Method and apparatus for channel state information reporting with model inference in wireless communication systems

Abstract

Methods and apparatuses for Channel State Information (CSI) reporting with model inference are provided. The method includes receiving, from a Base Station (BS), a first CSI report configuration including at least one first parameter indicating model inference, a first CSI resource setting and a second CSI resource setting, where each of the first CSI resource setting and the second CSI resource setting includes a first association identifier, performing at least one first measurement on at least one first resource indicated by the first CSI resource setting, and transmitting, to the BS, a first CSI report containing at least one first predicted value, where the at least one first predicted value is associated with at least one second resource indicated by the second CSI resource setting.

Description

METHOD AND APPARATUS FOR CHANNEL STATE INFORMATION REPORTING WITH MODEL INFERENCE IN WIRELESS COMMUNICATION SYSTEMS

[0001] The present disclosure is related to wireless communication and, more specifically, to methods and apparatuses for Channel State Information (CSI) reporting with model inference in wireless communication systems.

[0002] Various efforts have been made to improve different aspects of wireless communication for the cellular wireless communication systems, such as the 5thGeneration (5G) New Radio (NR) system, by improving data rate, latency, reliability, and mobility. The 5G NR system is designed to provide flexibility and configurability to optimize network services and types, accommodating various use cases, such as enhanced Mobile Broadband (eMBB), massive Machine-Type Communication (mMTC), and Ultra-Reliable and Low-Latency Communication (URLLC). As the demand for radio access continues to increase, however, there exists a need for further improvements in the art.Summery of Invention

[0003] The present disclosure is related to methods and apparatuses for Channel State Information (CSI) reporting with model inference in wireless communication systems.

[0004] According to a first aspect of the present disclosure, a User Equipment (UE) is provided. The UE includes at least one processor and at least one non-transitory computer-readable medium coupled to the at least one processor and storing one or more computer-executable instructions that, when executed by the at least one processor, cause the UE to receive, from a Base Station (BS), a first Channel State Information (CSI) report configuration including at least one first parameter indicating model inference, a first CSI resource setting and a second CSI resource setting, where each of the first CSI resource setting and the second CSI resource setting includes a first association identifier, perform at least one first measurement on at least one first resource indicated by the first CSI resource setting, and transmit, to the BS, a first CSI report containing at least one first predicted value, where the at least one first predicted value is associated with at least one second resource indicated by the second CSI resource setting.

[0005] In some implementations of the first aspect of the present disclosure, the one or more computer-executable instructions, when executed by the at least one processor, further cause the UE to receive, from the BS, a second CSI report configuration including at least one second parameter indicating model inference, a third CSI resource setting and a fourth CSI resource setting, where each of the third CSI resource setting and the fourth CSI resource setting includes a second association identifier different from the first association identifier, perform at least one second measurement on at least one third resource indicated by the third CSI resource setting, and transmit, to the BS, a second CSI report containing at least one second predicted value, where the at least one second predicted value is associated with at least one fourth resource indicated by the fourth CSI resource setting.

[0006] In some implementations of the first aspect of the present disclosure, the first CSI resource setting indicates a first CSI-Reference Signal (CSI-RS) resource set, the second CSI resource setting indicates a second CSI-RS resource set, and the first association identifier associates the first CSI-RS resource set with the second CSI-RS resource set for model inference operations.

[0007] In some implementations of the first aspect of the present disclosure, the at least one first predicted value includes at least one predicted Reference Signal Received Power (RSRP) value and at least one predicted resource identifier corresponding to the at least one second resource.

[0008] In some implementations of the first aspect of the present disclosure, the at least one predicted RSRP value includes K predicted RSRP values for K highest predicted beams, and the at least one predicted resource identifier includes K predicted resource identifiers corresponding to the K highest predicted beams, where K is a positive integer indicated by the first CSI report configuration.

[0009] In some implementations of the first aspect of the present disclosure, the one or more computer-executable instructions, when executed by the at least one processor, further cause the UE to apply an Artificial Intelligence / Machine Learning (AI / ML) model using the at least one first measurement as model input to generate the at least one first predicted value for the at least one second resource.

[0010] In some implementations of the first aspect of the present disclosure, the one or more computer-executable instructions, when executed by the at least one processor, further cause the UE to determine whether the first CSI report configuration indicates spatial domain prediction or temporal domain prediction based on presence or absence of at least one temporal domain specific parameter in the first CSI report configuration.

[0011] In some implementations of the first aspect of the present disclosure, the at least one temporal domain specific parameter includes at least one of a length of a time observation window, a time duration between adjacent future time instances, or a number of inference results to report, each inference result corresponding to a different future time instance.

[0012] According to a second aspect of the present disclosure, a method performed by a User Equipment (UE) for Channel State Information (CSI) reporting with model inference is provided. The method includes receiving, from a Base Station (BS), a first CSI report configuration including at least one first parameter indicating model inference, a first CSI resource setting and a second CSI resource setting, where each of the first CSI resource setting and the second CSI resource setting includes a first association identifier, performing at least one first measurement on at least one first resource indicated by the first CSI resource setting, and transmitting, to the BS, a first CSI report containing at least one first predicted value, where the at least one first predicted value is associated with at least one second resource indicated by the second CSI resource setting.

[0013] According to a third aspect of the present disclosure, a Base Station (BS) is provided. The BS includes at least one processor and at least one non-transitory computer-readable medium coupled to the at least one processor and storing one or more computer-executable instructions that, when executed by the at least one processor, cause the BS to transmit, to a User Equipment (UE), a first Channel State Information (CSI) report configuration including at least one first parameter indicating model inference, a first CSI resource setting and a second CSI resource setting, where each of the first CSI resource setting and the second CSI resource setting includes a first association identifier, transmit at least one first reference signal on at least one first resource indicated by the first CSI resource setting, and receive, from the UE, a first CSI report containing at least one first predicted value, where the at least one first predicted value is associated with at least one second resource indicated by the second CSI resource setting.

[0014] In some implementations of the third aspect of the present disclosure, the one or more computer-executable instructions, when executed by the at least one processor, further cause the BS to transmit, to the UE, a second CSI report configuration including at least one second parameter indicating model inference, a third CSI resource setting and a fourth CSI resource setting, where each of the third CSI resource setting and the fourth CSI resource setting includes a second association identifier different from the first association identifier, transmit at least one second reference signal on at least one third resource indicated by the third CSI resource setting, and receive, from the UE, a second CSI report containing at least one second predicted value, where the at least one second predicted value is associated with at least one fourth resource indicated by the fourth CSI resource setting.

[0015] In some implementations of the third aspect of the present disclosure, the first CSI resource setting indicates a first CSI Reference Signal (CSI-RS) resource set, the second CSI resource setting indicates a second CSI-RS resource set, and the first association identifier associates the first CSI-RS resource set with the second CSI-RS resource set for model inference operations.

[0016] In some implementations of the third aspect of the present disclosure, the at least one first predicted value includes at least one predicted Reference Signal Received Power (RSRP) value and at least one predicted resource identifier corresponding to the at least one second resource.

[0017] In some implementations of the third aspect of the present disclosure, the at least one predicted RSRP value includes K predicted RSRP values for K highest predicted beams, and the at least one predicted resource identifier includes K predicted resource identifiers corresponding to the K highest predicted beams, where K is a positive integer indicated by the first CSI report configuration.

[0018] In some implementations of the third aspect of the present disclosure, the first CSI report configuration includes at least one temporal domain specific parameter when indicating temporal domain prediction, where the at least one temporal domain specific parameter includes at least one of a length of a time observation window, a time duration between adjacent future time instances, or a number of inference results to report, each inference result corresponding to a different future time instance.

[0019] Aspects of the present disclosure are best understood from the following detailed disclosure when read with the accompanying drawings. Various features are not drawn to scale. Dimensions of various features may be arbitrarily increased or reduced for clarity of discussion.

[0020] FIG. 1 is a flowchart illustrating a method / process for CSI reporting with model inference, according to an example implementation of the present disclosure.

[0021] FIG. 2 is a flowchart illustrating a method / process for handling CSI reporting, according to an example implementation of the present disclosure.

[0022] FIG. 3 is a block diagram illustrating node for wireless communications, in accordance with various aspects of the present disclosure.

[0023] Some of the abbreviations in the present application are defined as follows and, unless otherwise specified, the abbreviations have the following meanings: Abbreviation　　　　　　　　Full name AC　　　　　　　　　　　　　　　　Additional Condition AI / ML　　　　　　　　　　　　Artificial Intelligence / Machine Learning ARFCN　　　　　　　　　　　　Absolute Radio Frequency Channel Number BA　　　　　　　　　　　　　　　　Bandwidth Adaptation BFD　　　　　　　　　　　　　　　　Beam Failure Detection BFR　　　　　　　　　　　　　　　　Beam Failure Recovery BM　　　　　　　　　　　　　　　　Beam Management BS　　　　　　　　　　　　　　　　Base Station BWP　　　　　　　　　　　　Bandwidth Part CA　　　　　　　　　　　　　　　　Carrier Aggregation CBD　　　　　　　　　　　　Candidate Beam Detection CD-SSB　　　　　　　　　　　　Cell-Defining SSB CE　　　　　　　　　　　　　　　　Control Element CORESET　　　　　　　　Control Resource Set CQI　　　　　　　　　　　　　　　　Channel Quality Indicator C-RNTI　　　　　　　　　　　　Cell-Radio Network Temporary Identifier CSI　　　　　　　　　　　　　　　　Channel State Information CSI-RS　　　　　　　　　　　　Channel State Information-Reference Signal DC　　　　　　　　　　　　　　　　Dual Connectivity DCI　　　　　　　　　　　　　　　　Downlink Control Information DL　　　　　　　　　　　　　　　　Downlink DRX　　　　　　　　　　　　Discontinuous Reception FR　　　　　　　　　　　　　　　　Frequency Range ID　　　　　　　　　　　　　　　　IDentifier IE　　　　　　　　　　　　　　　　Information Elements LBT　　　　　　　　　　　　　　　　Listen Before Talk LCM　　　　　　　　　　　　LifeCycle Management MAC　　　　　　　　　　　　Medium Access Control MAC CE　　　　　　　　　　　　Medium Access Control Control Element MCG　　　　　　　　　　　　Master Cell Group MIB　　　　　　　　　　　　　　　　Master Information Block MIMO　　　　　　　　　　　　Multiple Input Multiple Output NR　　　　　　　　　　　　　　　　New Radio NW　　　　　　　　　　　　　　　　Network NW-AC　　　　　　　　　　　　Network-side Additional Condition NZP　　　　　　　　　　　　　　　　Non-Zero-Power OD-SSB　　　　　　　　　　　　On-Demand SSB PBCH　　　　　　　　　　　　Physical Broadcast CHannel PCell　　　　　　　　　　　　Primacy Cell PDCCH　　　　　　　　　　　　Physical Downlink Control CHannel PDSCH　　　　　　　　　　　　Physical Downlink Shared CHannel PHY　　　　　　　　　　　　Physical Layer PRACH　　　　　　　　　　　　Physical Random Access CHannel PS　　　　　　　　　　　　　　　　Power Saving PUCCH　　　　　　　　　　　　Physical Uplink Control CHannel PUSCH　　　　　　　　　　　　Physical Uplink Shared CHannel RA　　　　　　　　　　　　　　　　Random Access RAR　　　　　　　　　　　　Random Access Response RB　　　　　　　　　　　　　　　　Resource Block RE　　　　　　　　　　　　　　　　Resource Element Rel　　　　　　　　　　　　　　　　Release RNTI　　　　　　　　　　　　Radio Network Temporary Identifier RLM　　　　　　　　　　　　Radio Link Monitoring RRC　　　　　　　　　　　　Radio Resource Control RRM　　　　　　　　　　　　Radio Resource Management SCell　　　　　　　　　　　　Secondary Cell SCG　　　　　　　　　　　　　　　　Secondary Cell Group SCS　　　　　　　　　　　　　　　　SubCarrier Spacing SMTC　　　　　　　　　　　　SSB Measurement Timing Configuration SR　　　　　　　　　　　　　　　　Scheduling Request SS　　　　　　　　　　　　　　　　Search Space SSB　　　　　　　　　　　　　　　　Synchronization Signal Block UCI　　　　　　　　　　　　　　　　Uplink Control Information UE　　　　　　　　　　　　　　　　User Equipment UL　　　　　　　　　　　　　　　　Uplink

[0024] The following contains specific information related to implementations of the present disclosure. The drawings and their accompanying detailed disclosure are merely directed to implementations. However, the present disclosure is not limited to these implementations. Other variations and implementations of the present disclosure will be obvious to those skilled in the art.

[0025] Unless noted otherwise, like or corresponding elements among the drawings may be indicated by like or corresponding reference numerals. Moreover, the drawings and illustrations in the present disclosure are generally not to scale and are not intended to correspond to actual relative dimensions.

[0026] For consistency and ease of understanding, like features may be identified (although, in some examples, not illustrated) by the same numerals in the drawings. However, the features in different implementations may be different in other respects and shall not be narrowly confined to what is illustrated in the drawings.

[0027] References to “one implementation,” “an implementation,” “example implementation,” “various implementations,” “some implementations,” “implementations of the present application,” etc., may indicate that the implementation(s) of the present application so described may include a particular feature, structure, or characteristic, but not every possible implementation of the present application necessarily includes the particular feature, structure, or characteristic. Further, repeated use of the phrase “in one implementation,” or “in an example implementation,” “an implementation,” do not necessarily refer to the same implementation, although they may. Moreover, any use of phrases like “implementations” in connection with “the present application” are never meant to characterize that all implementations of the present application must include the particular feature, structure, or characteristic, and should instead be understood to mean “at least some implementations of the present application” includes the stated particular feature, structure, or characteristic.

[0028] The term “coupled” is defined as connected, whether directly or indirectly through intervening components, and is not necessarily limited to physical connections. The term “comprising,” when utilized, means “including, but not necessarily limited to”; it specifically indicates open-ended inclusion or membership in the so-described combination, group, series, and the equivalent.

[0029] The expression “at least one of A, B and C” or “at least one of the following: A, B and C” means “only A, or only B, or only C, or any combination of A, B and C.” The terms “system” and “network” may be used interchangeably. The term “and / or” is only an association relationship for describing associated objects and represents that three relationships may exist such that A and / or B may indicate that A exists alone, A and B exist at the same time, or B exists alone. The character “ / ” generally represents that the associated objects are in an “or” relationship.

[0030] For the purposes of explanation and non-limitation, specific details, such as functional entities, techniques, protocols, and standards, are set forth for providing an understanding of the disclosed technology. In other examples, detailed disclosure of well-known methods, technologies, systems, and architectures are omitted so as not to obscure the present disclosure with unnecessary details.

[0031] Persons skilled in the art will immediately recognize that any network function(s) or algorithm(s) disclosed may be implemented by hardware, software, or a combination of software and hardware. Disclosed functions may correspond to modules which may be software, hardware, firmware, or any combination thereof.

[0032] A software implementation may include computer executable instructions stored on a computer-readable medium, such as memory or other type of storage devices. One or more microprocessors or general-purpose computers with communication processing capability may be programmed with corresponding executable instructions and perform the disclosed network function(s) or algorithm(s).

[0033] The microprocessors or general-purpose computers may include Application-Specific Integrated Circuits (ASICs), programmable logic arrays, and / or one or more Digital Signal Processor (DSPs). Although some of the disclosed implementations are oriented to software installed and executing on computer hardware, alternative implementations implemented as firmware, as hardware, or as a combination of hardware and software are well within the scope of the present disclosure. The computer-readable medium includes but is not limited to Random Access Memory (RAM), Read Only Memory (ROM), Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), flash memory, Compact Disc Read-Only Memory (CD-ROM), magnetic cassettes, magnetic tape, magnetic disk storage, or any other equivalent medium capable of storing computer-readable instructions.

[0034] A radio communication network architecture such as a Long-Term Evolution (LTE) system, an LTE-Advanced (LTE-A) system, an LTE-Advanced Pro system, or a 5G NR Radio Access Network (RAN) typically includes at least one base station (BS), at least one UE, and one or more optional network elements that provide connection within a network. The UE communicates with the network such as a Core Network (CN), an Evolved Packet Core (EPC) network, an Evolved Universal Terrestrial RAN (E-UTRAN), a 5G Core (5GC), or an internet via a RAN established by one or more BSs.

[0035] A UE may include, but is not limited to, a mobile station, a mobile terminal or device, or a user communication radio terminal. The UE may be a portable radio equipment that includes, but is not limited to, a mobile phone, a tablet, a wearable device, a sensor, a vehicle, or a Personal Digital Assistant (PDA) with wireless communication capability. The UE is configured to receive and transmit signals over an air interface to one or more cells in a RAN. A UE may be referred to as a PHY / MAC / RLC / PDCP / SDAP entity. The PHY / MAC / RLC / PDCP / SDAP entity may be referred to as the UE.

[0036] The BS may be configured to provide communication services according to at least a Radio Access Technology (RAT) such as Worldwide Interoperability for Microwave Access (WiMAX), Global System for Mobile communications (GSM) that is often referred to as 2G, GSM Enhanced Data rates for GSM Evolution (EDGE) RAN (GERAN), General Packet Radio Service (GPRS), Universal Mobile Telecommunication System (UMTS) that is often referred to as 3G based on basic wideband-code division multiple access (W-CDMA), high-speed packet access (HSPA), LTE, LTE-A, evolved LTE (eLTE) that is LTE connected to 5GC, NR (often referred to as 5G), and / or LTE-A Pro. However, the scope of the present disclosure is not limited to these protocols.

[0037] The BS may include, but is not limited to, a node B (NB) in the UMTS, an evolved node B (eNB) in LTE or LTE-A, a radio network controller (RNC) in UMTS, a BS controller (BSC) in the GSM / GERAN, an ng-eNB in an Evolved Universal Terrestrial Radio Access (E-UTRA) BS in connection with 5GC, a next generation Node B (gNB) in the 5G-RAN, or any other apparatus capable of controlling radio communication and managing radio resources within a cell. The BS may serve one or more UEs via a radio interface.

[0038] The BS may be operable to provide radio coverage to a specific geographical area using multiple cells forming the RAN. The BS may support the operations of the cells. Each cell may be operable to provide services to at least one UE within its radio coverage.

[0039] Each cell (often referred to as a serving cell) may provide services to serve one or more UEs within its radio coverage, such that each cell schedules the DL (and optionally UL resources) to at least one UE within its radio coverage for DL (and optionally UL packet transmissions from the UE). The BS may communicate with one or more UEs in the radio communication system via the plurality of cells.

[0040] A cell may allocate sidelink (SL) resources for supporting the Proximity Service (ProSe) or Vehicle to Everything (V2X) service. Each cell may have overlapped coverage areas with other cells.

[0041] In Multi-RAT Dual Connectivity (MR-DC) cases, the primary cell of a Master Cell Group (MCG) or a Secondary Cell Group (SCG) may be referred to as a Special Cell (SpCell). A Primary Cell (PCell) may include the SpCell of an MCG. A Primary SCG Cell (PSCell) may include the SpCell of an SCG. MCG may include a group of serving cells associated with the Master Node (MN), including the SpCell and optionally one or more Secondary Cells (SCells). An SCG may include a group of serving cells associated with the Secondary Node (SN), including the SpCell and optionally one or more SCells.

[0042] As described above, the frame structure for NR supports flexible configurations for accommodating various next generation (e.g., 5G) communication requirements, such as Enhanced Mobile Broadband (eMBB), Massive Machine Type Communication (mMTC), and Ultra-Reliable and Low-Latency Communication (URLLC), while fulfilling high reliability, high data rate, and low latency requirements. The Orthogonal Frequency-Division Multiplexing (OFDM) technology in the 3GPP may serve as a baseline for an NR waveform. The scalable OFDM numerology, such as adaptive sub-carrier spacing, channel bandwidth, and Cyclic Prefix (CP), may also be used.

[0043] Two coding schemes may be considered for NR, specifically Low-Density Parity-Check (LDPC) code and Polar Code. The coding scheme adaption may be configured based on channel conditions and / or service applications.

[0044] At least the DL transmission data, a guard period, and UL transmission data should be included in a transmission time interval (TTI) of a single NR frame. The respective portions of the DL transmission data, the guard period, and the UL transmission data should also be configurable based on, for example, the network dynamics of NR. SL resources may also be provided in an NR frame to support ProSe services or V2X services.

[0045] Any two or more of the following paragraphs, (sub)-bullets, points, actions, behaviors, terms, or claims described in the present disclosure may be combined logically, reasonably, and properly to form a specific method.

[0046] Any sentence, paragraph, (sub)-bullet, point, action, behaviors, terms, or claims described in the present disclosure may be implemented independently and separately to form a specific method.

[0047] Dependency, e.g., “based on”, “more specifically”, “preferably”, “in one embodiment”, “in some implementations”, etc., in the present disclosure is just one possible example which would not restrict the specific method.

[0048] “A and / or B” in the present disclosure may refer to either A or B, both A and B, or at least one of A and B.

[0049] In this disclosure, “X / Y” may encompass the meanings of “X or Y,” “X and Y,” and “X and / or Y,” as indicated by two or more of the sentences, paragraphs, sub-bullets, points, actions, behaviors, terms, alternatives, aspects, examples, embodiments, or claims described in the following invention(s).

[0050] One aspect of the present disclosure may be applied in various contexts, including communications, communication equipment (such as mobile telephone apparatus, base station apparatus, wireless LAN apparatus, and / or sensor devices), integrated circuits (such as communication chips), and software programs, among others.

[0051] The terms “an antenna port” and “antenna ports,” as discussed in the present disclosure, may refer to “an antenna port used for transmission of PUSCH(s) / PUCCH(s)” and “antenna ports used for transmission of PUSCH(s) / PUCCH(s),” respectively.

[0052] Some of the terms, definitions, and / or abbreviations included in the present disclosure may either be sourced from existing documents (such as those from ETSI, ITU, or other sources) or may be newly created by experts from the 3GPP whenever there was a need for a precise vocabulary.

[0053] Examples of some selected terms in the present disclosure are provided as follows.

[0054] The User Equipment (UE) may be referred to as a PHY / MAC / RLC / PDCP / SDAP entity. The PHY / MAC / RLC / PDCP / SDAP entity may be referred to as the UE. In some implementations, the UE may represent any device capable of wireless communication with the network.

[0055] The Network (NW) may include a network node, a TRP, a cell (e.g., SpCell, PCell, PSCell, and / or SCell), an eNB, a gNB, and / or a base station. The network components may work together to provide comprehensive coverage and service to UEs.

[0056] The Serving Cell may be a PCell, a PSCell, or an SCell. The serving cell may be an activated or a deactivated serving cell, depending on the current configuration and operational requirements.

[0057] The Special Cell (SpCell) may have specific characteristics depending on the connectivity mode. For Dual Connectivity operation, the term Special Cell may refer to the PCell of the MCG or the PSCell of the SCG depending on whether the MAC entity is associated with the MCG or the SCG, respectively. Otherwise, the term Special Cell may refer to the PCell. A Special Cell may support PUCCH transmission and contention-based Random Access and may always be activated.

[0058] The DCI may include downlink control information, and there may be various DCI formats used in a PDCCH. The DCI format may be a predefined format in which the downlink control information may be packed / formed and transmitted in a PDCCH.

[0059] The BWP may represent a subset of the total cell bandwidth of a cell. The Bandwidth Adaptation (BA) may be achieved by configuring the UE with BWP(s) and instructing the UE which of the configured BWPs is currently the active one. To enable a BA on the PCell, the gNB may configure the UE with UL and DL BWP(s). To enable the BA on SCells in case of CA, the gNB may configure the UE with one or more DL BWPs. For the PCell, the initial BWP may be the BWP used for an initial access. For the SCell(s), the initial BWP may be the BWP configured for the UE to operate after an SCell activation. The UE may be configured with a first active uplink BWP by the firstActiveUplinkBWP IE. If the first active uplink BWP is configured for an SpCell, the firstActiveUplinkBWP IE field may contain the identifier (ID) of the UL BWP to be activated upon performing the RRC (re)configuration. If the field is absent, the RRC (re)configuration may not impose a BWP switching. If the first active uplink BWP is configured for an SCell, the firstActiveUplinkBWP IE field may contain the ID of the uplink bandwidth part to be used upon the MAC-activation of an SCell.

[0060] The TCI state may include parameters for configuring a QCL relationship between one or more DL reference signals and a target reference signal set. For example, a target reference signal set may include the DMRS ports of a PDSCH, a PDCCH, a PUCCH, or a PUSCH. The reference signals may include UL or DL reference signals. In NR Rel-15 / 16, the TCI state may be used for a DL QCL indication, whereas the spatial relation information may be used for providing the UL spatial transmission filter information for the UL signal(s) or channel(s). A TCI state may include the information similar to the spatial relation information, which may be used for UL transmission. From the UL perspective, a TCI state may provide the UL beam information that may indicate the relationship between a UL transmission and the DL or UL reference signals (e.g., the CSI-RS, the SSB, the SRS, and the PTRS).

[0061] The beam may refer to a spatial domain filtering. In some implementations, the spatial filtering may be applied in the analog domain by adjusting a phase and / or an amplitude of a signal before being transmitted by a corresponding antenna element. In other implementations, the spatial filtering may be applied in the digital domain by the MIMO technique in the wireless communication system. For example, when a UE makes a PUSCH transmission by using a specific beam, this may imply that the UE makes the PUSCH transmission by using the specific spatial / digital domain filter. The beam may also be represented as an antenna, an antenna port, an antenna element, a group of antennas, a group of antenna ports, or a group of antenna elements. The beam may also be formed by a certain reference signal resource. The beam may be equivalent to a spatial domain filter through which the EM wave is radiated.

[0062] The AI / ML Model may be a data-driven algorithm that applies AI / ML techniques to generate a set of outputs based on a set of inputs. The model may be trained using historical data to learn patterns and relationships that may enable accurate predictions.

[0063] The Network-side model may be an AI / ML Model whose inference is performed entirely at the network. The UE-side model may be an AI / ML Model whose inference is performed entirely at the UE. This distinction may be important for understanding where the computational burden resides and how the system may scale.

[0064] The AI / ML model inference may be a process of using a trained AI / ML model to produce a set of outputs based on a set of inputs. The inference process may occur in real-time to support dynamic beam management decisions.

[0065] The model performance monitoring may be a procedure that monitors the inference performance of the AI / ML model. Model performance monitoring may involve the evaluation and assessment of AI / ML models implemented on the UE side at least in BM-Case1 and BM-Case2. Model performance monitoring may be categorized into two types: Type 1 and Type 2.

[0066] Type 1 performance monitoring may include the configuration and signaling from the gNB to the UE for measurement and reporting purposes. The UE may operate differently in this context. In Option 1, which may be NW-side performance monitoring, the UE may send reports to the NW to facilitate performance metric calculations. Alternatively, in Option 2, which may be UE-assisted performance monitoring, the UE may calculate performance metric(s) and may report them to the NW based on these metric(s). Additionally, there may be indications from the NW for the UE to perform LCM operations.

[0067] Type 2 performance monitoring may involve indications, requests, or reports from the UE to the gNB for performance observation, although such communications may not always be necessary. This type may also include necessary signaling from the gNB for performance monitoring-related measurements and reports. Furthermore, when focused on UE-side model monitoring, the UE may have the authority to make decisions regarding model selection, activation, deactivation, switching, or fallback operations, along with mechanisms to help the UE determine whether a particular functionality or model remains suitable for the operations of the UE.

[0068] The AI / ML model training may be a process to train an AI / ML Model by learning the input / output relationship in a data-driven manner and obtain the trained AI / ML Model for inference. The training process may utilize historical measurement data to develop accurate prediction capabilities.

[0069] The association identifier (ID) may enable the UE to assume the similar properties of a DL Tx beam or beam set / list associated with the same Association ID. This mechanism may ensure consistency between training and inference phases.

[0070] Unless stated specifically, "model" in this disclosure may refer to "AI / ML model". Unless stated specifically, the models described in this disclosure may refer to UE-side model. Unless stated specifically, "inference" in this disclosure may refer to "model inference". Unless stated specifically, "monitoring" in this disclosure may refer to "model performance monitoring", and the monitoring may further refer to "Type 1 performance monitoring Option 2" where the UE may send reports of performance metric(s) to the NW.

[0071] In the present disclosure, although the term “gNB” may have been used throughout the document, it should be understood that the term “gNB” may be replaced by any other type of BS (e.g., an eNB).

[0072] A Synchronization Signal Block (SSB) may include, or consist of, a Primary Synchronization Signal (PSS), a Secondary Synchronization Signal (SSS), and a Physical Broadcast Channel (PBCH) payload. The PSS and the SSS may be pseudo-random sequences with a length equal to 127. The PBCH payload may include a Master Information Block (MIB) (24 bits in total) and an 8 bits payload. The information in the MIB may include: a System Frame Number (SFN), a sub-carrier space (SCS), a DeModulation Reference Signal (DMRS) configuration, an Access Control, and mainly a configuration for a System Information Block 1 (SIB1) acquisition. By decoding the PSS and the SSS, a User Equipment (UE) may be able to identify a Physical Cell Identity (PCI) for a corresponding cell and may determine the symbol boundary. Consequently, with the decoding of the PBCH, the UE may determine the frame boundary and may try to decode the Physical Downlink Control Channel

[0073] CSI Framework

[0074] The Channel State Information (CSI) framework may include several configurations that may enable comprehensive channel measurement and reporting capabilities in wireless communication systems.

[0075] The NZP-CSI-RS resource (NZP-CSI-RS-Resource) may define the time-frequency resources of the NZP-CSI-RS resource within a slot. In some implementations, these resources may be organized to provide optimal coverage and measurement accuracy for the UE.

[0076] The NZP-CSI-RS resource set (NZP-CSI-RS-ResourceSet) may include a set of NZP-CSI-RS resources. The resource set may provide flexibility in configuring multiple resources for different measurement purposes.

[0077] The SSB CSI resource set may include a set of SSBs to be used for measurement. The SSB resources may enable the UE to perform synchronization and initial beam acquisition procedures.

[0078] The CSI resource setting (CSI-ResourceConfig) may include a set of NZP-CSI-RS resource sets for channel measurement. In some implementations, the CSI resource setting may serve as a container that may group related resources for specific measurement operations.

[0079] The CSI report configuration (CSI-ReportConfig) may define how the UE may report the CSI to the network. The configuration may specify various parameters that may control the reporting behavior and content.

[0080] The time-domain property of each NZP-CSI-RS resource and / or NZP-CSI-RS resource set and / or CSI resource setting may be one of periodic, semi-persistent, and aperiodic. In some implementations, the periodic configuration may enable regular measurements at fixed intervals, while the semi-persistent configuration may provide flexibility through activation and deactivation mechanisms. The aperiodic configuration may allow on-demand measurements triggered by the network.

[0081] AI / ML for Air Interface (AMA) for BM

[0082] In the realm of air interface design, Artificial Intelligence (AI) and Machine Learning (ML) may significantly enhance beam management for advanced communication systems, particularly in millimeter-wave (mmWave) communications. The effective beam management may involve predicting optimal beam alignments to ensure robust signal strength and coverage.

[0083] In some implementations, two types of resource sets may be defined for AMA BM. Set A may represent the target beam set that the UE may need to predict, while Set B may represent the measurement beam set that may provide input data for the AI / ML model. A UE capable of AMA BM may perform measurement on Set B and may use the UE-side model to infer the best beam(s) in Set A. This approach may enable efficient beam prediction without requiring exhaustive measurements on all possible beams.

[0084] The AMA BM may support two distinct use cases that may address different operational scenarios. In BM-Case1, the AI / ML techniques may predict optimal beam directions for Set A of beams in the spatial domain, using measurement results from Set B of beams. This spatial domain prediction may be particularly useful for stationary or slowly moving UEs where the primary challenge may be identifying the best beam direction. In contrast, BM-Case2 may predict beam changes for Set A of beams in the temporal domain, utilizing historical data and usage patterns from Set B of beams to enable timely adjustments to beam configurations.

[0085] Timers

[0086] The timer mechanism may provide precise control over various time-sensitive operations in the communication system. A timer may be running once the timer is started, until the timer is stopped or until the timer expires; otherwise, the timer may not be running. A timer may be started if the timer is not running or may be restarted if the timer is running. A timer may always be started or restarted from the initial value of the timer. The duration of a timer may not be updated until the timer is stopped or expires (e.g., due to BWP switching).

[0087] When the MAC entity applies zero value for a timer, the timer may be started and may immediately expire unless explicitly stated otherwise. Similarly, when the UE applies zero value for a timer, the timer may be started and may immediately expire unless explicitly stated otherwise. This behavior may ensure deterministic operation across different implementations.

[0088] CSI Framework to Support Beam Management for AI / ML for Interface

[0089] The integration of AI / ML capabilities into the CSI framework may represent a significant advancement in wireless communication technology. This integration may enable intelligent beam management that may adapt to changing channel conditions and user mobility patterns.

[0090] In some implementations, a CSI report configuration may indicate one or more of the following cases: model inference for BM-Case1, model performance monitoring for BM-Case1, model inference for BM-Case2, and model performance monitoring for BM-Case2. The flexibility to support multiple cases may allow the network to optimize the configuration based on specific deployment scenarios and performance requirements.

[0091] In some implementations, a single CSI report configuration may indicate that the CSI report is for both model inference and model performance monitoring together for either BM-Case1 or BM-Case2. This combined approach may reduce signaling overhead while ensuring comprehensive monitoring of AI / ML operations. If a UE is configured with such CSI report configuration, the UE may send one or more CSI reports to the NW based on the CSI report configuration, where a single CSI report may include inference result(s) and / or performance metric(s) of a single model or multiple models. Whether the report is for a single model or multiple models may be up to UE implementation, providing flexibility for different UE capabilities.

[0092] In some implementations, to use the CSI framework to support AMA BM, each operation may be indicated by a CSI report configuration in which some specific parameters are configured as a specific combination. This parameter-based approach may enable precise control over AI / ML operations without requiring extensive signaling changes.

[0093] In some implementations, a "resource set" in this disclosure may be provided in a CSI resource setting, e.g., CSI-ResourceConfig. A CSI resource setting may include one or more NZP CSI-RS resource sets (NZP-CSI-RS-ResourceSet) and / or one or more SSB resources (CSI-SSB-ResourceSet). A resource set may include one or more resources. "Resources" in this disclosure may be provided by a subset of the resources of a resource set, allowing for flexible resource allocation.

[0094] Design on CSI Report Configuration and CSI Resource Setting

[0095] The design of CSI report configuration and resource setting may form the foundation for enabling AI / ML-enhanced beam management in wireless systems. This design may need to balance flexibility, efficiency, and backward compatibility.

[0096] In some implementations, the NW may configure a CSI report configuration to a UE to configure the UE to perform AI / ML operation(s), e.g., training data collection, model inference, model performance monitoring. The CSI report configuration may include one or more parameters / contents that may precisely define the required operations.

[0097] In some implementations, the CSI report configuration may include the resource set for Set B. The resource set may be provided by an RRC parameter (e.g., resourcesForChannelMeasurement). To generate inference result(s), e.g., model output from UE side model(s), the UE may perform measurement on the resources in the resource set for Set B, with the measurement results as model input. This measurement-to-inference pipeline may enable real-time beam prediction. To generate performance metric(s), the UE may at least perform measurement on the resources in the resource set for Set B, with the measurement results as model input and / or may calculate the performance metric(s) based on the model output / inference result(s) and the measurement results performed on the resources or resource set for monitoring.

[0098] In some implementations, the CSI report configuration may include the resource set for Set A. The resource set may be provided by a new parameter, e.g., separate from the resource set for Set B. The structure of this new parameter may be CSI-ResourceConfig. The resource set configuration associated with Set A or Set B may include an association ID that may maintain consistency between different operational phases.

[0099] In some implementations, the CSI report configuration may include the resource set or resources for monitoring. The resource set or resources may be provided by a new parameter, e.g., a parameter other than the existing parameter (e.g., resourcesForChannelMeasurement). The resource set configuration associated with the resource set or resources for monitoring may include an association ID to ensure proper correlation with training data.

[0100] In some implementations, the resources or resource set for monitoring may be a subset of the resource set for Set A. The subset information may be indicated by one or more parameters provided in the CSI report configuration. This subset relationship may optimize monitoring efficiency by focusing on the most relevant resources.

[0101] In some implementations, the one or more parameters indicating the subset information may be a bitmap, where each bit in the bitmap may indicate a resource within the resource set for Set A. The size of the bitmap may be a predefined value or may be the size of the resource set for Set A, providing flexibility for different deployment scenarios.

[0102] In some implementations, the one or more parameters indicating the subset information may be a list of integers, where each integer in the list may indicate a resource within the resource set for Set A and may include the resource set or resources for monitoring. For example, an integer in the list may be the order in the resource set for Set A, enabling precise resource selection.

[0103] In some implementations, the one or more parameters indicating the subset information may be an integer number (e.g., N) indicating the number of the resources within the resource set for Set A. The UE may select N resources within the resource set for Set A via UE implementation, allowing for implementation-specific optimizations.

[0104] In some implementations, if the subset information is provided, the UE may determine that this CSI report configuration indicates the resources used for monitoring. This determination may trigger specific monitoring procedures at the UE.

[0105] In some implementations, the resources or resource set for monitoring may be totally or exactly the same as the resource set for Set A. This configuration may simplify the monitoring process when comprehensive performance evaluation is required.

[0106] In some implementations, the resource set for Set A and / or resource set for Set B and / or resources / resource set for monitoring may be configured within a single CSI resource setting, e.g., CSI-ResourceConfig. This unified configuration approach may reduce signaling overhead and simplify resource management. In some implementations, the RRC configuration structure may follow specific alternatives as described below.

[0107] As a first alternative configuration approach (Alt 1), the resource set for Set A and the resource set or resources for model performance monitoring may be configured by the same parameter. This approach may minimize the configuration complexity while maintaining flexibility. The configuration structure shown in Table 1 may illustrate how the csi-RS-ResourceSetListSetAandMonitoring-r19 parameter may be integrated into the existing CSI-ResourceConfig structure. The structure may include the nzp-CSI-RS-SSB-r19 sequence that may contain both NZP CSI-RS resource sets and SSB resource sets, providing comprehensive resource configuration capabilities. The optional nature of these parameters, indicated by "OPTIONAL -- Need R", may ensure backward compatibility with existing implementations that may not support AI / ML operations.

[0108] In some implementations, the RRC IE / field / parameter (e.g., csi-RS-ResourceSetList) may be the resource set for Set B. The UE may use this resource set to perform measurements that serve as the input for the AI / ML model inference process.

[0109] In some implementations, if the RRC IE / field / parameter (e.g., csi-RS-ResourceSetListSetAandMonitoring) is provided in the CSI resource setting, the UE may determine that this CSI resource setting provides the resource set for Set A and the resources or resource set for monitoring. This unified configuration approach may enable the NW to efficiently indicate both the target resource set for inference results and the resources for performance evaluation within a single parameter structure.

[0110] In some implementations, the RRC IE / field / parameter (e.g., csi-RS-ResourceSetListSetAandMonitoring-r19) may be the resource set for Set A when the CSI report configuration is used for inference. For example, the CSI report configuration may be included in the inference configuration, or the CSI report configuration may indicate that the purpose is for inference. When configured for inference operation, the UE may interpret this parameter as identifying the resources where the predicted beams from the AI / ML model output will be mapped.

[0111] In some implementations, the same RRC IE / field / parameter (e.g., csi-RS-ResourceSetListSetAandMonitoring-r19) may be the resource set or resources for monitoring when the CSI report configuration is used for monitoring. For example, the CSI report configuration may be included in the AI / ML model monitoring configuration, or the CSI report configuration may indicate that the purpose is for AI / ML model monitoring. This dual-purpose design may allow the NW to reuse the same parameter structure for different operational modes, thereby reducing signaling overhead.

[0112] As a second alternative configuration approach (Alt 2), the resources or resource set for monitoring and the resource set for Set A may be provided by separate NZP CSI-RS resource sets within a single IE (e.g., CSI-ResourceConfig), as illustrated in Table 2. This separation may provide more flexibility in resource allocation, allowing the NW to independently configure the monitoring resources and the Set A resources based on different performance requirements or network conditions. The separate configuration may enable the UE to distinguish between resources used for inference output mapping and resources specifically designated for performance metric calculation.

[0113] In some implementations, if the RRC IE / field / parameter (e.g., csi-RS-ResourceSetListSetA) is provided in the CSI resource setting, the UE may determine that this CSI resource setting provides the resource set for Set A. The presence of this parameter may enable the UE to identify the specific resources where the AI / ML model inference results will be mapped for beam prediction operations.

[0114] In some implementations, if the RRC IE / field / parameter (e.g., csi-RS-ResourceSetListMonitoring) is provided in the CSI resource setting, the UE may determine that this CSI resource setting provides the resource set for monitoring. The configuration of this separate parameter may allow the NW to explicitly designate resources for performance metric calculation, independent of the Set A resources used for inference output.

[0115] In some implementations, the RRC IE / field / parameter (e.g., csi-RS-ResourceSetListSetA-r19) may be the resource set for Set A when the CSI report configuration is used for inference. For example, the CSI report configuration may be included in the inference configuration, or the CSI report configuration may indicate that the purpose is for inference. The UE may utilize this resource set to map the beam predictions generated by the AI / ML model, where the predicted beams correspond to resources within Set A based on measurements performed on Set B.

[0116] In some implementations, the RRC IE / field / parameter (e.g., csi-RS-ResourceSetListMonitoring-r19) may be the resource set or resources for monitoring when the CSI report configuration is used for monitoring. For example, the CSI report configuration may be included in the AI / ML model monitoring configuration, or the CSI report configuration may indicate that the purpose is for AI / ML model monitoring. The UE may perform measurements on these monitoring resources to calculate performance metrics by comparing the actual measurements with the inference results.

[0117] As a third alternative configuration approach (Alt 3), the resources or resource set for monitoring may be a subset of the resource set for Set A. The subset information may be indicated by one or more parameters provided in the CSI report configuration, as shown in Table 3. The subset-based approach may provide flexibility in selecting specific resources from Set A for performance evaluation while reducing the signaling overhead compared to configuring completely separate resource sets. The NW may dynamically adjust the monitoring scope by modifying the subset parameters without reconfiguring the entire Set A resource set. The subset indication may enable efficient performance monitoring by focusing on the most relevant resources for model evaluation while maintaining consistency with the inference operation resources.

[0118] In some implementations, the one or more parameters indicating the subset information, such as the subsetInformationForMonitoring parameter, may be a bitmap, where each bit in the bitmap may represent a resource within the resource set for Set A. The size of the bitmap may be a predefined value or may be the size of the resource set for Set A. This bitmap configuration may allow the network to efficiently indicate which specific resources within Set A may be used for monitoring purposes.

[0119] In some implementations, the one or more parameters indicating the subset information may be a list of integers, where each integer in the list may indicate a resource within the resource set for Set A and may form part of the resource set or resources for monitoring. In some implementations, an integer in the list may represent the order of the resource in the resource set for Set A. This list-based approach may provide flexibility in selecting non-contiguous resources for monitoring.

[0120] In some implementations, the one or more parameters indicating the subset information may be an integer number, denoted as N, which may indicate the number of the resources within the resource set for Set A. The UE may select N resources within the resource set for Set A via UE implementation. This approach may allow the UE to have autonomy in selecting the most appropriate resources for monitoring based on the current channel conditions or other implementation-specific criteria.

[0121] In some implementations, the parameter indicating the subset information may be an integer, and the UE may consider the resources or resource sets whose order in the resource list for Set A may be before or after the integer to be the resources or resource sets for monitoring. In some implementations, if the integer is 7, then the UE may consider the first to the seventh resources or resource sets, or alternatively the seventh to the last resources or resource sets, in the resource list for Set A to be the resources or resource sets for monitoring.

[0122] In some implementations, the parameter indicating the subset information may be an integer, and the UE may consider the resources or resource sets in the resource list for Set A whose resource or resource set IDs may be less than or greater than the integer to be the resources or resource sets for monitoring. In some implementations, if the integer is 7 and Set A includes resource sets with ID values 4, 6, 8, and 10, then the UE may consider the resource sets with ID 4 and 6, or alternatively with ID 8 and 10, to be the resources or resource sets for monitoring.

[0123] In some implementations, the parameter may indicate the number of resources K as subset information, and the UE may select the K resources from Set A for monitoring. The selection criteria may be up to UE implementation or may be further configured by other parameters. The other parameters may include an ID where each ID may represent one selection approach. The selection approach may include random selection, Top K predicted L1-RSRP selection, and other selection methods that may be appropriate for the current network conditions.

[0124] In some implementations, if the subset information is provided in a CSI report configuration, the UE may determine that this CSI report setting may indicate the resources used for monitoring. If a CSI report configuration includes the subset information, the CSI RS resources or CSI RS resource set associated with the CSI report configuration may be implicitly indicated to a UE for model monitoring.

[0125] In some implementations, if an RRC IE / field / parameter, such as the csi-RS-ResourceSetListSetAorMonitoring-r19 IE, is provided, the RRC IE / field / parameter may be the resource set for Set A when the CSI report configuration is used for inference. The CSI report configuration may be included in the inference configuration, or the CSI report configuration may indicate that the purpose is for inference, though not limited to these scenarios.

[0126] In some implementations, if the csi-RS-ResourceSetListSetAorMonitoring-r19 IE and the subset information for monitoring are provided, the csi-RS-ResourceSetListSetAorMonitoring-r19 IE may be the resource set or resources for monitoring when the CSI report configuration is used for monitoring. The CSI report configuration may be included in the AI / ML model monitoring configuration, or the CSI report configuration may indicate that the purpose is for AI / ML model monitoring, though not limited to these scenarios.

[0127] In some implementations, the subset of Set A for monitoring and the components may vary, and the adaptation may rely on RRC reconfiguration or some specific conditions. In some implementations, the parameter may indicate more than one value, such as K1 and K2, so that the subset size may change, and the UE may perform monitoring based on different resource sets while the UE is in DRX active or DRX inactive period. This dynamic adaptation may allow the system to optimize monitoring resources based on the UE's current power state.

[0128] Report Quantity Indicator Configurations

[0129] In some implementations, the CSI report configuration may include a report quantity indicator that may specify the type of AI / ML operation to be performed.

[0130] In some implementations, following Options 1-2, the report quantity indicator may indicate one report quantity among (1) model inference for BM-Case1, (2) model performance monitoring for BM-Case1, (3) model inference for BM-Case2, and (4) model performance monitoring for BM-Case2. This configuration may allow the network to precisely specify which AI / ML operation the UE may perform.

[0131] In some implementations, following Options 1-3 and 3-3, the report quantity indicator may indicate one report quantity among (1) BM-Case1, (2) BM-Case2, or (3) both cases. This approach may provide flexibility in beam management scenarios where both spatial and temporal predictions may be beneficial.

[0132] In some implementations, following Option 1-4, the report quantity indicator may indicate one report quantity among (1) inference and (2) monitoring, which may simplify the configuration while maintaining operational flexibility.

[0133] In some implementations, following Option 2-1, the report quantity indicator may indicate the report quantity "inference and monitoring" as a combined operation, recognizing that model performance may benefit from concurrent inference and monitoring activities.

[0134] In some implementations, following Option 2-2, the report quantity indicator may indicate one report quantity among (1) model inference and model performance monitoring for BM-Case1, and (2) model inference and model performance monitoring for BM-Case2.

[0135] In some implementations, following Option 3-2, the report quantity indicator may indicate one report quantity among six possible combinations: (1) model inference for BM-Case1, (2) model performance monitoring for BM-Case1, (3) model inference and model performance monitoring for BM-Case1, (4) model inference for BM-Case2, (5) model performance monitoring for BM-Case2, and (6) model inference and model performance monitoring for BM-Case2.

[0136] In some implementations, following Option 3-4, the report quantity indicator may indicate one report quantity among (1) inference, (2) monitoring, and (3) inference and monitoring combined.

[0137] In some implementations, following Option 4-1, the report quantity indicator may indicate one report quantity among (1) AI / ML functionality deactivated and (2) AI / ML functionality activated, which may provide a high-level control mechanism for AI / ML operations.

[0138] In some implementations, the CSI report configuration may include an association ID. The association ID may represent one or a set of Network-side Additional Conditions (NW-ACs). This association ID may facilitate the linking between Set A and Set B resources to ensure consistency between training and inference phases.

[0139] In some implementations, the CSI report configuration may include one or more BM-Case2-related or BM-Case2-specific parameters that may be essential for temporal domain predictions.

[0140] In some implementations, the one or more BM-Case2-related or BM-Case2-specific parameters may include a length of time observation window. The length of time observation window may be used to indicate the time interval during which the UE may perform measurement on CSI-RS or SSB and may derive the inference results from the measurement results via AI / ML model for BM-Case2.

[0141] In some implementations, the one or more BM-Case2-related or BM-Case2-specific parameters may include a time duration between two adjacent future time instances. The time unit of the time duration may be symbols, slots, or subframes, though not limited to these units.

[0142] In some implementations, the one or more BM-Case2-related or BM-Case2-specific parameters may include a number of inference results to report, where each inference result may be in a future time instance, denoted by N. The UE may report the inference results related to each future time instance, such as predicted Top-K beams where K is greater than or equal to 1, and the corresponding L1-RSRP values. In some implementations, the UE may report N inference results to the gNB, where each inference result may include the information related to each future time inference, such as predicted Top-K beams and the corresponding L1-RSRP values.

[0143] In some implementations, the one or more BM-Case2-related or BM-Case2-specific parameters may include a maximum number of inference results to report, where each inference result may be in a future time instance, denoted by N. The UE may determine the number of inference results by the implementation of the UE but without exceeding the value N.

[0144] In some implementations, the CSI report configuration may include a parameter or a set of parameters indicating the type of this CSI report configuration. The type of the CSI report may be periodic, semi-persistent on PUCCH, semi-persistent on PUSCH, or aperiodic.

[0145] In some implementations, if the CSI report configuration is periodic, the UE may send the CSI report to the NW periodically, where the periodicity may be indicated by the CSI report configuration.

[0146] In some implementations, if the CSI report configuration is semi-persistent on PUCCH, the UE may receive a MAC CE indicating activation of the CSI report configuration from the NW, and then the UE may send the CSI report to the NW periodically until the UE receives another MAC CE indicating deactivation of the CSI report configuration, where the periodicity may be indicated by the CSI report configuration.

[0147] In some implementations, if the CSI report configuration is semi-persistent on PUSCH, the UE may receive a DCI that may activate the CSI report configuration, and then the UE may send the CSI report to the NW periodically until the UE receives another DCI indicating deactivation of the CSI report configuration, where the periodicity may be indicated by the CSI report configuration.

[0148] In some implementations, if the CSI report configuration is aperiodic, the UE may receive a DCI that may trigger the CSI report configuration, and then the UE may send the CSI report after an offset indicated by the CSI report configuration.

[0149] In some implementations, the CSI report configuration may include one or more parameters indicating the type of this CSI report configuration for AI / ML operations, such as model inference and model performance monitoring. The type of the CSI report may be periodic, semi-persistent on PUCCH, semi-persistent on PUSCH, or aperiodic.

[0150] In some implementations, if the CSI report configuration indicates both model inference and model performance monitoring, the one or more parameters may indicate the types of CSI report used for model inference and model performance monitoring, respectively. The type of CSI report for model inference and the type of CSI report for model performance monitoring may be configured independently. In some implementations, the type of CSI report for model inference and the type of CSI report for model performance monitoring may be the same or different, providing flexibility in how these operations are scheduled and reported.

[0151] In some implementations, if the CSI report is periodic, the UE may send the CSI report to the NW periodically, where the periodicity may be indicated by the CSI report configuration.

[0152] In some implementations, if the CSI report is semi-persistent on PUCCH, the UE may receive a MAC CE indicating activation of the CSI report configuration from the NW, and then the UE may send the CSI report to the NW periodically until the UE receives another MAC CE indicating deactivation of the CSI report configuration, where the periodicity may be indicated by the CSI report configuration.

[0153] In some implementations, if the CSI report is semi-persistent on PUSCH, the UE may receive a DCI that may activate the CSI report configuration, and then the UE may send the CSI report to the NW periodically until the UE receives another DCI indicating deactivation of the CSI report configuration, where the periodicity may be indicated by the CSI report configuration.

[0154] In some implementations, if the CSI report is aperiodic, the UE may receive a DCI that may trigger the CSI report configuration, and then the UE may send the CSI report after an offset indicated by the CSI report configuration.

[0155] In some implementations, the UE may recognize, consider, or know that the CSI report configuration indicates one or more of the AI / ML operations, such as model inference and model performance monitoring, if a specific combination of parameters in the CSI report configuration is configured. This recognition mechanism may enable the UE to properly interpret the network's intent and perform the appropriate AI / ML operations.

[0156] In some implementations, a single CSI report configuration may indicate the UE to perform model inference and model performance monitoring, which may streamline the configuration process and ensure coordinated operation of these related functions.

[0157] In some implementations, a single CSI report configuration may explicitly indicate the UE to perform model inference or performance monitoring, or both, based on the CSI RS resources or CSI RS resource sets included in the CSI report configuration. In some implementations, there may be an indicator included in the CSI report that may indicate the UE to perform model inference, model performance monitoring, or both model inference and model performance monitoring.

[0158] In some implementations, if the CSI report configuration indicates model inference for BM-Case1, the UE may perform measurement on the resource set provided by the parameter, such as the resourcesForChannelMeasurement parameter, indicating the resource set for Set B in the CSI report configuration. The UE may use the measurement results as model input and may generate model output, such as inference results, which may be the beam IDs or the differential L1-RSRP values of the predicted beams. The UE may report the inference results to the NW via a CSI report.

[0159] In some implementations, if the CSI report configuration indicates model inference for BM-Case2, the UE may perform measurement on the resource set provided by the parameter, such as the resourcesForChannelMeasurement parameter, indicating the resource set for Set B in the CSI report configuration. The UE may use the measurement results as model input and may generate model output as inference results, which may be the beam IDs or the differential L1-RSRP values of the predicted beams for future time instances.

[0160] In some implementations, if the CSI report configuration indicates model performance monitoring, the UE may perform measurement on the resource set for Set B or the resource set or resources for monitoring, or both, provided by one or more parameters in the CSI report configuration or in the CSI resource setting indicated by the CSI report configuration, following the design in this disclosure. The UE may use the measurement results of the resource set or resources for monitoring and the inference results generated based on the measurement results of the resource set for Set B to calculate performance metrics. These performance metrics may provide valuable feedback to the network about the accuracy and reliability of the AI / ML model predictions, enabling continuous improvement of the beam management system.

[0161] In the following sections, the methods to indicate the combination of BM-Case1 / BM-Case2 and inference / monitoring operations will be addressed. These indication mechanisms may provide the flexibility needed for the network to configure AI / ML-enhanced beam management operations according to different deployment scenarios and performance requirements.

[0162] Option 1: Configurability between {BM-Case1, BM-Case2} and between {inference, monitoring}

[0163] In some implementations, when a UE receives a CSI report configuration from the NW, the UE may need to know which operation among model inference for BM-Case1, model performance monitoring for BM-Case1, model inference for BM-Case2, and model performance monitoring for BM-Case2 the NW indicates the UE to perform by the CSI report configuration. This determination may be essential for the UE to properly execute the intended AI / ML operations and provide the appropriate feedback to the network.

[0164] Option 1-1: Implicit indication to differentiate {BM-Case1, BM-Case2} and {inference, monitoring}

[0165] In some implementations, the NW may configure a CSI report configuration to a UE. The NW may indicate the UE to perform BM-Case1 or BM-Case2 by implicit indication in the CSI report configuration and may indicate the UE to perform model inference or model performance monitoring by an implicit indication in the CSI report configuration. This implicit indication approach may reduce signaling overhead while maintaining configuration flexibility.

[0166] In some implementations, regarding BM-Case1 / BM-Case2 implicit differentiation, if one or more BM-Case2-related or BM-Case2-specific parameters are provided or present in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform AI / ML operations for BM-Case2, or the UE may determine to perform AI / ML operations for BM-Case2. In some implementations, if BM-Case2-related or BM-Case2-specific parameters are absent in the CSI report configuration, or if the UE has already initiated or activated AI / ML functionalities or models, or been indicated by the NW to activate AI / ML functionalities or models, or if the UE has applicable AI / ML functionalities or models, the UE may determine or consider that the CSI report configuration indicates the UE to perform AI / ML operations for BM-Case1, or the UE may determine to perform AI / ML operations for BM-Case1.

[0167] In some implementations, as a first alternative for inference / monitoring implicit differentiation, the resource set or resources for monitoring may be a subset of the resource set for Set A. The subset information may follow the design principles described in this disclosure. The subset information may be indicated by a parameter, such as a bitmap or a list of integers. The subset information may be indicated in the CSI report configuration or a CSI resource setting.

[0168] In some implementations, for inference only operation, if the resource set for Set A is provided and the resource set or resources for monitoring, namely the subset information, is not provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference only, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0169] In some implementations, for monitoring only operation, if the resource set for Set A is not provided and the resource set or resources for monitoring, namely the subset information, is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring only, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0170] In some implementations, as a second alternative for inference / monitoring implicit differentiation, the resource set or resources for monitoring and the resource set for Set A may be separately provided. The resource set or resources for monitoring and the resource set for Set A may be configured in the CSI report configuration or in the CSI resource setting, such as the CSI-ResourceConfig IE, indicated in the CSI report configuration.

[0171] In some implementations, for inference only operation under the second alternative, if the resource set for Set A is provided and the resource set or resources for monitoring is not provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference only, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0172] In some implementations, for monitoring only operation under the second alternative, if the resource set for Set A is not provided and the resource set or resources for monitoring is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring only, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0173] Option 1-2: Explicit indication to differentiate {BM-Case1, BM-Case2} and {inference, monitoring}

[0174] In some implementations, the NW may configure a CSI report configuration to a UE. The NW may indicate the UE to perform BM-Case1 or BM-Case2 by explicit indication in the CSI report configuration and may indicate the UE to perform model inference or model performance monitoring by an explicit indication in the CSI report configuration. This explicit indication approach may provide clear and unambiguous signaling of the intended AI / ML operations.

[0175] In some implementations, for BM-Case1 with inference operation, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a first value, such as 'bm1-inference', indicating model inference for BM-Case1, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference for BM-Case1, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0176] In some implementations, for BM-Case1 with monitoring operation, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a second value, such as 'bm1-monitoring', indicating model performance monitoring for BM-Case1, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring for BM-Case1, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0177] In some implementations, for BM-Case2 with inference operation, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a third value, such as 'bm2-inference', indicating model inference for BM-Case2, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference for BM-Case2, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0178] In some implementations, for BM-Case2 with monitoring operation, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a fourth value, such as 'bm2-monitoring', indicating model performance monitoring for BM-Case2, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring for BM-Case2, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0179] Option 1-3: Explicit indication to differentiate {BM-Case1, BM-Case2} and implicit indication to differentiate {inference, monitoring}

[0180] In some implementations, the NW may configure a CSI report configuration to a UE. The NW may indicate the UE to perform BM-Case1 or BM-Case2 by explicit indication in the CSI report configuration and may indicate the UE to perform model inference or model performance monitoring by an implicit indication in the CSI report configuration. This hybrid approach may balance signaling efficiency with configuration clarity.

[0181] In some implementations, for explicit BM-Case1 indication, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a first value, such as 'bm1', indicating operations for BM-Case1, the UE may determine or consider that the CSI report configuration indicates the UE to perform operations for BM-Case1, or the UE may send one or more CSI reports including relevant results to the NW based on the CSI report configuration.

[0182] In some implementations, for explicit BM-Case2 indication, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a second value, such as 'bm2', indicating operations for BM-Case2, the UE may determine or consider that the CSI report configuration indicates the UE to perform operations for BM-Case2, or the UE may send one or more CSI reports including relevant results to the NW based on the CSI report configuration.

[0183] In some implementations, as a first alternative for inference / monitoring implicit differentiation under Option 1-3, the resource set or resources for monitoring may be a subset of the resource set for Set A. The subset information may follow the design principles in this disclosure. The subset information may be indicated by a parameter, such as a bitmap or a list of integers. The subset information may be indicated in the CSI report configuration or a CSI resource setting.

[0184] In some implementations, for inference only operation under Option 1-3, if the resource set for Set A is provided and the resource set or resources for monitoring, namely the subset information, is not provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference only, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0185] In some implementations, for monitoring only operation under Option 1-3, if the resource set for Set A is not provided and the resource set or resources for monitoring, namely the subset information, is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring only, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0186] In some implementations, for combined inference and monitoring operation under Option 1-3, if the resource set for Set A is provided and the resource set or resources for monitoring, namely the subset information, is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0187] In some implementations, as a second alternative for inference / monitoring implicit differentiation under Option 1-3, the resource set or resources for monitoring and the resource set for Set A may be separately provided. The resource set or resources for monitoring and the resource set for Set A may be configured in the CSI report configuration or in the CSI resource setting, such as the CSI-ResourceConfig IE, indicated in the CSI report configuration.

[0188] In some implementations, for inference only operation under the second alternative of Option 1-3, if the resource set for Set A is provided and the resource set or resources for monitoring is not provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference only, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0189] In some implementations, for monitoring only operation under the second alternative of Option 1-3, if the resource set for Set A is not provided and the resource set or resources for monitoring is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring only, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0190] In some implementations, for combined inference and monitoring operation under the second alternative of Option 1-3, if the resource set for Set A is provided and the resource set or resources for monitoring is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0191] Option 1-4: Implicit indication to differentiate {BM-Case1, BM-Case2} and explicit indication to differentiate {inference, monitoring}

[0192] In some implementations, the NW may configure a CSI report configuration to a UE. The NW may indicate the UE to perform BM-Case1 or BM-Case2 by implicit indication in the CSI report configuration and may indicate the UE to perform model inference or model performance monitoring by an explicit indication in the CSI report configuration. This approach may provide precise control over the type of AI / ML operation while allowing flexible beam management case determination based on the presence of specific parameters.

[0193] In some implementations, continuing from Option 1-4, regarding BM-Case1 / BM-Case2 implicit differentiation, if one or more BM-Case2-related or BM-Case2-specific parameters are provided or present in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform AI / ML operations for BM-Case2, or the UE may determine to perform AI / ML operations for BM-Case2. In some implementations, if BM-Case2-related or BM-Case2-specific parameters are absent in the CSI report configuration, or if the UE has already initiated or activated AI / ML functionalities or models, or been indicated by the NW to activate AI / ML functionalities or models, or if the UE has applicable AI / ML functionalities or models, the UE may determine or consider that the CSI report configuration indicates the UE to perform AI / ML operations for BM-Case1, or the UE may determine to perform AI / ML operations for BM-Case1.

[0194] In some implementations, for explicit inference / monitoring differentiation, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a first value, such as 'inference', indicating model inference, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference operation, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration. In some implementations, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a second value, such as 'monitoring', indicating model performance monitoring, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring operation, or the UE may send one or more CSI reports including model performance metrics to the NW based on the CSI report configuration.

[0195] In some implementations, for BM-Case1 with inference operation, if BM-Case2-related or BM-Case2-specific parameters are absent in the CSI report configuration and if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a first value, such as 'inference', indicating model inference, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference for BM-Case1, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0196] In some implementations, for BM-Case1 with monitoring operation, if BM-Case2-related or BM-Case2-specific parameters are absent in the CSI report configuration and if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a second value, such as 'monitoring', indicating model performance monitoring, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring, such as calculating the performance metrics, for BM-Case1, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0197] In some implementations, for BM-Case2 with inference operation, if one or more BM-Case2-related or BM-Case2-specific parameters are provided or present in the CSI report configuration and if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a first value, such as 'inference', indicating model inference, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference for BM-Case2, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0198] In some implementations, for BM-Case2 with monitoring operation, if one or more BM-Case2-related or BM-Case2-specific parameters are provided or present in the CSI report configuration and if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a second value, such as 'monitoring', indicating model performance monitoring, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring, such as calculating the performance metrics, for BM-Case2, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0199] Option 2: Configurability between {BM-Case1, BM-Case2} with model inference and model performance monitoring always performed together

[0200] In some implementations, the design philosophy may recognize that model inference may benefit from being performed together with model performance monitoring because without model performance monitoring, the quality of the inference results may not be adequately assured. Therefore, in some implementations, when a UE receives a CSI report configuration from the NW, the UE may need to know which operation among model inference and model performance monitoring for BM-Case1, and model inference and model performance monitoring for BM-Case2, the NW indicates the UE to perform by the CSI report configuration.

[0201] Option 2-1: Implicit indication to differentiate {BM-Case1, BM-Case2}

[0202] In some implementations, the NW may configure a CSI report configuration to a UE. The NW may indicate the UE to perform BM-Case1 or BM-Case2 by implicit indication in the CSI report configuration and may indicate the UE to perform model inference and model performance monitoring together. The UE may send one or more CSI reports including the inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include inference results and performance metrics.

[0203] In some implementations, for BM-Case1 with combined inference and monitoring, if BM-Case2-related or BM-Case2-specific parameters are absent in the CSI report configuration and if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a value, such as 'aiml', indicating performing model inference and model performance monitoring, or if the UE has already initiated or activated AI / ML functionalities or models or been indicated by the NW to activate AI / ML functionalities or models, or if the UE has applicable AI / ML functionalities or models, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring for BM-Case1, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0204] In some implementations, if the report type is set to aperiodic, such as when the reporting is triggered while one or more conditions are met, the UE may report both results, namely inference results and performance metrics, even though the condition may be specifically met only for either inference result or performance monitoring. This approach may ensure comprehensive feedback to the network regardless of which specific trigger condition was satisfied.

[0205] In some implementations, for BM-Case2 with combined inference and monitoring, if one or more BM-Case2-related or BM-Case2-specific parameters are provided or present in the CSI report configuration and if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a value, such as 'aiml', indicating model inference and model performance monitoring, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring for BM-Case2, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0206] In some implementations, if the report type is set to aperiodic for BM-Case2 operations, such as when the reporting is triggered while one or more conditions are met, the UE may report both results, namely inference results and performance metrics, even though the condition may be specifically met only for either inference result or performance monitoring.

[0207] Option 2-2: Explicit indication to differentiate {BM-Case1, BM-Case2}

[0208] In some implementations, the NW may configure a CSI report configuration to a UE. The NW may indicate the UE to perform BM-Case1 or BM-Case2 by explicit indication in the CSI report configuration and may indicate the UE to perform model inference and model performance monitoring together. The UE may send one or more CSI reports including the inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include inference results and performance metrics.

[0209] In some implementations, for BM-Case1 with combined inference and monitoring, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a value, such as 'bm1', indicating AI / ML operations including model inference and model performance monitoring for BM-Case1, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring for BM-Case1, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include inference results and performance metrics.

[0210] In some implementations, for BM-Case2 with combined inference and monitoring, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a value, such as 'bm2', indicating AI / ML operations including model inference and model performance monitoring for BM-Case2, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring for BM-Case2, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include inference results and performance metrics.

[0211] Option 3: Configurability between {BM-Case1, BM-Case2} and between {inference, monitoring, inference and monitoring}

[0212] In some implementations, the system may provide maximum flexibility by recognizing that while model inference may benefit from model performance monitoring, the NW may, based on the implementation or decision of the NW, configure the UE to perform model inference only, model performance monitoring only, or model inference and model performance monitoring together. Therefore, in some implementations, when a UE receives a CSI report configuration from the NW, the UE may need to know which operation among model inference for BM-Case1, model performance monitoring for BM-Case1, model inference and model performance monitoring for BM-Case1, model inference for BM-Case2, model performance monitoring for BM-Case2, and model inference and model performance monitoring for BM-Case2, the NW indicates the UE to perform by the CSI report configuration.

[0213] Option 3-1: Implicit indication to differentiate {BM-Case1, BM-Case2} and {inference, monitoring, inference and monitoring}

[0214] In some implementations, the NW may configure a CSI report configuration to a UE. The NW may indicate the UE to perform BM-Case1 or BM-Case2 by implicit indication in the CSI report configuration and may indicate the UE to perform model inference or model performance monitoring or model inference and model performance monitoring by implicit indication in the CSI report configuration.

[0215] In some implementations, regarding BM-Case1 / BM-Case2 implicit differentiation under Option 3-1, if one or more BM-Case2-related or BM-Case2-specific parameters are provided or present in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform AI / ML operations for BM-Case2, or the UE may determine to perform AI / ML operations for BM-Case2. In some implementations, if BM-Case2-related or BM-Case2-specific parameters are absent in the CSI report configuration, or if the UE has already initiated or activated AI / ML functionalities or models or been indicated by the NW to activate AI / ML functionalities or models, or if the UE has applicable AI / ML functionalities or models, the UE may determine or consider that the CSI report configuration indicates the UE to perform AI / ML operations for BM-Case1, or the UE may determine to perform AI / ML operations for BM-Case1.

[0216] In some implementations, as a first alternative for inference / inference+monitoring implicit differentiation, the resource set or resources for monitoring and the resource set for Set A may be totally the same, such as provided by a resourcesForSetAandMonitoring parameter.

[0217] In some implementations, for inference only operation under the first alternative, if the resource set for Set A and monitoring is not provided in the CSI report configuration, or if the UE has already initiated or activated AI / ML functionalities or models or been indicated by the NW to activate AI / ML functionalities or models, or if the UE has applicable AI / ML functionalities or models, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference operation only, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0218] In some implementations, for combined inference and monitoring operation under the first alternative, if the resource set for Set A and monitoring is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0219] In some implementations, as a second alternative for inference / monitoring / inference+monitoring implicit differentiation, the resource set or resources for monitoring may be a subset of the resource set for Set A. The subset information may follow the design principles in this disclosure. The subset information may be indicated by a parameter, such as a bitmap or a list of integers. The subset information may be indicated in the CSI report configuration or a CSI resource setting.

[0220] In some implementations, for inference only operation under the second alternative, if the resource set for Set A is provided and the resource set or resources for monitoring, namely the subset information, is not provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference only, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0221] In some implementations, for monitoring only operation under the second alternative, if the resource set for Set A is not provided and the resource set or resources for monitoring, namely the subset information, is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring only, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0222] In some implementations, for combined inference and monitoring operation under the second alternative, if the resource set for Set A is provided and the resource set or resources for monitoring, namely the subset information, is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0223] In some implementations, as a third alternative for inference / monitoring / inference+monitoring implicit differentiation, the resource set or resources for monitoring and the resource set for Set A may be separately provided. The resource set or resources for monitoring and the resource set for Set A may be configured in the CSI report configuration or in the CSI resource setting, such as the CSI-ResourceConfig IE, indicated in the CSI report configuration.

[0224] In some implementations, for inference only operation under the third alternative, if the resource set for Set A is provided and the resource set or resources for monitoring is not provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference only, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0225] In some implementations, for monitoring only operation under the third alternative, if the resource set for Set A is not provided and the resource set or resources for monitoring is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring only, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0226] In some implementations, for combined inference and monitoring operation under the third alternative, if the resource set for Set A is provided and the resource set or resources for monitoring is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0227] Option 3-2: Explicit indication to differentiate {BM-Case1, BM-Case2} and {inference, monitoring, inference and monitoring}

[0228] In some implementations, the NW may configure a CSI report configuration to a UE. The NW may indicate the UE to perform BM-Case1 or BM-Case2 by explicit indication in the CSI report configuration and may indicate the UE to perform model inference or model performance monitoring by an explicit indication in the CSI report configuration.

[0229] In some implementations, for BM-Case1 with inference operation, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a first value, such as 'bm1-inference', indicating model inference for BM-Case1, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference for BM-Case1, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0230] In some implementations, for BM-Case1 with monitoring operation, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a second value, such as 'bm1-monitoring', indicating model performance monitoring for BM-Case1, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring for BM-Case1, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0231] In some implementations, for BM-Case1 with combined inference and monitoring operation, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a third value, such as 'bm1-inference-monitoring', indicating model inference and model performance monitoring for BM-Case1, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring for BM-Case1, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0232] In some implementations, for BM-Case2 with inference operation, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a fourth value, such as 'bm2-inference', indicating model inference for BM-Case2, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference for BM-Case2, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0233] In some implementations, for BM-Case2 with monitoring operation, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a fifth value, such as 'bm2-monitoring', indicating model performance monitoring for BM-Case2, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring for BM-Case2, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0234] In some implementations, for BM-Case2 with combined inference and monitoring operation, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a sixth value, such as 'bm2-inference-monitoring', indicating model inference and model performance monitoring for BM-Case2, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring for BM-Case2, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0235] Option 3-3: Explicit indication to differentiate {BM-Case1, BM-Case2} and implicit indication to differentiate {inference, monitoring, inference and monitoring}

[0236] In some implementations, for explicit BM-Case1 indication, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a first value, such as 'bm1', indicating operations for BM-Case1, the UE may determine or consider that the CSI report configuration indicates the UE to perform operations for BM-Case1, or the UE may send one or more CSI reports including relevant results to the NW based on the CSI report configuration.

[0237] In some implementations, for explicit BM-Case2 indication, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a second value, such as 'bm2', indicating operations for BM-Case2, the UE may determine or consider that the CSI report configuration indicates the UE to perform operations for BM-Case2, or the UE may send one or more CSI reports including relevant results to the NW based on the CSI report configuration.

[0238] In some implementations, as a first alternative for inference / inference+monitoring implicit differentiation, the resource set or resources for monitoring and the resource set for Set A may be totally the same, such as provided by an RRC IE / field / parameter included in the CSI report configuration, such as the resourcesForSetAandMonitoring IE.

[0239] In some implementations, for inference only operation under the first alternative, if the resource set for Set A and monitoring is not provided in the CSI report configuration, or if the UE has already initiated or activated AI / ML functionalities or models or been indicated by the NW to activate AI / ML functionalities or models, or if the UE has applicable AI / ML functionalities or models, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference operation only, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0240] In some implementations, for combined inference and monitoring operation under the first alternative, if the resource set for Set A and monitoring is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0241] In some implementations, as a second alternative for inference / monitoring / inference+monitoring implicit differentiation, the resource set or resources for monitoring may be a subset of the resource set for Set A. The subset information may follow the design principles in this disclosure. The subset information may be indicated by a parameter, such as a bitmap or a list of integers. The subset information may be indicated in the CSI report configuration or a CSI resource setting.

[0242] In some implementations, for inference only operation under the second alternative, if the resource set for Set A is provided and the resource set or resources for monitoring, namely the subset information, is not provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference only, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0243] In some implementations, for monitoring only operation under the second alternative, if the resource set for Set A is not provided and the resource set or resources for monitoring, namely the subset information, is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring only, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0244] In some implementations, for combined inference and monitoring operation under the second alternative, if the resource set for Set A is provided and the resource set or resources for monitoring, namely the subset information, is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0245] In some implementations, as a third alternative for inference / monitoring / inference+monitoring implicit differentiation, the resource set or resources for monitoring and the resource set for Set A may be separately provided. The resource set or resources for monitoring and the resource set for Set A may be configured in the CSI report configuration or in the CSI resource setting, such as the CSI-ResourceConfig IE, indicated in the CSI report configuration.

[0246] In some implementations, for inference only operation under the third alternative, if the resource set for Set A is provided and the resource set or resources for monitoring is not provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference only, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0247] In some implementations, for monitoring only operation under the third alternative, if the resource set for Set A is not provided and the resource set or resources for monitoring is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring only, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0248] In some implementations, for combined inference and monitoring operation under the third alternative, if the resource set for Set A is provided and the resource set or resources for monitoring is provided in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0249] Option 3-4: Implicit indication to differentiate {BM-Case1, BM-Case2} and explicit indication to differentiate {inference, monitoring, inference and monitoring}

[0250] In some implementations, the NW may configure a CSI report configuration to a UE. The NW may indicate the UE to perform BM-Case1 or BM-Case2 by implicit indication in the CSI report configuration and may indicate the UE to perform model inference or model performance monitoring by an explicit indication in the CSI report configuration.

[0251] In some implementations, regarding BM-Case1 / BM-Case2 implicit differentiation, if one or more BM-Case2-related or BM-Case2-specific parameters are provided or present in the CSI report configuration, the UE may determine or consider that the CSI report configuration indicates the UE to perform AI / ML operations for BM-Case2, or the UE may determine to perform AI / ML operations for BM-Case2. In some implementations, if BM-Case2-related or BM-Case2-specific parameters are absent in the CSI report configuration, or if the UE has already initiated or activated AI / ML functionalities or models or been indicated by the NW to activate AI / ML functionalities or models, or if the UE has applicable AI / ML functionalities or models, the UE may determine or consider that the CSI report configuration indicates the UE to perform AI / ML operations for BM-Case1, or the UE may determine to perform AI / ML operations for BM-Case1.

[0252] In some implementations, for explicit inference / monitoring / inference+monitoring differentiation, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a first value, such as 'inference', indicating model inference, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference operation, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration. In some implementations, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a second value, such as 'monitoring', indicating model performance monitoring, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring operation, or the UE may send one or more CSI reports including model performance metrics to the NW based on the CSI report configuration. In some implementations, if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a third value, such as 'inference-monitoring', indicating model inference and model performance monitoring, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring operation, or the UE may send one or more CSI reports including inference results and model performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0253] In some implementations, for BM-Case1 with inference operation, if BM-Case2-related or BM-Case2-specific parameters are absent in the CSI report configuration and if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a first value, such as 'inference', indicating model inference, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference for BM-Case1, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0254] In some implementations, for BM-Case1 with monitoring operation, if BM-Case2-related or BM-Case2-specific parameters are absent in the CSI report configuration and if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a second value, such as 'monitoring', indicating model performance monitoring, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring, such as calculating the performance metrics, for BM-Case1, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0255] In some implementations, for BM-Case1 with combined inference and monitoring operation, if BM-Case2-related or BM-Case2-specific parameters are absent in the CSI report configuration and if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a third value, such as 'inference-monitoring', indicating model inference and model performance monitoring, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring, such as calculating the performance metrics, for BM-Case1, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0256] In some implementations, for BM-Case2 with inference operation, if one or more BM-Case2-related or BM-Case2-specific parameters are provided or present in the CSI report configuration and if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a first value, such as 'inference', indicating model inference, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference for BM-Case2, or the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration.

[0257] In some implementations, for BM-Case2 with monitoring operation, if one or more BM-Case2-related or BM-Case2-specific parameters are provided or present in the CSI report configuration and if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a second value, such as 'monitoring', indicating model performance monitoring, the UE may determine or consider that the CSI report configuration indicates the UE to perform model performance monitoring, such as calculating the performance metrics, for BM-Case2, or the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration.

[0258] In some implementations, for BM-Case2 with combined inference and monitoring operation, if BM-Case2-related or BM-Case2-specific parameters are provided or present in the CSI report configuration and if the report quantity indicator, such as the reportQuantity-r19 IE, indicates a third value, such as 'inference-monitoring', indicating model inference and model performance monitoring, the UE may determine or consider that the CSI report configuration indicates the UE to perform model inference and model performance monitoring, such as calculating the performance metrics, for BM-Case2, or the UE may send one or more CSI reports including inference results and performance metrics to the NW based on the CSI report configuration, where each CSI report may include both inference results and performance metrics.

[0259] Contents of CSI report for {BM-Case1, BM-Case2} and {inference, monitoring}

[0260] In some implementations, if a UE receives a CSI report configuration, the UE may determine whether this CSI report configuration indicates BM-Case1 or BM-Case2, and model inference or model performance monitoring, or both, based on the designs described in the previous sections. The determination process may utilize the various indication mechanisms outlined in Options 1 through 3, allowing the UE to properly interpret the network's intent and generate the appropriate CSI reports.

[0261] In some implementations, for inference results for BM-Case1, if the UE is to report inference results for BM-Case1, the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration. The CSI report may include the predicted RSRP of the Top K beams and the beam IDs, where the beam ID may be the resource ID within the resource set. The CSI report may include the predicted RSRP of the Top K beams if the beam is not configured for corresponding measurement and may include measured L1-RSRP if the beam is configured for corresponding measurement. The value of K may be indicated in the CSI report configuration. The predicted RSRP may be based on the model output generated by the AI / ML model for spatial domain prediction.

[0262] In some implementations, for inference results for BM-Case2, if the UE is to report inference results for BM-Case2, the UE may send one or more CSI reports including inference results to the NW based on the CSI report configuration. The CSI report may include the predicted RSRP of the best beams and the beam ID, where the beam ID may be the resource ID within the resource set, at N future time instances. The value of N may be indicated in the CSI report configuration. The predicted RSRP may be based on the model output generated by the AI / ML model for temporal domain prediction.

[0263] In some implementations, for performance metrics for BM-Case1 and BM-Case2, if the UE is to report performance metrics for BM-Case1 or BM-Case2, the UE may send one or more CSI reports including performance metrics to the NW based on the CSI report configuration. The UE may calculate the performance metrics based on measurement of the resource set for Set B serving as model input, and the predicted RSRP or beam ID serving as model output, and the resource set or resources for monitoring.

[0264] In some implementations, as a first alternative, the performance metrics may be the Top 1 or Top K beam prediction accuracy, with or without margin, by comparing the prediction results and the Top 1 or Top K beam based on the measurements from the resource set or resources for monitoring. This accuracy metric may provide direct feedback on the model's prediction performance.

[0265] In some implementations, as a second alternative, the performance metrics may be the L1-RSRP difference information based on actual measurement of the L1-RSRP of one or more of Top K predicted beams, and L1-RSRP measurements from a resource set or resources for monitoring. This difference information may help assess the quality of the predicted beams.

[0266] In some implementations, as a third alternative, the performance metrics may be the RSRP difference information between the predicted RSRP and measured L1-RSRP of corresponding beams of a resource set or resources for monitoring. This comparison may provide insight into the accuracy of RSRP predictions.

[0267] In some implementations, as a fourth alternative, the performance metrics may be the probability information of the predicted beams to be the Top 1 or Top K beam. This probabilistic information may help the network understand the confidence level of the predictions.

[0268] Use association ID to link a resource set for Set B to the resource set for Set A

[0269] In some implementations, the system may utilize association IDs to achieve consistency between training and inference phases. Model training for a first model may rely on a first training data set, or model training for a second model may rely on a second training data set. When performing data collection for model training, the UE may need to perform measurement on the resource set for Set A or the resource set for Set B, or both. When performing model inference, the UE may need to know the property, such as the NW-AC, of the beams for which the UE may measure and feed the measurement results to the model input, and which training data set was used. This consistency may be achieved by configuring association IDs to the resource set for Set A and the resource set for Set B.

[0270] In some implementations, when a UE is indicated to perform model inference, as described in this disclosure, the UE may check the association ID configured to resource set for Set B, such as indicated in the CSI report configuration indicating the UE to perform model inference, to know which resource set for Set A the resource set for Set B is paired with. By this mechanism, the NW may configure to the UE a CSI report configuration indicating model inference, in which only the resource set for Set B is provided and the resource set for Set A is not provided. The UE may know the resource set for Set A by checking the association ID. This approach may save the signaling overhead for configuring model inference.

[0271] In some implementations, the association ID mentioned in this disclosure may be an integer value that uniquely identifies the association between resource sets.

[0272] In some implementations, a UE may be configured with multiple CSI resource settings. Each CSI resource setting may be configured with an association ID. Any two CSI resource settings may be associated if the two CSI resource settings are configured with the same association ID. The two CSI resource settings may be a resource set for Set A and a resource set for Set B, establishing a pairing for AI / ML operations.

[0273] In some implementations, the association ID may be configured in a CSI resource configuration. A CSI resource configuration may include an association ID, a CSI resource setting for Set A, and a CSI resource setting for Set B. The UE may associate the association ID with both the CSI resource setting for Set A and the CSI resource setting for Set B.

[0274] In some implementations, the association ID may be configured in a CSI report configuration. A CSI report configuration may include an association ID, a CSI resource setting for Set A, or a CSI resource setting for monitoring, or a CSI resource setting for Set B, or any combination thereof. The UE may associate the association ID with the CSI resource setting for Set A, the CSI resource setting for monitoring, and the CSI resource setting for Set B. The UE is not expected to receive a report configuration where the resource set for measurement of Set B, inference of Set A, and the resource set for monitoring is associated with different association IDs. In some implementations, one RRC IE / field / parameter may be used to indicate an association ID to a CSI resource setting. The RRC IE / field / parameter may be optionally configured if the UE is indicated to perform AI / ML operation.

[0275] In some implementations, in addition to CSI RS resource set ID, such as the nzp-CSI-ResourceSetId IE, each CSI RS resource set may be further associated with an association ID. The UE may perform AI / ML operation, such as model inference, model performance monitoring, or both model inference and model performance monitoring, based on the CSI report configuration. In some implementations, if the CSI report configuration configures 'inference', 'bm1', and 'association ID = 2', the UE may perform model inference for BM-Case1 and report the inference result based on the CSI RS resource sets configured with 'association ID = 2'. The UE may be configured with a CSI RS resource set to be model input for model inference. More specifically, in addition to association ID configured for Set A, the UE may be further configured with a CSI RS resource set ID for Set B.

[0276] In some implementations, the NW may configure a CSI report configuration to a UE where the CSI report configuration may only provide the resource set for Set B. The CSI report configuration may not provide the resource set for Set A, nor the resources or resource set for monitoring. The CSI report configuration may indicate the UE to perform model inference. The UE may perform measurement on the resources in the resource set for Set B or let the measurement results be the model input, or both. The UE may map the model output to the resource set or CSI resource setting configured with an association ID identical to the association ID of the resource set for Set B. The UE may perform model inference and report the inference results with the following operations: the UE may measure the resource set for Set B, in which an association ID is configured, and the UE may send a CSI report to the NW, where the CSI report may include the Top-1 or Top-K predicted RSRPs of the predicted beam or resource IDs of the resources in the resource set or the CSI resource setting configured with the same (value of) association ID.

[0277] FIG. 1 is a flowchart illustrating method / process 100 for CSI reporting with model inference, according to an example implementation of the present disclosure. Although actions 102, 104, and 106 are illustrated, as separate actions, represented as independent blocks in FIG. 1, these separately illustrated actions should not be construed as to be necessarily order-dependent. The order in which the actions are performed in FIG. 1 is not intended to be construed as a limitation, and any number of the disclosed blocks may be combined in any order to implement the method, or an alternative method. Each of actions 102, 104, and 106 may be performed independent of the other actions, and may be omitted in some implementations of the present disclosure. Moreover, method / process 100 may be combined with other procedures / methods described in the present disclosure. Process 100 may be performed by a UE, with each action of process 100 corresponding to an operation executed by the UE.

[0278] In action 102, the UE may receive, from a BS, a first CSI report configuration including at least one first parameter indicating model inference, a first CSI resource setting and a second CSI resource setting, where each of the first CSI resource setting and the second CSI resource setting includes the same association identifier (e.g., a first association identifier). In action 104, the UE may perform at least one first measurement on at least one first resource indicated by the first CSI resource setting. In action 106, the UE may transmit, to the BS, a first CSI report containing at least one first predicted value, where the at least one first predicted value may be associated with at least one second resource indicated by the second CSI resource setting.

[0279] In some implementations, the UE may receive, from the BS, a second CSI report configuration including at least one second parameter indicating model inference, a third CSI resource setting and a fourth CSI resource setting, where each of the third CSI resource setting and the fourth CSI resource setting includes a second association identifier different from the first association identifier. The UE may further perform at least one second measurement on at least one third resource indicated by the third CSI resource setting, and transmit, to the BS, a second CSI report containing at least one second predicted value, where the at least one second predicted value is associated with at least one fourth resource indicated by the fourth CSI resource setting.

[0280] In some implementations, the first CSI resource setting may indicate a first CSI-RS resource set, the second CSI resource setting may indicate a second CSI-RS resource set, and the first association identifier may associate the first CSI-RS resource set with the second CSI-RS resource set for model inference operations.

[0281] In some implementations, the at least one first predicted value may include at least one predicted RSRP value and at least one predicted resource identifier corresponding to the at least one second resource.

[0282] In some implementations, the at least one predicted RSRP value may include K predicted RSRP values for K highest predicted beams, and the at least one predicted resource identifier may include K predicted resource identifiers corresponding to the K highest predicted beams, where K is a positive integer indicated by the first CSI report configuration.

[0283] In some implementations, the UE may apply an AI / ML model using the at least one first measurement as model input to generate the at least one first predicted value for the at least one second resource.

[0284] In some implementations, the UE may determine whether the first CSI report configuration indicates spatial domain prediction or temporal domain prediction based on presence or absence of at least one temporal domain specific parameter in the first CSI report configuration.

[0285] In some implementations, the at least one temporal domain specific parameter may include at least one of a length of a time observation window, a time duration between adjacent future time instances, or a number of inference results to report. Each inference result may correspond to a different future time instance.

[0286] The extensive technical descriptions provided throughout this disclosure present specific implementation examples for various aspects of process 100, as well as additional implementations / embodiments that can be combined with or built upon the foundation established by process 100. For example, the detailed configuration options, including Options 1 through 3 with their various sub-options for differentiating between BM-Case1 / BM-Case2 and inference / monitoring operations, provide concrete examples of how the first parameter indicating model inference in process 100 can be implemented through different signaling mechanisms. The multiple RRC configuration structures (Alt 1, Alt 2, and Alt 3) demonstrate specific ways to realize the first and second CSI resource settings of process 100, offering flexibility in how these resource configurations can be organized and signaled. The various approaches for indicating subset information for monitoring resources, whether through bitmaps, integer lists, or resource counts, exemplify how the framework can extend beyond the basic elements of process 100 to support performance monitoring capabilities. The temporal domain specific parameters detailed in the specification, including time observation windows and future time instance configurations, provide concrete implementations of how process 100 can distinguish between spatial and temporal domain predictions. Furthermore, the comprehensive treatment of association ID mechanisms, multi-model management architectures, and CSI report content specifications illustrate additional features that can be integrated with process 100 to create a complete AI / ML-enhanced beam management system. It should be noted that unless specifically indicated otherwise, the various implementation approaches and configuration options described in this disclosure can also be implemented independently of process 100, providing flexibility for different system architectures and deployment scenarios where alternative methods for CSI reporting with model inference may be employed.

[0287] Furthermore, process 100 may provide an efficient framework for AI / ML-enhanced beam management in wireless communication systems through its use of association identifiers that link measurement resources with inference target resources, eliminating the need to explicitly configure both resource sets for each inference operation and thereby reducing signaling overhead. The method enables the UE to perform measurements on one set of resources and generate predicted values for another set of resources without exhaustive channel measurements, reducing measurement burden while providing comprehensive beam quality information to the network. By supporting multiple CSI report configurations with distinct association identifiers, the framework allows simultaneous deployment of different AI / ML models optimized for various channel conditions, with each model uniquely identified through its association identifier for flexible model selection. The generation of K predicted RSRP values with corresponding resource identifiers provides ranked beam predictions that enable robust beam selection decisions, where K can be configured to balance prediction comprehensiveness against reporting efficiency. The method's ability to differentiate between spatial and temporal domain predictions through the presence or absence of temporal-specific parameters enables a unified framework that addresses both immediate beam selection for stationary scenarios and predictive beam tracking for high-mobility cases, simplifying implementation while maintaining operational flexibility across diverse deployment conditions.

[0288] It should also be noted that the network device, such as the BS, may perform methods / actions corresponding to those performed by the UE. For example, the receiving actions performed by the UE may correspond to the transmitting / configuring actions of the network device; the transmitting actions performed by the UE may correspond to the receiving actions of the network device. That is, the network device and the UE may have reciprocally aligned roles in transmission and reception, as illustrated in FIG. 2.

[0289] FIG. 2 is a flowchart illustrating method / process 200 for handling CSI reporting, according to an example implementation of the present disclosure. Although actions 202, 204, and 206 are illustrated, as separate actions, represented as independent blocks in FIG. 2, these separately illustrated actions should not be construed as to be necessarily order-dependent. The order in which the actions are performed in FIG. 2 is not intended to be construed as a limitation, and any number of the disclosed blocks may be combined in any order to implement the method, or an alternative method. Each of actions 202, 204, and 206 may be performed independent of the other actions, and may be omitted in some implementations of the present disclosure. Moreover, method / process 200 may be combined with other procedures / methods described in the present disclosure. Process 200 may be performed by a BS, with each action of process 200 corresponding to an operation executed by the BS.

[0290] In action 202, the BS may transmit, to a UE, a first CSI report configuration including at least one first parameter indicating model inference, a first CSI resource setting and a second CSI resource setting, where each of the first CSI resource setting and the second CSI resource setting may include the same association identifier (e.g., a first association identifier). In action 204, the BS may transmit at least one first reference signal on at least one first resource indicated by the first CSI resource setting. In action 206, the BS may receive, from the UE, a first CSI report containing at least one first predicted value, where the at least one first predicted value may be associated with at least one second resource indicated by the second CSI resource setting.

[0291] FIG. 3 is a block diagram illustrating node 300 for wireless communications, in accordance with various aspects of the present disclosure. As illustrated in FIG. 3, node 300 may include transceiver 320, processor 328, memory 334, one or more presentation components 338, and at least one antenna 336. Node 300 may also include a radio frequency (RF) spectrum band module, a BS communications module, a network communications module, and a system communications management module, Input / Output (I / O) ports, I / O components, and a power supply (not illustrated in FIG. 3).

[0292] Each of the components may directly or indirectly communicate with each other over one or more buses 340. Node 300 may be a UE or a BS that performs various functions disclosed with reference to FIGS. 1-2.

[0293] Transceiver 320 has transmitter 322 (e.g., transmitting / transmission circuitry) and receiver 324 (e.g., receiving / reception circuitry) and may be configured to transmit and / or receive time and / or frequency resource partitioning information. Transceiver 320 may be configured to transmit in different types of subframes and slots including, but not limited to, usable, non-usable, and flexibly usable subframes and slot formats. Transceiver 320 may be configured to receive data and control channels.

[0294] Node 300 may include a variety of computer-readable media. Computer-readable media may be any available media that may be accessed by node 300 and include volatile (and / or non-volatile) media and removable (and / or non-removable) media.

[0295] The computer-readable media may include computer-storage media and communication media. Computer-storage media may include both volatile (and / or non-volatile media), and removable (and / or non-removable) media implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules, or data.

[0296] Computer-storage media may include RAM, ROM, EPROM, EEPROM, flash memory (or other memory technology), CD-ROM, Digital Versatile Disks (DVD) (or other optical disk storage), magnetic cassettes, magnetic tape, magnetic disk storage (or other magnetic storage devices), etc. Computer-storage media may not include a propagated data signal. Communication media may typically embody computer-readable instructions, data structures, program modules, or other data in a modulated data signal, such as a carrier wave, or other transport mechanisms and include any information delivery media.

[0297] The term “modulated data signal” may mean a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. Communication media may include wired media, such as a wired network or direct-wired connection, and wireless media, such as acoustic, RF, infrared, and other wireless media. Combinations of any of the aforementioned listed components should also be included within the scope of computer-readable media.

[0298] Memory 334 may include computer-storage media in the form of volatile and / or non-volatile memory. Memory 334 may be removable, non-removable, or a combination thereof. Example memory may include solid-state memory, hard drives, optical-disc drives, etc. As illustrated in FIG. 3, memory 334 may store a computer-readable and / or computer-executable instructions 332 (e.g., software codes) that are configured to, when executed, cause processor 328 to perform various functions disclosed herein, for example, with reference to FIGS. 1-2. Alternatively, instructions 332 may not be directly executable by processor 328 but may be configured to cause node 300 (e.g., when compiled and executed) to perform various functions disclosed herein.

[0299] Processor 328 (e.g., having processing circuitry) may include an intelligent hardware device, e.g., a Central Processing Unit (CPU), a microcontroller, an ASIC, etc. Processor 328 may include memory. Processor 328 may process data 330 and instructions 332 received from memory 334, and information transmitted and received via transceiver 320, the baseband communications module, and / or the network communications module. Processor 328 may also process information to send to transceiver 320 for transmission via antenna 336 to the network communications module for transmission to a CN.

[0300] One or more presentation components 338 may present data indications to a person or another device. Examples of presentation components 338 may include a display device, a speaker, a printing component, a vibrating component, etc.

[0301] In view of the present disclosure, it is obvious that various techniques may be used for implementing the disclosed concepts without departing from the scope of those concepts. Moreover, while the concepts have been disclosed with specific reference to certain implementations, a person of ordinary skill in the art may recognize that changes may be made in form and detail without departing from the scope of those concepts. As such, the disclosed implementations are to be considered in all respects as illustrative and not restrictive. It should also be understood that the present disclosure is not limited to the particular implementations disclosed and many rearrangements, modifications, and substitutions are possible without departing from the scope of the present disclosure.

Claims

A User Equipment (UE), the UE comprising:　　　　at least one processor; and　　　　at least one non-transitory computer-readable medium coupled to the at least one processor and storing one or more computer-executable instructions that, when executed by the at least one processor, cause the UE to:　　　　receive, from a Base Station (BS), a first Channel State Information (CSI) report configuration comprising at least one first parameter indicating model inference, a first CSI resource setting and a second CSI resource setting, wherein each of the first CSI resource setting and the second CSI resource setting comprises a first association identifier;　　　　perform at least one first measurement on at least one first resource indicated by the first CSI resource setting; and　　　　transmit, to the BS, a first CSI report containing at least one first predicted value, wherein the at least one first predicted value is associated with at least one second resource indicated by the second CSI resource setting.The UE of claim 1, wherein the one or more computer-executable instructions, when executed by the at least one processor, further cause the UE to:　　　　receive, from the BS, a second CSI report configuration comprising at least one second parameter indicating model inference, a third CSI resource setting and a fourth CSI resource setting, wherein each of the third CSI resource setting and the fourth CSI resource setting comprises a second association identifier different from the first association identifier;　　　　perform at least one second measurement on at least one third resource indicated by the third CSI resource setting; and　　　　transmit, to the BS, a second CSI report containing at least one second predicted value, wherein the at least one second predicted value is associated with at least one fourth resource indicated by the fourth CSI resource setting.The UE of claim 1, wherein the first CSI resource setting indicates a first CSI-Reference Signal (CSI-RS) resource set, the second CSI resource setting indicates a second CSI-RS resource set, and the first association identifier associates the first CSI-RS resource set with the second CSI-RS resource set for model inference operations.The UE of claim 1, wherein the at least one first predicted value comprises at least one predicted Reference Signal Received Power (RSRP) value and at least one predicted resource identifier corresponding to the at least one second resource.The UE of claim 4, wherein the at least one predicted RSRP value comprises K predicted RSRP values for K highest predicted beams, and the at least one predicted resource identifier comprises K predicted resource identifiers corresponding to the K highest predicted beams, wherein K is a positive integer indicated by the first CSI report configuration.The UE of claim 1, wherein the one or more computer-executable instructions, when executed by the at least one processor, further cause the UE to:　　　　apply an Artificial Intelligence / Machine Learning (AI / ML) model using the at least one first measurement as model input to generate the at least one first predicted value for the at least one second resource.The UE of claim 1, wherein the one or more computer-executable instructions, when executed by the at least one processor, further cause the UE to:　　　　determine whether the first CSI report configuration indicates spatial domain prediction or temporal domain prediction based on presence or absence of at least one temporal domain specific parameter in the first CSI report configuration.The UE of claim 7, wherein the at least one temporal domain specific parameter comprises at least one of:　　　　a length of a time observation window;　　　　a time duration between adjacent future time instances; or　　　　a number of inference results to report, each inference result corresponding to a different future time instance.A method performed by a User Equipment (UE) for Channel State Information (CSI) reporting with model inference, comprising:　　　　receiving, from a Base Station (BS), a first CSI report configuration comprising at least one first parameter indicating model inference, a first CSI resource setting and a second CSI resource setting, wherein each of the first CSI resource setting and the second CSI resource setting comprises a first association identifier;　　　　performing at least one first measurement on at least one first resource indicated by the first CSI resource setting; and　　　　transmitting, to the BS, a first CSI report containing at least one first predicted value, wherein the at least one first predicted value is associated with at least one second resource indicated by the second CSI resource setting.A Base Station (BS), the BS comprising:　　　　at least one processor; and　　　　at least one non-transitory computer-readable medium coupled to the at least one processor and storing one or more computer-executable instructions that, when executed by the at least one processor, cause the BS to:　　　　transmit, to a User Equipment (UE), a first Channel State Information (CSI) report configuration comprising at least one first parameter indicating model inference, a first CSI resource setting and a second CSI resource setting, wherein each of the first CSI resource setting and the second CSI resource setting comprises a first association identifier;　　　　transmit at least one first reference signal on at least one first resource indicated by the first CSI resource setting; and　　　　receive, from the UE, a first CSI report containing at least one first predicted value, wherein the at least one first predicted value is associated with at least one second resource indicated by the second CSI resource setting.The BS of claim 10, wherein the one or more computer-executable instructions, when executed by the at least one processor, further cause the BS to:　　　　transmit, to the UE, a second CSI report configuration comprising at least one second parameter indicating model inference, a third CSI resource setting and a fourth CSI resource setting, wherein each of the third CSI resource setting and the fourth CSI resource setting comprises a second association identifier different from the first association identifier;　　　　transmit at least one second reference signal on at least one third resource indicated by the third CSI resource setting; and　　　　receive, from the UE, a second CSI report containing at least one second predicted value, wherein the at least one second predicted value is associated with at least one fourth resource indicated by the fourth CSI resource setting.The BS of claim 10, wherein the first CSI resource setting indicates a first CSI Reference Signal (CSI-RS) resource set, the second CSI resource setting indicates a second CSI-RS resource set, and the first association identifier associates the first CSI-RS resource set with the second CSI-RS resource set for model inference operations.The BS of claim 10, wherein the at least one first predicted value comprises at least one predicted Reference Signal Received Power (RSRP) value and at least one predicted resource identifier corresponding to the at least one second resource.The BS of claim 13, wherein the at least one predicted RSRP value comprises K predicted RSRP values for K highest predicted beams, and the at least one predicted resource identifier comprises K predicted resource identifiers corresponding to the K highest predicted beams, wherein K is a positive integer indicated by the first CSI report configuration.The BS of claim 10, wherein the first CSI report configuration includes at least one temporal domain specific parameter when indicating temporal domain prediction, wherein the at least one temporal domain specific parameter comprises at least one of:　　　　a length of a time observation window;　　　　a time duration between adjacent future time instances; or　　　　a number of inference results to report, each inference result corresponding to a different future time instance.

Images