Method, apparatus and computer program

Abstract

There is provided an apparatus comprising means for receiving (700, 714), from a user equipment, measurement data relating to one or more cells and / or one or more beams of a network; determining (716), based on the measurement data, a prediction comprising at least one sequence of one or more beams predicted to have the highest signal quality for the user equipment at the respective one or more times; and sending (718) the prediction to the user equipment.

Description

Technical Field

[0001] The present application relates to a method, apparatus, system and computer program, and in particular, but not exclusively, to predicting a beam sequence to be used by a user equipment in the future. Background Art

[0002] A communication system can be viewed as a facility for enabling communication sessions between two or more entities (such as user terminals, base stations, and / or other nodes) by providing carrier waves between the various entities involved in the communication path. For example, a communication system can be provided with the aid of a communication network and one or more compatible communication devices. A communication session may include, for example, data communications for carrying communications such as voice, video, electronic mail (email), text messages, multimedia, and / or content data. Non-limiting examples of the services provided include two-way or multi-way calls, data communications or multimedia services, and access to data network systems such as the Internet.

[0003] In a wireless communication system, at least a portion of a communication session between at least two stations takes place over a wireless link. Examples of wireless systems include public land mobile networks (PLMNs), satellite-based communication systems, and different wireless local area networks, such as wireless local area networks (WLANs). Some wireless systems may be divided into cells, and are therefore often referred to as cellular systems.

[0004] A user can access a communication system with the help of a suitable communication device or terminal. The user's communication device may be referred to as user equipment (UE) or user device. The communication device is equipped with appropriate signal receiving and sending means for achieving communication, such as achieving access to a communication network or communicating directly with other users. The communication device can access a carrier provided by a site (e.g., a base station of a cell) and send and / or receive communications on the carrier.

[0005] Communication systems and associated equipment typically operate according to given standards or specifications that specify what the various entities associated with the system are allowed to do and how this should be achieved. Communication protocols and / or parameters to be used for the connection are also typically defined. An example of a communication system is the Universal Terrestrial Radio Access Network (UTRAN (3G Radio)). Other examples of communication systems are the Long Term Evolution (LTE) of the Universal Mobile Telecommunications System (UMTS) radio access technology and the so-called 5G or New Radio (NR) networks. NR is being standardized by the 3rd Generation Partnership Project (3GPP). Summary of the invention

[0006] According to one aspect, an apparatus is provided, comprising components for: receiving measurement data relating to one or more cells and / or one or more beams of a network from a user equipment; determining a prediction based on the measurement data, the prediction comprising at least one sequence of one or more beams, the one or more beams being predicted to have the highest signal quality for the user equipment at corresponding one or more time moments; and sending the prediction to the user equipment.

[0007] The means may be configured to: send an indication to the user equipment that beam sequence prediction is available for the user equipment, wherein the receiving is performed in response to sending the indication that beam sequence prediction is available.

[0008] The prediction may include at least a first sequence of one or more beams, the one or more beams being a sequence of one or more beams predicted to have the highest signal quality for the user equipment at corresponding one or more moments, and a second sequence of one or more beams at the corresponding one or more moments, wherein at least one of the one or more beams is different between the first sequence and the second sequence.

[0009] Transmitting the prediction may include transmitting at least one confidence interval associated with a corresponding sequence of the at least one sequence.

[0010] The confidence intervals can be different for different moments in the sequence.

[0011] Deterministic predictions can be performed by machine learning algorithms or adaptive algorithms.

[0012] The means may be configured to receive an indication from a user equipment indicating that a prediction for at least one of the one or more time instants is inaccurate.

[0013] The indication may include an indication that the prediction is inaccurate by exceeding a confidence interval.

[0014] The indication may include further measurement data for one or more beams at one or more time instances where the prediction was inaccurate.

[0015] The component can be used to: perform model retraining based on the indication.

[0016] Components may be configured to: determine a new prediction based at least in part on the indication; and send the new prediction to the user equipment.

[0017] Components can be used to: Ensure that the forecast is accurate.

[0018] Determining that the prediction is accurate may include at least one of: receiving an indication from the user device that the prediction is accurate; or, after a certain period of time, not receiving an indication from the user device that the prediction is inaccurate.

[0019] Determining that the prediction is accurate may include determining that the prediction is accurate within a confidence interval.

[0020] The indication that the prediction was accurate may also include information about the sequence of one or more beams that the user equipment has used.

[0021] The component can be used to: in response to determining that the prediction is accurate and based on the prediction, send a signaling control element command to the user equipment to trigger a beam or cell switch from a first beam in time interval t to a second beam in time interval t+1 for the user equipment.

[0022] According to one aspect, a device is provided, comprising components for: sending measurement data related to one or more cells and / or one or more beams of the network to a network node; and receiving a prediction from the network node, the prediction comprising at least one sequence of one or more beams, the one or more beams being predicted to have the highest signal quality for the device at corresponding one or more time instants, wherein the prediction is at least partially based on the measurement data.

[0023] The means may be configured to receive an indication from a network node that beam sequence prediction is available, wherein the sending is performed in response to receiving the indication that beam sequence prediction is available.

[0024] At least one sequence may include at least a first sequence of one or more beams, the one or more beams being a sequence of one or more beams predicted to have the highest signal quality for a user device at corresponding one or more time moments, and a second sequence of one or more beams, wherein at least one of the one or more beams is different between the first sequence and the second sequence.

[0025] Receiving the prediction may include receiving a confidence interval associated with a corresponding sequence of the at least one sequence.

[0026] The confidence intervals can be different for different moments in the sequence.

[0027] The component may be configured to: acquire further measurement data for at least one of the one or more beams for at least one of the corresponding one or more time instances; and determine whether the prediction is accurate based on a comparison between the further measurement data and the prediction.

[0028] The means may be configured to, in response to determining that the prediction for at least one of the one or more time instances is inaccurate, send an indication to the network node indicating that the prediction for at least one of the one or more time instances is inaccurate.

[0029] The indication may include an indication that the prediction is inaccurate by exceeding a confidence interval.

[0030] The indication may comprise further measurement data of at least one of the one or more beams at one of the one or more time instances at which the prediction was inaccurate.

[0031] Components may be configured to receive a new prediction from a network node, wherein the new prediction is based at least in part on an indication that the prediction is inaccurate.

[0032] The component can be used to: in response to determining that the prediction is accurate, send an indication to the network node that the prediction is accurate; or avoid sending an indication to the network node that the prediction is inaccurate, wherein the network node is configured to: assume that the prediction is accurate if the device does not send an indication to the network node that the prediction is inaccurate within a certain period of time.

[0033] Determining that the prediction is accurate may include determining that the prediction is accurate within a confidence interval.

[0034] The indication that the prediction was accurate may also include information about the trajectory followed by the user device.

[0035] The component may be configured to: receive a signaling control element command from a network node to trigger a beam switch from a first beam at time interval t to a second beam at time interval t+1; and perform the beam switch based on the signaling control element command.

[0036] The component may be configured to perform a beam switch from a first beam at time interval t to a second beam at time interval t+1 based on the prediction if the prediction is accurate.

[0037] According to one aspect, an apparatus is provided, comprising at least one processor and at least one memory comprising computer program code, 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: receive measurement data relating to one or more cells and / or one or more beams of a network from a user equipment; determine a prediction based on the measurement data, the prediction comprising at least one sequence of one or more beams, the one or more beams being predicted to have the highest signal quality for the user equipment at corresponding one or more time instants; and send the prediction to the user equipment.

[0038] At least one memory and at least one processor may be configured to cause the apparatus to: send an indication to a user equipment that beam sequence prediction is available for the user equipment, wherein the at least one memory and at least one processor may be configured to: cause the apparatus to receive measurement data in response to sending the indication that beam sequence prediction is available.

[0039] The prediction may include at least a first sequence of one or more beams, the one or more beams being a sequence of one or more beams predicted to have the highest signal quality for the user equipment at corresponding one or more moments, and a second sequence of one or more beams at the corresponding one or more moments, wherein at least one of the one or more beams is different between the first sequence and the second sequence.

[0040] Transmitting the prediction may include transmitting at least one confidence interval associated with a corresponding sequence of the at least one sequence.

[0041] The confidence intervals can be different for different moments in the sequence.

[0042] Deterministic predictions can be performed by machine learning algorithms or adaptive algorithms.

[0043] The at least one memory and the at least one processor may be configured to cause the apparatus to: receive an indication from a user equipment indicating that a prediction for at least one of the one or more time instants is inaccurate.

[0044] The indication may include an indication that the prediction is inaccurate by exceeding a confidence interval.

[0045] The indication may include further measurement data for one or more beams at one or more time instances where the prediction was inaccurate.

[0046] At least one memory and at least one processor may be configured to cause the apparatus to: perform model retraining based on the indication.

[0047] The at least one memory and the at least one processor may be configured to cause the apparatus to: determine a new prediction based at least in part on the indication; and send the new prediction to the user equipment.

[0048] The at least one memory and the at least one processor may be configured to cause the apparatus to: determine that the prediction is accurate.

[0049] At least one memory and at least one processor may be configured to cause the apparatus to determine that the prediction is accurate by performing at least one of: receiving an indication from the user device that the prediction is accurate; or after a certain period of time, not receiving an indication from the user device that the prediction is inaccurate.

[0050] The at least one memory and the at least one processor may be configured to cause the apparatus to determine that the prediction is accurate by determining that the prediction is accurate within a confidence interval.

[0051] The indication that the prediction was accurate may also include information about the sequence of one or more beams that the user equipment has used.

[0052] At least one memory and at least one processor can be configured to cause the device to: in response to determining that the prediction is accurate and based on the prediction, send a signaling control element command to the user equipment to trigger a beam or cell switch from a first beam in time interval t to a second beam in time interval t+1 for the user equipment.

[0053] According to one aspect, a device is provided, comprising at least one processor and at least one memory comprising computer program code, wherein the at least one memory and the computer program code are configured to, together with the at least one processor, cause the device to at least: send measurement data relating to one or more cells and / or one or more beams of a network to a network node; and receive a prediction from the network node, the prediction comprising at least one sequence of one or more beams, the one or more beams being predicted to have the highest signal quality for the device at corresponding one or more time instants, wherein the prediction is based at least in part on the measurement data.

[0054] At least one memory and at least one processor may be configured to cause the apparatus to: receive an indication from a network node that beam sequence prediction is available, wherein the at least one memory and at least one processor may be configured to: cause the apparatus to send measurement data in response to receiving the indication that beam sequence prediction is available.

[0055] At least one sequence may include at least a first sequence of one or more beams, the one or more beams being a sequence of one or more beams predicted to have the highest signal quality for a user device at corresponding one or more time moments, and a second sequence of one or more beams, wherein at least one of the one or more beams is different between the first sequence and the second sequence.

[0056] The at least one memory and the at least one processor may be configured to cause the apparatus to receive the prediction by receiving a confidence interval associated with a respective sequence of the at least one sequence.

[0057] The confidence intervals can be different for different moments in the sequence.

[0058] At least one memory and at least one processor can be configured to enable the device to: obtain additional measurement data for at least one of the one or more beams for at least one of the corresponding one or more moments; and determine whether the prediction is accurate based on a comparison between the additional measurement data and the prediction.

[0059] At least one memory and at least one processor may be configured to cause the apparatus to, in response to determining that the prediction for at least one of the one or more time instants is inaccurate, send an indication to a network node indicating that the prediction for at least one of the one or more time instants is inaccurate.

[0060] The indication may include an indication that the prediction is inaccurate by exceeding a confidence interval.

[0061] The indication may comprise further measurement data of at least one of the one or more beams at one of the one or more time instances at which the prediction was inaccurate.

[0062] The at least one memory and the at least one processor may be configured to cause the apparatus to: receive a new prediction from a network node, wherein the new prediction is based at least in part on an indication that the prediction is inaccurate.

[0063] At least one memory and at least one processor may be configured to cause the apparatus to: in response to determining that a prediction is accurate: send an indication to a network node that the prediction is accurate; or avoid sending an indication to the network node that the prediction is inaccurate, wherein the network node is configured to: assume that the prediction is accurate if the apparatus does not send an indication to the network node that the prediction is inaccurate within a certain period of time.

[0064] The at least one memory and the at least one processor may be configured to cause the apparatus to determine that the prediction is accurate by determining that the prediction is accurate within a confidence interval.

[0065] The indication that the prediction was accurate may also include information about the trajectory followed by the user device.

[0066] At least one memory and at least one processor may be configured to cause the device to: receive a signaling control element command from a network node to trigger a beam switch from a first beam in time interval t to a second beam in time interval t+1; and perform the beam switch based on the signaling control element command.

[0067] The at least one memory and the at least one processor may be configured to cause the apparatus to: if the prediction is accurate, perform beam switching from a first beam at time interval t to a second beam at time interval t+1 based on the prediction.

[0068] According to one aspect, a method is provided, comprising: receiving measurement data relating to one or more cells and / or one or more beams of a network from a user equipment; determining a prediction based on the measurement data, the prediction comprising at least one sequence of one or more beams, the one or more beams being predicted to have the highest signal quality for the user equipment at corresponding one or more time moments; and sending the prediction to the user equipment.

[0069] The method may include sending an indication to the user equipment that beam sequence prediction is available for the user equipment, wherein the receiving is performed in response to sending the indication that beam sequence prediction is available.

[0070] The prediction may include at least a first sequence of one or more beams, the one or more beams being a sequence of one or more beams predicted to have the highest signal quality for the user equipment at corresponding one or more moments, and a second sequence of one or more beams at the corresponding one or more moments, wherein at least one of the one or more beams is different between the first sequence and the second sequence.

[0071] Transmitting the prediction may include transmitting at least one confidence interval associated with a corresponding sequence of the at least one sequence.

[0072] The confidence intervals can be different for different moments in the sequence.

[0073] Deterministic predictions can be performed by machine learning algorithms or adaptive algorithms.

[0074] The method may include receiving an indication from a user equipment indicating that a prediction for at least one of the one or more time instances is inaccurate.

[0075] The indication may include an indication that the prediction is inaccurate by exceeding a confidence interval.

[0076] The indication may include further measurement data for one or more beams at one or more time instances where the prediction was inaccurate.

[0077] The method may include: performing model retraining based on the indication.

[0078] The method may include determining a new prediction based at least in part on the indication; and sending the new prediction to the user equipment.

[0079] The method may include determining that the prediction is accurate.

[0080] Determining that the prediction is accurate may include at least one of: receiving an indication from the user device that the prediction is accurate; or, after a certain period of time, not receiving an indication from the user device that the prediction is inaccurate.

[0081] Determining that the prediction is accurate may include determining that the prediction is accurate within a confidence interval.

[0082] The indication that the prediction was accurate may also include information about the sequence of one or more beams that the user equipment has used.

[0083] The method may include, in response to determining that the prediction is accurate and based on the prediction, sending a signaling control element command to the user equipment to trigger a beam or cell switch from a first beam at time interval t to a second beam at time interval t+1 for the user equipment.

[0084] According to one aspect, a method is provided, comprising: sending measurement data relating to one or more cells and / or one or more beams of a network to a network node; and receiving a prediction from the network node, the prediction comprising at least one sequence of one or more beams, the one or more beams being predicted to have the highest signal quality for a device at corresponding one or more time instants, wherein the prediction is based at least in part on the measurement data.

[0085] The method may include receiving an indication from a network node that beam sequence prediction is available, wherein the sending is performed in response to receiving the indication that beam sequence prediction is available.

[0086] At least one sequence may include at least a first sequence of one or more beams, the one or more beams being a sequence of one or more beams predicted to have the highest signal quality for a user device at corresponding one or more time moments, and a second sequence of one or more beams, wherein at least one of the one or more beams is different between the first sequence and the second sequence.

[0087] Receiving the prediction may include receiving a confidence interval associated with a corresponding sequence of the at least one sequence.

[0088] The confidence intervals can be different for different moments in the sequence.

[0089] The method may include acquiring further measurement data for at least one of the one or more beams for at least one of the corresponding one or more time instances; and determining whether the prediction is accurate based on a comparison between the further measurement data and the prediction.

[0090] The method may include, in response to determining that the prediction for at least one of the one or more time instances is inaccurate, sending an indication to a network node indicating that the prediction for at least one of the one or more time instances is inaccurate.

[0091] The indication may include an indication that the prediction is inaccurate by exceeding a confidence interval.

[0092] The indication may comprise further measurement data of at least one of the one or more beams at one of the one or more time instances at which the prediction was inaccurate.

[0093] The method may include receiving a new prediction from a network node, wherein the new prediction is based at least in part on an indication that the prediction is inaccurate.

[0094] The method may include: in response to determining that the prediction is accurate, sending an indication to the network node that the prediction is accurate; or avoiding sending an indication to the network node that the prediction is inaccurate, wherein the network node is configured to: assume that the prediction is accurate if the device does not send an indication to the network node that the prediction is inaccurate within a certain period of time.

[0095] Determining that the prediction is accurate may include determining that the prediction is accurate within a confidence interval.

[0096] The indication that the prediction was accurate may also include information about the trajectory followed by the user device.

[0097] The method may include: receiving a signaling control element command from a network node to trigger a beam switch from a first beam at a time interval t to a second beam at a time interval t+1; and performing the beam switch based on the signaling control element command.

[0098] The method may include, if the prediction is accurate, performing a beam switch from a first beam at time interval t to a second beam at time interval t+1 based on the prediction.

[0099] According to one aspect, a computer-readable medium is provided, comprising program instructions for causing an apparatus to perform at least the following: receiving measurement data relating to one or more cells and / or one or more beams of a network from a user device; determining a prediction based on the measurement data, the prediction comprising at least one sequence of one or more beams, the one or more beams being predicted to have the highest signal quality for the user device at corresponding one or more time instants; and sending the prediction to the user device.

[0100] The program instructions may be configured to cause the apparatus to execute: sending an indication to a user equipment that beam sequence prediction is available for the user equipment, wherein the receiving is performed in response to sending the indication that beam sequence prediction is available.

[0101] The prediction may include at least a first sequence of one or more beams, the one or more beams being a sequence of one or more beams predicted to have the highest signal quality for the user equipment at corresponding one or more moments, and a second sequence of one or more beams at the corresponding one or more moments, wherein at least one of the one or more beams is different between the first sequence and the second sequence.

[0102] Transmitting the prediction may include transmitting at least one confidence interval associated with a corresponding sequence of the at least one sequence.

[0103] The confidence intervals can be different for different moments in the sequence.

[0104] Deterministic predictions can be performed by machine learning algorithms or adaptive algorithms.

[0105] The program instructions may be configured to cause the apparatus to perform: receiving an indication from a user equipment that a prediction for at least one of the one or more time instants is inaccurate.

[0106] The indication may include an indication that the prediction is inaccurate by exceeding a confidence interval.

[0107] The indication may include further measurement data for one or more beams at one or more time instances where the prediction was inaccurate.

[0108] The program instructions can be used to cause the device to: perform model retraining based on the indication.

[0109] The program instructions may be configured to cause the apparatus to perform: determining a new prediction based at least in part on the indication; and sending the new prediction to the user equipment.

[0110] The program instructions may be used to cause the apparatus to perform: determining that the prediction is accurate.

[0111] Determining that the prediction is accurate may include at least one of: receiving an indication from the user device that the prediction is accurate; or, after a certain period of time, not receiving an indication from the user device that the prediction is inaccurate.

[0112] Determining that the prediction is accurate may include determining that the prediction is accurate within a confidence interval.

[0113] The indication that the prediction was accurate may also include information about the sequence of one or more beams that the user equipment has used.

[0114] The program instructions can be used to cause the device to execute: in response to determining that the prediction is accurate and based on the prediction, send a signaling control element command to the user equipment to trigger a beam or cell switch from a first beam in time interval t to a second beam in time interval t+1 for the user equipment.

[0115] According to one aspect, a computer-readable medium is provided, comprising program instructions for causing an apparatus to at least perform: sending measurement data relating to one or more cells and / or one or more beams of a network to a network node; and receiving a prediction from the network node, the prediction comprising at least one sequence of one or more beams, the one or more beams being predicted to have the highest signal quality for the apparatus at corresponding one or more time instants, wherein the prediction is based at least in part on the measurement data.

[0116] The program instructions may be used to cause the apparatus to perform: receiving an indication from a network node that beam sequence prediction is available, wherein the sending is performed in response to receiving the indication that beam sequence prediction is available.

[0117] At least one sequence may include at least a first sequence of one or more beams, the one or more beams being a sequence of one or more beams predicted to have the highest signal quality for a user device at corresponding one or more time moments, and a second sequence of one or more beams, wherein at least one of the one or more beams is different between the first sequence and the second sequence.

[0118] Receiving the prediction may include receiving a confidence interval associated with a corresponding sequence of the at least one sequence.

[0119] The confidence intervals can be different for different moments in the sequence.

[0120] The program instructions can be used to cause the device to execute: obtaining additional measurement data for at least one of the one or more beams for at least one of the corresponding one or more time moments; and determining whether the prediction is accurate based on a comparison between the additional measurement data and the prediction.

[0121] The program instructions may be used to cause the apparatus to perform: in response to determining that the prediction for at least one of the one or more time instants is inaccurate, sending an indication to a network node indicating that the prediction for at least one of the one or more time instants is inaccurate.

[0122] The indication may include an indication that the prediction is inaccurate by exceeding a confidence interval.

[0123] The indication may comprise further measurement data of at least one of the one or more beams at one of the one or more time instances at which the prediction was inaccurate.

[0124] The program instructions may be used to cause the apparatus to perform: receiving a new prediction from a network node, wherein the new prediction is based at least in part on an indication that the prediction is inaccurate.

[0125] The program instructions can be used to cause the device to execute: in response to determining that the prediction is accurate, send an indication to the network node that the prediction is accurate; or avoid sending an indication to the network node that the prediction is inaccurate, wherein the network node is configured to: assume that the prediction is accurate if the device does not send an indication to the network node that the prediction is inaccurate within a certain period of time.

[0126] Determining that the prediction is accurate may include determining that the prediction is accurate within a confidence interval.

[0127] The indication that the prediction was accurate may also include information about the trajectory followed by the user device.

[0128] The program instructions may be used to cause the apparatus to execute: receiving a signaling control element command from a network node to trigger beam switching from a first beam at time interval t to a second beam at time interval t+1; and performing beam switching based on the signaling control element command.

[0129] The program instructions may be used to cause the apparatus to execute: if the prediction is accurate, performing a beam switch from a first beam at time interval t to a second beam at time interval t+1 based on the prediction.

[0130] According to one aspect, a non-transitory computer readable medium is provided, comprising program instructions for causing an apparatus to at least perform a method according to any of the preceding aspects.

[0131] In the above, many different embodiments have been described. It should be understood that further embodiments can be provided by combining any two or more of the above embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0132] Embodiments will now be described, by way of example only, with reference to the accompanying drawings, in which:

[0133] Figure 1 shows a representation of a network system according to some example embodiments;

[0134] Figure 2 shows a representation of a control device according to some example embodiments;

[0135] Figure 3 shows representations of apparatus according to some example embodiments;

[0136] Figure 4 An example signaling diagram for L1 / 2 inter-cell mobility is shown;

[0137] Figure 5 A diagram showing an example predictive mobility scenario is shown;

[0138] Figure 6 Methods according to some examples are shown;

[0139] Figures 7 to 9 shows signaling interactions according to some examples; and

[0140] Fig.10 An example neural network architecture is shown. DETAILED DESCRIPTION

[0141] In the following, certain embodiments are explained with reference to a mobile communication device capable of communicating via a wireless cellular system and a mobile communication system serving such a mobile communication device. Figure 1 , Figure 2 and Figure 3 Certain general principles of wireless communication systems, their access systems, and mobile communication devices are briefly explained to help understand the technology underlying the described examples.

[0142] Figure 1 A schematic diagram of a 5G system (5GS) is shown. The 5GS may consist of a terminal or user equipment (UE), a 5G radio access network (5GRAN) or a next generation radio access network (NG-RAN), a 5G core network (5GC), one or more application functions (AFs), and one or more data networks (DNs).

[0143] 5G-RAN may include one or more gNodeBs (GNBs), or one or more gNodeB (GNB) distributed unit functions connected to one or more gNodeB (GNB) centralized unit functions.

[0144] 5GC may include the following entities: network slice selection function (NSSF); network exposure function (NEF); network storage function (NRF); policy control function (PCF); unified data management (UDM); application function (AF); authentication server function (AUSF); access and mobility management function (AMF); and session management function (SMF).

[0145] Figure 2 Shows the control Figure 1 An example of a control device 200 for a 5GRAN or 5GC function shown in FIG. The control device may include at least one random access memory (RAM) 211a, at least one read-only memory (ROM) 211b, at least one processor 212, 213, and an input / output interface 214. At least one processor 212, 213 may be coupled to the RAM 211a and the ROM 211b. At least one processor 212, 213 may be configured to execute appropriate software code 215. For example, the software code 215 may allow one or more steps to be performed to implement one or more aspects of the present invention. The software code 215 may be stored in the ROM 211b. The control device 200 may be interconnected with another control device 200 that controls another function of the 5GRAN or 5GC. In some embodiments, each function of the 5GRAN or 5GC includes a control device 200. In an alternative embodiment, two or more functions of the 5GRAN or 5GC may share a control device.

[0146] Figure 3 An example of a terminal 300 is shown, for example Figure 1The terminal 300 is shown. The terminal 300 may be provided by any device capable of sending and receiving radio signals. Non-limiting examples include user equipment, a mobile station (MS) or a mobile device, such as a mobile phone or a so-called "smartphone", a computer equipped with a wireless interface card or other wireless interface facilities (e.g., a USB dongle), a personal data assistant (PDA) or tablet computer with wireless communication capabilities, a machine type communication (MTC) device, an Internet of Things (IoT) type communication device, or any combination of these devices. The terminal 300 may provide data communication for carrying communications, for example, the communication may be one or more of voice, electronic mail (email), text messages, multimedia, data, machine data, etc.

[0147] The terminal 300 may receive signals via the air or radio interface 307 by suitable means for receiving, and may transmit signals by suitable means for transmitting radio signals. Figure 3 In the mobile device, the transceiver device is schematically represented by a block 306. The transceiver device 306 may be provided, for example, by a radio component and an associated antenna arrangement. The antenna arrangement may be provided internally or externally to the mobile device.

[0148] The terminal 300 may be equipped with at least one processor 301, at least one read-only memory ROM 302a, at least one random access memory RAM 302b, and other possible components 303 for performing the tasks it is designed to perform (including controlling access to and communicating with access systems and other communication devices) with the assistance of software and hardware. At least one processor 301 is coupled to RAM 302b and ROM 302a. At least one processor 301 can be configured to execute appropriate software code 308. For example, software code 308 can allow one or more aspects of the present invention to be performed. Software code 308 can be stored in ROM 302a.

[0149] The processor, memory and other related control devices may be arranged on an appropriate circuit board and / or in a chipset. This feature is indicated by reference numeral 304. The device may optionally have a user interface, such as a keyboard 305, a touch-sensitive screen or touchpad, a combination thereof, etc. Depending on the type of device, one or more of a display, a speaker and a microphone may also be optionally provided.

[0150] One of the goals of 3GPP is to enhance mobility through L1 / 2 inter-cell mobility. Unlike the L3 mobility procedure where the handover between two cells is determined by the RRC layer, L1 / 2 inter-cell mobility is configured by the Centralized Unit (CU) and executed by the MAC layer terminating at the Distributed Unit (DU).

[0151] Figure 4An example implementation of the L1 / 2 inter-cell mobility signaling diagram from the serving cell in DU1 to the target cell in DU2 is shown, where each DU is connected to a central unit (CU), which is a so-called cross-DU but within the CU scenario. The diagram also applies to the cell change scenario within the same DU (i.e., DU1 is the same as DU2, i.e., switching from the serving cell in DU1 to the target cell in DU1, meaning that the UE is still in the same DU).

[0152] At step 400, the UE sends a measurement report to DU1. At 402, DU1 forwards the measurement report to the CU. At 404a, the CU sends a UE context setup request to DU1 and sends the same request to DU2 at 404b. In response, at 406a, DU1 sends a UE context setup response to the CU and at 406b, DU2 sends a UE context setup response to the CU.

[0153] At 408, the CU generates an RRC reconfiguration message based on the measurement report for L1 cell change, which contains the configuration of the cell prepared in the target DU (i.e., DU2). At 410, the CU sends the RRC reconfiguration message to the UE, and the UE confirms that the RRC reconfiguration is completed by sending an RRC reconfiguration complete message to the CU at 412.

[0154] Therefore, steps 400 to 412 may represent a preparation phase for L1 / 2 inter-cell mobility, in which the network decides to configure potential target cells (in DU2) for L1 / 2 inter-cell mobility based on measurement reports received from the UE.

[0155] After confirming the RRC reconfiguration to the network at step 412 , at step 414 , the UE starts to periodically report the L1 beam measurement results of the serving cell and the candidate target cells.

[0156] Once it is determined that there is a target candidate cell whose radio link beam measurement result is better than that of the serving cell, for example, L1-RSRP measured by the target beam > L1-RSRP measured by the serving beam + an offset such as trigger time (TTT), at 416, the serving cell (i.e., DU1) sends a signaling control element (e.g., MAC CE) or an L1 message to trigger a cell change to the target candidate cell.

[0157] At step 418, the UE performs handover from the serving cell to the target cell.

[0158] One benefit of L1 inter-cell mobility compared to baseline handover and conditional handover is that the interruption during handover execution and the execution time of handover / cell change can be significantly reduced because the UE does not need to perform higher layer (RRC, PDCP) reconfiguration and in some scenarios the UE does not need to use the random access channel procedure when connecting to the target cell.

[0159] 3GPP TS 38.214 describes the NR beam measurement reporting framework. Specifically, a channel state information (CSI) framework is defined, and the time and frequency resources used by the UE to report CSI are controlled by the gNB. CSI may include channel quality indicator (CQI), precoding matrix indicator (PMI), CSI-RS resource indicator (CRI), SS / PBCH block resource indicator (SSBRI), layer indicator (LI), rank indicator (RI), L1-RSRP or L1-SINR. That is, CSI may include L1 measurements.

[0160] CSI reporting can be controlled using CSI reporting configurations, where each reporting configuration CSI-ReportConfig can be associated with a single downlink BWP for channel measurement (indicated by the higher layer parameter BWP-Id) given in the associated CSI-ResourceConfig. CSI-ReportConfig may include parameters for a CSI reporting band, such as: codebook configuration including codebook subset restrictions, time domain behavior, frequency granularity of CQI and PMI, measurement restriction configuration, and CSI-related quantities to be reported by the UE, such as layer indicator (LI), L1-RSRP, L1-SINR, CRI, and SSBRI (SSB resource indicator).

[0161] The time domain behavior of CSI-ReportConfig may be indicated by the higher layer parameter reportConfigType and may be set to "aperiodic", "semi-persistent on PUCCH", "semi-persistent on PUSCH" or "periodic". For "periodic" and "semi-persistent on PUCCH" / "semi-persistent on PUSCH" CSI reporting, the configured periodicity and slot offset apply in the parameter hierarchy that configures the UL BWP for CSI reports to be transmitted. The higher layer parameter reportQuantity may indicate the CSI-related, L1-RSRP-related or L1-SINR-related quantity to be reported.

[0162] Each CSI-ResourceConfig may include a configuration of a list of S≥1 CSI resource sets (given by the higher layer parameter csi-RS-ResourceSetList), the list consisting of references to (multiple) NZP CSI-RS resource sets and / or (multiple) SS / PBCH block sets, or the list consisting of references to (multiple) CSI-IM resource sets. Each CSI resource setting may be located in a DL BWP identified by the higher layer parameter BWP-id, and all CSI resource settings linked to a CSI report setting have the same DL BWP.

[0163] In the 3GPP study project “Study on Artificial Intelligence (AI) / Machine Learning (ML) over NR Air Interface”, several goals were achieved.

[0164] These goals include focusing on use cases for CSI feedback enhancement, such as reducing overhead, improving accuracy, prediction, and beam management, such as beam prediction in the time and / or spatial domain to reduce overhead and latency, and improve beam selection accuracy.

[0165] Also included are objectives related to physical layer aspects, which involve providing specific use cases as well as regulatory impact at the UE and gNB collaboration level, such as new signalling (e.g. signalling of AI / ML UE capabilities), signalling aspects for training and validation data assistance, assistance information, measurements and feedback.

[0166] Another objective is related to AI / ML models, terminology and descriptions, aiming to identify common and specific features of the framework study to identify different levels of cooperation between UE and gNB related to the selected use cases, such as different UE / gNB cooperation levels for separate or joint ML operations.

[0167] Mobility management is a scheme used in wireless communications to ensure service continuity for UEs during mobile operations by minimizing dropped calls, RLF, unnecessary handovers, and ping-pong effects. In addition, for applications with strict QoS requirements (such as reliability, latency, etc.), QoE may be sensitive to handover performance, so mobility management should avoid handover failures and reduce delays during handovers.

[0168] However, for some traditional methods, it is difficult to achieve near-zero failure switching based on trial-and-error solutions.

[0169] A RAN study on NR and EN-DC data collection enhancements was initiated in RAN3#110e with the overall goal of studying the high-level principles and functional framework for enabling AI in the RAN, including AI functions and the required inputs and outputs of ML algorithms.

[0170] One use case defined in TR 37.817 Section 5.3 "Mobility Optimization" aims to improve the mobility performance of UEs by using AI / ML solutions. For example, an ML model located in the network can receive radio measurements from UEs, predicted resource states of neighboring RAN nodes, and UE trajectory predictions as inputs. These inputs can be used to predict the handover target node in advance, thereby greatly improving the chances of a successful handover for the UE.

[0171] UE trajectory prediction is discussed in 3GPP RAN3. These predictions may include the UE's latitude, longitude, altitude, cell ID and beam ID for a period of time in the future.

[0172] As mentioned above, inter-cell mTRP transmission and L1-centric inter-cell mobility may rely on L1 beam measurement reports (ie, CSI measurement reports) of serving cells and non-serving cells.

[0173] The UE may report the serving cell and non-serving cell as part of the periodic L1 beam measurement report. The reporting period can be as short as 5ms, allowing the network to collect measurement data from the UE more frequently compared to the minimum reporting period (120ms) for cell / beam quality reporting for L3 mobility.

[0174] However, L1 beam reporting increases the signaling overhead of the system and the power consumption of the UE.

[0175] The signaling overhead caused by L1 beam reporting on the air interface can be alleviated with the help of trajectory prediction.

[0176] In the present disclosure, UE trajectory refers to the specific path / sequence of the UE through the cells and beams of the wireless network, which can be characterized by the physical cell identifier (PCI) and the synchronization signal block (SSB) beam index, and can be further refined in the network by CSI-RS based signals.

[0177] Predicting the future trajectory of the UE helps reduce the number of target cells for which measurements need to be reported.

[0178] However, while it is possible to generate statistical predictions about UE trajectories based on data collected over time (e.g., several beam IDs), making accurate trajectory predictions for a single UE may be difficult.

[0179] Figure 5 An example of the problem of a mobile terminal moving on a street with an intersection is shown. Figure 5 The beam radiation pattern of each cell and the position of the UE for different time steps are shown in FIG.

[0180] The problem is that an incorrect trajectory prediction at one moment will reduce the accuracy of future predictions, because the predictions are processed in a chained manner, and the current prediction results can assist the next beam prediction.

[0181] For example, from Figure 5 , the network may predict that in 90% of cases, multiple UEs with similar previous trajectories on the street will access cell 2 after passing cell 1 (i.e., the UE goes straight along the road), and in 10% of cases, they will access cell 3 (i.e., the UE turns right at the intersection). Then, the network's prediction of the best future beam index at time steps [t, t+1, t+2, t+3, t+4, t+5] becomes [1, 2, 3, 4, 5, 5] respectively, which corresponds to the predicted best beam for the UE's straight path along the road.

[0182] However, if Figure 5 As shown in Figure 1, when the UE turns right, the optimal beam indices at different time steps should be [1,2,3,6,7,8]. Since the prediction at time step t+3 is wrong and the ML model uses the predictions at time t, t+1, t+2, and t+3 to perform the prediction at time t+4, it is likely that the prediction at time t+4 will also be wrong. Therefore, the accuracy of future predictions will be reduced due to the reliance on the wrong prediction for future predictions.

[0183] In summary, beam measurements based on inter-cell low-level mobility may increase signaling overhead on the air interface and may need to be optimized (reduced) without affecting the beam selection error rate. Although machine learning methods based on UE trajectory prediction can be used to reduce signaling overhead, incorrect predictions may lead to mobility failures (for example, if the network tells the UE to switch to beam 5 of cell 2 at time t+4, but the UE trajectory has changed, such as Figure 5 As shown, the signal quality of beam 7 in cell 3 is better, which will lead to switching failure) and increase the service interruption time, and also require additional signaling to handle the recovery of the connection.

[0184] refer to Figure 6 , methods according to some examples are shown.

[0185] At 600, the method includes receiving, from a user equipment, measurement data related to one or more cells and / or one or more beams of a network.

[0186] At 602, the method includes determining a prediction based on measurement data, the prediction including at least one sequence of one or more beams predicted to have a highest signal quality for a user equipment at respective one or more time instances.

[0187] At 604, the method includes sending the prediction to the user equipment.

[0188] At 606, the method includes sending measurement data related to one or more cells and / or one or more beams of the network to a network node.

[0189] At 608, the method includes receiving a prediction from a network node, the prediction including at least one sequence of one or more beams predicted to have a highest signal quality for the device at respective one or more time instances, wherein the prediction is based at least in part on the measurement data.

[0190] Some examples may provide a solution to predict the correct sequence of future beams without compromising beam selection error rate and mobility failure rate, while reducing signaling overhead. Some examples utilize predictive inter-cell beam management methods to reduce CSI measurement overhead (L1 beam reporting) and avoid mobility failures by ensuring that the network is aware of changes in UE trajectory within a certain confidence interval (e.g., mean square error threshold (depending on scenario, deployment, and configuration)).

[0191] In some examples, the network (e.g., DU for L1 mobility) may predict the future N best beam indices / L1-RSRP values ​​and share the prediction with the UE.

[0192] The prediction output may include the PCI and SSB-RS / CSI-RS index of the cell. The network may share new predictions with the UE based on past measurement reports (e.g., CSI measurement reports) received from the UE, and potentially based on other UEs identified / selected as being along the same UE trajectory or similar geographic movement path (e.g., based on statistical distribution of historical data).

[0193] In some examples, the UE may evaluate the prediction received from the network and trigger a CSI measurement report for L1 beam measurement if the prediction for the current time step does not match the UE's current measurement. In some examples, if the prediction for the current time step matches the UE's current measurement, no measurement report is sent.

[0194] Once the network receives the report from the UE, it can revise its previous estimate of future beams based on the UE’s instructions. The network can then make inferences based on the measurement data from the UE to generate new predictions.

[0195] If the network's prediction is correct, the UE's beam change can be managed by the network through signaling CE beam switch commands for intra-cell beam management and signaling CE triggered switch messages for L1 inter-cell mobility. As long as the network's prediction of the best beam index / L1-RSRP measurement is correct, the UE can perform a beam change.

[0196] The data collected by the network based on incorrect predictions can be used to facilitate model updates / retraining. Thus, the network can store updated measurement data for subsequent model training. In some examples, the network can initiate model updates / retraining, for example, if the network frequently receives incorrect predictions, an update / retraining can be triggered.

[0197] In some examples, the network can share with the UE not only the best predicted trajectory in the beam domain, but also other possible trajectories of the UE (e.g., the most likely trajectory determined based on other UEs and historical data). Figure 5 In this case, the network can share the predicted trajectory options [1,2,3,4,5,5] and [1,2,3,6,7,8] with the UE. In this case, the UE can indicate to the network which predicted trajectory option it follows based on the actual measurements.

[0198] refer to Figure 7 , showing signaling interactions according to some examples.

[0199] At 700 , once an RRC measurement event is triggered, the UE sends a measurement report including measurement results (eg, L3 measurement results) to the serving cell under DU1 .

[0200] At 702, DU1 forwards the measurement result to CU.

[0201] At 704, based on the measurement results, the CU may configure the cells under DU1 and DU2. For example, the CU may send a UE context setup request to DU1 at 704a and send the request to DU2 at 704b. The CU does this to activate inter-cell beam management / L1 inter-cell mobility.

[0202] In some examples, the UE context setup request may include a request to learn the machine learning capabilities of the DU and whether it is capable of performing predictive beam management / L1 mobility.

[0203] At 706a, DU1 sends a UE context setup response to the CU; at 706b, DU2 sends a UE context setup response to the CU. Optionally, if there is a request at 704a and 704b respectively, the UE context setup response may include information indicating predictive beam management / L1 mobility capabilities.

[0204] In some examples, this information may include information about the input data required for the model, the output type, and the prediction window (e.g., how many beams can be predicted ahead).

[0205] Based on the received predictive beam management / L1 mobility capabilities of the cell, the CU generates an indication that RRC reconfiguration and predictive mobility are enabled at 708. This indication may be performed under the CSI-ReportConfig information element.

[0206] The CU sends an RRC reconfiguration message to the UE along with an indication that predictive mobility is enabled at 710. The network may also include other additional information, such as information about synchronization points (or events) at which the UE must send L1 measurement reports to ensure that it remains synchronized / aligned with the network with respect to the predicted and measured beams.

[0207] At 712, the UE sends an RRC reconfiguration complete message to the CU, indicating that the UE has received the RRC reconfiguration.

[0208] At 714, the UE starts reporting N L1-RSRP measurements of serving cells and / or non-serving cells as indicated in the RRC reconfiguration. The measurement report may include, for example, beam measurements of cells configured for inter-cell beam management / L1 inter-cell mobility. The network may use it (as a seed) for initial prediction.

[0209] At 716, the serving cell under DU1 is predicted based on the data collected from the UE. The prediction can be used to identify the cell or beam that the UE is expected to pass through (ie, the predicted UE trajectory passes through the identified beam).

[0210] For example, at time step t, the network predicts the next 10 best beams with a granularity of 10 milliseconds. The prediction granularity can also be different for each step, representing the predicted dwell time for each beam index (for example, because the UE may not move uniformly across the predicted beams).

[0211] The prediction output may include the PCI of the cell as well as the SSB-RS / CSI-RS index and / or the RSRP value of each SSB-RS / CSI-RS. More details of the ML model output are provided below.

[0212] At 718, the serving cell under DU1 sends the prediction output, ie, the prediction of the next best beam index / RSRP value, to the UE.

[0213] As part of the prediction, the network can also send a confidence interval for the prediction. The UE can use this to determine whether the network's prediction is within a tolerable range based on actual measurements.

[0214] For example, if the confidence interval of the RSRP value prediction at time step t+2 is 1 dB, and the predicted RSRP value of a certain beam index at t+2 differs by less than 1 dB from the RSRP value measured at t+2 for the same beam index, the UE may not trigger a report. In addition, the UE switches to those beams predicted (or recommended) by the network.

[0215] The following steps may depend on the accuracy of the UE's prediction based on the comparison of UE measurements with the network prediction. Three different cases are considered:

[0216] Case 1: Wrong prediction

[0217] Case 2: Correct prediction and network controlled mobility

[0218] Case 3: Correct prediction and UE controls mobility

[0219] refer to Figure 8 , shows an example signaling interaction for Case 1 above, i.e., a case where the prediction is incorrect. Figure 8 The signaling interaction can be described in the above Figure 7 After the relevant step 718.

[0220] The UE performs one or more beam measurements at 800. As long as the network prediction matches the actual measurement, the UE may not report beam measurement results for the configured cells.

[0221] The UE may continue to make measurements on both serving and non-serving cells, but may exclude L1 beam measurements as long as the actual measurements match the network predictions.

[0222] If the network predicts the best future beam index and the predicted best beam index at time t+X is not equal to the measured best beam index at time t+X, an L1 report is triggered.

[0223] For example, refer to Figure 5 For example, the network may predict the best beam prediction index to be [1,2,3,4,5,5]. The UE may trigger a report at time t+3 because the measured best beam belongs to cell 3 and beam 6 (rather than cell 2 and beam 4) to indicate that the predicted beam is wrong and include the actual measurement value in the report.

[0224] In some examples, if the network predicts the best beam measurement value (e.g., L1-RSRP) and the measured beam measurement value is not within the confidence interval predicted by the network (signaled by the network when providing measurement configuration parameters), an L1 report is triggered.

[0225] The network can share, for example, a 95% confidence interval along with the prediction. The confidence interval for each beam prediction can be different, depending on the data statistics the network has collected for that particular beam.

[0226] In some examples, if the UE does not receive any predictions about KPIs (e.g., CQI, PMI, RI, etc.) from the network, it may still send a CSI report including these KPIs.

[0227] At 802 , the UE determines, based on the previously received prediction and the measurement results obtained at 800 , that there is a mismatch between the network prediction and the actual measurement.

[0228] The UE sends an indication to the network that the prediction is incorrect at 804. The indication may include actual measurements performed by the UE indicating that the network's prediction at, for example, time step t+T is wrong.

[0229] In some examples, the indication is sent when the mismatch between the network prediction and the actual measurement is greater than a confidence interval for the prediction.

[0230] In some examples, where the network performs signal quality value prediction, the UE can share the mismatch that occurs between [tN, t] (e.g., t_err). This can indicate to the network how much it needs to adjust the prediction to obtain more accurate measurement data. In some examples, if t_err is within the predicted confidence interval, the UE can ignore the difference - that is, the UE does not send the indication at 804.

[0231] In some examples, where the network performs signal quality value prediction, the UE can keep a history of each beam measurement with a length of N, such as the past 10 training frames. If the signal quality prediction for this beam is wrong, the UE can share the historical beam measurement values ​​between [tN, t], where t is the current time step.

[0232] At 806, the network can perform a model update / retraining based on the indication received from the UE at 804. That is, in response to the incorrect prediction feedback, the network can take actions to improve its ML model. In some examples, the network can store the indication, optionally including the measurement data, for later use in model training.

[0233] In some examples, if the network frequently receives incorrect predictions in succession, such as receiving more than a certain number of indications, or receiving more than a certain number of indications within a certain period of time, the network can initiate a model update (e.g., switching to another model) / retraining.

[0234] At 808, the network shares the new prediction with the UE based on the received measurement results.

[0235] In some examples, if the UE measurement results indicate that beam switching is required, the DU may also indicate beam switching in the new prediction. For example, the network may predict that the UE observes [1,2,2,2,2] as the best beam index, but if the UE's actual measurement results indicate that the best beam sequence observed by the UE is [1,2,2,3,3], the network may send a new prediction to the UE including switching to beam 3.

[0236] refer to Fig. 9 , shows an example signaling interaction under the above situation 2 or 3, that is, the situation where the prediction is correct. Fig. 9 The signaling interaction can be described in the above Figure 7 After the relevant step 718.

[0237] Steps 900 to 908 correspond to the previously described case 2, where the network's prediction is correct and the mobility is controlled by the network. It should be understood that in some examples where steps 900 to 908 are performed, steps 910 to 912 may not be performed.

[0238] At 900, the UE performs one or more beam measurements and, based on these measurements, determines at 902 that the prediction is correct. For example, the network may predict that at a specific time, such as time t+40ms (t is the predicted time step), the UE will observe a specific beam, such as SSB-RS index 5 of PCI 2, as the strongest beam measurement value. Based on the prediction, the UE performs beam measurements at a specific time, such as time step t+40ms, and measures that the specific beam, such as SSB-RS index 5 of PCI 2, is the strongest beam index, thereby verifying the prediction.

[0239] Optionally, the UE may send an indication of the prediction status at 904. For example, the UE may send an indication of whether the prediction is correct.

[0240] In some examples, a 1-bit indication may be used to indicate that the network prediction is accurate. This may allow the network to trigger a beam switch based on the prediction.

[0241] In some examples, if the network's prediction is still within the confidence interval, the UE may send information about the mismatch between the network prediction and the UE measurement.

[0242] If multiple trajectories are shared with the UE, the UE may also include in the indication the trajectory information that it followed.

[0243] At 906, the network determines that the prediction is correct.

[0244] For example, the network may receive an indication at 904 indicating that the prediction is correct (e.g., a 1-bit indication, an indication of length k bits, where k ≥ 1, an indication providing at least two different values ​​to refer to a correct or incorrect prediction, an indication providing a probability p that the UE is still on the predicted path, where the probability p ≥ thres1 indicates that the predicted path is correct and thres2 indicates that the predicted path is incorrect). Alternatively, if the UE is configured to report L1-RSRP beam measurements only when the prediction is incorrect, as long as the UE does not indicate a prediction error at time step t+40ms, the network can assume that the prediction at this time step t+40ms is correct.

[0245] At 908, the serving cell sends a signaling control element (eg, MAC CE) command to the UE to trigger beam / cell switching.

[0246] In some examples, a signaling CE command is sent when the network determines that the prediction at time step t is not equal to the prediction at time step t+1 (t!=t+1), that is, the network determines that there is a beam switching event.

[0247] At 910, the UE notifies of the received beam switch.

[0248] In some examples, if the measurement at time t+X does not match the prediction at time t+X, the UE may reject the beam switch, and the UE may inform the network of the actual measurement value of the current beam.

[0249] Step 912 corresponds to the case 3 described above, where the network's prediction is correct and the mobility is controlled by the UE. It should be understood that in some examples of performing step 912, steps 900 to 910 may not be performed.

[0250] At 912, the UE determines a beam switch / change based on a prediction received from the network. That is, the UE may determine a beam switch without receiving a signaling CE message sent by the network to perform. In some examples, the UE may determine to perform a beam switch in response to determining that an actual beam measurement matches a network prediction.

[0251] At 914, in response to the signaling CE command received at 908, or in response to the UE determining at 912 that beam switching needs to be performed, the UE performs a beam switching operation.

[0252] For example, the UE may determine whether the predicted beam at time step t+1 belongs to the beam of the pre-prepared / target cell. When the UE determines that the predicted beam belongs to the corresponding beam, it performs a random access procedure to access the target beam of the pre-prepared / target cell, thereby completing the beam switching.

[0253] At 916, the UE indicates to the network that the beam switching operation has been successfully performed. The UE may provide this indication using L1 / L2 ACK for the new beam uplink after successfully decoding the PDCCH on the previous serving beam.

[0254] As mentioned above, in some examples, the network may also share different trajectories with the UE. For example, the network may share trajectories in a sequence of beam IDs, so two trajectories may be indicated by two sequences of beam IDs (e.g., the first trajectory is [1,2,3,4,5,5] and the second trajectory is [1,2,3,6,7,8]).

[0255] In some examples, the beams and cells involved in the trajectory are pre-prepared by the network. That is, the network expects the UE to connect to the beam specified by the trajectory.

[0256] In some examples, as long as the UE follows a trajectory in the trajectory list, it does not need to send reports to the network about its serving beam or neighboring beam measurements. However, the UE can also provide channel state information, such as beam precoding information, to the network on demand.

[0257] In some examples, the UE may also switch to a new beam according to the trajectory in the manner described above. For example, the network may provide a signaling CE at a predicted time, or the UE may indicate that the switching time has arrived through an indicator.

[0258] If the UE wishes to select a cell / beam that is not in the trace list, it can start sending complete neighbor beam / cell measurement reports.

[0259] In some examples, the network can order the trace indications in its trace list. This ordering indicates that the first trace in the list has cells ready, but the cells of other traces are not yet ready. When the UE decides not to follow the first trace, it indicates to the network the corresponding index of the trace in the list that it will follow.

[0260] For example, given a first trajectory [1,2,3,4,5,5] and a second trajectory [1,2,3,6,7,8], the first three beams [1,2,3] are the same for both. Once the UE realizes that it is going to 6 instead of 4 after beam / cell 3, it indicates to the network "select trajectory 2". In response to receiving this indication, the network can prepare the cells / beams on the second trajectory for the UE and send an indication to the UE that these cells / beams are ready. After the UE receives the confirmation from the network, it can assume that the cells in path 2 are ready.

[0261] In some examples, no ML model is needed. The network can obtain a list of trajectories by collecting the past serving cell / beam sequences of the UE and sorting these trajectories by frequency of occurrence (e.g., which beams the UE usually follows after connecting to beam A).

[0262] In some examples, inter-cell handover can also be organized by pre-configured RACH grants. In this way, the network does not need to send signaling CE to trigger the UE's change, but the UE initiates the HO (e.g., case 3).

[0263] As explained above, in some examples, an ML framework can be used to predict UE trajectories, identifying a series of cells or beams that the UE is expected to traverse. That is, an ML model in the network can be used to predict the best beam index and cell ID for a given UE connected to a cell.

[0264] For example, the model can generate the predicted optimal SSB-RS index and physical cell ID (PCI) for the UE at time step t based on past measurement data. In some examples, a long short-term memory (LSTM) network can be applied to predict the future from sequences of varying lengths. Since LSTM networks are good at learning long-term relationships in data, they are more suitable for time series prediction, and they can also solve the gradient vanishing problem in RNNs by introducing special gates (e.g., forget gates, input gates).

[0265] Fig.10 An example of a neural network architecture for a model is shown.

[0266] Input Matrix In the past time T, there are D features, namely (eg, input type), where t is the predicted time and i is the UE ID (eg, C-RNTI).

[0267] The size of the input matrix depends on the granularity of the input. For example, if the time span T is 1 second and the granularity is 100ms, 10 stacked input features D will be generated, where D is also the number of features.

[0268] exist Fig.10 In the example, there are two LSTM layers followed by a softmax output layer. It should be understood that in other examples, different configurations can be used. Each LSTM unit receives an input x t , generating unit state And output a hidden state value This value can be used in other LSTM cells.

[0269] The input / forget / output gates of the LSTM unit can use the sigmoid activation function σ. The output value of the sigmoid function ranges between 0 and 1. The tanh activation function is used for the unit state c t and the hidden state h t , the output value range is between -1 and 1. The output layer can use softmax or sigmoid activation function to calculate the probability of each label, such as the probability that a beam / cell becomes the best beam / cell at time t+T.

[0270] The input data for the model may include the following information:

[0271] UE measurement reports, such as RSRP, RSRQ or SINR measurements;

[0272] The correct prediction at the previous time step;

[0273] UE velocity information (e.g., calculated based on observed Doppler shift);

[0274] UE location information (optional, if available);

[0275] Current / predicted resources of the serving cell and neighboring cells (optional, if available).

[0276] UE measurement reports may help the network understand the channel attenuation over time and can be used for initial prediction. As long as the model prediction (verified by the UE) is correct, the network may not need the use / transmission / consideration / evaluation of UE measurement reports.

[0277] The correct prediction may include the best beam index and PCI. As long as the model prediction is verified to be correct based on the information provided by the UE, the network can use the early prediction as a factual basis to predict future situations.

[0278] The UE location information may include the coordinates and speed of the UE.

[0279] The current / predicted resources of the serving cell and neighboring cells can be used by the model to understand the traffic load information of different beams / cells, which affects the inter-cell / beam interference and the best beam index / RSRP observed by the UE.

[0280] The model is able to perform multi-step forecasts. That is, the model can use the prediction results of the previous step as input for the subsequent predictions.

[0281] Based on the above input data, the model can perform at least the following types of predictions:

[0282] Prediction of the best beam index and PCI for a given UE;

[0283] Prediction of RSRP value of reference signal index configured in the cell.

[0284] To predict the optimal beam index and PCI, the ML model generates an output vector for the prediction at time t+T in is the best SSB-RS index, is the physical cell ID of UE i at time t+T. It should be noted that and For example, it is the beam of this PCI. In the case of multi-step prediction, the output may be a matrix Where g is the granularity of the prediction (e.g., 100 ms), and Ng = T. So, if T is 1000 ms and g is 100 ms, the ML model may generate 10 prediction results showing the best beam / cell combination for UE i.

[0285] To predict the RSRP value, the ML model can generate the output vector in It is the prediction of the RSRP value of the SSB-RS beam index with cell ID c at t+T. is the estimated RSRP measurement of SSB-RS index 0 of cell c observed by UE i at time step t+T. In the case of multi-step prediction, the output may be a matrix Where g is the prediction granularity, for example 100ms, and Ng=T.

[0286] For model training, the above input data and UE feedback are recorded over time and collected in the network entity that performs the training. Since beam management (including intra-cell and inter-cell beam management) can be handled by gNB or gNB distributed unit, inference can also be done in gNB or gNB-DU.

[0287] As an example use case, suppose Figure 5 In the scenario, at time step t, cell 1 predicts the best beam index for the UE to be {t:1,t+1:2,t+2:3,t+3:4,t+4:5,t+5:5}.

[0288] The UE moves along the street, continuously measuring the serving cell and the non-serving cell, and does not initiate a report as long as the prediction is correct.

[0289] At time step t+3, the UE turns right and measures that the signal of beam id 6 in cell 3 is the strongest. The UE compares the actual measurement value with the network's predicted value at t+3 (beam id 4).

[0290] Since the network prediction is wrong, the UE informs the network of the wrong prediction along with other beam measurements at t+3. The network updates the prediction based on the correct measurement received at t+3 and the UE's trajectory, for example, to {t:1, t+1:2, t+2:3, t+3:6}, and then predicts that the beam index at t+4 is 7 and t+5 is 8, and informs the UE of the new prediction.

[0291] Therefore, in some examples, an adaptive inter-cell beam management process is provided to reduce the CSI measurement overhead. The network is able to collect feedback on its predictions (getting the real situation from the UE) and retrain the model. The use of a network-driven ML approach means that there is no need to implement ML capabilities on the UE side. For multi-step predictions, the network only uses the past correct predictions as input for the next round of predictions. With the method described above, the network can avoid consecutive wrong predictions, thereby preventing mobility failures and long outages.

[0292] In some examples, an apparatus is provided that includes components for: receiving measurement data related to one or more cells and / or one or more beams of a network from a user device; determining a prediction based on the measurement data, the prediction including at least one sequence of one or more beams, the one or more beams being predicted to have the highest signal quality for the user device at corresponding one or more time moments; and sending the prediction to the user device.

[0293] In some examples, the apparatus may include at least one processor and at least one memory storing computer program code, the memory and the computer program code being configured to, together with the at least one processor, cause the apparatus to at least: receive measurement data relating to one or more cells and / or one or more beams of a network from a user device; determine a prediction based on the measurement data, the prediction comprising at least one sequence of one or more beams, the one or more beams being predicted to have the highest signal quality for the user device at corresponding one or more time instants; and send the prediction to the user device.

[0294] In some examples, a device is also provided that includes components for: sending measurement data related to one or more cells and / or one or more beams of the network to a network node; and receiving a prediction from the network node, the prediction including at least one sequence of one or more beams, the one or more beams being predicted to have the highest signal quality for the device at corresponding one or more time moments, wherein the prediction is based at least in part on the measurement data.

[0295] In some examples, the device may include at least one processor and at least one memory storing computer program code, the memory and the computer program code configured to, together with the at least one processor, cause the device to at least: send measurement data related to one or more cells and / or one or more beams of the network to a network node; and receive a prediction from the network node, the prediction including at least one sequence of one or more beams, the one or more beams being predicted to have the highest signal quality for the device at corresponding one or more time instants, wherein the prediction is based at least in part on the measurement data.

[0296] It should be understood that these devices may include or be coupled to other units or modules, such as radio components or radio heads for transmission and / or reception. Although these devices are described as one entity, different modules and memories may be implemented in one or more physical or logical entities.

[0297] It should be noted that although some embodiments are described in conjunction with 5G networks, similar principles can also be applied to other networks and communication systems. Therefore, although certain embodiments are described above by way of example and with reference to specific example architectures of wireless networks, technologies and standards, these embodiments can also be applied to any other suitable communication system forms not shown or described herein.

[0298] It is also noted that while the foregoing describes exemplary embodiments, many variations and modifications may be made to the disclosed solution without departing from the scope of the present invention.

[0299] In general, various embodiments may be implemented by hardware, dedicated circuits, software, logic, or any combination thereof. Some aspects of the disclosure may be implemented by hardware, while other aspects may be implemented by firmware or software executed by a controller, microprocessor, or other computing device, but the disclosure is not limited thereto. Although various aspects of the disclosure may be shown and described in block diagrams, flow charts, or other graphical representations, it should be understood that the blocks, devices, systems, techniques, or methods described herein may be implemented by (as non-limiting examples) hardware, software, firmware, dedicated circuits or logic, general-purpose hardware or controllers or other computing devices, or some combination thereof.

[0300] In this application, the term "circuitry" may refer to one or more, or all, of the following:

[0301] (a) implemented only in hardware circuits (e.g., implemented using only analog and / or digital circuits);

[0302] (b) a combination of hardware circuitry and software, such as (where applicable):

[0303] (i) a combination of analog and / or digital hardware circuits and software / firmware;

[0304] (ii) any portion of a hardware processor working in conjunction with software (including a digital signal processor), software and memory to enable a device such as a mobile phone or server to perform various functions;

[0305] (c) Hardware circuits and / or processors (such as a microprocessor or a portion of a microprocessor) that require software (such as firmware) for operation but may not be present when the software is not required for operation.

[0306] This definition of "circuitry" applies to all uses of the term in this application, including in any claims. As another example, in this application, "circuitry" also covers only a hardware circuit or a processor (or multiple processors), or a portion of a hardware circuit or processor and its (or their) supporting software and / or firmware. For example, if relevant to a particular claim element, "circuitry" also covers a baseband integrated circuit or processor integrated circuit of a mobile device, or a similar integrated circuit in a server, cellular network device, or other computing or networking equipment.

[0307] The embodiments of the present disclosure may be implemented by computer software executable by a data processor of a mobile device, for example, in a processor entity, or by hardware, or by a combination of software and hardware. Computer software or programs, also referred to as program products, include software routines, applets and / or macros, which may be stored in a data storage medium readable by any device and which contain program instructions for performing specific tasks. A computer program product may include one or more computer executable components that are configured to perform embodiments when the program is run. One or more computer executable components may be at least one piece of software code or a portion thereof.

[0308] In this regard, it should also be noted that the logic flow blocks in the Figures may represent program steps, interconnected logic circuits, blocks and functions, or a combination of program steps and logic circuits, blocks and functions. The software may be stored on physical media such as memory chips, storage blocks implemented within the processor, magnetic media such as hard disk or floppy disks, and optical media such as DVDs and their data variants, CDs, etc. These physical media are non-transitory media.

[0309] The memory may be of any type suitable for the local technical environment and may be implemented using any suitable data storage technology, such as semiconductor-based storage devices, magnetic storage devices and systems, optical storage devices and systems, fixed memory, and removable memory. The data processor may be of any type suitable for the local technical environment and may include, as non-limiting examples, general-purpose computers, special-purpose computers, microprocessors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), FPGAs, gate-level circuits, and processors based on multi-core processor architectures.

[0310] Embodiments of the present disclosure may be practiced in various components such as integrated circuit modules. Integrated circuit design is generally a highly automated process. Complex and powerful software tools are available for converting a logic level design into a semiconductor circuit design that can be etched on a semiconductor substrate.

[0311] The scope of protection sought by each embodiment of the present disclosure is defined by the independent claims. The embodiments and features in this specification that do not fall within the scope of the independent claims should be interpreted as examples that help understand each embodiment of the present disclosure.

[0312] The foregoing has been a comprehensive and detailed description of exemplary embodiments of the present disclosure by way of non-limiting examples. However, after reading the foregoing in conjunction with the accompanying drawings and the appended claims, a person skilled in the relevant art may discover various modifications and adjustments. However, all such and similar modifications to the teachings of the present disclosure will still fall within the scope of the present invention as defined by the appended claims. In fact, there are additional embodiments, including combinations of one or more embodiments with other embodiments previously discussed.

Claims

1. A device comprising components for: receiving, from a user equipment, measurement data relating to one or more cells and / or one or more beams of a network; determining, based on the measurement data, a prediction comprising at least one sequence of one or more beams predicted to have a highest signal quality for the user equipment at respective one or more time instances; and The prediction is sent to the user equipment.

2. An apparatus according to claim 1, wherein the prediction comprises at least a first sequence of one or more beams, the one or more beams being the sequence of one or more beams predicted to have the highest signal quality for the user equipment at the corresponding one or more moments, and a second sequence of one or more beams at the corresponding one or more moments, wherein at least one of the one or more beams is different between the first sequence and the second sequence.

3. A device according to any preceding claim, wherein the component is used to: An indication is received from the user equipment, the indication indicating that the prediction for at least one of the one or more time instances is inaccurate.

4. The device according to claim 4, wherein the component is used to: Based on the instructions, model retraining is performed.

5. The device according to claim 3 or 4, wherein the component is used for: determining a new forecast based at least in part on the indication; and The new prediction is sent to the user equipment.

6. The device according to any one of claims 1 to 2, wherein the component is used for: Determine whether the prediction is accurate.

7. The apparatus of claim 6, wherein determining that the prediction is accurate comprises at least one of: receiving an indication from the user equipment that the prediction is accurate; or After a certain period of time, no indication is received from the user equipment that the prediction is inaccurate.

8. The apparatus of claim 7, wherein the indication that the prediction is accurate further comprises information about a sequence of one or more beams that the user equipment has used.

9. The device according to any one of claims 6 to 8, wherein the component is used to: In response to determining that the prediction is accurate and based on the prediction, a signaling control element command is sent to the user equipment to trigger a beam or cell switch from a first beam in time interval t to a second beam in time interval t+1 for the user equipment.

10. An apparatus comprising components for: sending measurement data relating to one or more cells and / or one or more beams of the network to a network node; and A prediction is received from the network node, the prediction comprising at least one sequence of one or more beams predicted to have the highest signal quality for the device at respective one or more time instances, wherein the prediction is based at least in part on the measurement data.

11. An apparatus according to claim 10, wherein the at least one sequence comprises at least a first sequence of one or more beams, the one or more beams being the sequence of one or more beams predicted to have the highest signal quality for the user equipment at the corresponding one or more time moments, and a second sequence of one or more beams, wherein at least one of the one or more beams is different between the first sequence and the second sequence.

12. The device according to any one of claims 10 to 11, wherein the component is used to: acquiring further measurement data for at least one of the one or more beams for at least one of the corresponding one or more time instances; and Based on a comparison between the further measurement data and the prediction, it is determined whether the prediction is accurate.

13. The device according to claim 12, wherein the component is used to: In response to determining that the prediction for at least one of the one or more time instances is inaccurate, sending an indication to the network node, the indication indicating that the prediction for at least one of the one or more time instances is inaccurate.

14. The apparatus of claim 12, wherein the component is configured to: in response to determining that the prediction is accurate, sending an indication to the network node that the prediction is accurate; or Refraining from sending an indication to the network node that the prediction is inaccurate, wherein the network node is configured to assume that the prediction is accurate if the apparatus does not send an indication to the network node that the prediction is inaccurate within a certain time.

15. The device according to claim 14, wherein the component is used to: receiving a signaling control element command from the network node to trigger a beam switch from a first beam at time interval t to a second beam at time interval t+1; and Based on the signaling control element command, the beam switching is performed.

16. The device according to any one of claims 14 to 15, wherein the component is used to: If the prediction is accurate, beam switching from a first beam at time interval t to a second beam at time interval t+1 is performed based on the prediction.

Images