Time domain overlap prediction with UE model monitoring enhancement

Abstract

In accordance with an example embodiment of the present invention, at least one method and an apparatus are provided to predict a beam prediction output using a prediction sequence of beam measurement outputs for at least one observation window associated with at least one prediction window providing reference signal received power measurements, the beam prediction output comprises a Top-1 or Top-K beam identifier or predicted reference signal received power of a beam subset associated with the beam prediction output; identifying that a prediction window of the at least one prediction window overlaps at least a portion of an observation window of the at least one observation window over a period of time; the beam prediction output during the overlapping portion is compared with a measurement of the observation window during the overlapping portion to perform long-term real-time monitoring to determine at least one prediction output for use in the next moment in time, where the determination is based on a prediction reference signal received power of the at least one prediction output being above a threshold.

Description

Technical Field

[0001] The teachings according to exemplary embodiments of the present invention generally relate to a new machine learning dedicated bearer for machine learning or artificial intelligence related data exchange to generate a new user equipment side model with prediction quality indication, and more specifically, to a new machine learning dedicated bearer for machine learning or artificial intelligence related data exchange to generate a new user equipment side model with prediction quality indication using overlapping sliding window prediction. Background Art

[0002] This section is intended to provide a background or context to the inventions described in the claims. The description herein may include concepts that could be employed, but not necessarily concepts that have been previously conceived or employed. Therefore, unless otherwise indicated herein, the content described in this section is not prior art with the description and claims of this application and is not admitted to be prior art by inclusion in this section.

[0003] Certain abbreviations that may be found in the description and / or drawings are defined as follows:

[0004] 5G - Fifth Generation

[0005] AI - Artificial Intelligence

[0006] CSI - Channel State Information

[0007] DL - Downlink

[0008] gNB-5G / NR base station

[0009] LSTM - Long Short-Term Memory

[0010] ML - Machine Learning

[0011] NR-New Radio

[0012] NN - Neural Network

[0013] RAN - Radio Access Network

[0014] RS-reference signal

[0015] Rx - Receiver

[0016] SSB - Synchronization Signal Block

[0017] TRP - Launch Point

[0018] Tx-Transmitter

[0019] UE - User Equipment

[0020] In many transmission systems, beamforming transmissions can be identified and employed to provide higher beamforming gain and improved coverage. One process used by a base station or user equipment for beamforming is beam scanning. In beam scanning, the beamforming direction is obtained by measuring the signal in different beam directions. The beam scanning process includes processes such as performing signal quality measurements, performing beam selection, and sending quality indicators and reference signals in different beam directions.

[0021] The exemplary embodiments of the present invention provide improved operating methods for at least the above-mentioned operations. Summary of the invention

[0022] This section includes examples of possible implementations and is not meant to be limiting.

[0023] In another exemplary aspect of the present invention, there is provided an apparatus, such as a user equipment side or network side apparatus, comprising: at least one processor; and at least one non-volatile memory storing instructions, which, when executed by the at least one processor, causes the apparatus to at least: predict a beam prediction output using a beam measurement output prediction sequence of at least one observation window associated with at least one prediction window providing a reference signal received power measurement, wherein the beam prediction output comprises a Top-1 or Top-K beam identifier or predicted reference signal received power of a beam subset associated with the beam prediction output; identify that a prediction window of the at least one prediction window overlaps at least a portion of an observation window of the at least one observation window over a time period; and compare the beam prediction output during the overlapping portion with a measured value of the observation window during the overlapping portion to perform long-term real-time monitoring, thereby determining at least one prediction output for use at a next moment, wherein the determination is based on the predicted reference signal received power of the at least one prediction output being higher than a threshold.

[0024] In another exemplary aspect of the present invention, a method is provided, comprising: predicting a beam prediction output using a predicted sequence of beam measurement outputs of at least one observation window associated with at least one prediction window providing a reference signal received power measurement, wherein the beam prediction output comprises a Top-1 or Top-K beam identifier or predicted reference signal received power of a beam subset associated with the beam prediction output; identifying that a prediction window of the at least one prediction window overlaps at least a portion of an observation window of the at least one observation window over a time period; and comparing the beam prediction output during the overlapping portion with a measured value of the observation window during the overlapping portion to perform long-term real-time monitoring, thereby determining at least one prediction output for use at a next moment, wherein the determination is based on the predicted reference signal received power of the at least one prediction output being higher than a threshold.

[0025] A further exemplary embodiment is an apparatus and method, comprising: the apparatus and method of the preceding paragraphs, wherein the at least one observation window comprises at least one sliding observation window for identifying a reference signal received power, and wherein the at least one prediction window comprises at least one sliding prediction window, wherein the identifying and comparing is performed using network configuration information from a communications network, which comprises a channel state information reference signal associated with an overlapping portion of the observation window, wherein the network configuration information comprises an indication of a size of the at least one sliding prediction window and provides configuration for the apparatus to predict one of a downlink receive beam pair or a downlink transmit beam pair as a new serving beam, wherein the network configuration information provides configuration for the apparatus The method comprises the steps of: configuring the downlink transmission beam identifier to predict one of the downlink transmission beam identifiers or a subset of the downlink transmission beam identifiers or one of the downlink transmission beam identifiers as a new service beam, wherein the prediction window overlaps with at least one next observation window in the time period, wherein the comparison comprises comparing the beam prediction outputs of more than one observation window with the prediction window during the overlapping portion to perform the long-term real-time monitoring, wherein for each prediction in the at least one prediction output, obtaining at least one downlink transmission beam identification reference signal received power or at least one downlink transmission and reception beam pair identification reference signal received power provided by machine learning determination, wherein the machine learning determination is to determine the predicted new service beam. a physical uplink shared channel, a physical uplink control channel, a physical downlink shared channel, or a physical downlink control channel, wherein at least one of the downlink receive beam or the downlink transmit beam pair is from at least one of the downlink receive beam pair observation window or the downlink transmit beam pair observation window, wherein the reference signal received power measurement is provided from time t to time tN, wherein the determining that the first prediction window overlaps is during time t to t+K, wherein the predicted first prediction window with the predicted output is during time t to t+L, wherein the predicted beam measurement sequence is consistent with the reference signal received power measurement sequence based on the reference signal received power measurement sequence. wherein information is received from the communications network comprising an indication of a channel state information reference signal according to an overlapping portion of at least one of more than one sliding prediction time windows or at least one observation time window simultaneously, wherein training is performed and inference is run to obtain a set of reference signal received power measurements of reference signal received power beams predicted, wherein at least one prediction output is based on a measurement layer 1 reference signal received power measurement, wherein at least one prediction output is based on a measurement layer 1 reference signal and auxiliary information, the auxiliary information comprising a beam identity, a user equipment location, and a line of sight or non-line of sight indication used as an input to a neural network based artificial intelligence / machine learning (AI / ML) model,wherein the prediction sequence is a reference signal received power over time, and / or wherein comparing the beam prediction output with the measurement value of the overlap portion is using a beam measurement prediction sequence from the at least one observation window during time t to t+M.

[0026] A non-transitory computer readable medium storing program code, the program code being executed by at least one processor to perform at least the methods described in the above paragraphs.

[0027] In another exemplary aspect of the present invention, a device is provided, comprising: a device for predicting a beam prediction output using a predicted sequence of beam measurement outputs of at least one observation window associated with at least one prediction window for providing a reference signal received power measurement, wherein the beam prediction output comprises a Top-1 or Top-K beam identifier or a predicted reference signal received power of a beam subset associated with the beam prediction output; a device for identifying that a prediction window of the at least one prediction window overlaps with at least a portion of an observation window of the at least one observation window over a time period; and a device for comparing the beam prediction output during the overlapping portion with the observation window measurement value during the overlapping portion to perform long-term real-time monitoring, thereby determining at least one prediction output for use at a next moment, wherein the determination is based on the predicted reference signal received power of the at least one prediction output being higher than a threshold.

[0028] The communication system includes a network side device and a user equipment side device for performing the above operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The above and other aspects, features and advantages of various embodiments of the present application will become more apparent from the following detailed description with reference to the accompanying drawings, in which similar reference numerals are used to designate similar or equivalent elements. The accompanying drawings are drawn to facilitate a better understanding of the embodiments of the present application and are not necessarily drawn to scale:

[0030] Figure 1 A synchronization signal block and related channel state information reference signal for downlink (DL) beam management or a downlink beam pair are shown;

[0031] Figure 2 User equipment beamforming with a multi-panel configuration is shown;

[0032] Figure 3 The temporal / time domain prediction is shown;

[0033] Figure 4 shows the time-domain beam prediction (TBP) when (i) the prediction windows overlap each other and (ii) the observation window overlaps some part of the prediction window;

[0034] Figure 5 The real-time monitoring capability of user equipment based on time-domain beam prediction is shown;

[0035] Figure 6 shows a signaling diagram illustrating enhancements to the "real-time" monitoring procedure and configuration for UE-side DL Tx prediction;

[0036] Figure 7 shows a method that can be performed by a device according to an exemplary embodiment of the present invention; and

[0037] Figure 8 A high-level block diagram is shown of various devices useful for implementing various aspects of the present invention. DETAILED DESCRIPTION

[0038] In exemplary embodiments of the present invention, at least one method and apparatus are proposed for a new user equipment side model with prediction quality indication using overlapping sliding window prediction.

[0039] AI / ML-based beam management

[0040] To support new AI / ML-enabled radio interfaces in next-generation cellular systems, 3GPP is currently studying a new study item (RP-213599) for Release 18. AI / ML-based beam management targets spatial and / or temporal beam prediction to reduce overhead and latency.

[0041] The beam management use case is further investigated in spatial and / or time domain beam prediction. The scope of spatial beam prediction (BM-Case1) is to predict the best downlink Top-1 or Top-K Tx beams, or the best downlink Tx-Rx beam pair at different spatial locations. Time domain beam prediction (BM-Case2) aims to predict the most likely beam to be used at the next moment, e.g., beam prediction in the spatial domain (BM-Case1).

[0042] SI range for beam management with AI / ML

[0043] In a study on artificial intelligence (AI) / machine learning (ML) for NR air interfaces, the agreed standards for this application include a use case that needs to be focused on:

[0044] The initial set of use cases includes:

[0045] o CSI feedback enhancements, e.g., reduced overhead, improved accuracy and prediction [RAN1];

[0046] o Beam management, e.g., beam prediction in the time and / or spatial domain to reduce overhead and latency and improve beam selection accuracy [RAN1];

[0047] o Improve positioning accuracy in different scenarios, including for example those with severe NLOS conditions

[0048] [RAN1] scene;

[0049] Finalize representative sub-use cases for each use case through RAN#98 for characterization and baseline performance evaluation:

[0050] o The AI / ML approaches for the selected sub-use cases need to be diverse enough to support the various requirements for the gNB-UE collaboration levels

[0051] Note: The sole purpose of selecting use cases for this study is to develop a framework for applying AI / ML to the air interface for these and other use cases. This selection itself is not intended to provide any indication of the prospects for any future specification project.

[0052] Study the 3GPP framework for over-the-air AI / ML corresponding to each target use case in terms of aspects such as performance, complexity and potential regulatory impact.

[0053] In the agreement:

[0054] AI / ML feature identification and feature-based LCM for UE-side model and / or UE part of dual-side model:

[0055] o Capabilities are AI / ML enabled features / FGs enabled by configuration, where configuration is based on

[0056] The conditions indicated by the UE capability are supported;

[0057] o Accordingly, the function-based LCM operates based on at least one configuration of AI / ML-enabled features / FGs or a specific configuration of AI / ML-enabled features / FGs:

[0058] FFS: Supports signaling for function-based LCM operations, such as activation / deactivation / fallback / switching of AI / ML functions;

[0059] FFS: whether / how to address additional conditions (such as scenarios, sites, and datasets) to help the UE side to transparently operate the model at the functional level (without identifying the model);

[0060] ·FFS: other aspects that may constitute function;

[0061] o FFS: What aspects should be specified as functionally available features / conditions for FG will be discussed in each sub-use case agenda;

[0062] AI / ML model identification and model ID-based LCM for UE-side model and / or UE part of dual-side model:

[0063] o Model ID-based LCM runs based on the identified model, where the model can be associated with specific configurations / conditions related to UE capabilities of AI / ML-enabled features / FGs and additional conditions (such as scenarios, sites, and datasets) determined / identified between the UE side and the NW side;

[0064] o FFS: Which aspects should be considered as additional conditions during the model identification process and how to incorporate them into the model description information will be discussed in the agenda of each sub-use case;

[0065] o FFS: the relationship between functions and models, e.g., whether the model can be identified with reference to the functions;

[0066] o FFS: Relationship between function-based LCM and model ID-based LCM.

[0067] Note: The applicability of function-based LCM and model ID-based LCM is discussed separately.

[0068] Prior to the above agreement, the following agreement had been reached

[0069] For UE side model And the UE part of the bilateral model:

[0070] - For AI / ML function identification

[0071] o Reuse the traditional 3GPP feature framework as a starting point for discussions.

[0072] o The UE represents the functions / functionality supported by a given sub-use case.

[0073] · Take the UE capability report as a starting point;

[0074] -For AI / ML model identification

[0075] oModels are identified at the network by model ID. UE indicates supported AI / ML models;

[0076] -In function-based LCM

[0077] o Network instructions are activated / deactivated via 3GPP signaling (such as RRC, MAC-CE, DCI) /

[0078] Rollback / switch AI / ML functions;

[0079] o The model may not be identifiable at the network and the UE may perform model level LCM:

[0080] · Investigate whether and to what extent NWs should understand / interact with LCM at the model level;

[0081] -In model ID-based LCM, the model is identified at the network and the network / UE can activate / deactivate / select / switch individual AI / ML models through the model ID.

[0082] What needs further investigation is the relationship between function identification and model identification, as well as the impact of performance monitoring and RAN4.

[0083] AI / ML Beam Management

[0084] Typically, the UE will be configured with at least one beam (reference signal) as a reference for receiving / transmitting data and control channels. The UE may be configured with one or more PDCCH channels and one or more PDSCH channels received on one or more DL beams. The UE may be capable of beamforming (i.e., it is capable of forming UL / DL beams for TX and RX) and may also use omnidirectional transmission and reception.

[0085] In 3GPP NR 5G, a cell (a gNB can have one or more cells) area can be covered by one or more beams provided by one or more transmit / receive points (TRPs), including TRP#1, ... TRP#X.

[0086] Each beam may carry an identifier, enabling the UE to identify the beam and perform measurements (e.g., received power, RSRP reference signal received power), and other related measurements associated with the specific identifier. Each SSB may be identified by an identifier carried by the SSB block. In addition, the SSB beam may be further used for training for downlink measurement signals for beam management.

[0087] For downlink measurement signals for beam management, NZP-CSI-RS (non-zero power channel state information reference signal) can be configured. Therefore, beam management from the downlink aspect can be performed using SSB and CSI-RS signals. As an example, SSB can be used to train more UEs, which are configured to report the N highest quality SSB beams, and the network can further configure the UE to report the M highest quality CSI-RS (#0...#K) beams associated with a specific SSB (such as SSB#0). This association can be configured by the network.

[0088] Figure 1 The synchronization signal block and the related channel state information reference signal used for downlink (DL) beam management or downlink beam pair are shown. Figure 1 As shown, there is a transmission point (TRP) that outputs synchronization signal blocks #0-#N and related channel state information reference signals RS#0-#K. Figure 1 In the middle is the MT having the MT beam including the best RX beam.

[0089] UE beamforming with multi-panel configuration

[0090] NR 5G also supports UE beamforming. At low frequencies, the UE may not use beamforming and may use omnidirectional beams (e.g., equal transmit and / or receive gain in all directions). However, at high frequencies, i.e., above 6 GHz, the UE may have one or more antenna panels that form one or more beams, such as Figure 2 shown.

[0091] The channel between the NW node and the UE changes continuously and the UE can report measurements performed under highly variable channel conditions. To avoid signal attenuation, the UE can also decide to switch antenna panels. Therefore, the measurements will be performed by one or more Rx antenna panels.

[0092] If the model is located at the UE side, the NW can use the model input / output measurements reported from multiple UE panels since it is not aware of the UE switching panels.

[0093] Figure 2 FIG. 4 shows a user equipment beamforming with a multi-panel configuration. Figure 2 As shown, there is a UE beamforming with TX / RX beams (index) 0...k and antenna configuration of 1...N antennas for Y panels. This beamforming uses TX / RX beams (index) 0...k and antenna configuration of 1...N antennas for Y panels.

[0094] In RAN WG#1#112bis-e AI 9.2.1, the following consensus was reached on the functional identification and model identification of the basic framework of the 3GPP framework for AI / ML.

[0095] Time domain beam prediction

[0096] For the beam prediction use case, the gNB or UE can predict the best beam in the spatial domain (BM-Case1) and the time domain (BM-Case2). The gNB or UE can perform the prediction by running inference or performing training.

[0097] In time domain beam prediction (BM-Case2), the gNB or UE can perform training or run inference to predict the set A beam or the Top-K of the set A beam at the future time. In BM-Case2, the gNB or UE can use the historical measurement results of the set B beam to predict the set A beam, where (i) the set B beam is a subset of the set A beam; (ii) the set B beam is different from the set A beam, that is, the set B beam represents a wide beam set and the set A beam represents a narrow beam set; (iii) the set B beam is the same as the set A beam.

[0098] Figure 3 The time domain / temporal domain prediction is shown. Figure 3 As shown, the transmissions from the gNB over time are used to measure information such as RSRP and location at different times to form a prediction output. Figure 3 As shown, beamforming is performed over time from the gNB using the measured RSRP and optional location information T1...Tn, where the ML model predicts over time using CRI / SSBR+RSRP.

[0099] For the UE side model, the UE can perform prediction and report the predicted set A beam id(s) / RSRP or the Top-K of set A beam id(s) / RSRP to the NW at future time steps.

[0100] In addition, for the NW side model, the NW predicts the Set A beam id(s) / RSRP or the Top-K of Set A beam id(s) / RSRP in future time steps using the history of Set B beam measurements reported by the UE.

[0101] In the standard when this application was proposed, although the UE side model and UE monitoring were introduced, there was no "real-time" monitoring mechanism or UE reporting "real-time" monitoring capability. The exemplary embodiments of the present invention can at least solve these deficiencies.

[0102] Before describing in detail the example embodiments disclosed herein, reference is made to Figure 8 Detailed description is provided to illustrate simplified block diagrams of various electronic devices that are suitable for use in implementing the exemplary embodiments of this invention.

[0103] Figure 8 is a block diagram of one possible and non-limiting system in which example embodiments may be practiced.

[0104] Steering Figure 8 , which shows a block diagram of one possible and non-limiting example in which an example embodiment may be practiced. The figure shows a user equipment (UE) 110 and a wireless access LTE, 5G or 6G network base station, namely a gNB 170 and an NCE / MME / GW 190. Figure 8In the example of FIG. 1 , a user equipment (UE) 110 wirelessly communicates with a wireless network 100. A UE is a wireless device that can access the wireless network 100. The UE 110 includes one or more processors 120, one or more memories 125, and one or more transceivers 130 interconnected by one or more buses 127. Each of the one or more transceivers 130 includes a receiver Rx 132 and a transmitter Tx 133. The one or more buses 127 can be address, data, or control buses, and can also include any interconnection mechanism, such as a series of lines on a motherboard or integrated circuit, optical fiber or other optical communication devices, etc. The one or more transceivers 130 are connected to one or more antennas 128. The one or more memories 125 include computer program code 123. The UE 110 includes an output module 140, which includes one or both parts of 140-1 and / or 140-2, and can be implemented in a variety of ways. The output module 140 can be implemented in hardware as an output module 140-1, such as as part of one or more processors 120. Output module 140-1 may also be implemented as an integrated circuit or by other hardware such as a programmable gate array. In another example, output module 140 may be implemented as output module 140-2, which is implemented as computer program code 123 and executed by one or more processors 120. For example, one or more memories 125 and computer program code 123 may be configured to, together with one or more processors 120, cause user equipment 110 to perform one or more operations described herein. UE 110 communicates with gNB 170 via wireless link 111.

[0105] In this example, gNB 170 is a base station that provides access to wireless network 100 for wireless devices such as UE 110. For example, gNB 170 may be a base station for 5G, also known as New Radio (NR). In 5G, gNB 170 may be a NG-RAN node, which is defined as a gNB or ng-eNB. A gNB is a node that provides NR user plane and control plane protocol terminations to a UE and is connected to a 5GC (e.g., NCE / MME / GW 190) via an NG interface. An NG-eNB is a node that provides E-UTRA user plane and control plane protocol terminations to a UE and is connected to a 5GC via an NG interface. An NG-RAN node may include multiple gNBs, which may also include a central unit (CU) (gNB-CU) 196 and a distributed unit (DU) (gNB-DU), where DU 195 is shown. Note that a DU may include or be coupled to and control a radio unit (RU). The gNB-CU is a logical node that hosts the radio resource control (RRC), SDAP, and PDCP protocols of a gNB or the RRC and PDCP protocols of an en-gNB that controls the operation of one or more gNB-DUs. The gNB-CU terminates at the F1 interface connected to the gNB-DU. The F1 interface is illustrated as reference numeral 198, although reference numeral 198 also illustrates links between remote elements of the gNB 170 and centralized elements of the gNB 170, such as the link between the gNB-CU 196 and the gNB-DU 195. The gNB-DU is a logical node that hosts the RLC, MAC, and PHY layers of a gNB or en-gNB, the operation of which is controlled in part by the gNB-CU. One gNB-CU supports one or more cells. One cell is supported by only one gNB-DU. The gNB-DU terminates at the F1 interface 198 connected to the gNB-CU. Note that DU 195 is considered to include transceiver 160, e.g., as part of a RU, but in some of the examples, transceiver 160 may be part of a separate RU, e.g., controlled by and connected to DU 195. gNB 170 may also be an eNB (evolved NodeB) base station for LTE (Long Term Evolution), or any other suitable base station or node.

[0106] The gNB 170 includes one or more processors 152, one or more memories 155, one or more network interfaces (N / WI / F) 161, and one or more transceivers 160 interconnected via one or more buses 157. Each of the one or more transceivers 160 includes a receiver Rx 162 and a transmitter Tx 163. The one or more transceivers 160 are connected to one or more antennas 158. The one or more memories 155 include computer program code 153. The CU 196 may include a processor 152, a memory 155, and a network interface 161. It should be noted that the DU 195 may also include its own memory, processor, and / or other hardware, but is not shown.

[0107] The gNB 170 includes an output module 150, which includes a portion or both of 150-1 and / or 150-2, which can be implemented in a variety of ways. The output module 150 can be implemented in hardware as the output module 150-1, such as as part of one or more processors 152. The output module 150-1 can also be implemented as an integrated circuit or through other hardware such as a programmable gate array. In another example, the output module 150 can be implemented as an output module 150-2, which is implemented as computer program code 153 and executed by one or more processors 152. For example, one or more memories 155 and computer program code 153 are configured to, together with one or more processors 152, cause the gNB 170 to perform one or more operations described herein. It should be noted that the functionality of the output module 150 can be distributed, such as distributed between the DU 195 and the CU 196, or can be implemented only in the DU 195.

[0108] One or more network interfaces 161 communicate over a network, such as via links 176 and 131. Two or more gNBs 170 may communicate using link 176. Link 176 may be wired or wireless or both, and may implement, for example, an Xn interface for 5G, an X2 interface for LTE, or other suitable interfaces for other standards.

[0109] The one or more buses 157 may be address, data, or control buses and may include any interconnection mechanism, such as a series of wires on a motherboard or integrated circuit, optical fiber or other optical communication device, wireless channel, etc. For example, the one or more transceivers 160 may be implemented as a remote radio head (RRH) 195 for LTE or a distributed unit (DU) 195 receiver for a gNB implementation for 5G, other elements of the gNB 170 may be physically located at a different location from the RRH / DU, and the one or more buses 157 may be implemented in part as, for example, fiber optic cables or other suitable network connections to connect other elements of the gNB 170 (e.g., central unit (CU), gNB-CU) to the RRH / DU 195. Reference numeral 198 also indicates these suitable network links.

[0110] It should be noted that the description here indicates that a "cell" performs functions, but it should be clear that the devices that constitute the cell can also perform these functions. The cell constitutes a part of the base station. That is, each base station can have multiple cells. For example, there can be three cells for a single carrier frequency and associated bandwidth, each cell covering one-third of a 360-degree area, so that the coverage area of ​​a single base station can be approximately elliptical or circular. In addition, each cell can correspond to a single carrier, and the base station can use multiple carriers. Therefore, if each carrier has three 120-degree cells and two carriers, the base station has a total of 6 cells.

[0111] The wireless network 100 may include one or more network elements 190, which may include core network functions. In addition, the NCE / MME / GW 190 may, for example, perform access & mobility management functions (AMF), location management functions (LMF), mobile management entities (MME), network control elements (NCE), policy control functions (PCF), serving gateways (SGW), session management functions (SMF), and unified data management (UDM). The NCE / MME / GW 190 provides connections to other networks such as telephone networks and / or data communication networks (such as the Internet) via one or more links 181. Such core network functions for 5G may include access and mobility management functions (AMF(S)) and / or user plane functions (UPF(S)) and / or session management functions (SMF(S)). Core network functions for LTE may include MME (mobility management entity) / SGW (serving gateway) functions. These are only example functions that the NCE / MME / GW 190 may support, and it should be noted that both 5G and LTE functions may be supported. The gNB 170 is coupled to the network element 190 via a link 131. The link 131 may be implemented as, for example, an NG interface for 5G, an S1 interface for LTE, or other suitable interface for other standards. The network element 190 includes one or more processors 175, one or more memories 171, and one or more network interfaces (N / WI / F) 180, interconnected by one or more buses 185. The one or more memories 171 include computer program code 173. The one or more memories 171 and the computer program code 173 are configured to, together with the one or more processors 175, cause the network element 190 to perform one or more operations.

[0112] The wireless network 100 can implement network virtualization, which is the process of combining hardware and software network resources and network functions into a single, software-based management entity (virtual network). Network virtualization involves platform virtualization, usually in combination with resource virtualization. Network virtualization can be divided into external virtualization and internal virtualization. External virtualization refers to merging many networks or parts of networks into a virtual unit, and internal virtualization refers to providing network-like functions to software containers on a single system. It should be noted that the virtualized entities produced by network virtualization are still implemented to some extent using hardware such as processors 152 or 175 and memories 155 and 171, and these virtualized entities will also produce technical effects.

[0113] Computer readable memories 125, 155, and 171 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 memory, flash memory, magnetic memory and systems, optical memory devices and systems, fixed memory, and removable memory. Computer readable memories 125, 155, and 171 may be devices for performing storage functions. Processors 120, 152, and 175 may be of any type suitable for the local technical environment and may include one or more of a general-purpose computer, a special-purpose computer, a microprocessor, a digital signal processor (DSP), and a processor based on a multi-core processor architecture, as non-limiting examples. Processors 120, 152, and 175 may be used to perform functions, such as devices for controlling UE 110, gNB 170, NCE / MME / GW 190, and other functions described herein.

[0114] In general, various embodiments of the user device 110 may include, but are not limited to, cellular phones such as smart phones, tablet computers, personal digital assistants (PDAs) with wireless communication capabilities, portable computers with wireless communication capabilities, image capture devices such as digital cameras with wireless communication capabilities, gaming devices with wireless communication capabilities, music storage and playback devices with wireless communication capabilities, Internet devices that allow wireless Internet access and browsing, tablet computers with wireless communication capabilities, and portable units or terminals that include a combination of the above functions.

[0115] One or more of output modules 140-1, 140-2, 150-1, and 150-2 may be configured to implement a high-level grammar for compressed representation of a neural network based on the examples described herein. Computer program code 173 may also be configured to implement a high-level grammar for compressed representation of a neural network based on the examples described herein.

[0116] In addition, various embodiments of any of these devices may be used with a UE vehicle, a high altitude platform station, or any other such node associated with a terrestrial network, or with any drone-type radio or radio in an aircraft or other airborne device, or with a vessel traveling on water, such as a boat.

[0117] As mentioned above, in the standard discussion of this application, although the UE side model and UE monitoring are introduced, there is no "real-time" monitoring mechanism or UE report on "real-time" monitoring capability. The exemplary embodiments of the present invention can at least solve these deficiencies.

[0118] BM-Case2 is applicable to the conditions of the functions associated with BM-Case1, which assumes DL Tx beam prediction and supports Top-KDL TX beam prediction, as follows:

[0119] o K=1,2,4,[8]

[0120] - SSB-based and / or CSI-RS-based RSRP measurements, support for predicting the best K NZP CSI-RS resource is defined here.

[0121] Set B conditions:

[0122] oMeasured DL RS (SSB, CSI-RS)

[0123] - Define support for RSRP measurements using SSB and / or CSI-RS;

[0124] o Dimension of the measured DL RS set (4, 8, 12,

[16] ):

[0125] - indicates the minimum number of NZP-CSI-RS resources that the UE should measure and use to predict the best K NZP CSI-RS resource;

[0126] oMeasured DL RS set period (40ms, 80ms):

[0127] - indicates the minimum duration for which the NZP-CSI-RS resource should be measured;

[0128] o DL RS set mode for measurement (e.g. fixed, preconfigured list, random)

[0129] - indicates restrictions on the conditions of set B.

[0130] Set A conditions:

[0131] o Predicted DL RS (CSI-RS)

[0132] - Define support for predicting CSI-RS resources

[0133] o Predicted DL RS set dimensions (12, 16, 32, 64)

[0134] - indicates the maximum number of NZP-CSI-RS resources that should be configured to predict the NZP-CSI-RS resource set

[0135] o Predicted DL RS set - number of future instances (40 ms, 80 ms)

[0136] - indicates the maximum duration (or number of future instances compared to the measurement period) for which the NZP-CSI-RS resources can be predicted based on set B.

[0137] NW side performance monitoring conditions

[0138] oSupports the measurement of predicted DL RS sets (full set A, partial set A)

[0139] - Define support for measuring NZP-CSI-RS resources corresponding to set A.

[0140] oMeasurement period (100 ms, 200 ms)

[0141] -Indicates the minimum period when supporting NZP-CSI-RS resources corresponding to set A.

[0142] Conditions to support ML functions

[0143] oThe maximum number of supported functions (1, 2, 4, 8, )

[0144] - indicates the maximum number of functions that can be configured for the UE (e.g., the number of parameter combinations for enabling ML-enabled features)

[0145] oDelay in activating the function (2ms, 4ms, ...).

[0146] - Indicates the delay time required to activate or switch a function

[0147] oGeneralization conditions of the function (yes, no)

[0148] - Indicates that the UE supports any function that considers the configuration of 1-4 parameter combinations .

[0149] Furthermore, example embodiments of the present invention propose specific necessary signaling between the NW and the UE to identify “real-time” monitoring characteristic conditions, especially in time domain beam prediction, which, to the best of our knowledge, has not been studied so far.

[0150] For DL ​​Tx and DL Tx-Rx beam-pair time-domain beam prediction (TBP), the prediction output will be predicted as an output sequence of multiple consecutive time steps. The prediction can be performed by the NW or the UE. In the current 3GPP discussion, when the NW or UE performs time-domain / temporal prediction, the prediction output is an output sequence called a "prediction window" using an input sequence collected during an "observation window".

[0151] During LCM operation, how to monitor the time domain prediction output so that:

[0152] o When the UE performs temporal domain beam prediction and the UE performs model monitoring, how the UE compares the prediction with the ground truth in real time; and

[0153] o How to configure the UE to switch between Model#A and Model#B or back to legacy mode when performance degradation is predicted.

[0154] Figure 4The time domain beam prediction (TBP) is shown as follows: (i) the prediction windows overlap each other, (ii) the observation window overlaps some parts of the prediction window

[0155] exist Figure 4 In the AI / ML model, the AI / ML model is located on the UE side. The UE performs L1-RSRP measurements and creates sliding windows and marks data in the data preprocessing part. Therefore, the UE will have multiple L1-RSRP sliding windows or a cascade of L1-RSRP sliding windows with auxiliary information (such as UE location, Tx beam angle, LoS / nLos indication, etc.). The input to the neural network can be L1-RSRP or a cascade of L1-RSRP with auxiliary information, and these windows include RSRP observation window #1, observation window #P, observation window #Q, prediction window #1 with beam ID / RSRP sequence, prediction window #P with beam ID / RSRP sequence, and prediction window #Q with beam ID / RSRP sequence.

[0156] The UE uses the sequence of aggregate B-beam RSRP / LI-RSRP measurements such that the aggregate B-beam RSRP / L1-RSRP measurements are used for observation window #1 (measurements from time tN to time t-1) to predict prediction window #1 (prediction output from time t to t+L).

[0157] The prediction output includes the Top-1 or Top-K beam ID of SetA or the predicted RSRP of the beams of SetA.

[0158] Therefore, prediction window #1 and prediction window #P start from time t+K ′ 1 to time t+L overlap each other

[0159] How does the UE handle the overlapping prediction parts?

[0160] When considering observation window #1 and prediction window #1 and observation window #P and prediction window #P, there is an overlap between observation window #P and prediction window #1 from time t to t+K:

[0161] This will result in the UE being able to perform "real time" monitoring. The steps for UE reporting and how the NW configures the UE to perform "real time" monitoring need to be addressed.

[0162] Furthermore, when considering observation window #1 and prediction window #1 and observation window #Q and prediction window #Q, observation window #1 and prediction window #Q have an overlapping portion from time t to time t+M'.

[0163] The UE can perform "real-time" monitoring, where the UE can compare the predicted output with the actual conditions (measurements) on the ground during the period t to t+M'. This results in the UE having the ability to perform "long-term real-time" monitoring.

[0164] In one embodiment, the "real-time" monitoring capability includes that the UE can simultaneously perform sensing for data collection and perform prediction.

[0165] In addition, if Figure 4 As shown, there is no solution on how to deal with the situation where the prediction sequence outputs "multiple prediction windows" overlap each other.

[0166] At least Figure 4 and Figure 5 In the present invention, some novel features introduced by the exemplary embodiments of the present invention are highly reflected, which are summarized as follows:

[0167] For each prediction, obtain at least one DL Tx beam ID / RSRP or at least one DL Tx-Rx beam pair ID / RSRP, which is provided by the ML to determine whether the predicted new serving DL Tx beam or DL ​​Tx-Rx beam pair will be configured for at least one of (PUSCH / PUCCH / PDSCH / PDCCH),

[0168] For example, if the prediction provides at least one beam ID in the DLTop1 or Top-K beam that is already a serving beam, the NW may not update one or more serving beams to the UE. The determination may be made separately for each channel, or the same beam may be used for each of the DL / UL channels).

[0169] In an example embodiment of the present invention, a UE-side model may be assumed. Therefore, the network may need to estimate the reliability of the UE prediction.

[0170] In order to determine whether the UE prediction is reliable, monitoring-based LCM operation needs to be considered. In particular, a "real-time" monitoring mechanism is introduced.

[0171] like Figure 5 As shown, the example embodiment of the present application introduces a "real-time" monitoring method for time domain beam prediction (TBP) on the UE side, wherein the UE can verify the predicted output with the actual situation on the ground in real time.

[0172] The UE may verify the monitoring output based on at least quality metrics, namely (i) L1-RSRP difference (difference between predicted L1-RSRP and measured L1-RSRP), (ii) accuracy of Top-1 or Top-K beam ID prediction compared with the actual situation.

[0173] Top-1 or top-K beam prediction or L1-RSRP prediction within the time window, where:

[0174] o UE prediction can determine if a beam fails (e.g., link communication between UE and NW is unreliable such that L1-RSRP prediction < threshold, or L1-RSRP prediction for Top-1 / Top-K beams < threshold). This beam failure assessment can be used for example for PDCCH beams.

[0175] The UE performs L1-RSRP measurement corresponding to a DL Tx beam or a DL Tx-Rx beam pair. The UE may determine the Rx beam corresponding to the received Rx antenna panel to be used for DL ​​Tx beam measurement. The UE shall perform L1-RSRP measurement corresponding to the specified / defined DL Rx beam.

[0176] The UE then feeds the L1-RSRP as input into the neural network for DL ​​TX beam prediction. These can provide L1-RSRP or L1-RSRP and auxiliary information such as UE location, LoS or nLoS indication, UE Rx panel ID for Rx beam, etc.

[0177] The UE feeds L1-RSRP and DL Rx beam information (i.e. Rx beam ID) as input into the neural network for DLTx-Rx beam pair prediction. In addition, these can feed L1-RSRP and DL Rx beam information or L1-RSRP and DL Rx beam information including auxiliary information such as UE location, LoS or nLoS indication, Rx panel ID of UE Rx beam, etc.

[0178] exist Figure 5 In the example, through the overlapping part of observation window #Q and prediction window #1, the UE will have the ability to perform "long-term real-time" monitoring, that is, the UE performs measurement and prediction at the same time.

[0179] With this feature, the UE can efficiently report monitoring output over a long period of time.

[0180] The exemplary embodiments in this application solve the new configuration from NW to UE and the new indication reporting problem from UE to NW in the present invention.

[0181] UE-side time domain beam prediction monitoring mechanism

[0182] In this section, DL Tx time beam prediction is provided, in which monitoring based on UE-side LCM operation is addressed.

[0183] DL TX beam prediction

[0184] exist Figure 5In the present invention, signaling and configuration from NW to UE for DL ​​Tx beam prediction are provided, wherein the UE can determine the DL Rx beam with respect to the DL Rx panel for measuring the DL Tx beam. The UE will use the determined DL Rx beam to measure the received signal received power (RSRP) from the DL Tx beam. The UE will then use the sequence of RSRP / L1-RSRP over time as input to the neural network.

[0185] Figure 6 The NW configuration process, reporting from UE and the "real-time monitoring" mechanism are described.

[0186] Figure 6 A signaling diagram is shown illustrating the enhancements to the “real-time” monitoring procedure and the configuration to be used for UE-side DL Tx prediction.

[0187] Figure 6 The signaling diagram illustrates the enhancements to the "real-time" monitoring procedure and the configuration on the UE side for DL ​​Tx prediction.

[0188] Some key steps of the invention are as follows: Figure 6 Detailed instructions include Figure 6 Additional examples of these steps are shown below:

[0189] 1. In Figure 6 In the beginning step of the RRC configuration is determined. During the RRC configuration, the UE is configured to report the capability to perform both measurement and prediction simultaneously;

[0190] 2. The NW then transmits one or more CSI-RS or SSB reference signals (RS) that scan between each Tx beam;

[0191] 3.UE determines the DL Rx beam to measure the DL Tx beam. The UE will measure the RSRP / L1-RSRP from the Tx beam;

[0192] 4. According to an exemplary embodiment of the present application, NW indicates the size of the prediction window and configures the UE to perform DL Tx beam prediction;

[0193] 5. The UE then performs data pre-processing on the input and marks the output for DL ​​Tx beam prediction based on L1-RSRP measurement;

[0194] 6. According to an exemplary embodiment of the present application, the NW transmits the CSI-RS according to the overlapping part of the observation window and the prediction window;

[0195] 7. Then, the UE will perform training or run inference to obtain the Top-1 or Top-K or L1-RSRP of the set A beam prediction. Then, according to the exemplary embodiment of the present application, the UE compares the predicted output with the measured value of the overlapped part (the overlapped part of the observation window and the prediction window);

[0196] 8. According to an example embodiment of the present application, the UE reports "real-time monitoring" outputs from the overlapping parts of the observation and prediction windows, which can be (i) metrics, i.e., accuracy of predictions and / or (ii) events, e.g., if predicted RSRP or predicted L1-RSRP>threshold.

[0197] DL Tx-Rx beam pair prediction

[0198] exist Figure 6 In the present invention, signaling and configuration from NW to UE for DL ​​Tx-Rx beam pair prediction are provided, wherein the UE can determine the DL Rx beam with respect to the DL Rx panel for measuring the DL Tx beam.

[0199] The UE will then use the determined DL Rx beam to measure the received signal received power (RSRP) of the DL Tx-Rx beam pair.

[0200] In beam pair prediction, the UE will use the RSRP / L1-RSRP sequence and the Rx beam information from the received / best Rx panel as input to the neural network for time beam prediction.

[0201] NW configuration process, reports from UE and monitoring-based LCM process such as Figure 6 shown.

[0202] Figure 6 : A signaling diagram is shown illustrating the improvements to the "real-time" monitoring process and the configuration to be used for UE-side DL Tx-Rx beam pair prediction.

[0203] Some key steps of the invention are as follows: Figure 6 as shown in steps 4, 6, 7, and 8 in .

[0204] Details as Figure 6 As shown below:

[0205] 1. In Figure 6 In the beginning step of the RRC configuration is determined. During the RRC configuration, the UE is configured to report the capability to perform measurements and perform predictions simultaneously;

[0206] 2. The NW then transmits one or more CSI-RS or SSB reference signals (RS) that scan between Tx beams;

[0207] 3. The UE determines the DL Rx beam to measure the DL Tx-Rx beam pair. The UE will measure the RSRP / L1-RSRP from the Tx-Rx beam pair;

[0208] 4. According to an exemplary embodiment of the present application, NW indicates the size of the prediction window and configures the UE to perform DL Tx-Rx beam pair prediction;

[0209] 5. The UE then performs data preprocessing on the input and outputs the prediction for DL ​​Tx-Rx beam pairs based on the L1-RSRP measurement flag;

[0210] 6. According to an exemplary embodiment of the present application, the NW transmits the CSI-RS according to the overlapping part of the observation window and the prediction window;

[0211] 7. The UE will then perform training or run inference to obtain the predicted Top-1 or Top-K or L1-RSRP of the set A DL Tx-Rx beam pairs. Then, according to an exemplary embodiment of the present application, the UE compares the predicted output (in the overlap window) with the measured value (the observed value during the overlap part);

[0212] 8. Then, according to an example embodiment of the present application, the UE reports a "real-time monitoring" output, which can be (i) a metric, i.e., the accuracy of the prediction and / or (ii) an event, e.g., if the predicted RSRP or predicted L1-RSRP>threshold.

[0213] UE-side AI / ML model

[0214] In one embodiment, the UE may perform prediction by performing training or running inference. To obtain the set A beam ID or the Top1 / Top-K beam ID of set A, the UE may be configured to perform time domain beam prediction using LSTM or Conv-LSTM or Transformer models.

[0215] After the UE obtains the prediction output, the UE can perform "real-time" monitoring with supported capabilities and then report the "real-time" monitoring output to the NW.

[0216] In one embodiment, "real-time" monitoring capability includes the UE potentially performing sensing and performing prediction for data collection simultaneously.

[0217] In another embodiment, if the monitoring metrics and / or events are not met, the NW may configure the UE to update the AI / ML model or configure the UE to download a new model.

[0218] Figure 7 It is shown that the method according to the exemplary embodiment of the present invention can be performed by a device.

[0219] Figure 7 The diagram illustrates a device that can be used by, for example, but not limited to, a network device (e.g., Figure 8 The operation performed by the UE 110 in FIG. Figure 7 As shown in block 710 of , a beam prediction output is predicted using a prediction sequence of beam measurement outputs for at least one observation window associated with at least one prediction window providing a reference signal received power measurement. Figure 7 As shown in block 720 of , the beam prediction output includes a Top-1 or Top-K beam identifier or predicted reference signal received power of a beam subset associated with the beam prediction output. Figure 7 As shown in block 730 of , it is identified that a prediction window of at least one prediction window overlaps at least a portion of an observation window of at least one observation window over a period of time. Figure 7 As shown in block 740 of , the beam prediction output during the overlapped portion is compared with the measured value of the observation window during the overlapped portion to perform long-term real-time monitoring, thereby determining at least one prediction output for use at the next moment. Figure 7 As shown in block 750, the determination is based on a predicted reference signal received power of at least one prediction output being greater than a threshold.

[0220] According to the example embodiment described in the preceding paragraph, wherein the at least one observation window comprises at least one sliding observation window identifying the reference signal received power, and wherein the at least one prediction window comprises at least one sliding prediction window.

[0221] According to the example embodiment described in the preceding paragraphs, the identifying and comparing uses network configuration information from the communication network, which includes a channel state information reference signal associated with the overlapping portion of the observation window.

[0222] According to the example embodiment described in the above paragraphs, the network configuration information includes an indication of at least one sliding prediction window size, and provides a configuration to the device to predict one of the downlink receive beam pair or the downlink transmit beam pair as a new serving beam.

[0223] According to the example embodiment described in the above paragraph, the network configuration information provides a configuration for the device to predict a downlink transmission beam identifier or one of a subset of downlink transmission beam identifiers or one of the downlink transmission beam identifiers as a new serving beam

[0224] According to the example embodiment described in the above paragraph, the prediction window overlaps with at least one next observation window over a period of time, and the comparison includes comparing the beam prediction output of more than one observation window with the prediction window during the overlapping portion to perform long-term real-time monitoring.

[0225] According to the example embodiment described in the above paragraph, for each prediction of at least one prediction output, at least one downlink transmit beam identification reference signal received power or at least one downlink transmit and receive beam pair identification reference signal received power provided by machine learning determination is obtained.

[0226] According to the example embodiment described in the above paragraphs, the machine learning determination is to determine whether a predicted new service downlink transmit beam or a downlink transmit and receive beam pair identification beam pair will be configured as at least one of a physical uplink shared channel, a physical uplink control channel, a physical downlink shared channel, or a physical downlink control channel.

[0227] According to the example embodiment described in the above paragraph, at least one of the downlink receive beam or the downlink transmit beam pair is from at least one of the downlink receive beam pair observation window or the downlink transmit beam pair observation window.

[0228] According to the example embodiment described in the above paragraphs, wherein the reference signal received power measurement value is provided from time t to time tN, wherein the determination that the first prediction window is overlapping is during time t to t+K, and wherein the first prediction window with the predicted output is predicted from time t to t+L.

[0229] According to the example embodiment described in the above paragraphs, the predicted beam measurement sequence is associated with a sliding window based on reference signal received power measurement.

[0230] According to the example embodiment described in the above paragraph, the information received from the communication network includes an indication of a channel state information reference signal according to an overlapping portion of at least one of more than one sliding prediction time windows or at least one simultaneous observation time window.

[0231] According to the example embodiments described in the above paragraphs, training is performed and inference is run to obtain a set of reference signal received powers for measuring reference signal received powers in beam prediction.

[0232] According to the example embodiment described in the above paragraphs, at least one predicted output is based on measuring layer 1 reference signal received power measurements.

[0233] According to the example embodiment described in the above paragraph, at least one prediction output is based on a layer 1 reference signal and auxiliary information, which includes beam identification, user equipment location and line-of-sight or non-line-of-sight indication used as input to a neural network-based AI / ML artificial intelligence / machine learning model.

[0234] According to the example embodiment described in the above paragraph, the prediction sequence is the reference signal received power over time.

[0235] According to the example embodiment described in the above paragraph, comparing the beam prediction output with the measurement value of the overlapping portion is performed using a beam measurement prediction sequence from at least one observation window during time t to t+M.

[0236] Non-transitory computer readable medium (as shown in FIG. Figure 8 The memory 125 stores program code (computer program code 123 and / or Figure 8 The output module 140-2 in the program code is executed by at least one processor ( Figure 8 The processor 120 and / or output module 140-1) in the processor 120 and / or output module 140-1) are executed to perform at least the operations described in the above paragraphs.

[0237] According to an exemplary embodiment of the present invention, a device is provided, comprising: a device for predicting a beam prediction output using a predicted sequence of beam measurement outputs of at least one observation window associated with at least one prediction window for providing a reference signal received power measurement, wherein the beam prediction output comprises a Top-1 or Top-K beam identifier or a predicted reference signal received power of a beam subset associated with the beam prediction output; a device for identifying that a prediction window of at least one prediction window overlaps at least a portion of an observation window of at least one observation window within a time period; and a device for comparing the beam prediction output during the overlapping portion with the observation window measurement value during the overlapping portion to perform long-term real-time monitoring, thereby determining at least one prediction output for use at a next moment, wherein the determination is based on the predicted reference signal received power of at least one prediction output being higher than a threshold.

[0238] In the examples of the invention described in the above paragraphs, wherein at least the means for using, identifying and comparing comprises a non-transitory computer readable medium [ Figure 8 125], which is encoded with at least one processor [ Figure 8 A computer program executable by the processor 120 and / or the output module 140-1] Figure 8 Computer program code 123 and / or output module 140-2].

[0239] In addition, according to an exemplary embodiment of the present invention, there is a circuit for performing operations. Such a circuit may include any type of circuit, including a content encoding circuit, a content decoding circuit, a processing circuit, an image generation circuit, a data analysis circuit, etc. In addition, the circuit may also include a discrete circuit, an application-specific integrated circuit (ASIC) and / or a field programmable gate array circuit (FPGA), etc., and a processor specifically configured by software to perform the corresponding functions, or a dual-core processor with software and a corresponding digital signal processor, etc.). In addition, the necessary inputs and outputs of the circuit, the functions performed by the circuit, and the interconnection of the circuit with other components (possibly via inputs and outputs) are also provided, and these components may include other circuits to perform the exemplary embodiments of the present invention described herein.

[0240] According to an example embodiment of the present invention, as in the present application, the provided "circuit" may include at least one or more or all of the following:

[0241] (a) hardware-only circuit implementation (such as implementation in analog and / or digital circuits only);

[0242] (b) a combination of hardware circuitry and software such as (where applicable);

[0243] (i) a combination of analog and / or digital hardware circuits and software / firmware; and

[0244] (ii) any portion of a hardware processor with software (including a digital signal processor), software and memory that work together to enable a device such as a mobile phone or server to perform various functions, such as the functions or operations according to the example embodiments of the present invention disclosed herein); and

[0245] (c) A hardware circuit and / or processor, such as a microprocessor or portion of a microprocessor, that requires software (e.g., firmware) to operate, but the software may not be present when not required to operate.

[0246] According to example embodiments of the present invention, according to example embodiments of the present invention disclosed in this application, there are suitable circuits for performing at least the novel operations, and "circuitry" as used herein refers to at least the following:

[0247] (a) hardware-only circuit implementations (such as implementations in analog and / or digital circuits only); and

[0248] (b) a combination of circuitry and software (and / or firmware), such as (as applicable): (i) a combination of processors or (ii) portions of processors / software (including digital signal processors), software and memory, which work together to enable a device such as a mobile phone or server to perform various functions); and

[0249] (c) Circuitry that requires software or firmware to operate, such as a microprocessor or portion of a microprocessor, even if the software or firmware is not actually present.

[0250] This definition of circuitry applies to all uses of the term in this application, including in any claims. As a further example, as used in this application, the term "circuitry" would also cover an implementation of only a processor (or multiple processors) or a portion of a processor and its (or their) accompanying software and / or firmware. For example, and where applicable to a specific claim element, the term "circuitry" would also cover a baseband integrated circuit or application processor integrated circuit for a mobile device, or a similar integrated circuit in a server, cellular network device, or network device.

[0251] In general, various embodiments may be implemented in hardware or dedicated circuits, software, logic, or any combination thereof. For example, certain aspects may be implemented in hardware, while other aspects may be implemented in firmware or software, which may be executed by a controller, microprocessor, or other computing device, but the invention is not limited thereto. Although various aspects of the invention may be illustrated and described using block diagrams, flow charts, or other graphical representations, it is well known that the blocks, devices, systems, techniques, or methods described herein may be implemented in hardware, software, firmware, dedicated circuits or logic, general purpose hardware or controllers or other computing devices, or some combination thereof, as non-limiting examples.

[0252] Embodiments of the present invention may be practiced in various components such as integrated circuit modules. The design of integrated circuits is essentially a highly automated process. Complex and powerful software tools can convert logic-level designs into semiconductor circuit designs for etching into shape on semiconductor substrates.

[0253] The word "exemplary" is used herein to mean "serving as an example, instance, or illustration." Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments. All embodiments described in this detailed description are exemplary embodiments intended to enable persons skilled in the art to make or use the invention, rather than to limit the scope of the invention which is defined by the claims.

[0254] The foregoing description provides by way of exemplary and non-limiting examples a full and informative description of the best method and apparatus presently contemplated by the inventors for carrying out the invention. However, various modifications and adaptations may be apparent to those skilled in the relevant art in view of the foregoing description, read in conjunction with the accompanying drawings and the appended claims. Nevertheless, all such and similar modifications to the teachings of the exemplary embodiments of the invention will still fall within the scope of protection of the invention.

[0255] It should be noted that the terms "connected," "coupled," or any variations thereof, refer to any connection or coupling between two or more elements, whether direct or indirect, and may include the presence of one or more intermediate elements "connected" or "coupled" together between two elements. The coupling or connection between elements may be physical, logical, or a combination thereof. Herein, two elements may be considered to be "connected" or "coupled" together through the use of one or more wires, cables, and / or printed electrical connections, as well as through the use of electromagnetic energy, such as electromagnetic energy having wavelengths in the radio frequency region, the microwave region, and the optical (visible and invisible) region, as a few non-limiting and non-exhaustive examples.

[0256] In addition, some features of the preferred embodiments of the present invention may be used preferably without the corresponding use of other features.Therefore, the above description should be regarded as only illustrative of the principles of the present invention, rather than limiting thereof.

Claims

1. A device for communication, comprising: at least one processor; as well as at least one memory storing instructions that, when executed by the at least one processor, cause the apparatus to at least: predicting a beam prediction output using a beam measurement output prediction sequence for at least one observation window associated with at least one prediction window providing a reference signal received power measurement, The beam prediction output includes a Top-1 or Top-K beam identifier or a predicted reference signal received power of a beam subset associated with the beam prediction output; identifying that a prediction window of the at least one prediction window overlaps at least a portion of an observation window of the at least one observation window over a time period; comparing the beam prediction output during the overlapped portion with the measured values ​​of the observation window during the overlapped portion to perform long-term real-time monitoring, thereby determining at least one prediction output for use at a next moment, The determination is based on the predicted reference signal received power of the at least one prediction output being higher than a threshold.

2. The device according to claim 1, wherein: The at least one observation window comprises at least one sliding observation window identifying a reference signal received power, and wherein the at least one prediction window comprises at least one sliding prediction window.

3. The device as claimed in claim 2, wherein: The identifying and comparing uses network configuration information from a communication network, including a channel state information reference signal associated with an overlapping portion of the observation window.

4. The device as claimed in claim 3, wherein: The network configuration information includes an indication of a size of the at least one sliding prediction window and provides a configuration to the apparatus for predicting one of the downlink receive beam pair or the downlink transmit beam pair as a new serving beam.

5. The device according to claim 3, wherein: The network configuration information provides configuration to the apparatus to predict a downlink transmit beam identifier, one of a subset of downlink transmit beam identifiers, or one of the downlink transmit beam identifiers as a new serving beam.

6. The device according to claim 1, wherein: The prediction window overlaps with at least one next observation window over the time period, and The comparing comprises comparing the beam prediction outputs of more than one observation window with the prediction window during the overlapping portion to perform the long-term real-time monitoring.

7. The device of claim 1, wherein: The at least one non-transitory memory stores instructions that are executed by the at least one processor to cause the apparatus to at least: For each prediction in the at least one prediction output, obtain at least one downlink transmit beam identification reference signal received power or at least one downlink transmit and receive beam pair identification reference signal received power provided by the machine learning determination.

8. The device according to claim 7, wherein: The machine learning determination is to determine whether a predicted new serving downlink transmit beam or a downlink transmit and receive beam pair identifies whether the beam pair will be configured for at least one of a physical uplink shared channel, a physical uplink control channel, a physical downlink shared channel, or a physical downlink control channel.

9. The device according to claim 7, wherein: At least one of the downlink receive beam or the downlink transmit beam pair is from at least one of the downlink receive beam pair observation window or the downlink transmit beam pair observation window.

10. The device of claim 1, wherein: The reference signal received power measurement is provided from time t to time tN, wherein the determining the first prediction window overlap is during time t to t+K, wherein the prediction of the first prediction window with the predicted output is during time t to t+L.

11. The device of claim 1, wherein: The predicted beam measurement sequence is associated with a sliding window based on reference signal received power measurements.

12. The device of claim 1, wherein: The at least one non-transitory memory stores instructions that are executed by the at least one processor to cause the apparatus to at least: Information is received from the communication network, comprising an indication of a channel state information reference signal according to an overlapping portion of at least one of more than one sliding prediction time windows or at least one simultaneous observation time window.

13. The device of claim 12, wherein: The at least one non-transitory memory stores instructions that are executed by the at least one processor to cause the apparatus to at least: Perform training and run inference to obtain a set of reference signal received power measurements for beam predictions.

14. The device of claim 1, wherein: The at least one predicted output is based on measuring a layer 1 reference signal received power measurement.

15. The device of claim 14, wherein: The at least one prediction output is based on the measurement layer 1 reference signal and auxiliary information, including beam identification, user equipment location, and line-of-sight or non-line-of-sight indication for use as input to a neural network-based AI / ML artificial intelligence / machine learning model.

16. The device of claim 1, wherein: The prediction sequence is the reference signal received power over time.

17. The device of claim 1, wherein: Comparing the beam prediction output with the measurements of the overlapped portion is performed using a predicted sequence of beam measurements from the at least one observation window during time t to t+M.

18. A method for communication, comprising: predicting a beam prediction output using a prediction sequence of beam measurement outputs for at least one observation window associated with at least one prediction window providing a reference signal received power measurement, The beam prediction output includes a Top-1 or Top-K beam identifier or a predicted reference signal received power of a beam subset associated with the beam prediction output; identifying that a prediction window of the at least one prediction window overlaps at least a portion of an observation window of the at least one observation window over a period of time; and comparing the beam prediction output during the overlapped portion with the measured values ​​of the observation window during the overlapped portion to perform long-term real-time monitoring, thereby determining at least one prediction output for use at a next moment, The determination is based on the predicted reference signal received power of the at least one prediction output being higher than a threshold.

19. The method of claim 18, wherein: The at least one observation window comprises at least one sliding observation window identifying a reference signal received power, and wherein the at least one prediction window comprises at least one sliding prediction window.

20. The method of claim 19, wherein: The identifying and comparing uses network configuration information from a communications network including a channel state information reference signal associated with an overlapping portion of the observation window.

21. The method of claim 20, wherein: The network configuration information includes an indication of the size of the at least one sliding prediction window and provides a configuration for the apparatus to predict one of the downlink receive beam pair or the downlink transmit beam pair as a new serving beam.

22. The method of claim 20, wherein: The network configuration information configures the device to predict a downlink transmit beam identifier, one of a subset of downlink transmit beam identifiers, or one of the downlink transmit beam identifiers as a new serving beam.

23. The method of claim 18, wherein: The prediction window overlaps with at least one next observation window over the time period, and Wherein, the comparing comprises comparing the beam prediction output of more than one observation window with the prediction window during the overlapping portion to perform the long-term real-time monitoring.

24. The method of claim 18, comprising: For each prediction of at least one prediction output, at least one downlink transmit beam identification reference signal received power or at least one downlink transmit and receive beam pair identification reference signal received power provided by machine learning determination is obtained.

25. The method of claim 24, wherein: The machine learning determination is to determine whether the predicted new serving downlink transmit beam or downlink transmit and receive beam pair identification beam pair will be configured as at least one of a physical uplink shared channel, a physical uplink control channel, a physical downlink shared channel or a physical downlink control channel.

26. The method of claim 24, wherein: At least one of the downlink receive beam or the downlink transmit beam pair is from at least one of a downlink receive beam pair observation window or a downlink transmit beam pair observation window.

27. The method of claim 18, wherein: The reference signal received power measurement is provided from time t to time tN, wherein the determining of the first prediction window overlap is during time t to t+K, wherein the prediction of the first prediction window with the predicted output is during time t to t+L.

28. The method of claim 18, wherein: The predicted sequence of beam measurements is associated with a sliding window based on reference signal received power measurements.

29. The method of claim 18, comprising: Information is received from the communication network, comprising an indication of a channel state information reference signal according to an overlapping portion of at least one of the more than one sliding prediction time windows or at least one simultaneous observation time window.

30. The method of claim 29, comprising: Perform training and run inference to obtain a set of reference signal received power measurements for beam predictions.

31. The method of claim 18, wherein: The at least one predicted output is based on measuring a layer 1 reference signal received power measurement.

32. The method of claim 31, wherein: The at least one prediction output is based on a layer 1 reference signal and auxiliary information including beam identification, user equipment location and line-of-sight or non-line-of-sight indication for use as input to a neural network-based AI / ML artificial intelligence / machine learning model.

33. The method of claim 18, wherein: The prediction sequence is the reference signal received power over time.

34. The method of claim 18, wherein: The comparing of the beam prediction output with the measurements of the overlapped portion is performed using a predicted sequence of beam measurements from the at least one observation window during time t to t+M.

35. A computer program product comprising instructions for causing an apparatus to perform at least the method according to any one of claims 18 to 34.

36. A computer readable medium comprising instructions for causing an apparatus to perform at least the method according to any one of claims 18 to 34.

Images