media can be utilized. The computer-readable media can be computer-readable storage media. The computer-readable storage media can be storage devices. Storage devices can be, for example and without limitation, electronic, magnetic, optical, electromagnetic, holographic, micromechanical, or semiconductor systems, apparatuses, or devices, or any suitable combination of the foregoing.

[0056] More specific examples (a non-exhaustive list) of the storage device would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory ("RAM"), a read-only memory ("ROM"), an erasable programmable read-only memory ("EPROM" or Flash memory), a portable compact disc read-only memory ("CD-ROM"), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer-readable storage medium can be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

[0057] Code for carrying out operations for embodiments can be any number of lines and any combination of programming languages, including an object oriented programming language such as Python, Ruby, Java, Smalltalk, C++, or the like, and conventional procedural programming languages, such as the "C" programming language, or the like, and / or machine languages such as assembly languages. The code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network ("LAN"), wireless LAN ("WLAN"), or a wide area network ("WAN"), or the like, or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider ("ISP")). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, application specific circuitry, or field programmable gate array ("FPGA") circuitry, or other hardware can execute the

[0058] Furthermore, the described features, structures, or characteristics of the embodiments can be combined in any suitable manner. In the following description, numerous specific details are provided, such as examples of programming, software modules, user selections, network transactions, database queries, database structures, hardware modules, hardware circuits, hardware chips, etc., to provide a thorough understanding of the embodiments. One skilled in the relevant art will recognize, however, that the embodiments can be practiced without one or more of the specific details, or with other methods, components, materials, and so forth. In other instances, well-known structures, materials, or operations are not shown or described in detail in order to avoid obscuring aspects of the embodiments. Reference throughout this specification to "an embodiment," "embodiments," "one embodiment," or similar language means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Thus, appearances of the phrase "in one embodiment" or similar language throughout this specification may, but do not necessarily, all refer to the same embodiment. The terms "including," "comprising," "having," and variations thereof mean "including but not limited to," unless expressly specified otherwise. An enumerated list of items does not imply that any or all of the items are mutually exclusive, unless expressly specified otherwise. The terms "a," "an," and "the" also mean "one or more," unless expressly specified otherwise.

[0059] The code can also be stored in a storage device that can direct a computer, other programmable data processing apparatus, or other devices to function in a particular manner, such that the instructions stored in the storage device produce an article of manufacture including instructions which implement the function / act specified in the flowchart and / or block diagram block or blocks.

[0060] The code can also be stored in a storage device that can direct a computer, other programmable data processing apparatus, or other devices to function in a particular manner, such that the instructions stored in the storage device produce an article of manufacture including instructions which implement the function / act specified in the flowchart and / or block diagram block or blocks.

[0061] The code can also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other devices to produce a computer implemented process such that the code which execute on the computer or other programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0062] The flow and / or block diagrams in the drawings represent possible architectures, functionalizations, and operations of devices, systems, methods, and program products according to various embodiments. In this regard, each block in the flow and / or block diagrams can represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks can sometimes be executed in the reverse order, depending on the functionality involved. Other steps and methods can be conceived that are equivalent in function, logic, or effect to those illustrated, with the scope of the present disclosure intended to include all such steps and methods.

[0063] It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks can sometimes be executed in the reverse order, depending on the functionality involved. Other steps and methods can be conceived that are equivalent in function, logic, or effect to those illustrated, with the scope of the present disclosure intended to include all such steps and methods.

[0064] Although various arrow types and line types can be employed in the flow and / or block diagrams, these are understood to be merely illustrative of the logical flows of the depicted embodiments. For example, an arrow can indicate a waiting or monitoring period of time of unspecified duration between enumerated steps of the depicted embodiment. It will also be noted that each block of the block diagrams and / or flowchart illustrations, and combinations of blocks in the block diagrams and / or flowchart illustrations, can be implemented by special purpose hardware-based systems that perform the specified functions or acts, or combinations of special purpose hardware and code.

[0065] The detailed description set forth below, in connection with the appended drawings and embodiments described herinin, is intended as a description of various configurations and is not intended to represent the only configurations in which the concepts described herein can be practiced. The detailed description includes specific details for the purpose of providing a thorough understanding of the various concepts. However, it will be apparent to those skilled in the art that these concepts can be practiced without these specific details. For example, while 3GPP terminology from, e.g., 5G NR, can be used to exemplify embodiments herein in the present disclosure, this should not be seen as limiting the scope of the present disclosure.

[0066] The present disclosure relates to a wireless communication system, which can be a 5G NR wireless communication system, for example. More specifically, it denotes a RAN of a wireless communication system for exchanging data with UEs via radio signals. For example, the RAN can transmit data to a UE (downlink, DL), e.g., data received from a core network (CN). The RAN can also receive data from a UE (uplink, UL), which can be forwarded to the CN.

[0067] In the illustrated example, the RAN comprises one base station, BS. Of course, the RAN can comprise more than one BS to increase the coverage of the wireless communication system. Depending on the implemented wireless communication standard, each of these BSs can be referred to as NB, eNodeB (or eNB), gNodeB (or gNB in case of a 5G NR wireless communication system), access point, etc.

[0068] The UE is located in the coverage of the BS. The coverage of the BS corresponds for example to an area in which the UE can decode a PDCCH transmitted by the BS.

[0069] Examples of a wireless device suitable for implementing any of the methods performed at the UE discussed in the present disclosure correspond to a device providing wireless connectivity with a RAN of a wireless communication system and that can be used to exchange data with this RAN. Such a wireless device can be comprised in a UE. The UE can be for example a cellular phone, a wireless modem, a wireless communication device, a handheld device, a laptop computer, etc. The UE can also be an Internet of Things (IoT) device, like a wireless camera, a smart sensor, a smart meter, smart glasses, a vehicle (manned or unmanned), a global positioning system device, etc., or any other device that can run an application requiring the exchange of data with a remote recipient via a wireless device.

[0070] The wireless device comprises one or more processors and one or more memories. The one or more processors can comprise for example a Central Processing Unit (CPU), a Digital Signal Processor (DSP), a Field-Programmable Gate Array (FPGA), an Application-Specific Integrated Circuit (ASIC), etc. The one or more memories can comprise any type of computer-readable volatile and non-volatile memory (magnetic hard disk, solid state disk, optical disk, electronic memory, etc.). The one or more memories can store a computer program product in the form of a set of program code instructions to be executed by the one or more processors to implement all or part of the steps of the method for exchanging data performed at the UE side according to any of the embodiments disclosed herein.

[0071] The wireless device can also comprise a main radio, MR, unit. The MR unit corresponds to the main wireless communication unit of the wireless device for exchanging data with the BSs of the RAN using radio signals. The MR unit can implement one or more wireless communication protocols and can be for example a 3G, 4G, 5G, NR, WiFi, WiMax, etc. transceiver, etc. In a preferred embodiment, the MR unit corresponds to a 5G NR wireless communication unit.

[0072] The description of the elements in each figure can refer to the elements in the preceding figures. Identical numbers in all figures refer to the same elements, including alternative embodiments of the same elements.

[0073] Figure 1 An example table of CM modes is shown. CM modes are defined and preconfigured for a set of mode categories based on using MUSIM in the UE for ML operations. Based on the lifecycle management operations of ML, the mode categories can be baseline mode or enhanced mode. Different CM modes are then utilized to index ML operations for configuration. When the UE has different numbers of USIMs, combinations of multiple CM modes are configured.

[0074] Figure 2 An example table of MUSIM based CM modes is shown. MUSIM can provide multiple radio links between the UE and different gNBs. AI / ML model operations can be extended to multiple radio links, which are simultaneously connected with separate gNBs / networks through independent wireless resources. Depending on different model operation modes, multiple radio links can be cooperative to handle model operations between the UE and different gNBs across MUSIM. In the case of CM operations, target benefits include balancing signaling overhead across radio links and reducing latency by utilizing multiple link based synchronized model operations, etc.

[0075] Figure 3 A flow diagram of enabling / disabling CM for MUSIM is shown. This shows whether to enable cooperative ML depending on USIM configuration. For example, when using single USIM, CM is not operated. In contrast, MUSIM enables CM. An indication message is transmitted for enabling MUSIM based CM, and is transmitted using L1 / L2 signaling. For example, a single bit can be used for indication, with bit 0 for enabling CM and bit 1 for disabling CM.

[0076] Figure 4 A flow diagram of enabling sequential / parallel CM for MUSIM is shown. Depending on the UE Tx / Rx capability of modem / RF chain supporting DSDS / DSDA, it is determined whether the type of cooperative ML is sequential CM or parallel CM. For example, DSDS is used for sequential CM, and DSDA is used for parallel CM. For sequential CM, any activated USIM is run as connected mode, only one CM mode is run at a time, and any associated CM mode in the combination can be sequentially operated through the same USIM or different USIMs. For parallel CM, two or more CM modes are run in parallel when multiple USIMs in MUSIM are simultaneously activated as connected mode. Thus, a preconfigured combination of CM modes for MUSIM can then be performed.

[0077] Figure 5Flow diagrams are shown for primary and supplemental CMs for MUSIM. CM operation modes are assigned to each USIM, where the CM modes are mapped to MUSIM based on different criteria such as ML support configuration, radio link conditions, etc. For example, a primary CM mode is assigned to USIM1 (for mandatory ML operation, e.g., model inference), and a supplemental CM mode is assigned to USIM2 (for mandatory / optional ML operation, e.g., model training). Depending on the activated CM type and primary CM mode, the supplemental CM mode associated with the primary CM mode can be pre-configured in order to reduce model adaptation latency and / or balance signaling overhead across MUSIM radio links. For example, a CM mode combination for MUSIM can be pre-configured for the primary and supplemental CM modes in advance, where then an index value of the CM mode is indicated (through DCI, MAC CE, or RRC signaling).

[0078] Figure 6 An example table is shown that sets a priority level of CM modes. Based on a bilateral or UE-side model setup, a preferred CM mode combination is used to configure ML operation for a MUSIM device. As step #1, a CM mode is determined for multiple radio links of each MUSIM. And in step #2, for each CM mode, a priority level of each mode is indicated in order to prioritize a primary ML operation on a certain dedicated USIM link. For example, if 1 bit is used, bit 0 indicates a high priority and bit 1 indicates a low priority. The priority level of the CM mode can be dynamically switched or updated. The number of priority levels can be more than two, and each CM mode with a priority is associated with an assigned USIM.

[0079] Figure 7 A signaling flow of priority levels of CM modes is shown. When two radio links are used for CM operation, one of the links can need to switch to other ML operation with a separate dedicated model activation (e.g., ML operation interruption). In this case, when one of the two links becomes unavailable due to other services, the priority level of CM mode is used to determine the CM mode with priority. Thus, the priority of the CM mode combination of two USIMs can be pre-configured, such as a higher priority CM mode (through a primary link) and a lower priority CM mode (through a non-primary link). The CM mode priority level is also used for UE MUSIM behavior for switching activities. If there are more than two USIMs, the priority level of the CM mode for each USIM can be configured.

[0080] Figure 8The signaling flow for CMs between PLMNs (Public Land Mobile Networks) is illustrated. In this diagram, it is assumed that the two USIMs belong to separate PLMNs (e.g., between PLMNs). DSDA-based cooperative ML is configured between the gNB and the UE. Candidate combinations of CM modes for USIM1 and USIM2 are generated between the primary link gNB and the UE for model operation. The final CM mode combination is confirmed after receiving the CM mode selection from the non-primary link gNB. A similar process can be applied to DSDS scenarios, but the radio link needs to be used in a time-multiplexed manner.

[0081] Figure 9 The signaling flow for a CM within a PLMN is illustrated. In this diagram, the two USIMs belong to the same PLMN (e.g., within the PLMN). DSDA-based cooperative ML is configured between the gNB and the UE. Based on the CM support capabilities reported by the UE, the CM mode combination for USIM1 and USIM2 is determined for model operation. After confirming the CM operation settings, model operation begins. In this scenario, different RAN nodes supporting UEs with MUSIMs can communicate with each other via the X2 interface for CM operation-related signaling.

[0082] Figure 10 The signaling flow for activating a CM using trigger events is illustrated. A CM with a MUSIM can be activated from a non-cooperative ML using a single USIM. A list of trigger events (e.g., traffic load) for activating cooperative MLs can be pre-configured, allowing the CM mode to be determined across each USIM. Requests for CMs can be determined by the network side or by the UE side, depending on the implementation-specific use case.

[0083] Figure 11 The diagram illustrates a signaling flow for activating the CM using model splitting. To improve signaling overhead imbalance and ML processing latency, the ML model used for operations between the network and the UE can be split into M1 and M2, where M1 is the ML model to be processed via USIM1, and M2 is the ML model to be processed via USIM2. Alternatively, the datasets used for model training / inference / monitoring can be split into multiple dataset groups to be processed through each USIM link.

[0084] Key advantages of the described method include: reducing latency and signaling overhead imbalance through joint ML operations based on varying numbers of MUSIM-based radio links.

[0085] This application aims to provide fundamental interoperability mechanisms and data flow for AI / ML support in radio access network collaboration, particularly in multi-USIM-based ML operations.

[0086] Based on the proposed invention, the behavior of gNB-UE can be significantly improved in potential scenarios by supporting AI / ML operations in wireless communication through joint ML operations.

Claims

1. A method for data-driven model signaling supporting collaborative operation of multiple USIMs and AI / ML models, comprising: Enable / disable collaborative ML (CM) operations via multiple USIMs (MUSIM). Different CM types (e.g., sequential CM and parallel CM) are determined based on the UE's DSDS / DSDA support capabilities of the modem / RF chain. The different CM operation modes assigned to each USIM are executed based on the mapping relationship between CM modes and USIMs. Determine the primary CM operation mode and supplementary CM operation mode for each USIM.

2. The method of claim 1, wherein CM patterns are defined and pre-configured for a set of pattern categories, wherein the ML-based lifecycle management operations, the pattern categories may be baseline patterns or enhanced patterns, and then different CM patterns are used to index ML operations for configuration.

3. The method of claim 2, wherein the plurality of radio links may cooperate to process model operation between the UE and different gNBs across MUSIM, wherein different model operation modes may be configured.

4. The method of claim 1, wherein CM operation is enabled or disabled depending on the MUSIM configuration, wherein an indication message for enabling / disabling MUSIM-based CM is transmitted via L1 / L2 signaling.

5. The method of claim 1, wherein determining sequential CM or parallel CM based on the UE Tx / Rx capability supporting DSDS / DSDA in the modem / RF chain comprises: Execution sequence CM, where only one CM mode is run at a time when any active USIM is used as the connection mode, and any associated CM modes in the combination can be executed sequentially through the same USIM or different USIMs; Execute parallel CMs, where two or more CM modes are executed in parallel when MUSIM is simultaneously activated in connected mode, so that a combination of CM modes pre-configured with MUSIM can then be executed.

6. The method of claim 1, wherein primary CM and supplementary CM are performed for the MUSIM, wherein CM operating modes are assigned to each USSIM, and wherein CM modes are mapped to the MUSIM based on different criteria (e.g., ML support configuration and / or radio link conditions), comprising: Assign the primary CM mode to USIM1 (for forcing ML operations, such as model inference), and... Assign the supplementary CM mode to USIM2 (for mandatory / optional ML operations, such as model training). Pre-configure supplementary CM modes associated with the primary CM mode for CM mode combinations, wherein the index value of the CM mode is then indicated via DCI and / or MAC CE and / or RRC signaling.

7. The method of claim 6, wherein setting the priority level of the CM mode for the preferred CM mode combination includes: Determine the CM mode corresponding to MUSIM. Indicates the priority level of each CM mode corresponding to MUSIM. On a dedicated USIM link, primary ML operations are processed first, where the priority level of CM mode can be dynamically switched or updated. The number of pre-configured priority levels can be more than two, and each CM mode with priority is associated with the assigned USIM.

8. The method of claim 7, wherein when any used radio link becomes unavailable due to other services, determining a CM mode with priority based on the priority of the CM mode includes: Pre-configure high-priority CM mode (via the main link) and low-priority CM mode (via a non-main link). Set the CM mode priority level for the handover operations of each MUSIM link.

9. The method according to any of the preceding claims, wherein configuring DSDA / DSDS-based collaborative ML between the gNB and the UE comprises: Generate candidate combinations of CM patterns with USIM1 and USIM2 for model operations. Receive CM mode selection results from non-primary links Execute the final CM pattern combination.

10. The method according to any of the preceding claims, wherein a triggering event is used to activate CM operation, wherein a list of triggering events (e.g., traffic load) for activating CM is used to activate a CM from a non-cooperative ML employing a single USIM, and the request for CM can be determined by the network side or by the UE side for a use case specific to the implementation.

11. The method according to any of the preceding claims, wherein CM operation is activated by model splitting, wherein the ML model for operation between the network and the UE is split and the split model is processed through each MUSIM; similarly, the dataset for model training / inference / monitoring can also be split into multiple dataset groups to be processed through each USIM link.

12. An apparatus for data-driven model signaling supporting collaborative operation of multiple USIMs and AI / ML models, the apparatus comprising a wireless transceiver and a processor coupled to a memory, the memory storing computer program instructions configured to implement the steps of the method as claimed in claims 1 to 11.

13. An apparatus for data-driven model signaling supporting collaborative operation of AI / ML models implemented by a gNB for multiple USIMs, the apparatus comprising a wireless transceiver and a processor coupled to a memory, the memory storing computer program instructions configured to implement the steps of the method as claimed in claims 1 to 11.

14. A user equipment comprising the device according to claim 12.

15. A base station comprising the device according to claim 13.

16. A wireless communication system, wherein, The gNB includes a processor coupled to a memory storing computer program instructions configured to implement the steps of the method as claimed in claims 1 to 11; The user equipment (UE) includes a processor coupled to a memory storing computer program instructions configured to implement the steps of the method as described in claims 1 to 11.

Images