Apparatus for beam management

By using machine learning models in 5G NR beam management for beam prediction and optimized channel status information reporting, the overhead and delay in beam selection process are solved, and throughput and scheduling efficiency are improved.

CN119096480APending Publication Date: 2024-12-06NOKIA TECHNOLOGIES OY
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202380036401.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2022-04-29
Filing Date
2023-03-21
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

In the process of 5G NR beam management, the existing technology has great overhead and delays, resulting in throughput and scheduling delay problems. In particular, the base station needs to send a large number of reference signals when selecting the best beam, resulting in waste of resources and time consumption.

Method used

The machine learning model is used to perform beam prediction. By measuring reference signals and generating sorting lists, channel status information reports are optimized, unnecessary beam scanning and measurement are reduced, and beam selection efficiency is improved.

Benefits of technology

It reduces the overhead and delay of base stations and user equipment in the beam selection process, improves overall cell throughput and scheduling efficiency, reduces delay, and optimizes the beam management process.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119096480A_ABST
    Figure CN119096480A_ABST
Patent Text Reader

Abstract

An apparatus comprising a module for transmitting an indication of beam prediction capability in a spatial domain of the apparatus; a module for receiving an indication of a likelihood of a beam for at least one beam prediction mode in the spatial domain; means for receiving an indication of a beam prediction mode in the spatial domain; means for measuring one or more reference signals from the beam to obtain a set of measurements; means for using the measurement results as input to a machine learning model configured to generate an ordered list of beams of at least the measured one or more reference signals using the input and using the likelihood of the beams; a module for generating a channel state information report based on the ordered list of beams; and a module for transmitting a report comprising an index of the beams of the ordered list.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to a beam management process, device and method. Background Art

[0002] The first release of the 5G Advanced standard will be 3GPP Rel (Release)-18. One of the work items / study items (WI / SI) concerns Artificial Intelligence (AI) / Machine Learning (ML) for the NR (New Radio) air interface. In this study, the goal is to explore the benefits of enhancing the air interface with features that enable improved support for AI / ML-based algorithms, thereby increasing performance and / or reducing complexity / overhead. The enhanced performance here depends on the use case considered and can, for example, improve throughput, robustness, accuracy or reliability.

[0003] Through the NR beam measurement and reporting process, the user equipment (UE) reports the strongest beam, but the process of selecting the strongest beam is usually associated with additional overhead and delay. In addition, the NR beam reporting framework is not optimized for selecting the best beam for 5G / NR base stations (gNB) to improve overall cell-level throughput and reduce scheduling delays in a more dynamic way. In particular, the following issues can be identified:

[0004] Greater Overhead: The gNB has to send a large number of reference signals, such as synchronization signal blocks (SSBs) and channel state information reference signals (CSI-RS), to find the best beam towards the UE, which causes overhead issues because each beam is associated with a different SSB or CSI-RS resource.

[0005] Higher latency: The time required for the gNB and UE to complete beam scanning and refinement to establish the best beam pair (for both Tx and Rx) is typically based on multiple rounds of measurements (P1, P2, and P3 processes).

[0006] Lower throughput and scheduling delays: gNBs have limitations in scheduling UEs as they may only be able to use one beam at a given time (e.g., time slot), which causes additional latency when supporting UEs within a cell and often results in reduced overall throughput. Summary of the invention

[0007] Now, an improved method and a technical device for implementing the method have been invented, which alleviate the above-mentioned problems. Various aspects include a method, an apparatus and a non-transitory computer-readable medium comprising a computer program or a signal stored therein, characterized by what is stated in the independent claims. Various details of the embodiments are disclosed in the dependent claims and the corresponding figures and description.

[0008] The independent claims define the scope of protection sought by various embodiments of the invention. Embodiments and features described in this specification that do not fall within the scope of the independent claims, if any, are to be construed as examples that aid in understanding the various embodiments of the invention.

[0009] According to a first aspect, there is provided an apparatus comprising:

[0010] means for transmitting an indication of beam prediction capability in the spatial domain of the apparatus;

[0011] means for receiving an indication of a likelihood of a beam for at least one beam prediction mode in the spatial domain;

[0012] means for receiving an indication of a beam prediction mode in the spatial domain;

[0013] means for measuring one or more reference signals from the beam to obtain a set of measurements;

[0014] means for using the measurements as inputs to a machine learning model and using the likelihood of the beams according to the beam prediction pattern in the spatial domain, wherein the machine learning model is configured to use the inputs to generate a ranked list of at least the measured beams of one or more reference signals;

[0015] means for generating a channel state information report based on the ranked list of beams; and

[0016] Means for sending the report comprising an index of the beam of the ordered list.

[0017] According to an embodiment, the device comprises:

[0018] Means for receiving two sets of RS indices having a first configuration defining a first set of RS indices for beams on which signal measurements are to be performed and a second configuration defining a second set of RS indices for beams to be considered in prediction;

[0019] means for measuring one or more reference signals from beams of said first set to obtain a set of measurements; and

[0020] A module for using the measurements as input to a machine learning model and using the likelihood of the beam based on the beam prediction pattern in the spatial domain, wherein the machine learning model is configured to use the input to generate a ranked list of the beams of the first set and the second set.

[0021] According to an embodiment, the machine learning model for beam prediction in the spatial domain is configured to construct the ranked list based on one or more of the following:

[0022] The best and strongest beam,

[0023] Best for beamforming, or

[0024] Optimal beam with additional metrics.

[0025] According to an embodiment, the means for generating a channel state information report is configured to include in the report one or more of the following:

[0026] Beam index,

[0027] Beam reception quality indicator,

[0028] Applicable UE panel indicators,

[0029] Rank indicator,

[0030] Precoding matrix indicator,

[0031] Channel quality information, and

[0032] Layer indicator.

[0033] According to an embodiment, the apparatus comprises means for receiving an update of a context for beam prediction in the spatial domain.

[0034] According to an embodiment, the mode includes at least the following:

[0035] The best beam prediction in the spatial domain is the beam prediction for the strongest beam,

[0036] The best beam prediction in the spatial domain is the beam prediction for the most efficient beam for scheduling.

[0037] According to an embodiment, the device includes a module for updating the machine learning model; and a module for sending the updated machine learning model to a network element.

[0038] The apparatus according to the second aspect comprises at least one processor and at least one memory having computer program code stored thereon, the at least one memory and the computer program code being configured to, together with the at least one processor, cause the apparatus to at least perform:

[0039] sending an indication of a beam prediction capability of the apparatus in a spatial domain;

[0040] receiving an indication of a likelihood of a beam for at least one beam prediction mode in the spatial domain;

[0041] receiving an indication of a beam prediction mode in a spatial domain;

[0042] measuring one or more reference signals from the beam to obtain a set of measurement results;

[0043] using the measurements as inputs to a machine learning model and using the likelihood of the beams according to the beam prediction pattern in the spatial domain, wherein the machine learning model is configured to use the inputs to generate a ranked list of the beams of at least the measured one or more reference signals;

[0044] generating a channel state information report based on the ranked list of beams; and

[0045] The report is sent including an index of the beam of the ordered list.

[0046] The method according to the third aspect comprises:

[0047] sending an indication of a beam prediction capability of the apparatus in a spatial domain;

[0048] receiving an indication of a likelihood of a beam for at least one beam prediction mode in the spatial domain;

[0049] receiving an indication of a beam prediction mode in a spatial domain;

[0050] measuring one or more reference signals from the beam to obtain a set of measurement results;

[0051] using the measurements as inputs to a machine learning model and using the likelihood of the beams according to the beam prediction pattern in the spatial domain, wherein the machine learning model is configured to use the inputs to generate a ranked list of the beams of at least the measured one or more reference signals;

[0052] generating a channel state information report based on the ranked list of beams; and

[0053] The report is sent including the index of the beam of the sorted list.

[0054] According to other aspects, a computer-readable storage medium includes code for use by an apparatus, which, when executed by a processor, causes the apparatus to perform the above-described method. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] For a more complete understanding of example embodiments, reference is now made to the following description taken in conjunction with the accompanying drawings, in which:

[0056] Figure 1 shows a schematic block diagram of an apparatus for combining a dual SIM / MUSIM arrangement according to an embodiment;

[0057] Figure 2 schematically illustrates a layout of an apparatus according to an example embodiment;

[0058] Figure 3 A portion of an exemplary radio access network is shown;

[0059] Figure 4 Describes an example of a gNB-UE signaling framework for implementing ML-based beam prediction;

[0060] Figure 5 An exemplary scenario for predicting the ranking of the best beams at a UE is shown;

[0061] Figure 6 An example showing how a scheduler uses ML-based beam reports from a UE and updates beam likelihoods;

[0062] Figure 7 Some advantages of beam likelihood information are shown; and

[0063] Figure 8 A flow chart of beam management according to an embodiment is shown. DETAILED DESCRIPTION

[0064] Suitable devices and possible mechanisms for performing beam scanning and mapping operations are described in more detail below. Although the following focuses on 5G networks, the embodiments further described below are by no means limited to implementation only in such networks, but are applicable to any network that supports beam scanning and mapping operations.

[0065] In this regard, first refer to Figure 1 and Figure 2 ,in Figure 1 A schematic block diagram of an exemplary apparatus or electronic device 50 is shown, which may incorporate arrangements according to embodiments. Figure 2 The layout of the device according to the exemplary embodiment is shown. Figure 1 and Figure 2 Components in .

[0066] The electronic device 50 may be, for example, a mobile terminal or user equipment of a wireless communication system. The device 50 may include a housing 30 for combining and protecting the device. The device 50 may also include a display 32 and a keyboard 34. Instead of a keyboard, the user interface may be implemented as a virtual keyboard or data input system as part of a touch-sensitive display.

[0067] The device may include a microphone 36 or any suitable audio input, which may be a digital or analog signal input. The device 50 may also include an audio output device, such as headphones 38, speakers, or any of analog audio or digital audio output connections. The device 50 may also include a battery 40 (or the device may be powered by any suitable mobile energy device, such as a solar cell, a fuel cell, or a clockwork generator). The device may also include a camera 42 capable of recording or capturing images and / or videos. The device 50 may also include an infrared port 41 for short-range line-of-sight communication with other devices. In other embodiments, the device 50 may also include any suitable short-range communication solution, such as a Bluetooth wireless connection or a USB / Firewire wired connection.

[0068] The device 50 may include a controller 56 or processor for controlling the device 50. The controller 56 may be connected to a memory 58, which may store user data and instructions for implementation on the controller 56. The memory may be a random access memory (RAM) and / or a read-only memory (ROM). The memory may store computer-readable, computer-executable software, including instructions that, when executed, cause the controller / processor to perform the various functions described herein. In some cases, the software may not be directly executable by the processor, but may cause the computer (e.g., when compiled and executed) to perform the functions described herein. The controller 56 may also be further connected to a codec circuit 54, which is suitable for performing encoding and decoding of audio and / or video data, or assisting the encoding and decoding performed by the controller.

[0069] The device 50 may include a radio interface circuit 52 connected to the controller and adapted to generate wireless communication signals, such as for communicating with a cellular communication network, a wireless communication system, or a wireless local area network. The device 50 may also include an antenna 44 connected to the radio interface circuit 52 for transmitting the radio frequency signals generated at the radio interface circuit 52 to other devices and receiving the radio frequency signals from other devices.

[0070] In the following, different exemplary embodiments will be described using a radio access architecture based on advanced long term evolution (LTE-Advanced (Advanced), LTE-A) or new radio (NR, 5G) as an example of an access architecture to which the embodiments can be applied, but the embodiments are not limited to such architectures. Those skilled in the art understand that by appropriately adjusting parameters and processes, the embodiments can also be applied to other types of communication networks with appropriate modules. Some examples of other options for applicable systems are Universal Mobile Telecommunications System (UMTS) Radio Access Network (UTRAN or E-UTRAN), Long Term Evolution (LTE, the same as E-UTRA), Wireless Local Area Network (WLAN or WiFi), Worldwide Interoperability for Microwave Access (WiMAX), Bluetooth Personal Communications Service (PCS), ZigBee Wideband Code Division Multiple Access (WCDMA), systems using Ultra-Wideband (UWB) technology, sensor networks, Mobile Ad Hoc Networks (MANETs), and Internet Protocol Multimedia Subsystem (IMS), or any combination thereof.

[0071] Figure 3 An example of a simplified system architecture is described, showing only some elements and functional entities, which are all logical units and whose implementation may differ from what is shown. Figure 3 The connections shown are logical connections; the actual physical connections may be different. It will be apparent to those skilled in the art that the system will typically include Figure 3 Other functions and structures than those shown. However, the embodiments are not limited to the systems given as examples, and a person skilled in the art may apply the solution to other communication systems having the necessary properties.

[0072] Figure 3 The example of FIG. 1 shows a portion of an exemplary radio access network.

[0073] Figure 3 User equipment 300 and 302 are shown, which are configured to be wirelessly connected to an access node (such as (e / g) NodeB (Node B)) 304 providing the cell on one or more communication channels in the cell. The physical link from the user equipment to the (e / g) NodeB is called an uplink or reverse link, while the physical link from the (e / g) NodeB to the user equipment is called a downlink or forward link. It should be understood that the (e / g) NodeB or its functions can be implemented by using any node, host, server or access point entity suitable for such purpose.

[0074] The communication system usually includes more than one (e / g)NodeB, in which case the (e / g)NodeBs may also be configured to communicate with each other via wired or wireless links designed for this purpose. These links can be used for signaling purposes. (e / g)NodeB is a computing device configured to control the radio resources of the communication system coupled thereto. NodeB may also be referred to as a base station, an access point, or any other type of interface device, including a relay station capable of operating in a wireless environment. (e / g)NodeB includes a transceiver or is coupled to a transceiver. A connection to an antenna unit is provided from the transceiver of the (e / g)NodeB, which establishes a bidirectional radio link to a user equipment. The antenna unit may include multiple antennas or antenna elements. (e / g)NodeB is also connected to a core network 310 (CN or next generation core NGC). Depending on the system, the counterpart on the CN side may be a service gateway (S-GW, routing and forwarding user data packets), a packet data network gateway (P-GW), or a mobile management entity (MME), etc., for providing a connection between a user equipment (UE) and an external packet data network. CN may include a network entity or node that may be referred to as a management entity. Examples of network entities include at least an Access and Mobility Management Function (AMF).

[0075] User equipment (also known as user equipment (UE), user terminal, terminal device, wireless device, mobile station (MS), etc.) shows a type of device to which resources on the air interface are allocated and assigned, so any features of the user equipment described in this document can be implemented with corresponding network devices, such as relay nodes, eNBs and gNBs. An example of such a relay node is a layer 3 relay (self-backhaul relay) towards a base station.

[0076] User equipment generally refers to a portable computing device, which includes wireless mobile communication devices that operate with or without a user identification module (SIM), including but not limited to the following types of devices: mobile stations (mobile phones), smart phones, personal digital assistants (PDAs), handheld devices (handsets), devices using wireless modems (alarm or measurement devices, etc.), laptop computers and / or touch screen computers, tablet computers, game consoles, notebook computers, and multimedia devices. It should be understood that user equipment may also be almost exclusively uplink-only devices, one example of which is a camera or video camera that loads images or video clips to a network. User equipment may also be a device that can operate in an Internet of Things (IoT) network, which is a scenario in which objects can transmit data over a network without the need for human-to-human or human-to-computer interaction. Therefore, a user device may be an IoT device. The user device may also utilize the cloud. In some applications, the user device may include a small portable device with a radio component (such as a watch, headphones, or glasses), and the calculation is performed in the cloud. The user device (or in some embodiments, a layer 3 relay node) is configured to perform one or more user device functions. A user device may also be called a subscriber unit, a mobile station, a remote terminal, an access terminal, a user terminal, or user equipment (UE), to name just a few names or devices.

[0077] The various techniques described herein may also be applied to cyber-physical systems (CPS) (systems of cooperating computing elements that control physical entities). CPS can enable the implementation and utilization of a large number of interconnected ICT devices (sensors, actuators, processors, microcontrollers, etc.) embedded in physical objects in different locations. Mobile cyber-physical systems, where the physical systems in question have inherent mobility, are a subclass of cyber-physical systems. Examples of mobile physical systems include mobile robots and electronic products transported by humans or animals.

[0078] Furthermore, although these devices are described as a single entity, different units, processors and / or memory units may also be implemented. Figure 1 Not all are shown).

[0079] 5G enables the use of multiple-input multiple-output (MIMO) antennas, more base stations or nodes than LTE (the so-called small cell concept), including macro sites operating in cooperation with smaller base stations, and adopting various radio technologies depending on service requirements, use cases and / or available spectrum. The access nodes of the radio network form transmit / receive (TX / Rx) points (TRPs), and it is expected that UEs access a network of multiple TRPs (e.g. macro cells, small cells, pico cells, femto cells, remote radio heads, relay nodes, etc.) that overlap at least partially. The access nodes can be equipped with massive MIMO antennas, i.e. very large antenna arrays consisting of, for example, hundreds of antenna elements, implemented in a single antenna panel or in multiple antenna panels, capable of communicating with UEs using multiple simultaneous radio beams. The UE can be equipped with a MIMO antenna with an antenna array consisting of, for example, tens of antenna elements, implemented in a single antenna panel or in multiple antenna panels. Thus, the UE can access one TRP using one beam, one TRP using multiple beams, multiple TRPs using one (common) beam or multiple TRPs using multiple beams.

[0080] 4G / LTE networks support some multi-TRP schemes, but in 5G NR, multi-TRP features are enhanced, such as sending multiple control signals through multiple TRPs, thereby improving link diversity gain. In addition, high carrier frequencies (such as mmWave) and massive MIMO antennas require new beam management processes for multi-TRP technology.

[0081] 5G mobile communications support a wide range of use cases and related applications, including video streaming, augmented reality, different ways of sharing data and various forms of machine type applications (such as (massive) machine type communication (mMTC), including vehicle safety, different sensors and real-time control. 5G is expected to have multiple radio interfaces, namely sub-6GHz, centimeter wave (cmWave) and millimeter wave (mmWave), and be able to integrate with existing traditional radio access technologies (such as LTE). At least in the early stages, the integration with LTE can be implemented as a system where macro coverage is provided by LTE and 5G radio interface access comes from small cells through aggregation to LTE. In other words, 5G plans to support inter-RAT operability (such as LTE-5G) and inter-RI operability (inter-radio interface operability, such as sub-6GHz-cmWave, sub-6GHz-cmWave-mmWave). One of the concepts considered for use in 5G networks is network slicing, in which multiple independent and dedicated virtual sub-networks (network instances) can be created within the same infrastructure to run services with different requirements for latency, reliability, throughput and mobility.

[0082] The frequency bands for 5G NR are divided into two frequency ranges: Frequency Range 1 (FR1) includes the bands below 6 GHz, which are the bands traditionally used by previous standards, but also new bands that are extended to cover potential new spectrum supply from 410 MHz to 7125 MHz, and Frequency Range 2 (FR2) includes the bands from 24.25 GHz to 52.6 GHz. FR2 thus includes bands in the millimeter wave range, which require somewhat different approaches in terms of radio resource management compared to bands in FR1 due to their shorter range and higher available bandwidth.

[0083] The current architecture in LTE networks is completely distributed in the radio and completely centralized in the core network. Low-latency applications and services in 5G require content to be close to the radio, which leads to local splitting and multi-access edge computing (MEC). 5G enables analysis and knowledge generation at the data source. This approach requires the use of resources that may not be continuously connected to the network, such as laptops, smartphones, tablets, and sensors. MEC provides a distributed computing environment for application and service hosting. It is also able to store and process content in close proximity to cellular subscribers for faster response times. Edge computing covers a wide range of technologies such as wireless sensor networks, mobile data collection, mobile signature analysis, collaborative distributed peer-to-peer networking and processing, and can also be divided into local cloud / fog computing and grid / mesh computing, dew computing, mobile edge computing, cloudlets, distributed data storage and retrieval, autonomous self-healing networks, remote cloud services, augmented reality and virtual reality, data caching, the Internet of Things (massive connectivity and / or latency critical), critical communications (autonomous driving vehicles, traffic safety, real-time analysis, time-critical control, healthcare applications).

[0084] The communication system can also communicate with other networks, such as the public switched telephone network or the Internet 312, or utilize services provided by them. The communication network can also support the use of cloud services, for example, at least part of the core network operations can be performed as a cloud service (this is in the Figure 3 The communication system may also include a central control entity or the like, providing facilities for networks of different operators to collaborate, for example, in spectrum sharing.

[0085] Edge cloud can be introduced into the radio access network (RAN) by leveraging network function virtualization (NFV) and software defined networking (SDN). Using edge cloud may mean performing access node operations at least partially in a server, host or node that is operably coupled to a remote radio head or base station including radio components. Node operations may also be distributed among multiple servers, nodes or hosts. The application of cloudRAN architecture enables RAN real-time functions to be performed on the RAN side (in distributed units (DUs)) and non-real-time functions to be performed in a centralized manner (in centralized units (CUs) 308).

[0086] It should also be understood that the distribution of labor between core network operations and base station operations may be different from LTE or even non-existent. Some other technological advances that may be used are big data and all-IP, which may change the way networks are built and managed. 5G (or New Radio (NR)) networks are designed to support multiple hierarchical structures, where MEC servers can be placed between the core and the base station or NodeB (gNB). It should be understood that MEC can also be applied to 4G networks. gNB is the next generation NodeB (or new NodeB) that supports 5G networks (i.e. NR).

[0087] 5G can also utilize non-terrestrial nodes 306, such as access nodes, to enhance or supplement the coverage of 5G services, such as by providing backhaul, wireless access for wireless devices, service continuity for machine-to-machine (M2M) communications, service continuity for Internet of Things (IoT) devices, service continuity for passengers on board, ensuring service availability for critical communications, and / or ensuring service availability for future railway / maritime / aviation communications. Non-terrestrial nodes can have a fixed position relative to the surface of the earth, or non-terrestrial nodes can be mobile non-terrestrial nodes that can move relative to the surface of the earth. Non-terrestrial nodes may include satellites and / or HAPS. Satellite communications can utilize geostationary earth orbit (GEO) satellite systems or low earth orbit (LEO) satellite systems, particularly giant constellations (systems in which hundreds of (nano) satellites are deployed). Each satellite in a giant constellation can cover several satellite-enabled network entities that create ground cells. Ground cells can be created by ground relay nodes 304 or by gNBs located on the ground or on satellites.

[0088] It is understood by those skilled in the art that the depicted system is only an example of a part of a radio access system, and in practice, the system may include multiple (e / g)NodeBs, a user equipment may access multiple radio cells, and the system may also include other devices, such as physical layer relay nodes or other network elements, etc. At least one of the (e / g) NodeBs may be a home (e / g) NodeB. In addition, in the geographical area of ​​the radio communication system, a variety of different types of radio cells and a plurality of radio cells may be provided. The radio cells may be macro cells (or umbrella cells), which are large cells, typically up to tens of kilometers in diameter, or smaller cells, such as micro cells, femto cells or pico cells. Figure 1 The (e / g)NodeB can provide any type of these cells. A cellular radio system can be implemented as a multi-layer network including several cells. Usually, in a multi-layer network, one access node provides one cell or multiple cells, so multiple (e / g)NodeBs are needed to provide this network structure.

[0089] In order to meet the needs of improving the deployment and performance of communication systems, the concept of "plug-and-play" (e / g) NodeB is introduced. Generally, a network that can use "plug-and-play" (e / g) NodeB includes, in addition to the home (e / g) NodeB (H(e / g) NodeBs), the home NodeB gateway or HNB-GW ( Figure 1 (not shown). An HNB gateway (HNB-GW) usually installed in an operator's network can aggregate services from a large number of HNBs back to the core network.

[0090] The Radio Resource Control (RRC) protocol is used in various wireless communication systems to define the air interface between the UE and the base station (such as eNB / gNB). 3GPP specifies the protocol in TS 36.331 for LTE and TS 38.331 for 5G. As far as RRC is concerned, the UE can operate in idle mode or connected mode in LTE and 5G, where the radio resources available to the UE depend on the mode the UE is currently in. In 5G, the UE can also operate in inactive mode. In RRC idle mode, the UE has no connection for communication, but the UE is able to listen to paging messages. In RRC connected mode, the UE can operate in different states, such as CELL_DCH (dedicated channel), CELL_FACH (forward access channel), CELL_PCH (cell paging channel) and URA_PCH (URA paging channel). The UE can communicate with the eNB / gNB through various logical channels such as broadcast control channel (BCCH), paging control channel (PCCH), common control channel (CCCH), dedicated control channel (DCCH) and dedicated traffic channel (DTCH).

[0091] The transitions between states are controlled by the state machine of RRC. When the UE is powered on, it is in disconnected mode / idle mode. The UE can transition to RRC connected mode by initial attach or connection establishment. If the UE is inactive for a short period of time, the eNB / gNB can suspend its session by moving to RRC inactive (Inactive) and resume its session by moving to RRC connected mode. The UE can move from RRC connected mode or from RRC inactive mode to RRC idle mode.

[0092] The actual user and control data from the network to the UE is sent through downlink physical channels. In 5G, the downlink physical channels include the physical downlink control channel (PDCCH) carrying the necessary downlink control information (DCI), the physical downlink shared channel (PDSCH) carrying user data and system information for users, and the physical broadcast channel (PBCH) carrying the necessary system information to enable the UE to access the 5G network.

[0093] User and control data from the UE to the network are sent through uplink physical channels. In 5G, the uplink physical channels include a physical uplink control channel (PUCCH) for uplink control information including HARQ feedback acknowledgment, scheduling requests, and downlink channel state information for link adaptation, a physical uplink shared channel (PUSCH) for uplink data transmission, and a physical random access channel (PRACH) for UE to request connection establishment called random access.

[0094] In the following, enhanced beamforming methods for AI / ML-assisted beam prediction in the spatial domain for multi-criteria beam optimization will be described in more detail according to various embodiments.

[0095] Figure 4 A simplified signaling diagram process performed between a user equipment (UE) and a base station gNB according to an embodiment of the present disclosure is shown. The UE reports 401 to the gNB an indication of the beam prediction capability in the spatial domain of the UE. The capability of applying the beam prediction in the spatial domain of the trained machine learning (ML) model may include one or more of the following aspects.

[0096] According to one aspect, the UE can measure a set of indicated reference signals (RS) and use these measurements as input to a trained ML model to predict the best beam (reference signal index (RS index)), wherein the best beam derived from the beam prediction (which is the output of the ML model) can include beams from both one set of reference signals and another set of reference signals. The other set of RS can be explicitly or implicitly indicated to the UE and / or known to the UE, but not measured by the UE.

[0097] The UE may also indicate Option 1 and / or Option 2 capabilities.

[0098] In option 1, the UE is able to apply the trained ML model for beam prediction in the spatial domain in different contexts, where the context includes at least the following.

[0099] - Mode 1: The "best beam" prediction in the spatial domain is understood as the beam prediction of the strongest beam. In this context, the strongest beams are those with the highest Layer 1 Received Signal Received Power (L1-SRP) or Layer 1 Signal to Interference and Noise Ratio (L1-SINR)

[0100] - Mode 2: The "best beam" prediction in the spatial domain is understood as the prediction of the most efficient beam for scheduling. In this case, such beams are those that may allow faster scheduling and are not necessarily the strongest beams from a power (e.g. LI-RSRP, LI-SINR) perspective.

[0101] The ML model used for beam prediction may switch from one mode to another with or without additional latency, and the ML model may not have any additional training, validation, or testing phases when switching operating modes.

[0102] In option 2, which is a variation of option 1, the UE is able to apply the trained ML model for beam prediction in the spatial domain, which also provides at least one additional output metric associated with each best beam, where the metric may allow identification of the strongest beam and / or the most efficient beam.

[0103] The UE may also indicate limitations and / or constraints or capabilities of the ML model input / output parameters, dimensions and other related aspects, where the input parameters include at least one additional parameter related to the beam likelihood (the likelihood of the gNB using a beam or statistics related to beam usage) of one set of RSs and / or another set of RSs.

[0104] The gNB receives the capability indication and checks 402 the content of the capability indication and may perform the following operations.

[0105] Based on the reported UE capabilities, the gNB may determine a trained ML model for beam prediction in the spatial domain and, based on the determination, configure the UE to use the trained ML model.

[0106] According to one aspect, for example, the ML model can be trained offline at the gNB, where the trained model can be transferred to the UE after training. According to another aspect, the ML model can be trained offline at the UE. The ML model can have a continuous learning process and can be trained at the gNB or the UE. If trained at the gNB, updates to the ML model can be sent to the UE. Any other collaborative mode for model deployment may also be feasible.

[0107] The gNB may then send 403 to the UE one or more configurations defining two sets of RS indices, wherein a first set of RS indices comprises measurement resources, i.e., a beam set for performing actual signal measurements, and a second set of RS indices is used for prediction resources, i.e., a beam set to be considered in the prediction.

[0108] The indication of the two sets of RS indices may occur explicitly (eg, via RRC configuration, a Media Access Control Element (MAC-CE) indication) or implicitly (eg, based on a set of predefined rules in the specification, where the two sets of RS indices are linked to each other).

[0109] According to an embodiment, the gNB may select RS indices to be included in the first set and the second set based on, for example, a determined direction of the UE relative to the gNB. Thus, the selection of beams may be performed among those beams directed toward the UE, and beams directed slightly away from the direction of the UE may be selected into the first set and the second set. Thus, beams directed in completely different directions (i.e., away from the UE) may not be selected into the first set or the second set. Thus, some unnecessary computational work may be avoided.

[0110] The gNB may also send 404 a configuration and / or indication to the UE that defines or updates at least one parameter related to the likelihood of a beam for the first set of RSs and / or the second set of RSs. In this context, the likelihood of a beam means the likelihood of the gNB using the beam or statistics related to beam usage.

[0111] According to one embodiment of the present disclosure, the likelihood of the beam may be configured within the same configuration in which the first set of RSs and / or the second set of RSs are configured. For example, if the first set of RSs and the second set of RSs are configured as channel measurement resource sets (resourcesForChannelMeasurement sets) in a channel state information (CSI) report configuration, the likelihood value may be associated with the resources in each channel measurement resource set.

[0112] According to another embodiment of the present disclosure, the likelihood of the beam may be updated through dynamic signaling (e.g., through a media access control element (MAC-CE) for the first set of RSs and / or the second set of RSs, or in a general manner for the RS level). For example, the MAC-CE may update the likelihood values ​​of one or more RSs, where the MAC-CE is specifically designed for this purpose.

[0113] The beam likelihood information may include the following information:

[0114] -Measured or expected beam usage information in the time domain over a specific time window. In other words, the number of times a beam has been scheduled or is expected to be scheduled divided by all scheduling counts within the time window.

[0115] - Measured or expected RB (resource block) load per beam. This allows the ML model to understand if UE traffic can be offloaded to a certain beam, or if the UE should be offloaded from the beam.

[0116] - Prioritization of beams. The UE can know if the gNB prioritizes certain beams due to higher priority users. This enables the ML model to understand which beams are more likely to be scheduled first.

[0117] When the UE receives the configuration of the first and second RS sets and the applicability likelihood of the beams, the UE may apply the trained model based on one of the following ways.

[0118] According to the first option, if the UE is able to apply the trained ML model in different contexts, the gNB may also send 405 a configuration or indication to the UE to use mode 1 or mode 2 for beam prediction. If mode 1 strongest beam is applied to beam prediction, the UE may use equal values ​​for beam likelihood in the input of the ML model, where the predicted result represents the strongest beam (the beam with the highest L1-RSRP or L1-SINR) and is not affected by the configured / updated beam likelihood. If mode 2 efficient scheduling is applied to beam prediction, the UE may use the latest value of beam likelihood in the input of the ML model, where the predicted result represents the most suitable beam (a beam that may allow faster scheduling, not necessarily the strongest beam) and is affected by the configured / updated beam likelihood.

[0119] According to the second option, the gNB provides at least one additional output metric associated with each best beam, the UE can use the latest value of the beam likelihood in the input of the ML model, and the predicted result can provide an additional metric for each beam for the best beam, wherein the metric allows identification of the strongest beam and the most suitable beam.

[0120] When the UE receives the above message from the gNB, the UE may perform 406 signal strength measurements (RS measurements) on the indicated first set of reference signals (RS) and use the results of these measurements as inputs to the trained ML model. The UE also uses the likelihood and the indicated operating mode as additional inputs to the ML model. The ML model also takes into account the indicated first set of reference signals and the indicated second set of reference signals.

[0121] When the UE has executed 407 the ML model to obtain a result, the output of the trained ML model may be a ranked list of at least a portion of the beams of the first set and the second set, the ranking being based on a predetermined criterion. According to an embodiment, the criterion comprises one or more of: the best strongest beam, the most suitable beam, or the best beam with an additional metric. For example, the ranked list may include beam indices of all beams of the first set and the second set, the beams being ranked with the most suitable beam being the highest and the least suitable beam being the lowest. The most suitable may mean, for example, the beam that is most likely to provide the minimum delay, the minimum power consumption, the most efficient channel utilization, the minimum interference, etc.

[0122] The UE uses the output of the ML model to generate 408 a channel state information (CSI) report, where the best beam is reported 409 to the gNB. The CSI report may include a beam index (CSI-RS resource indicator (CRI) and / or SSB resource indicator (SSBRI) with or without L1-RRP / L1-SINR) and CSI feedback, i.e., rank indicator (RI) / precoding matrix indicator (PMI) / channel quality information (CQI) / layer indicator (LI).

[0123] The following will refer to Figure 5 An exemplary scenario is described in more detail. In this scenario, a subset of narrow beam measurements is used to predict the ranking of the best narrow beam at the UE.

[0124] The gNB decides to select the beams marked x1, x2, x3 for measurement and the other beams marked y1, y2, y3 for prediction. The gNB sends a configuration, i.e., the first set of RS indices contains the RS indices for beams x1, x2 and x3, and the second set of RS indices contains the RS indices for beams y1, y2 and y3. Therefore, the UE obtains a configuration that defines a first subset of beams x1, x2 and x3 for measurement and a second subset of beams y1, y2 and y3 for prediction. The UE also receives likelihood information, which can be marked as L1, L2, L3, ... for mode 1 (strongest beam) and 0, 0, 0, ... for mode 2 (efficient scheduling), associated with each beam and operating mode (e.g., mode 1 or mode 2) to be considered in beam prediction. As Figure 5As shown, the UE uses the measurement results of the beams indicated as RS indices in the first set, the corresponding L1-RSRP values, and the likelihood values ​​corresponding to the indication pattern as inputs to the ML model. In this example, the UE performs ML model prediction locally and provides a CRI list associated with K beams. Figure 5 As shown, the list includes measured beams (solid fill) and predicted beams (hollow fill), for example, which are arranged in descending order from the most likely best beam (e.g., for mode 1, the most likely strongest beam) to the less likely best beam (e.g., for mode 1, the less likely strongest beam). Finally, the sorted beam list is added to the beams reporting uplink control information to be sent back to the gNB.

[0125] Figure 6 and Figure 7 An example implementation of the UE-side ML 610 is shown, which predicts the best beam for scheduling with the help of beam likelihood information (option 1, mode 2).

[0126] The inputs to the ML model 610 are beam likelihood information 612 from the gNB and internal channel measurements 614 of the UE. The predictions 616 are then used as feedback to the network to influence the gNB spatial beam selection / scheduling.

[0127] Figure 7 Some potential benefits of the beam likelihood information above are shown. In this example, the UE is measuring beam ID 3 as the strongest beam. With the baseline implementation, the UE will feed back the strongest beam portion of the CSI feedback and the gNB will then schedule the UE with the strongest beam. By doing this, the scheduling latency for all UEs will increase as the new beam must be scheduled in the time domain, sharing resources with two other beams.

[0128] However, the beam likelihood information provides the UE with the necessary information for the scheduler 600 to estimate the best beam, thereby improving throughput and / or delay. The example shows that beam ID 1 and ID 2 are equally scheduled in the time domain, while the residual block (RB) load of beam ID 1 is higher than beam ID 2. This information can be signaled to the UE as part of the beam likelihood information, for example as follows:

[0129] -Beam ID 1: 50% scheduling possibility and 60% RB load

[0130] -Beam ID 2: 50% scheduling possibility and 30% RB load

[0131] -Beam ID 3: 0% scheduling possibility and 0&RB load

[0132] With this information, combined with the UE’s internal channel measurements, the ML model can understand the following:

[0133] - How much radio resources are needed per beam to schedule users (UE channel measurement)

[0134] - How much radio resources are available per beam (RB load)

[0135] - What is the scheduling delay based on beam likelihood (scheduling possibility)?

[0136] In summary, the UE’s ML model can infer that it is better to recommend the gNB to schedule at beam ID 2, rather than utilizing the strongest beam. Therefore, not only will the user running ML benefit from the lower latency, but other UEs may also benefit from this scenario, as no new beam has to be scheduled in the time domain.

[0137] This method is Figure 8 The invention discloses a flowchart reflecting the operation of a terminal device (such as a user equipment (UE)), wherein the method includes sending (800) by the user equipment an indication of beam prediction capability in a spatial domain of the device; receiving (802) an indication of the likelihood of a beam for at least one beam prediction mode in the spatial domain; receiving (804) an indication of a beam prediction mode in the spatial domain; measuring (806) one or more reference signals from the beam to obtain a set of measurement results; using (808) the measurement results as input to a machine learning model and using the likelihood of the beam based on the beam prediction mode in the spatial domain, wherein the machine learning model is configured to use the input to generate a ranked list of beams of at least one or more reference signals measured; generating (810) a channel state information report based on the ranked list of beams; and sending (812) a report including an index of the beam of the ranked list.

[0138] According to an embodiment, the method includes receiving two sets of RS indices having a first configuration and a second configuration, the first configuration defining a first set of RS indices of beams on which signal measurement is to be performed, and the second configuration defining a second set of RS indices of beams to be considered in prediction; measuring one or more reference signals of beams from the first set to obtain a set of measurement results; and using the measurement results as input to a machine learning model and using the likelihood of the beams according to a beam prediction pattern in a spatial domain, wherein the machine learning model is configured to use the input to generate a sorted list of beams of the first set and the second set.

[0139] According to an embodiment, the machine learning model is configured to construct a ranked list based on one or more of: best strongest beam, best fit beam, or best beam with additional metrics.

[0140] According to an embodiment, the method includes using equal values ​​of the likelihood of the beam in the input of the machine learning model in the best strongest beam mode, wherein the predicted result represents the strongest beam, and using the latest value of the likelihood of the beam in the input of the machine learning model in the best fit beam mode, wherein the predicted result represents the best fit beam.

[0141] According to an embodiment, the method comprises including in the report one or more of:

[0142] - beam index,

[0143] - rank indicator,

[0144] - precoding matrix indicator,

[0145] - Channel quality information, and

[0146] - Layer indicator

[0147] -Beam reception quality (L1-RSRP / L1-SINR)

[0148] - Applicable UE panel indicators.

[0149] According to an embodiment, the method comprises receiving an indication of a context for beam prediction in the spatial domain.

[0150] According to an embodiment, the mode includes at least the following:

[0151] - the best beam prediction in the spatial domain is the beam prediction of the strongest beam,

[0152] - The best beam prediction in the spatial domain is the beam prediction for the most efficient beam for scheduling.

[0153] According to an embodiment, the method includes updating the machine learning model; and sending the updated machine learning model to the network element.

[0154] According to an embodiment, the beam likelihood information includes one or more of the following information:

[0155] -- measured or expected beam usage information in the time domain over a specific time window,

[0156] -- measured or expected resource block loading per beam,

[0157] --Beam priority.

[0158] According to an embodiment, the method comprises updating the likelihood of the beam by dynamic signaling for at least one of the first set of RS indices and / or the second set of RS indices.

[0159] According to an embodiment, the method includes switching a machine learning model used for beam prediction from one mode to another mode.

[0160] According to an embodiment, the method comprises performing the switching of the operating mode without any additional training, validation or testing phase.

[0161] According to an embodiment, the method includes indicating limitations, constraints or capabilities of machine learning model input / output parameters, dimensions and other related aspects, wherein the input parameters include at least one additional parameter related to consideration of the likelihood of the beam.

[0162] According to an embodiment, the method includes receiving likelihood information, which may be labeled L1, L2, L3, ... for mode 1 (strongest beam) and 0, 0, 0, ... for mode 2 (efficient scheduling), associated with each beam to be considered in beam prediction and an operating mode (e.g., mode 1 or mode 2).

[0163] According to one aspect, an apparatus, such as a UE, includes a module for sending an indication of a beam prediction capability in a spatial domain of the apparatus; a module for receiving an indication of a likelihood of a beam for at least one beam prediction mode in the spatial domain; a module for receiving an indication of a beam prediction mode in the spatial domain; a module for measuring one or more reference signals from a beam to obtain a set of measurement results; a module for using the measurement results as input to a machine learning model and using the likelihood of the beam according to the beam prediction mode in the spatial domain, wherein the machine learning model is configured to use the input to generate a ranked list of beams of at least one or more measured reference signals; a module for generating a channel state information report based on the ranked list of beams; and a module for sending a report including an index of the beam of the ranked list.

[0164] According to another aspect, an apparatus comprises a plurality of antenna panels, at least one processor and at least one memory, on which a computer program code is stored, and the at least one memory and the computer program code are configured to, together with the at least one processor, cause the apparatus to at least perform: sending an indication of a beam prediction capability of the apparatus in a spatial domain; receiving an indication of a likelihood of a beam for at least one beam prediction mode in the spatial domain; receiving an indication of a beam prediction mode in the spatial domain; measuring one or more reference signals from the beam to obtain a set of measurement results; using the measurement results as input to a machine learning model and using the likelihood of the beam according to the beam prediction mode in the spatial domain, wherein the machine learning model is configured to use the input to generate a sorted list of beams of at least one or more measured reference signals; generating a channel state information report based on the sorted list of beams; and sending a report including an index of the beam in the sorted list.

[0165] On the other hand, it relates to a computer program product stored on a non-transitory storage medium, comprising computer program code, which, when executed by at least one processor, causes the device to at least perform: sending an indication of the device's beam prediction capability in the spatial domain; receiving an indication of the likelihood of a beam for at least one beam prediction mode in the spatial domain; receiving an indication of a beam prediction mode in the spatial domain; measuring one or more reference signals from the beam to obtain a set of measurement results; using the measurement results as input to a machine learning model and using the likelihood of the beam based on the beam prediction mode in the spatial domain, wherein the machine learning model is configured to use the input to generate a ranked list of beams of at least one or more measured reference signals; generating a channel state information report based on the ranked list of beams; and sending a report including an index of the beam of the ranked list.

[0166] According to an embodiment, the device includes a computer program code, which is configured to cause the device to perform: receiving two sets of RS indices having a first configuration and a second configuration, the first configuration defining a first set of RS indices of beams on which signal measurement is to be performed, and the second configuration defining a second set of RS indices of beams to be considered in prediction; measuring one or more reference signals from the beams of the first set to obtain a set of measurement results; and using the measurement results as input to a machine learning model and using the likelihood of the beam according to a beam prediction pattern in the spatial domain, wherein the machine learning model is configured to use the input to generate a sorted list of beams of the first set and the second set.

[0167] According to an embodiment, the apparatus comprises a computer program code configured to cause the apparatus to perform: constructing a ranked list by a machine learning model of beam prediction in the spatial domain based on one or more of the following:

[0168] The best and strongest beam,

[0169] Best for beamforming, or

[0170] Optimal beam with additional metrics.

[0171] According to an embodiment, the apparatus comprises computer program code configured to cause the apparatus to execute:

[0172] In the best strongest beam mode equal values ​​of the likelihood of the beams are used at the input of the machine learning model, where the predicted result represents the strongest beam,

[0173] In Best Fit Beam mode the latest value of the likelihood of the beam is used as the input to the machine learning model, where the predicted result represents the best fit beam.

[0174] According to an embodiment, the apparatus comprises a computer program code configured to cause the apparatus to perform: including one or more of the following in a channel state information report:

[0175] Beam index,

[0176] Beam reception quality indicator,

[0177] Applicable UE panel indicators,

[0178] Rank indicator,

[0179] Precoding matrix indicator,

[0180] Channel quality information, and

[0181] Layer indicator.

[0182] According to an embodiment, the apparatus comprises computer program code configured to cause the apparatus to execute:

[0183] An indication of an update of a context for beam prediction in the spatial domain is received.

[0184] According to an embodiment, the apparatus comprises computer program code configured to cause the apparatus to execute:

[0185] Update machine learning models; and

[0186] The updated machine learning model is sent to the network element.

[0187] According to an embodiment, the apparatus comprises computer program code configured to cause the apparatus to execute:

[0188] The beam likelihood information includes one or more of the following information:

[0189] - measured or expected beam usage information in the time domain over a specific time window,

[0190] - measured or expected resource block loading per beam,

[0191] - Priority of the beam.

[0192] According to an embodiment, the apparatus comprises computer program code configured to cause the apparatus to execute:

[0193] The likelihood of the beam is updated by dynamic signaling for at least one of the first set of RS indexes and / or the second set of RS indexes.

[0194] According to an embodiment, the apparatus comprises computer program code configured to cause the apparatus to execute:

[0195] Switch the machine learning model used for beam prediction from one mode to another.

[0196] According to an embodiment, the apparatus comprises a computer program code configured to cause the apparatus to perform the switching of the operation mode without any additional training, validation or testing phase.

[0197] According to an embodiment, the apparatus comprises computer program code configured to cause the apparatus to execute:

[0198] Indicates limitations, constraints, or capabilities of machine learning model input / output parameters, dimensions, and other related aspects, where the input parameters include at least one additional parameter related to the likelihood of the beam being considered.

[0199] According to an embodiment, the apparatus comprises computer program code configured to cause the apparatus to execute:

[0200] Likelihood information for Mode 1 and for Mode 2 associated with each beam and operating mode to be considered in beam prediction is received.

[0201] In general, various embodiments of the present invention may be implemented in hardware or dedicated circuits or any combination thereof. Although various aspects of the present invention may be shown and described as block diagrams or using some other pictorial representations, it is well understood that these blocks, devices, systems, techniques or methods described herein may be implemented in hardware, software, firmware, dedicated circuits or logic, general purpose hardware or controllers or other computing devices or some combination thereof as non-limiting examples.

[0202] Embodiments of the present invention may be practiced in various components such as integrated circuit modules. The design of integrated circuits is generally a highly automated process. Complex and powerful software tools are available to convert logic-level designs into semiconductor circuit designs ready to be etched and formed on semiconductor substrates.

[0203] Programs, such as those offered by Synopsys Inc. of Mountain View, Calif., and Cadence Design Inc. of San Jose, Calif., use well-established design rules along with a library of pre-stored design modules to automatically route conductors and locate components on a semiconductor chip. Once the design of a semiconductor circuit is complete, the resulting design in a standardized electronic format (e.g., Opus, GDSII, etc.) can be transferred to a semiconductor fabrication facility or "fab" for manufacturing.

[0204] The foregoing description provides a complete and informative description of exemplary embodiments of the present invention by way of exemplary and non-limiting examples. However, various modifications and adaptations will become apparent to those skilled in the relevant art in view of the foregoing description when read in conjunction with the accompanying drawings and the appended claims. However, all such and similar modifications to the teachings of the present invention will still fall within the scope of the present invention.

Claims

1. A device comprising: means for transmitting an indication of beam prediction capability in the spatial domain of the apparatus; means for receiving an indication of a likelihood of a beam for at least one beam prediction mode in the spatial domain; means for receiving an indication of a beam prediction mode in the spatial domain; means for measuring one or more reference signals from the beam to obtain a set of measurements; means for using the measurements as inputs to a machine learning model and using the likelihood of the beams according to the beam prediction pattern in the spatial domain, wherein the machine learning model is configured to use the inputs to generate a ranked list of at least the measured beams of one or more reference signals; means for generating a channel state information report based on the ranked list of beams; as well as Means for sending the report comprising an index of the beam of the ranked list.

2. The device according to claim 1, comprising: Means for receiving two sets of RS indices having a first configuration defining a first set of RS indices for beams on which signal measurements are to be performed and a second configuration defining a second set of RS indices for beams to be considered in prediction; means for measuring one or more reference signals from beams of said first set to obtain a set of measurements; as well as A module for using the measurements as input to a machine learning model and using the likelihood of the beam based on the beam prediction pattern in the spatial domain, wherein the machine learning model is configured to use the input to generate a ranked list of the beams of the first set and the second set.

3. The device according to claim 1 or 2, wherein: The machine learning model for beam prediction in the spatial domain is configured to construct the ranked list based on one or more of: The best and strongest beam, Best for beamforming, or Optimal beam with additional metrics.

4. The device according to claim 3, wherein: The device comprises: means for using equal values ​​of the likelihoods of beams in the input to the machine learning model in the best strongest beam mode, wherein the predicted outcome represents the strongest beam, Means for using, in the best fit beam mode, a most recent value of the likelihood of a beam in the input to the machine learning model, wherein the predicted result represents the best fit beam.

5. The device according to any one of claims 1 to 4, wherein: The means for generating a channel state information report is configured to include in the report one or more of: Beam index, Beam reception quality indicator, Applicable UE panel indicators, Rank indicator, Precoding matrix indicator, Channel quality information, and Layer indicator.

6. The device according to any one of claims 1 to 5, comprising: Means for receiving an update of a context for beam prediction in the spatial domain.

7. The device according to claim 6, wherein: The mode includes at least the following: The best beam prediction in the spatial domain is the beam prediction for the strongest beam, The best beam prediction in the spatial domain is the beam prediction for the most efficient beam for scheduling.

8. The device according to any one of claims 1 to 7, comprising: A module for updating the machine learning model; as well as Means for sending the updated machine learning model to the network element.

9. The device according to any one of claims 1 to 8, wherein: The beam likelihood information includes one or more of the following information: - measured or expected beam usage information in the time domain over a specific time window, - measured or expected resource block loading per beam, - Priority of the beam.

10. The device according to any one of claims 1 to 9, comprising: Means for updating the likelihood of the beam through dynamic signaling for at least one of the first set of RS indexes and / or the second set of RS indexes.

11. The device according to any one of claims 1 to 10, comprising: A module for switching the machine learning model for beam prediction from one mode to another.

12. The apparatus of claim 11, the means for switching being configured to perform the switching of the operating mode without any additional training, validation, or testing phase.

13. The device according to any one of claims 1 to 12, comprising: A module for indicating limitations, constraints or capabilities of the machine learning model input / output parameters, dimensions and other related aspects, wherein the input parameters include at least one additional parameter related to the likelihood of the beam being considered.

14. The device according to any one of claims 1 to 13, comprising: Means for receiving likelihood information for mode 1 and for mode 2, said likelihood information being associated with each beam to be considered in said beam prediction and said mode of operation.

15. An apparatus comprising at least one processor and at least one memory, wherein the at least one memory has computer program code stored thereon, wherein the at least one memory and the computer program code are configured to, together with the at least one processor, cause the apparatus to at least perform: sending an indication of a beam prediction capability of the apparatus in a spatial domain; receiving an indication of a likelihood of a beam for at least one beam prediction mode in the spatial domain; receiving an indication of a beam prediction mode in a spatial domain; measuring one or more reference signals from the beam to obtain a set of measurement results; using the measurements as inputs to a machine learning model and using the likelihood of the beams according to the beam prediction pattern in the spatial domain, wherein the machine learning model is configured to use the inputs to generate a ranked list of at least the measured beams of one or more reference signals; generating a channel state information report based on the ranked list of beams; and The report is sent including an index of the beam of the ordered list.

16. The apparatus according to claim 15, wherein the at least one memory has computer program code stored thereon, the at least one memory and the computer program code being configured to, together with the at least one processor, cause the apparatus to at least perform: Receiving two sets of RS indices having a first configuration and a second configuration, the first configuration defining a first set of RS indices of beams on which signal measurements are to be performed, the second configuration defining a second set of RS indices of beams to be considered in prediction; measuring one or more reference signals from beams of the first set to obtain a set of measurements; as well as Using the measurements as input to a machine learning model and using the likelihood of the beams based on the beam prediction pattern in the spatial domain, wherein the machine learning model is configured to use the input to generate a ranked list of the beams of the first set and the second set.

Citation Information

Cited By

  • Life cycle management procedures for uplink beam prediction

    WO2026137116A1