Optimizing model performance of user equipments (UEs) based on collective intelligence at base stations

Abstract

Embodiments of the present disclosure disclose optimizing model performance of a UE based on collective intelligence at a base station. The base station transmits a first segment of CSI-RS to the plurality of UEs via a DL channel corresponding to each of the plurality of UEs in the area. The base station receives a compressed CSI feedback from each of the plurality of UEs, the compressed CSI feedback including the second segment of the predicted CSI-RS, and the one or more CSI parameters. The base station reconstructs a CSI-RS associated with each of the plurality of UEs based on the CSI feedback received from the corresponding UE and the first segment. The base station identifies one or more UEs from the plurality of UEs that provide the optimized CSI based on a predefined threshold using the reconstructed CSI-RS. Thereafter, the base station sends assistance information and model parameters associated with one or more UEs to each of the plurality of UEs for training the corresponding model. Thus, the present disclosure improves the KPI of the UE based on the model parameters and assistance information of the best performance UE (s).

Description

Cross-reference to related applications

[0001] This application claims the benefit of Indian Provisional Application No. 202341081613, filed on November 30, 2023, entitled “OPTIMIZING MODEL PERFORMANCE OFUSER EQUIPMENTS (UEs) BASED ON COLLECTIVE INTELLIGENCE AT BASE STATION”, the entire contents of which are expressly incorporated herein by reference. Technical Field

[0002] This disclosure generally relates to communication technologies, and more specifically to optimizing the model performance of a UE based on collective intelligence at a base station. Background Technology

[0003] In the 3rd Generation Partnership Project (3GPP), the Channel State Information Reference Signal (CSI-RS) is a reference signal used in the downlink (DL) direction of 5G New Radio (NR) for channel sounding purposes. It measures the characteristics of the wireless channel so that it can be used with correct modulation, code rate, beamforming, etc. In wireless communication, CSI represents the known channel properties of the communication link. CSI information describes how a signal propagates from the transmitter to the receiver and represents, for example, combined effects of scattering, fading, and power attenuation with distance. The transmitter is the base station, and the receiver is the user equipment (UE). Furthermore, CSI enables transmission to adapt to current channel conditions, which is crucial for achieving reliable communication with high data rates in multi-antenna systems.

[0004] The information disclosed in the Background section of this disclosure is intended only to enhance the understanding of the general background of this disclosure and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0005] Channel State Information (CSI) represents the state of a communication link from (multiple) transmitting sources to (multiple) receiving sources. Currently, base stations or next-generation Node Bs (gNBs) transmit CSI-RS periodically or aperiodically in the DL (Depth Link), and UEs use CSI-RS to measure CSI information such as rank, precoder matrix indicator, channel quality indicator, etc. In AI-ML-based CSI feedback enhancement use cases, all this CSI information is derived under the assumption of a common basis function to reduce overhead. However, a problem associated with CSI-RS signals is that UEs with the same capabilities in the same cell, reporting similar Reference Signal Received Power (RSRP) and ground truth (i.e., CSI-RS received from the gNB at the UE), often report different CSI feedbacks. This is possible because of differences in the training and / or inference algorithms used at the UE. Therefore, CSI prediction key performance indicators (KPIs) are affected, and this directly impacts the "CSI compression" KPI. Furthermore, if intermediate KPIs in CSI prediction are affected due to incorrect model parameters or UE training algorithms, the final KPI will also be affected. In light of the above discussion, it is necessary to optimize the UE model performance based on collective intelligence at the base station in order to overcome the aforementioned problems.

[0006] In one embodiment, a base station is disclosed. The base station is configured to: transmit a first segment of a Channel State Information Reference Signal (CSI-RS) to a plurality of UEs via a downlink physical channel corresponding to each UE in a region. The base station is configured to: receive compressed CSI feedback from each of the plurality of UEs, the compressed CSI feedback including a second segment of a predicted CSI-RS and one or more CSI parameters, wherein the compressed CSI feedback is predicted based on the first segment of the CSI-RS using a model associated with each of the plurality of UEs. The base station is configured to reconstruct the CSI-RS associated with each of the plurality of UEs based on the CSI feedback received from the corresponding UE and the first segment. The base station is configured to: identify one or more UEs from the plurality of UEs that provide optimized CSI using the reconstructed CSI-RS, based on a predefined threshold. Thereafter, the base station is configured to transmit auxiliary information and at least one of the parameters of a model associated with the one or more UEs to each of the plurality of UEs, excluding the one or more UEs, for training the corresponding model.

[0007] In another embodiment, a method is disclosed. The method includes transmitting a first segment of a Channel State Information Reference Signal (CSI-RS) to a plurality of UEs via a downlink physical channel corresponding to each UE in a region. The method includes receiving compressed CSI feedback from each UE, the compressed CSI feedback including a second segment of a predicted CSI-RS and one or more CSI parameters, wherein the compressed CSI feedback is predicted based on the first segment of the CSI-RS using a model associated with each UE. The method includes reconstructing the CSI-RS associated with each UE based on the CSI feedback received from the corresponding UE and the first segment. The method includes identifying one or more UEs providing optimized CSI from the plurality of UEs using the reconstructed CSI-RS, based on a predefined threshold. The method includes transmitting auxiliary information and at least one of the parameters of a model associated with the one or more UEs to each UE in the plurality of UEs, for training the corresponding model.

[0008] In another embodiment, a non-transitory computer-readable medium is disclosed. This non-transitory computer-readable medium is configured to transmit a first segment of a Channel State Information Reference Signal (CSI-RS) to a plurality of UEs via a downlink physical channel corresponding to each UE in a region. The non-transitory computer-readable medium is configured to receive compressed CSI feedback from each of the plurality of UEs, the compressed CSI feedback including a second segment of a predicted CSI-RS and one or more CSI parameters, wherein the compressed CSI feedback is predicted based on the first segment of the CSI-RS using a model associated with each of the plurality of UEs. The non-transitory computer-readable medium is configured to reconstruct the CSI-RS associated with each of the plurality of UEs based on the CSI feedback received from the corresponding UE and the first segment. The non-transitory computer-readable medium is configured to identify one or more UEs providing optimized CSI from the plurality of UEs using the reconstructed CSI-RS, based on a predefined threshold. Subsequently, the non-transitory computer reading medium is configured to send auxiliary information and at least one of the parameters of a model associated with one or more UEs to each of the plurality of UEs for training the corresponding model.

[0009] The above description of the invention is for illustrative purposes only and is not intended to limit the scope in any way. Other aspects, embodiments, and features will become apparent from the accompanying drawings and the following detailed description, in addition to the illustrative aspects, embodiments, and features described above. Attached Figure Description

[0010] The accompanying drawings, which are incorporated in and constitute a part of this disclosure, illustrate exemplary embodiments and, together with the specification, serve to explain the disclosed principles. The same numerals are used throughout the drawings to refer to similar features and components. Some embodiments of devices and / or methods according to this subject matter will now be described by way of example and with reference to the accompanying drawings, in which:

[0011] Figure 1 The illustration depicts an exemplary environment for optimizing the model performance of a UE based on collective intelligence at a base station, according to an embodiment of the present disclosure.

[0012] Figure 2 A detailed block diagram of a base station for optimizing the model performance of a UE based on collective intelligence at the base station, according to an embodiment of the present disclosure, is shown.

[0013] Figure 3 The illustration shows a sequence diagram of an embodiment according to the present disclosure, illustrating interaction with a base station and a UE to optimize the model performance of the UE based on collective intelligence at the base station;

[0014] Figure 4 The illustration shows a flowchart of an embodiment according to the present disclosure, illustrating an exemplary method for optimizing the model performance of a UE based on collective intelligence at the base station; and

[0015] Figure 5 A block diagram of an exemplary computer system for implementing embodiments consistent with this disclosure is illustrated.

[0016] Those skilled in the art will understand that any block diagram herein represents a conceptual diagram of an illustrative system embodying the principles of the subject matter. Similarly, it will be understood that any flowchart, diagrammatic flowchart, state transition diagram, pseudocode, etc., represents various processes that can be substantially represented in a computer-readable medium and executed by a computer or processor, whether or not such computer or processor is explicitly shown. Detailed Implementation

[0017] In this document, the term "exemplary" is used to mean "as an example, instance, or illustration." Any embodiment or implementation of the subject matter described herein as "exemplary" is not necessarily to be construed as more preferred or advantageous than other embodiments.

[0018] While this disclosure is readily available in various modifications and alternatives, specific embodiments thereof have been illustrated by way of example in the accompanying drawings and will be described in detail below. However, it should be understood that this is not intended to limit this disclosure to the specific forms disclosed; rather, this disclosure is intended to cover all modifications, equivalents, and alternatives falling within the spirit and scope of this disclosure.

[0019] The terms “comprises,” “comprising,” or any other variations thereof are intended to cover non-exclusive inclusion, such that an arrangement, apparatus, or method that includes a list of components or steps may include not only those listed components or steps, but also other components or steps not expressly listed or inherent to such an arrangement, apparatus, or method. In other words, without further constraints, one or more elements in an apparatus, system, or device that begins with “comprises … a (including a …)” do not exclude the presence of other elements or additional elements in the apparatus, system, or device.

[0020] In the following detailed description of embodiments of the present disclosure, reference is made to the accompanying drawings, which form part of this disclosure, illustrating specific embodiments in which the present disclosure may be practiced by way of illustration. These embodiments are described in sufficient detail to enable those skilled in the art to practice the present disclosure, and it should be understood that other embodiments may be utilized and changes may be made without departing from the scope of the present disclosure. Therefore, the following description should not be considered limiting.

[0021] It should be noted that, for ease of interpretation, this disclosure uses the terms and names defined in the 3GPP standards. More specifically, the terms “Channel State Information Reference Signal (CSI-RS)”, “Reference Signal (RS)”, “User Equipment (UE)”, “Base Station”, “Next Generation Node B (gNB)”, “Cell”, “New Radio (NR)”, “gNB Distributed Unit (DU)”, “gNB Centralized Unit (CU)”, “Key Performance Indicators (KPIs)”, etc., shall be interpreted in accordance with the provisions of the 3GPP standards.

[0022] As used herein, the term "model parameters" refers to information used to optimize the model performance of a UE based on collective intelligence at the base station, using model parameters and auxiliary information. More specifically, a first segment of the CSI-RS is transmitted to multiple UEs in the area via a downlink channel. In another embodiment, compressed CSI feedback is received, including a second segment of the predicted CSI-RS and CSI parameters. In one embodiment, the compressed CSI feedback is predicted using an artificial intelligence (AI) / machine learning (ML) model associated with multiple UEs. In another embodiment, the CSI-RS associated with each of the multiple UEs is reconstructed based on the CSI feedback and the first segment. In another embodiment, the UE providing optimized CSI is identified from the multiple UEs using the reconstructed CSI-RS based on a predefined threshold. In another embodiment, auxiliary information and model parameters associated with the UE are sent to the multiple UEs for training their corresponding models. The auxiliary information and model parameters of the UE providing optimized CSI are used to train other UEs in the area to improve the KPIs of all multiple UEs. In one embodiment, reference... Figures 1-5 A detailed explanation of how to optimize UE model performance using model parameters and auxiliary information based on collective intelligence at the base station.

[0023] Figure 1 The illustration depicts an exemplary environment 100 for optimizing UE model performance based on collective intelligence at the base station using model parameters and auxiliary information. The exemplary environment 100 includes a base station 101 and UEs 1021, 1022, ..., 102. n (In this document, it is alternatively referred to as multiple UEs 102). In one embodiment, base station 101 may include, but is not limited to, gNB, gNB-DU, gNB-CU, next-generation evolved NodeB (ng-eNB), and evolved NodeB (eNB). In one embodiment, multiple UEs 102 may include, but is not limited to, mobile phones, smartphones, etc. In one embodiment, base station 101 may interact with multiple UEs 102 to optimize the model performance of the UEs based on collective intelligence at base station 101.

[0024] In one embodiment, base station 101 transmits a first segment of a Channel State Information Reference Signal (CSI-RS) to the plurality of UEs 102 via a downlink physical channel corresponding to each UE 102 in the area. In one embodiment, the first segment may be a percentage value of the CSI-RS transmitted to the plurality of UEs 102. For example, the percentage value may be 60%, 40%, 30%, 90%, etc. In one embodiment, the first segment of the CSI-RS is transmitted to the plurality of UEs 102 periodically or aperiodically. Furthermore, base station 101 receives compressed CSI feedback from each of the plurality of UEs 102, which includes a predicted second segment of the CSI-RS, ground truth, and one or more CSI parameters. The compressed CSI feedback is predicted based on the first segment of the CSI-RS using a model associated with each of the plurality of UEs 102. In one embodiment, the compressed CSI feedback is received via a Physical Uplink Control Channel (PUCCH). In one embodiment, the compressed CSI feedback also includes key performance indicators (KPIs), auxiliary information, and model parameters corresponding to each of the plurality of UEs 102. In one embodiment, base station 101 is configured to reconstruct the CSI-RS associated with each of the plurality of UEs based on CSI feedback received from the corresponding UE and a first segment. Furthermore, base station 101 uses the reconstructed CSI-RS to identify one or more UEs from the plurality of UEs 102 that provide optimized CSI, based on a predefined threshold. In one embodiment, base station 101 compares the reconstructed CSI-RS associated with each of the plurality of UEs 102 with the predefined threshold. Furthermore, base station 101 identifies one or more UEs from the plurality of UEs 102 that provide optimized CSI based on this comparison. In one embodiment, the reconstructed CSI-RS of one or more UEs is greater than or equal to the predefined threshold. Subsequently, the base station sends auxiliary information and at least one of the parameters of a model associated with the one or more UEs to each of the plurality of UEs 102, excluding the one or more UEs, for training the corresponding model. In one embodiment, the auxiliary information for the one or more UEs includes spatial frequency, delay-Doppler, and domain compression parameters.

[0025] Figure 2 A detailed block diagram of a base station for optimizing the model performance of a UE based on collective intelligence at the base station, according to an embodiment of the present disclosure, is shown.

[0026] In one embodiment, such as Figure 2As shown, base station 101 may include processor 201, I / O interface 202, and memory 203. In some embodiments, memory 203 may be communicatively coupled to processor 201. Memory 203 stores instructions executable by processor 201, which, when executed, cause base station 101 to optimize UE model performance based on collective intelligence at the base station, as disclosed in this disclosure. In one embodiment, as... Figure 2 As shown, memory 203 may include one or more modules 204 and data 210. The one or more modules 204 may be configured to use data 210 to perform the processes of this disclosure to optimize the model performance of the UE based on collective intelligence at the base station. In one embodiment, each of the one or more modules 204 may be a hardware unit that may be external to memory 203 and coupled to base station 101.

[0027] This document describes in detail the data 210 in the memory 203 of the base station 101 and one or more modules 204.

[0028] In one implementation, one or more modules 204 may include, but are not limited to, a transmitting module 205, a receiving module 206, a reconfiguration module 207, an identification module 208, and one or more other modules 209 associated with the base station 101.

[0029] In one embodiment, the data 210 in the memory 203 may include signal data 211, feedback data 212, reconstructed data 213, model parameters 214, auxiliary information 215, and other data 216 associated with the base station 101.

[0030] In one embodiment, data 210 in memory 203 may be processed by one or more modules 204 of base station 101. In one embodiment, one or more modules 204 may be implemented as dedicated units, and when implemented in this way, the modules may be configured with the functions defined in this disclosure to produce novel hardware. As used herein, the term module may refer to application-specific integrated circuits (ASICs), electronic circuits, field-programmable gate arrays (FPGAs), programmable system-on-chip (PSoCs), combinational logic circuits, and / or other suitable components that provide the functions described above.

[0031] One or more modules 204 of this disclosure are used to optimize the model performance of the UE based on collective intelligence at the base station using model parameters and auxiliary information. One or more modules 204 may also include other modules 209 to perform various miscellaneous functions of the base station 101. It should be understood that such modules can be represented as a single module or a combination of different modules. One or more modules 204 and data 210 can be implemented in any base station for optimizing the model performance of the UE based on collective intelligence at the base station using model parameters and auxiliary information.

[0032] Signal data 211 may include information about a first segment of CSI-RS. In one embodiment, the first segment of CSI-RS is sent to multiple UEs 102.

[0033] Feedback data 212 may include information about compressed CSI feedback, which includes a second segment of the predicted CSI-RS and one or more CSI parameters.

[0034] The reconstructed data 213 may include information about the reconstructed CSI-RS associated with each of the plurality of UEs 102.

[0035] Model parameters 214 can include information such as the number of layers in the learning network and the number of nodes for each layer.

[0036] Auxiliary information 215 may include information such as spatial frequency, delay-Doppler, and domain compression parameters.

[0037] Other data 216 may store data generated by modules used to perform various functions of base station 101, including temporary data and temporary files.

[0038] In one embodiment, the transmitting module 205 of base station 101 is configured to transmit a first segment of CSI-RS to a plurality of UEs 102 located in an area. In one embodiment, the plurality of UEs 102 may reside in the same cell. CSI-RS is a reference signal used for channel sensing purposes in the downlink direction of 5G New Radio (NR) and is used to measure the characteristics of the radio channel so that it can use the correct modulation, code rate, beamforming, etc. The receiving module 206 of base station 101 is configured to receive compressed CSI feedback from each of the plurality of UEs 102. The compressed feedback may include, but is not limited to, a second segment of the predicted CSI-RS, ground truth, and one or more CSI parameters. In one embodiment, the compressed CSI feedback also includes KPIs, auxiliary information, and model parameters corresponding to each of the plurality of UEs 102. In one embodiment, each of the plurality of UEs 102 is configured to report important model parameters and their values, as well as intermediate KPIs of the predicted CSI-RS, to base station 101. In one embodiment, one or more parameters may include, but are not limited to, frequency, bandwidth, etc. In one embodiment, the second segment may be a percentage value, such as 40%, 60%, 30%, etc. In one embodiment, compressed CSI feedback is predicted based on a first segment of CSI-RS by a model associated with each of the plurality of UEs 102. In one embodiment, the model may be trained separately on the base station 101 side and on the plurality of UEs 102 side. In one embodiment, the CSI generation portion on the plurality of UEs 102 side and the CSI-RS reconstruction portion on the base station 101 side are trained respectively on the plurality of UEs 102 side and the base station 101 side. In one embodiment, separate training includes sequential training starting from training on the plurality of UEs 102 side. In another embodiment, separate training includes sequential training starting from training on the base station 101 side. In yet another embodiment, separate training includes parallel training on the plurality of UEs 102 side and on the base station 101 side. Furthermore, the reconstruction module 207 of the base station 101 is configured to reconstruct the CSI-RS associated with each of the plurality of UEs 102 based on the CSI feedback and the first segment received from the corresponding UE. In one embodiment, the identification module 208 uses the reconstructed CSI-RS to identify one or more UEs from the plurality of UEs 102 that provide optimized CSI based on a predefined threshold. In one embodiment, the identification module 208 identifies one or more UEs with the best intermediate KPI. In another embodiment, the identification module 208 is configured to compare the reconstructed CSI-RS associated with each of the plurality of UEs 102 with a predefined threshold. After comparing whether the reconstructed CSI-RS of one or more UEs is greater than or equal to the predefined threshold, the identification module 208 identifies one or more UEs that provide optimized CSI. In one embodiment, the predefined threshold is the ideal KPI.Subsequently, the transmission module 205 of base station 101 is configured to transmit at least one of auxiliary information and parameters (alternatively referred to as model parameters) of a model associated with the one or more UEs to each of the plurality of UEs 102, excluding one or more UEs. Upon receiving, the plurality of UEs 102, excluding one or more UEs, use the auxiliary information and model parameters to train their corresponding models to provide optimized CSI.

[0039] Figure 3 The illustration shows a sequence diagram of an embodiment according to the present disclosure, illustrating interaction with a base station and a UE to optimize the model performance of the UE based on collective intelligence at the base station. Figure 3 The diagram illustrates UE 301 with optimal KPIs, multiple UEs 302, a base station distributed unit 303 (optionally referred to as gNB-DU), and a base station centralized unit 304 (optionally referred to as gNB-CU). Initially, as... Figure 3As shown in step 305, UE 301 and multiple UEs 302 are in a Radio Resource Control (RRC) connection state, receiving CSI-RS in the downlink. In one embodiment, an RRC connection is established, and the base station has configured all the parameters required for communication between the UEs. In step 306, the RRC reconfiguration between gNB-CU 304 and the multiple UEs 302 ensures that all configuration-related details are provided to the multiple UEs. In one embodiment, UE 301 is part of the multiple UEs 302. In one embodiment, the RRC reconfiguration enables AI-ML-based CSI feedback enhancement use cases. Furthermore, in step 307, gNB-DU sends the percentage of CSI-RS to the multiple UEs 302 and UE 301. In step 308, UE 301 and the multiple UEs 302 utilize an AI-ML model and predict CSI-RS, and generate CSI feedback based on the transmitted ground truth CSI-RS. At step 309, UE 301 and multiple UEs 302 send compressed CSI feedback, including KPIs and model parameters, to gNB-DU 303 via PUCCH. In one embodiment, UE 301 and multiple UEs 302 report important model parameters, their values, and intermediate KPIs for the CSI prediction sub-use case to the network / gNB-DU 303. At step 310, gNB-DU identifies the UE 301 that provides optimized CSI based on a predefined threshold and the reconstructed CSI-RS. In one embodiment, gNB-DU monitors the intermediate KPIs reported by each UE and identifies the UE with the best intermediate KPI. Furthermore, in one embodiment, gNB-DU 303 sends a UE context modification request (UE Context Modification Request) message to gNB-CU 304 to initiate a reconfiguration of the parameters of the AI-ML model for CSI feedback enhancement. In another embodiment, gNB-CU 304 sends a UE context modification request (UE Context Modification Request) message to gNB-DU 303. Subsequently, at step 311, gNB-CU 304 sends auxiliary information and model parameters of UE 301 to multiple UEs 302 to train their corresponding models. In one embodiment, after receiving the auxiliary information and model parameters, the multiple UEs 302 send an RRC reconfiguration confirmation (RRC) to gNB-CU 304. In another embodiment, the multiple UEs 302 that receive the optimal model parameters and auxiliary information from gNB-CU 304 use them for training on their UE side.

[0040] Figure 4 A flowchart illustrating an embodiment of the present disclosure is shown, demonstrating an exemplary method for optimizing the model performance of a UE based on collective intelligence at a base station.

[0041] like Figure 4As shown, method 400 may include one or more blocks for performing procedures in base station 101. Method 400 may be described in the general context of computer-executable instructions. Typically, computer-executable instructions may include routines, programs, objects, components, data structures, procedures, modules, and functions that perform a particular function or implement a particular abstract data type.

[0042] The order in which method 400 is described is not intended to be construed as limiting, and any number of described method blocks can be combined in any order to implement the method. Furthermore, individual blocks can be removed from the method without departing from the scope of the subject matter herein. Moreover, the method can be implemented in any suitable hardware, software, firmware, or a combination thereof.

[0043] At box 401, the first segment of a Channel State Information Reference Signal (CSI-RS) is transmitted to multiple UEs via a downlink physical channel corresponding to each of the multiple user equipments (UEs) in the area.

[0044] At box 402, compressed CSI feedback is received from each of the plurality of UEs, the compressed CSI feedback including a second segment of the predicted CSI-RS and one or more CSI parameters, wherein the compressed CSI feedback is predicted based on the first segment of the CSI-RS using a model associated with each of the plurality of UEs.

[0045] At box 403, the CSI-RS associated with each of the multiple UEs is reconstructed based on the CSI feedback received from the corresponding UE and the first segment.

[0046] At box 404, the reconstructed CSI-RS is used to identify one or more UEs from multiple UEs that provide optimized CSI, based on a predefined threshold.

[0047] At box 405, auxiliary information and at least one of the parameters of the model associated with the one or more UEs are sent to each of the multiple UEs, excluding one or more UEs, for training the corresponding model. Computing System

[0048] Figure 5A block diagram of an exemplary computer system 500 for implementing embodiments consistent with this disclosure is illustrated. In one embodiment, the computer system 500 may be used to implement a base station 101. The computer system 500 may include a central processing unit (“CPU” or “processor”) 502. The processor 502 may include at least one data processor for performing processes in a virtual memory area network. The processor 502 may include dedicated processing units such as an integrated system (bus) controller, a memory management control unit, a floating-point unit, a graphics processing unit, a digital signal processing unit, etc.

[0049] The processor 502 can be configured to communicate with one or more input / output (I / O) devices 509 and 510 via I / O interface 501. I / O interface 501 can employ communication protocols / methods such as, but not limited to: audio, analog, digital, mono, RCA, stereo, IEEE-1394, serial bus, Universal Serial Bus (USB), infrared, PS / 2, BNC, coaxial, component, composite, digital video interface (DVI), high-definition multimedia interface (HDMI), RF antenna, S-Video, VGA, IEEE 802.n / b / g / n / x, Bluetooth, cellular (e.g., Code Division Multiple Access (CDMA), High Speed ​​Packet Access (HSPA+), Global System for Mobile Communications (GSM), Long Term Evolution (LTE), WiMax, etc.).

[0050] Using I / O interface 501, computer system 500 can communicate with one or more I / O devices 509 and 510. For example, input device 509 can be an antenna, keyboard, mouse, joystick, (infrared) remote control, camera, card reader, fax machine, dongle, biometric reader, microphone, touchscreen, touchpad, trackball, stylus, scanner, storage device, transceiver, video device / source, etc. Output device 510 can be a printer, fax machine, video display (e.g., cathode ray tube (CRT), liquid crystal display (LCD), light-emitting diode (LED), plasma display, plasma display panel (PDP), organic light-emitting diode display (OLED), etc.), audio speakers, etc.

[0051] In some embodiments, computer system 500 may include base station 101. Processor 502 may be configured to communicate with communication network 511 via network interface 503. Network interface 503 may communicate with communication network 511. Network interface 503 may employ connectivity protocols, including but not limited to: direct connection, Ethernet (e.g., twisted pair 10 / 100 / 1000 Base T), Transmission Control Protocol / Internet Protocol (TCP / IP), Token Ring, IEEE 802.11a / b / g / n / x, etc. Communication network 511 may include, but is not limited to: direct interconnect, local area network (LAN), wide area network (WAN), wireless network (e.g., using Wireless Application Protocol), Internet, etc. Using network interface 503 and communication network 511, computer system 500 may utilize imagery to communicate with multiple UEs 512 to optimize UE model performance based on collective intelligence at the base station. Network interface 503 can use connection protocols, including but not limited to: direct connection, Ethernet (e.g., twisted pair 10 / 100 / 1000 Base T), Transmission Control Protocol / Internet Protocol (TCP / IP), Token Ring, IEEE 802.11a / b / g / n / x, etc.

[0052] Communication networks 511 include, but are not limited to: direct interconnection, e-commerce networks, peer-to-peer (P2P) networks, local area networks (LANs), wide area networks (WANs), wireless networks (e.g., using Wireless Application Protocol), the Internet, Wi-Fi, etc. The first and second networks can be private or shared networks, representing the association of different types of networks communicating with each other using various protocols (e.g., Hypertext Transfer Protocol (HTTP), Transmission Control Protocol / Internet Protocol (TCP / IP), Wireless Application Protocol (WAP), etc.). Furthermore, the first and second networks can include various network devices, including routers, bridges, servers, computing devices, storage devices, etc.

[0053] In some embodiments, the processor 502 may be configured to connect to the memory 505 (e.g., via the storage interface 504) Figure 5 (RAM, ROM, etc., not shown) Communication. Storage interface 504 can be connected to memory 505, including but not limited to memory drives, removable disk drives, etc., which use connection protocols such as Serial Advanced Technology Attachment (SATA), Integrated Drive Electronics (IDE), IEEE-1394, Universal Serial Bus (USB), Fibre Channel, Small Computer System Interface (SCSI), etc. Memory drives may also include magnetic drums, disk drives, magneto-optical drives, optical disc drives, redundant arrays of independent disks (RAID), solid-state storage devices, solid-state drives, etc.

[0054] Memory 505 may store a collection of program or database components, including but not limited to user interface 506, operating system 507, etc. In some embodiments, computer system 500 may store user / application data, such as data, variables, records, etc., described in this disclosure. Such a database may be implemented as a fault-tolerant, relational, scalable, and secure database, such as Oracle® or Sybase®.

[0055] Operating system 507 facilitates resource management and operation of computer system 500. Examples of operating systems include, but are not limited to: APPLE MACINTOSH® OS X, UNIX®, and UNIX-like system distributions (e.g., BERKELEY SOFTWARE EDISTRIBUTION). TM (BSD), FreeBSD TM NETBSD TM OPENBSD TM etc.), LINUX DISTRIBUTIONS TM (e.g., REDHAT) TM UBUNTU TM KUBUNTU TM etc.), IBM TM OS / 2, Microsoft TM WINDOWS TM (XP) TM VISTA TM / 7 / 8, 10, etc.), APPLE® IOS TM Google® Android TM BLACKBERRY® OS, etc.

[0056] In some embodiments, computer system 500 may implement program components stored in web browser 508. Web browser 508 may be a hypertext viewing application, such as Microsoft Internet Explorer, Google Chrome, Mozilla Firefox, Apple Safari, etc. Secure web browsing may be provided using Hypertext Transfer Protocol Security (HTTPS), Secure Sockets Layer (SSL), Transport Layer Security (TLS), etc. Web browser 508 may utilize facilities such as AJAX, DHTML, Adobe Flash, JavaScript, Java, Application Programming Interface (API), etc. In some embodiments, computer system 500 may implement program components stored in mail server. Mail server may be an Internet mail server such as Microsoft Exchange. Mail server may utilize facilities such as ASP, ActiveX, ANSI C++ / C#, Microsoft .NET, Common Gateway Interface (CGI) scripts, Java, JavaScript, Perl, PHP, Python, WebObjects, etc. Mail server may utilize communication protocols such as Internet Message Access Protocol (IMAP), Message Application Programming Interface (MAPI), Microsoft Exchange, Post Office Protocol (POP), Simple Mail Transfer Protocol (SMTP), etc. In some embodiments, the computer system 500 may implement a program component for storing an email client. The email client may be an email viewing application, such as Apple Mail, Microsoft Entourage, Microsoft Outlook, Mozilla Thunderbird, etc.

[0057] refer to Figure 4 The disclosed methods or references Figures 1-3 One or more operations of the explained base station 101 can be implemented using software that includes computer-executable instructions stored on one or more computer-readable media (e.g., non-transitory computer-readable media, such as one or more optical discs, volatile memory components (e.g., DRAM or SRAM), or non-volatile memory or storage components (e.g., hard disk drives or solid-state non-volatile memory components, such as flash memory components), and executed on a computer (e.g., any suitable computer, such as a laptop, netbook, webbook, tablet computing device, smartphone, or other mobile computing device). Such software can, for example, be executed on a single local computer.

[0058] Furthermore, one or more computer-readable storage media can be used to implement embodiments consistent with this disclosure. A computer-readable storage medium refers to any type of physical memory on which processor-readable information or data can be stored. Therefore, a computer-readable storage medium can store instructions executable by one or more processors, including instructions for causing the processor to perform steps or stages consistent with the embodiments described herein. The term "computer-readable medium" should be understood to include tangible articles but exclude carrier waves and transient signals, i.e., non-transient signals. Examples include random access memory (RAM), read-only memory (ROM), volatile memory, non-volatile memory, hard disk drives, CD (optical disc) ROMs, DVDs, flash drives, magnetic disks, and any other known physical storage media.

[0059] The various embodiments of this disclosure offer numerous advantages. Embodiments of this disclosure enable the improvement of KPIs for all UEs based on model parameters and auxiliary information from (multiple) optimal-performing UEs. Furthermore, this disclosure can be used for all model training types, such as joint training of a two-sided model on a single-side / entity basis (i.e., network-side or UE-side), separate training of a two-sided model on both the network-side and UE-side, and joint training. Moreover, the gNB-DU of this disclosure can consider more than one UE KPI when deriving optimal-performing model parameters. This disclosure provides predictions with reduced overhead, improved accuracy, and enhanced CSI feedback.

[0060] Those skilled in the art will understand that, generally, the terms used herein are “open” terms (e.g., the term “comprising” should be interpreted as “including but not limited to,” the term “having” should be interpreted as “having at least,” and the term “comprising” should be interpreted as including but not limited to, etc.). For example, to aid understanding, a particular embodiment may include the use of introductory phrases “at least one” and “one or more” to introduce a statement. However, the use of such phrases should not be construed as implying that a statement introduced by the indefinite article “a” or “an” limits any particular portion of the description containing such an introduced statement to containing only the disclosure of one such statement, even when the statement includes the introductory phrases “one or more” or “at least one” and indefinite articles (such as “a” and “an”) (e.g., “a” and / or “an” should generally be interpreted as meaning “at least one” or “one or more”); the same applies to the use of definite articles used to introduce such statements. Furthermore, even if a particular part of the description is explicitly stated, those skilled in the art will recognize that such a statement should generally be interpreted as indicating at least the number of statements (e.g., a bare statement of “two statements” without other modifiers generally indicates at least two statements or two or more statements).

[0061] While various aspects and embodiments have been disclosed herein, other aspects and embodiments will be apparent to those skilled in the art. The various aspects and implementations disclosed herein are for illustrative purposes only and are not intended to be limiting; the true scope and spirit are indicated by the appended claims.

Claims

1. A base station (101) configured as follows: The first segment of a Channel State Information Reference Signal (CSI-RS) is transmitted to the plurality of UEs (102) via a downlink physical channel corresponding to each of the plurality of UEs (102) in the area; Compressed CSI feedback is received from each of the plurality of UEs, the compressed CSI feedback including a second segment of a predicted CSI-RS and one or more CSI parameters, wherein the compressed CSI feedback is predicted based on the first segment of the CSI-RS using a model associated with each of the plurality of UEs; Based on the CSI feedback received from the corresponding UE and the first segment, reconstruct the CSI-RS associated with each of the plurality of UEs; Using the reconstructed CSI-RS, one or more UEs providing optimized CSI are identified from the plurality of UEs based on a predefined threshold; as well as Send auxiliary information and at least one of the parameters of the model associated with the one or more UEs to each of the plurality of UEs, excluding the one or more UEs, for training the corresponding model.

2. The base station (101) according to claim 1, wherein the first segment of the CSI-RS is periodically transmitted to the plurality of UEs (102).

3. The base station (101) according to claim 1, wherein the first segment of the CSI-RS is transmitted aperiodically to the plurality of UEs (102).

4. The base station (101) according to claim 1, wherein the compressed CSI feedback is received via the Physical Uplink Control Channel (PUCCH).

5. The base station (101) according to claim 1, wherein the compressed CSI feedback further includes: Key performance indicators (KPIs), auxiliary information, and model parameters corresponding to each of the plurality of UEs (102).

6. The base station (101) according to claim 1 is configured as follows: The reconstructed CSI-RS associated with each of the plurality of UEs (102) is compared with the predefined threshold; and Based on the comparison, one or more UEs (102) are identified from the plurality of UEs (102) for providing the optimized CSI, wherein the reconstructed CSI-RS of the one or more UEs is greater than or equal to the predefined threshold.

7. The base station (101) according to claim 1, wherein the auxiliary information of the one or more UEs includes spatial frequency, delay-Doppler, and domain compression parameters.

8. A method comprising: The base station (101) transmits the first segment of the Channel State Information Reference Signal (CSI-RS) to the plurality of UEs (102) via a downlink physical channel corresponding to each of the plurality of UEs (102) in the area; The base station (101) receives compressed CSI feedback from each of the plurality of UEs (102), the compressed CSI feedback including a second segment of a predicted CSI-RS and one or more CSI parameters, wherein the compressed CSI feedback is predicted based on the first segment of the CSI-RS using a model associated with each of the plurality of UEs (102); The base station (101) reconstructs the CSI-RS associated with each of the plurality of UEs (102) based on the CSI feedback received from the corresponding UE and the first segment; The base station (101) uses the reconstructed CSI-RS to identify one or more UEs (102) that provide optimized CSI from the plurality of UEs based on a predefined threshold; as well as The base station (101) sends auxiliary information and at least one of the parameters of the model associated with the one or more UEs to each of the plurality of UEs (102) other than the one or more UEs, for training the corresponding model.

9. The method of claim 8, wherein the first segment of the CSI-RS is periodically transmitted to the plurality of UEs (102).

10. The method of claim 8, wherein the first segment of the CSI-RS is transmitted aperiodically to the plurality of UEs (102).

11. The method of claim 8, wherein the compressed CSI feedback is received via the Physical Uplink Control Channel (PUCCH).

12. The method of claim 8, wherein the compressed CSI feedback further comprises: Key performance indicators (KPIs), auxiliary information, and model parameters corresponding to each of the plurality of UEs (102).

13. The method of claim 8, comprising: The base station (101) compares the reconstructed CSI-RS associated with each of the plurality of UEs (102) with the predefined threshold; as well as The base station (101) identifies one or more UEs (102) from the plurality of UEs (102) based on the comparison, wherein the reconstructed CSI-RS of the one or more UEs is greater than or equal to the predefined threshold.

14. The method of claim 8, wherein the auxiliary information of the one or more UEs includes spatial frequency, delay-Doppler, and domain compression parameters.

15. A non-transitory computer-readable medium comprising instructions for performing operations, the operations including: The first segment of a Channel State Information Reference Signal (CSI-RS) is transmitted to the plurality of UEs (102) via a downlink physical channel corresponding to each of the plurality of UEs (102) in the area; Compressed CSI feedback is received from each of the plurality of UEs (102), the compressed CSI feedback including a second segment of a predicted CSI-RS and one or more CSI parameters, wherein the compressed CSI feedback is predicted based on the first segment of the CSI-RS using a model associated with each of the plurality of UEs (102); Based on the CSI feedback received from the corresponding UE and the first segment, reconstruct the CSI-RS associated with each of the plurality of UEs (102); Using the reconstructed CSI-RS, one or more UEs providing optimized CSI are identified from the plurality of UEs (102) based on a predefined threshold; as well as Send auxiliary information and at least one of the parameters of the model associated with the one or more UEs to each of the plurality of UEs (102) other than the one or more UEs, for training the corresponding model.

Images