[go: up one dir, main page]

WO2025089995A1 - Appareil et procédé de mise à jour d'un nœud de réseau - Google Patents

Appareil et procédé de mise à jour d'un nœud de réseau Download PDF

Info

Publication number
WO2025089995A1
WO2025089995A1 PCT/SE2023/051059 SE2023051059W WO2025089995A1 WO 2025089995 A1 WO2025089995 A1 WO 2025089995A1 SE 2023051059 W SE2023051059 W SE 2023051059W WO 2025089995 A1 WO2025089995 A1 WO 2025089995A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
cells
network
model
tiles
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
PCT/SE2023/051059
Other languages
English (en)
Inventor
Tahar ZANOUDA
Ali SHIBLI
Rémi BOURGERIE
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Telefonaktiebolaget LM Ericsson AB
Original Assignee
Telefonaktiebolaget LM Ericsson AB
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Telefonaktiebolaget LM Ericsson AB filed Critical Telefonaktiebolaget LM Ericsson AB
Priority to PCT/SE2023/051059 priority Critical patent/WO2025089995A1/fr
Publication of WO2025089995A1 publication Critical patent/WO2025089995A1/fr
Pending legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/08—Configuration management of networks or network elements
    • H04L41/0803—Configuration setting
    • H04L41/0813—Configuration setting characterised by the conditions triggering a change of settings
    • H04L41/082—Configuration setting characterised by the conditions triggering a change of settings the condition being updates or upgrades of network functionality
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/02—Arrangements for optimising operational condition
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/08—Testing, supervising or monitoring using real traffic
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/12—Discovery or management of network topologies
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/147—Network analysis or design for predicting network behaviour
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/16—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using machine learning or artificial intelligence

Definitions

  • the invention relates to an apparatus for updating a network node, a corresponding method, a corresponding computer program, and a corresponding computer readable storage medium.
  • a reliable communication infrastructure depends on performance of software network technologies, and hardware network technologies used by mobile operators.
  • WO 2020121084 A1 a system and method for improving machine learning model performance in a communications network is provided.
  • WO 2018164618 A1 a method for predicting a performance indicator for a service in a network is provided.
  • An object of the invention is to improve network performance of a network node in a communication network.
  • an apparatus for updating a network node based on one or more recommended network updates is provided.
  • the network node covering one or more coverage areas.
  • the apparatus is configured to receive data from the network node.
  • the data relates to the one or more coverage areas.
  • the apparatus is configured to compute a current network performance based on the received data.
  • the apparatus is configured to determine one or more recommended network updates of the network node.
  • the one or more recommended network updates are based on the computed current network performance.
  • the apparatus is configured to update the network node based on the one or more recommended network updates.
  • the network node is updated based on recommended network updates.
  • the recommended network updates are based on a computed current network performance of the network node within coverage areas.
  • the current network performance is computed based on data related to the coverage areas covered by the network node. Hence, network performance of the network node is improved within all the coverage areas.
  • the data comprises, for each one or more of the coverage areas, one or more of: network performance data; network topology; signal strength measurement data; satellite imagery data; geographical data; historical traffic data.
  • the computation of the current network performance comprises one or more of: dividing each of the one or more coverage areas into one or more tiles; associating each of the one or more tiles to the data; processing one or more machine learning features for each of the tiles based on the association of each of the one or more tiles; training a first Machine Learning, ML, model to generate one or more estimated signal strength measurements using the one or more machine learning features; generating a radio coverage representation of the one or more coverage areas using the one or more estimated signal strength measurements.
  • processing of the one or more machine learning features for each of the one or more tiles comprises one or more of: extracting, for each of the one or more tiles, the data related to each of the one or more tiles; calculating, for each of the one or more tiles a unique geographical identifier; calculating, for each of the one or more tiles an angular direction of the data measurement’s location corresponding to each tile location; calculating, for each of the one or more tiles, a geographical distance; arranging one or more vectors, each of the one or more vectors corresponds to each of the one or more tiles; encoding each of the one or more vectors; and splitting the one or more machine learning features into one or more K-folds.
  • training the first ML model to impute estimated one or more signal strength measurements using the one or more ML features comprises one or more of: training the ML model using the features of K-1 folds of the one or more K folds; and evaluating the ML model using the features of one fold of the one or more K folds.
  • the determination of the one or more recommended network updates comprises one or more of: determining one or more coverage holes based on the computed current network performance; determining one or more cells associated with the determined coverage holes; and computing an estimated network performance of the one or more recommended network updates based on the one or more coverage holes and the one or more cells.
  • computing an estimated network performance comprises one or more of: generating one or more structural links between the one or more cells, and one or more neighbouring cells to the one or more cells; generating one or more topologic vector representations for each of the one or more cells; generating one or more functional vector representations for each of the one or more cells; generating one or more semantic vector representations for each of the one or more cells; and running a temporal graph neural network model with a network graph as input of the temporal graph neural network model, and one or more predicted key performance indicators for each of the one or more cells as output of the temporal graph neural network model, the network graph is based on the one or more structural links, the one or more topologic vector representations, the one or more functional vector representations and the one or more semantic vector representations.
  • generating one or more topologic vector representations for each of the one or more areas comprises one or more of: training a second ML model to infer one or more topologic vector representations for each of the one or more cells; running the ML model with one or more satellite imagery data associated with the one or more cells as input, and the one or more topologic vector representations as output of the ML model; and determine, for each of the one or more cells, one or more similar topologic vector representations.
  • generating one or more functional vector representations for each of the one or more areas comprises extracting one or more geographical data associated to each of the one or more cells.
  • generating one or more semantic vector representations for each of the one or more areas comprises one or more of: extracting one or more historical traffic data associated to each of the one or more cells; training a third ML model to infer one or more semantic vector representations for each of the one or more cells; running the third ML model with the one or more historical traffic data associated with the one or more cells as input, and the one or more semantic vector representations as output of the third ML model; and determine, for each of the one or more cells, one or more similar semantic vector representations.
  • the update of the network node (210) comprises one or more of: comparing the one or more predicted KPIs to the data; and triggering an update of the network node if the one or more predicted KPIs show an improvement compared to the data.
  • a method for updating a network node based on one or more recommended network updates is provided.
  • the method is performance by an apparatus.
  • the method comprises receiving data from the network node, the data relates to the one or more coverage areas.
  • the method comprises computing a current network performance based on the received data.
  • the method comprises determining one or more recommended network updates of the network node, the one or more recommended network updates are based on the computed current network performance.
  • the method comprises updating the network node based on the one or more recommended network updates.
  • the data comprises, for each one or more of the coverage areas, one or more of: network performance data; network topology; signal strength measurement data; satellite imagery data; geographical data; and historical traffic data.
  • the computation of the current network performance comprises one or more of: dividing each of the one or more coverage areas into one or more tiles; associating each of the one or more tiles to the data; processing one or more machine learning features for each of the tiles based on the association of each of the one or more tiles; training a first Machine Learning, ML, model to generate one or more estimated signal strength measurements using the one or more machine learning features; and generating a radio coverage representation of the one or more coverage areas using the one or more estimated signal strength measurements.
  • processing of the one or more machine learning features for each of the one or more tiles comprises one or more of: extracting, for each of the one or more tiles, the data related to each of the one or more tiles; calculating, for each of the one or more tiles a unique geographical identifier; calculating, for each of the one or more tiles an angular direction of the data measurement’s location corresponding to each tile location; calculating, for each of the one or more tiles, a geographical distance; arranging one or more vectors, each of the one or more vectors corresponds to each of the one or more tiles; encoding each of the one or more vectors; and splitting the one or more machine learning features into one or more K-folds.
  • training the first ML model to impute estimated one or more signal strength measurements using the one or more ML features comprises one or more of: training the ML model using the features of K- 1 folds of the one or more K folds; and evaluating the ML model using the features of one fold of the one or more K folds.
  • the determining the one or more recommended network updates comprises one or more of: determining one or more coverage holes based on the computed current network performance; determining one or more cells associated with the determined coverage holes; and computing an estimated network performance of the one or more recommended network updates based on the one or more coverage holes and the one or more cells.
  • computing the current network performance comprises one or more of: generating one or more structural links between the one or more cells, and one or more neighbouring cells to the one or more cells; generating one or more topologic vector representations for each of the one or more cells; generating one or more functional vector representations for each of the one or more cells; generating one or more semantic vector representations for each of the one or more cells; and running a temporal graph neural network model with a network graph as input of the temporal graph neural network model, and one or more predicted key performance indicators, KPIs, for each of the one or more cells as output of the temporal graph neural network model, the network graph is based on the one or more structural links, the one or more topologic vector representations, the one or more functional vector representations and the one or more semantic vector representations.
  • generating the one or more topologic vector representations comprises one or more of: training a second ML model to infer one or more topologic vector representations for each of the one or more cells; running the ML model with one or more satellite imagery data associated with the one or more cells as input, and the one or more topologic vector representations as output of the ML model; and determine, for each of the one or more cells, one or more similar topologic vector representations.
  • generating the one or more functional vector representations comprises extracting one or more geographical data associated to each of the one or more cells.
  • generating the one or more semantic vector representations comprises one or more of: extracting one or more historical traffic data associated to each of the one or more cells; training a third ML model to infer one or more semantic vector representations for each of the one or more cells; running the third ML model with the one or more historical traffic data associated with the one or more cells as input, and the one or more semantic vector representations as output of the third ML model; and determine, for each of the one or more cells, one or more similar semantic vector representations.
  • updating the network node comprises one or more of: comparing the one or more predicted KPIs to the data; and triggering an update of the network node if the one or more predicted KPIs show an improvement compared to the data.
  • a computer program comprises instructions, which when executed on an apparatus, causes the apparatus to perform the method according to any one of the embodiments of the second aspect of the invention.
  • a computer readable storage medium comprises a computer program according to third aspect of the invention.
  • At least one or more embodiments advantageously enable to update a network node based on one or more recommended network updates.
  • At least one or more embodiments advantageously enable to update a network node based on the impact of the one or more recommended updates.
  • an update is triggered in the event the one or more recommended updates show an improvement compared to current data.
  • At least one or more embodiments advantageously enable to determine one or more recommended updates of a network node based on an estimated network performance of one or more coverage areas.
  • Figure 1 shows a method performed by an apparatus.
  • Figure 2 shows an embodiment of the apparatus and network node.
  • Figure 3 shows an embodiment of data.
  • Figure 4 shows an embodiment of one or more tiles.
  • Figure 5 shows an embodiment of the method.
  • Figure 6 shows an embodiment of the method.
  • Figure 7 shows an embodiment of a first Machine Learning model.
  • Figure 8 shows an embodiment of the method.
  • Figure 9 shows an embodiment of the method.
  • Figure 10 shows an embodiment of the method.
  • Figure 11 shows an embodiment of the method.
  • Figure 12 shows an embodiment of the method.
  • Figure 13 shows an embodiment of the method.
  • Figure 14 shows an embodiment of a second Machine Learning model.
  • Figure 15 shows an embodiment of the method.
  • Figure 16 shows an embodiment of the method.
  • Figure 17 shows an embodiment of a third Machine Learning model.
  • Figure 18 shows an embodiment of a Temporal Graph neural network model.
  • Figure 19 shows an embodiment of a network graph.
  • Figure 20 shows an embodiment of the method.
  • Figure 21 shows a block diagram of the apparatus.
  • Figure 22 shows a block diagram of the apparatus.
  • the solution to be disclosed, and its embodiments provides an apparatus for updating a network node based on one or more recommended network updates, the network node covering one or more coverage areas.
  • the apparatus is configured to receive data from the network node, the data is related to the one or more coverage areas.
  • the apparatus is configured to compute a current network performance based on the received data.
  • the apparatus is configured to determine one or more recommended network updates of the network node.
  • the one or more recommended network updates are based on the computed current network performance.
  • the apparatus is configured to update the network node based on the one or more recommended network updates.
  • the present invention in its embodiments allow to update a network node based on recommended network updates.
  • the recommended network updates are based on a computed current network performance of the network node within coverage areas.
  • the current network performance is computed based on data related to the coverage areas covered by the network node.
  • network performance of the network node is improved within all the coverage areas covered by the network node.
  • FIG. 1 a flowchart depicting an embodiment of a method 100 is provided.
  • the method 100 is performed by an apparatus 200.
  • the method 100 is used for updating a network node 210 based on one or more recommended network updates 250.
  • the apparatus 200 may be is a network node.
  • the network node refers to equipment capable, configured, arranged and/or operable to communicate directly or indirectly with a wireless device and/or with other network nodes or equipment in a wireless network to enable and/or provide wireless access to the wireless device and/or to perform other functions (e.g., administration) in the wireless network.
  • Examples of network nodes include, but are not limited to, access points (APs) (e.g., radio access points), base stations (BSs) (e.g., radio base stations, Node Bs, evolved Node Bs (eNBs) and NR NodeBs (gNBs)).
  • APs access points
  • BSs base stations
  • eNBs evolved Node Bs
  • gNBs NR NodeBs
  • Base stations may be categorized based on the amount of coverage they provide (or, stated differently, their transmit power level) and may then also be referred to as femto base stations, pico base stations, micro base stations, or macro base stations.
  • a base station may be a relay node or a relay donor node controlling a relay.
  • a network node may also include one or more (or all) parts of a distributed radio base station such as centralized digital units and/or remote radio units (RRUs), sometimes referred to as Remote Radio Heads (RRHs). Such remote radio units may or may not be integrated with an antenna as an antenna integrated radio.
  • RRUs remote radio units
  • RRHs Remote Radio Heads
  • Such remote radio units may or may not be integrated with an antenna as an antenna integrated radio.
  • Parts of a distributed radio base station may also be referred to as nodes in a distributed antenna system (DAS).
  • DAS distributed antenna system
  • network nodes include multi-standard radio (MSR) equipment such as MSR BSs, network controllers such as radio network controllers (RNCs) or base station controllers (BSCs), base transceiver stations (BTSs), transmission points, transmission nodes, multi-cell/multicast coordination entities (MCEs), core network nodes (e.g., MSCs, MMEs), O&M nodes, OSS nodes, SON nodes, positioning nodes (e.g., E-SMLCs), and/or MDTs.
  • MSR multi-standard radio
  • RNCs radio network controllers
  • BSCs base station controllers
  • BTSs base transceiver stations
  • transmission points transmission nodes
  • MCEs multi-cell/multicast coordination entities
  • core network nodes e.g., MSCs, MMEs
  • O&M nodes e.g., OSS nodes, SON nodes, positioning nodes (e.g., E-SMLCs), and/or MDTs.
  • network nodes may represent any suitable device (or group of devices) capable, configured, arranged, and/or operable to enable and/or provide a wireless device with access to the wireless network or to provide some service to a wireless device that has accessed the wireless network.
  • the network node 210 may be an equipment capable, configured, arranged and/or operable to communicate directly or indirectly with a wireless device and/or with other network nodes or equipment in a wireless network to enable and/or provide wireless access to the wireless device and/or to perform other functions (e.g., administration) in the wireless network.
  • network nodes include, but are not limited to, access points (APs) (e.g., radio access points), base stations (BSs) (e.g., radio base stations, Node Bs, evolved Node Bs (eNBs) and NR NodeBs (gNBs)).
  • APs access points
  • BSs base stations
  • eNBs evolved Node Bs
  • gNBs NR NodeBs
  • Base stations may be categorized based on the amount of coverage they provide (or, stated differently, their transmit power level) and may then also be referred to as femto base stations, pico base stations, micro base stations, or macro base stations.
  • a base station may be a relay node or a relay donor node controlling a relay.
  • a network node may also include one or more (or all) parts of a distributed radio base station such as centralized digital units and/or remote radio units (RRUs), sometimes referred to as Remote Radio Heads (RRHs). Such remote radio units may or may not be integrated with an antenna as an antenna integrated radio.
  • RRUs remote radio units
  • RRHs Remote Radio Heads
  • Such remote radio units may or may not be integrated with an antenna as an antenna integrated radio.
  • Parts of a distributed radio base station may also be referred to as nodes in a distributed antenna system (DAS).
  • DAS distributed antenna system
  • network nodes include multi-standard radio (MSR) equipment such as MSR BSs, network controllers such as radio network controllers (RNCs) or base station controllers (BSCs), base transceiver stations (BTSs), transmission points, transmission nodes, multi-cell/multicast coordination entities (MCEs), core network nodes (e.g., MSCs, MMEs), O&M nodes, OSS nodes, SON nodes, positioning nodes (e.g., E-SMLCs), and/or MDTs.
  • MSR multi-standard radio
  • RNCs radio network controllers
  • BSCs base station controllers
  • BTSs base transceiver stations
  • transmission points transmission nodes
  • MCEs multi-cell/multicast coordination entities
  • core network nodes e.g., MSCs, MMEs
  • O&M nodes e.g., OSS nodes, SON nodes, positioning nodes (e.g., E-SMLCs), and/or MDTs.
  • network nodes may represent any suitable device (or group of devices) capable, configured, arranged, and/or operable to enable and/or provide a wireless device with access to the wireless network or to provide some service to a wireless device that has accessed the wireless network.
  • the skilled person would understand that even if there is only one network node (e.g., the network node 210) illustrated in the Figure 2, it is possible to have one network node, or a plurality of network nodes (e.g., more than one).
  • the network node 210 may cover one or more coverage areas 220, 221 , 222.
  • the one or more coverage areas 220, 221 , 222 may be a geographical area covered by the network node 210.
  • the one or more coverage areas 220, 221 , 222 may be defined by geographical measurements expressed in latitude and/or longitude.
  • the first coverage area 220 is defined by a first latitude coordinate and a first longitudinal coordinate.
  • the second coverage area 221 is defined by a second latitude coordinate and a second longitudinal coordinate.
  • the third coverage area 222 is defined by a third latitude coordinate and a third longitudinal coordinate.
  • only three coverage areas are illustrated (e.g., a first coverage area 220, a second coverage area 221 , and a third coverage area 222), however the skilled person would understand that it is possible to have only one coverage area or a plurality of coverage areas (e.g., more than one).
  • the one or more coverage areas 220, 221 , 222 may be referred to as service area (e.g., area wherein the network node 210 provide telecommunication services) characterized by geographical boundaries, licensed amount of frequency spectrum, radio propagation conditions, population and functional land use that can evolve over time.
  • a coverage area is dependent on users needs in the geographical area, coverage area urban dynamics, the coverage area functional land use, the coverage area spatial topology - which could have an influence on signal quality and path loss within the coverage area.
  • KPIs Accessibility Key Performance Indicators
  • Integrity KPIs Integrity KPIs
  • Utilization KPIs Retainability KPIs
  • Mobility KPIs Mobility KPIs
  • Energy Efficiency KPIs Reliability KPIs
  • TS 28 554 V17.7.0 3GPP Technical Specification
  • the method 100 comprises receiving 110 data 230 from the network node 210.
  • the data 230 relates to each of the one or more coverage areas 220, 221 , 222.
  • the data 230 comprises data related to the first coverage area 220.
  • the data 230 comprises data related to the second coverage area 221 .
  • the data 230 comprises data related to the third coverage area 222.
  • the data 230 is related to the network performance within the one or more coverage areas 220, 221 , 222.
  • the data 230 may comprise one or more network performance data 310.
  • the data 230 may comprise network topology information 320.
  • the data 230 may comprise signal strength measurement data 330.
  • the data 230 may comprise satellite imagery data 340.
  • the data 230 may comprise geographical data 350.
  • the data 230 may comprise historical traffic data 360.
  • the network performance data 310 may correspond to measurement of performance for each of the one or more coverage areas 220, 221 , 222.
  • the measurement of the performance may comprises one or more of: utilization measurement, accessibility measurement, throughput measurement, and latency measurement.
  • Table 1 illustration of the network performance data 310.
  • the network performance data 310 may be collected from real ongoing performance of the network node 210.
  • the network performance data 310 may be collected by the network node 210.
  • the network performance data 310 may be related to an utilization measurement of each of the one or more coverage areas 220, 221 , 222.
  • the utilization measurement of each of the one or more coverage areas 220, 221 , 222 may correspond to the Utilization KPIs as defined, for example, in 3GPP TS 28 554 V17.7.0 section 6.4.
  • the network performance data 310 may be related to an accessibility measurement of each of the one or more coverage areas 220, 221 , 222.
  • the accessibility measurement of each of the one or more coverage areas 220, 221 , 222 may correspond to the Accessibility KPIs as defined, for example, in 3GPP TS28 554 V17.7.0 section 6.2.
  • the network performance data 310 may be related to a throughput measurement of each of the one or more coverage areas 220, 221 , 222.
  • the throughput measurement of each of the one or more coverage areas 220, 221 , 222 may be defined in amount of data transmitted over a communication channel in a given period of time in each of the one or more coverage areas 220, 221 , 222.
  • the network performance data 310 may be related to a latency measurement of each of the one or more coverage areas 220, 221 , 222.
  • the latency measurement of each of the one or more coverage areas 220, 221 , 222 may be defined by delay in a packet’s arrival within each of the one or more coverage areas 220, 221 , 222.
  • the network performance data 310 may be collected at a regular time interval (e.g., every one or more seconds, every one or more minutes, every one or more days, every one or more weeks, every one or more months, etc.).
  • the network performance data 310 may be organized in a time-series format.
  • the network topology information 320 may be related to each of the one or more coverage areas 220, 221 , 222.
  • the network topology information 320 may comprise one or more of: number of network nodes in each of the one or more coverage areas 220, 221 , 222, physical location of the network nodes in each of the one or more coverage areas 220, 221 , 222.
  • the physical location of the network nodes may be defined by latitude coordinates, and longitudinal coordinate.
  • the signal strength measurement data 330 may be related to each one of the one or more coverage areas 220, 221 , 222.
  • the signal strength measurement data 330 may be acquired from a Minimization Drive Test (MDT).
  • the MDT correspond to MDT as defined, for example, in 3GPP TS 37. 320 V17.5.0.
  • MDT Drive Test
  • Table 2 illustration of the signal strength measurement related to the first coverage area 220.
  • the signal strength measurement 330 related to each of the one or more coverage areas 220, 221 , 222 may comprise a hashed coverage area identifier.
  • a hashed coverage area identifier of the first coverage area 220 may be 1234AT.
  • the signal strength measurement 330 related to each of the one or more coverage areas 220, 221 , 222 may comprise a hashed sample identifier.
  • a hashed sample identifier for the first coverage area 220 may be 1234T8.
  • the signal strength measurement 330 related to each of the one or more coverage areas 220, 221 , 222 may comprise a measurement location.
  • a measurement location for the first coverage area may be an anonymised measurement latitude and an anonymised measurement longitude.
  • the signal strength measurement 330 related to each of the one or more coverage areas may comprise a Reference Signal Received Quality (RSRQ) for each of the one or more coverage areas 220, 221 , 222.
  • RSRQ Reference Signal Received Quality
  • the signal strength measurement 330 may comprise a Reference Signal Received Power (RSRP) for each of the one or more coverage areas 220, 221 , 222.
  • RSRP Reference Signal Received Power
  • an RSRP of the first coverage area 220 may be -93.
  • the signal strength measurement 330 may comprise a signal-to- noise ratio of a given signal (SINR/SNR) for each of the one or more coverage areas 220, 221 , 222.
  • SINR/SNR for the first coverage area 220 may be 3.
  • the signal strength measurement 330 may comprise a technology indication for each of the one or more coverage areas 220, 221 , 222.
  • the technology indication may indicate which communication technology is used, such as a 3GPP telecommunication (e.g., Third Generation (3G), Fourth Generation (4G) 5 th Generation (5G), or any future generation).
  • a technology indication for the first coverage area 220 may be LTE/4G.
  • the satellite imagery data 340 may correspond to one or more images of Earth.
  • the one or more images of Earth may be collected by one or more imaging satellites (e.g., Earth observation satellite, and/or Earth remote sensing satellite).
  • the one or more images of Earth may be provided by a third-party entity, such as Sentinel-2.
  • the one or more images of Earth may comprise one or more of: spatial characteristics, spectral characteristics, temporal characteristics, and/or radiometric characteristics.
  • the geographical data 350 may comprise point-of-interest location information for each of the one or more coverage areas 220, 221 , 222.
  • the point-of-interest location information may indicate a distinctive location, such as a restaurant, a hospital, or a stadium. Indeed, such a distinctive location may have particular network traffic patterns.
  • the geographical data 350 may comprise infrastructure data related to the point-of-interest location information.
  • the historical traffic data 360 may be associated to each of the one or more coverage areas 220, 221 , 222.
  • the historical traffic data 360 may correspond to a total amount of transferred data volume on a medium access control (MAC) sublayer in a Down Link Performance counter (DL PM).
  • MAC medium access control
  • DL PM Down Link Performance counter
  • the historical data 360 is extracted from the PM counter.
  • the historical data 360 may in a timeseries format. Even if there is only one line in Table 3, the skilled person would understand that it is possible to have a multitude of historical data at different timepoint. Indeed, each line of the Table 3 correspond to historical data associated with the first coverage area 220 at a particular time point.
  • the method 100 comprises computing 120 a current network performance 240 based on the received data 230.
  • the current network performance 240 is related to the one or more coverage areas 220, 221 , 222.
  • the step of computing 120 of the current network performance 240 comprises dividing 510 each of the one or more coverage areas 220, 221 , 222 into one or more tiles 410.
  • Figure 4 an illustration of the one or more tiles 410 is provided.
  • the first coverage area 220 is divided into four tiles 411 , 412, 413, 414.
  • the skilled person would understand that even if there are four tiles represented for the first coverage area 220, it is possible to have only one tile or to have a plurality of the tiles (e.g., more than one).
  • the one or more tiles 410 may all be of a same format.
  • Each of the one or more tiles 410 may be in a hexagon format.
  • Each of the one or more tiles 410 may be in a quadrilateral format.
  • Each of the one or more tiles 410 may be in a polygon format.
  • Each of the one or more tiles 410 may be represented by a latitude coordinate and a longitudinal coordinate.
  • a tile 411 of the one or more tile of the first coverage area 220, illustrated in Figure 4 may be represented by a latitude coordinate lalitude_coordinate_tile_411 and a longitudinal coordinate longitudinal_coordinate_tile_411.
  • the step of computing 120 of the current network performance 240 comprises associating 520 each of the one or more tiles 410 to the received data 230.
  • each of the one or more tiles 410 is associated to the data 230 related to the latitude coordinate and the longitudinal coordinate of each of the one or more tiles 410.
  • Each of the one or more tiles 410 is associated to the data 230 received in the step 110 of the method 100.
  • each of the tiles 411 , 412, 413, 414 of the first coverage area 220 are associated to the network performance data 310 related to the first coverage area 220, the network topology information 320 related to the first coverage area 220, the signal strength measurement data 330 related to the first coverage area 220, satellite imagery data 340 related to the first coverage area 220, the geographical data 350 related to the first coverage area 220, and/or the historical traffic data 360 related to the first coverage area 220.
  • the tile 411 of the first coverage area 220 may be associated to the data 230 related to the latitude coordinate lalitude_coordinate_tile_411 and the longitudinal coordinate longitudinal_coordinate_tile_411.
  • the tile 411 may be associated to network performance data 310 related to (lalitude_coordinate_tile_417; longitudinal_coordinate_tile_411), the network topology information 320 related to (lalitude_coordinate_tile_417; longitudinal_coordinate_tile_411), the signal strength measurement data 330 related to (lalitude_coordinate_tile_417; longitudinal_coordinate_tile_411), satellite imagery data 340 related to (lalitude_coordinate_tile_417; longitudinal_coordinate_tile_411), the geographical data 350 related to (lalitude_coordinate_tile_417; longitudinal_coordinate_tile_411), and/or the historical traffic data 360 related to (lalitude_coordinate_tile_41 T, longitudinal_coordinate_tile_411).
  • the step of computing 120 of the current network performance 240 comprises processing 530 one or more machine learning features 535 for each of the tiles 410.
  • the processing 530 is based on the association of each of the one or more tiles 410.
  • Figure 6 an embodiment of the processing 530 is provided.
  • processing 530 comprises extracting 610, for each of the one or tiles 410, the data 230 related to each of the one or more tiles 410.
  • the data 230 related to each of the one or more tiles 410 may comprise one or more of: the network performance data 310 related to the tile, the network topology information 320 related to the tile, the signal strength measurement data 330 related to the tile, the satellite imagery data 340 related to the tile, the geographical data 350 related to the tile, and/or the historical traffic data 360 related to the tile.
  • processing 530 comprises calculating 620, for each of the one or more tiles 410, a unique geographical identifier.
  • the unique geographical identifier may be obtained by using an “H3 method” as defined, for example, in Sahr, K., White, D., & Kimerling, A. J. (2003), Geodesic discrete global grid systems, Cartography and Geographic Information Science, 30(2), 121 -134.
  • each of the one or more tiles 410 may be identified by a unique geographical identifier corresponding to a region of the Earth.
  • the tile 411 may be identified by the unique geographical identifier unique_geographical_identifier_tile_411 obtained by using the H3 method on (lalitude_coordinate_tile_411 ; longitudinal_coordinate_tile_411).
  • processing 530 comprises calculating 630, for each of the one or more tiles 410, an angular direction of the data 230 measurement’s location corresponding to each of the tile location (e.g., latitude coordinate, and longitudinal coordinate).
  • the angular direction may vary between a value of 0 to a value of 360.
  • the calculation of the angular direction of the signal measurement’s location corresponding to the tile location may be performed by calculating a bearing azimuth between the measurement location of the data 230 related to each tile, and the tile location.
  • processing 530 comprises calculating 640, for each of the one or more tiles 410, a geographical distance between the measurement location of the data 230 related to each tile, and the tile location.
  • the processing 530 comprises extracting statistics on device brand usage. In particular, depending on the manufacturer (e.g., Apple, Samsung, etc%) of the device, a particular technology might be used.
  • processing 530 comprises arranging 650 one or more vectors.
  • Each of the one or more vectors corresponds to each of one or more tiles 410.
  • Each vector of the one or more vectors may comprise one or more of: utilization measurement, accessibility measurement, throughput measurement, latency measurement, number of network nodes, signal strength measurement, RSRQ, RSRP, SINR/SNR, technology indication, MCC, MNC, unique geographical identifier, angular direction, geographical distance.
  • Each of the one or more machine learning features 535 may correspond to a vector of the one or more vectors of the step 650.
  • processing 530 comprises encoding 660 each vector of the one or more vectors corresponding to each of the one or more machine learning features 535.
  • the encoding may correspond to one-hot vectorizing the vector.
  • processing 530 comprises splitting 670 the one or more machine learning features 535 into one or more K-folds 705.
  • An embodiment of the one or more K-folds is provided in Figure 7.
  • Each of the K-fold may comprise a subpart of the one or more machine learning features 535.
  • K may represent a numerical value
  • the steps 610-670 may be performed simultaneously. In an embodiment, the step 610 may be performed before the steps 620-670. In an embodiment, the steps 620-640 may be performed before the steps 650-670. In an embodiment, the steps 650-660 may be performed before the step 670.
  • the step of computing 120 of the current network performance 240 comprises training 540 a first Machine Learning (ML) model 710 to generate one or more estimated signal strength measurements 740. The first ML model 710 may be trained using the one or more machine learning features 535.
  • ML Machine Learning
  • the one or more estimated signal strength measurements 740 may comprises one or more of: estimated RSRQ, estimated RSRP, and/or estimated SINR/SNR.
  • training 540 comprises training 720 the first ML model 710 using the features of K-1 folds 725 of the one or more K folds 705.
  • the K-1 folds may be comprised in the K-folds 705.
  • training 540 comprises evaluating 730 the first ML model 710 using the features of one fold 735 of the one or more K-folds 705.
  • the one fold 735 is comprised in the one or more K folds 705, but is not comprised in the K-1 folds 725 used by training 540 the first ML model 710.
  • evaluating 730 may comprise comparing the one or more estimated signal strength measurements 740 to the signal strength measurements data 330.
  • the apparatus 200 may compare the estimated RSRP with the RSRP received in the step 110 of the method 100.
  • the apparatus 200 may compare the estimated RSRQ with the RSRQ received in the step 110 of the method 100.
  • the step of computing 120 of the current network performance 240 comprises generating 550 a radio coverage representation 810 of the one or more coverage areas 220, 221 , 222 using the one or more estimated signal measurements 740.
  • a radio coverage representation 810 comprises for each tile of the one or more tiles 410 the associated data 230, and the one or more estimated signal strength measurements 740.
  • the tile 411 is associated to the data related to (lalitude_coordinate_tile_411; longitudinal_coordinate_tile_411), and the tile 411 is associated to the one or more estimated signal strength signal measurements related to (lalitude_coordinate_tile_411; longitudinal_coordinate_tile_411).
  • Using the one or more estimated signal strength measurements 740 may allow, in the event there is missing and/ or erroneous data from the data 230, to impute the one or more estimated signal strength measurements 740. Imputing the one or more estimated signal strength measurements 740 may be performed for each of the one or more tiles 410 that may have missing data from the data 230. Imputing the one or more estimated signal strength measurements 740 may be performed for each of the one or more tiles 410 that may have erroneous data from the data 230. For example, in the event, tile 412 is not associated with signal strength measurement 330, then tile 412 will be imputed with estimated signal strength measurements (e.g., the one or more signal strength measurements 740 illustrated in Figure 7) generated from running the first ML model 710.
  • estimated signal strength measurements e.g., the one or more signal strength measurements 740 illustrated in Figure 710.
  • the method 100 comprises determining 130 one or more recommended network updates 250 for the network node 210.
  • the one or more recommended network updates 250 may be based on the computed current network performance 240 computed in the step 120 of the method 100, as described above.
  • the step of determining 130 comprises determining 910 one or more coverage holes 1010 based on the computed current network performance 240.
  • the determining 910 of the one or more coverage holes 1010 may be based on the generated radio coverage representation 810.
  • an embodiment of the one or more coverage holes 1010 is provided.
  • the one or more coverage holes 1010 may correspond to Earth regions where signal strength measurements and/or estimated signal strength measurement is below a threshold.
  • the threshold corresponds to a benchmark of acceptable signal strength.
  • the threshold may to set by a network engineer.
  • the step of determining 130 comprises determining 920 one or more cells 1020 associated with the determined coverage holes 1010.
  • Each of the one or more cells 1020 may be defined by one or more of: a latitude coordinate, a longitudinal coordinate, a licensed frequency spectrum, a radio propagation condition, and/or number of population (e.g., humans, vehicles, user equipments, etc.).
  • Each of the one or more cells 1020 may be defined a geographical shape. The geographical shape may be calculated by using the network node geographical position (e.g., latitude coordinate of the network node 210, longitudinal coordinate of the network node 210), cell azimuth, and/or cell range azimuth.
  • Cell azimuth refers to an horizontal clockwise antenna direction corresponding to north.
  • a cell range azimuth corresponds to a value set between 0-360 degrees.
  • Cell range refers to how far a signal strength is provided in the cell.
  • the one or more cells 1020 corresponds to one or more tiles 410.
  • the step of determining 130 comprises computing 930 an estimated network performance 1030 of the one or more recommended network updates 250.
  • the estimated network performance 1030 may be based on the one or more coverage holes 1010, and the one or more cells 1020.
  • the estimated network performance 1030 may correspond to an expected network performance of the network one or more coverage areas 220, 221 , 222 in the event the one or more recommended network updates 250 are performed.
  • an embodiment of the step 930 is provided.
  • the step of computing 930 comprises generating 1110 one or more structural links between the one or more cells 1020 and one or more neighbouring cells 1210 to the one or more cells 1020.
  • the one or more neighbouring cells 1210 may correspond to one or more tiles of the one or more tiles 410 that are neighbours to the one or more cells 1020.
  • the tile 412 may correspond to a cell of the one or more cells 1020
  • the tiles 411 , 413, and 414 may correspond to neighbouring cells of the one or more neighbouring cells 1210.
  • the one or more cells 1020 may be associated with the historical traffic data 360 that is related to them.
  • the one or more neighbouring cells 1210 may be associated with the historical traffic data 360 that is related to them. Indeed, in the step 520 as described hereabove, each of the one or more tiles 410 has been associated to its related data 230. For example, in the first coverage area 220, the tile 411 has been associated to the data 230 related to (latitude_coordinate_tile_411 ; longitude_coordinate_tile_411 ), the tile 412 has been associated to the data 230 related to (latitude_coordinate_tile_412; longitude_coordinate_tile_412), and the associated of the tiles 413, 414 has been performed in the same manner.
  • the one or more structural links correspond to logical mobility relations between the one or more cells 1020 and the one or more neighbouring cell.
  • the one or more structural links represent handover mobility links between the one or more cells 1020 and the one or more neighbouring cells.
  • the one or more structural links are generated from Mobility relations between the one or more cells 1020 and the one or more neighbouring cells.
  • a cell is deployed in a geographical area to provide service to users (e.g., user equipments).
  • One source cell implementing 3GPP LTE and/or NR technologies can be connected to neighbouring cells.
  • the neighbouring cells enable mobility functionalities when the users move.
  • the one or more structural links comprise Mobility MO classes.
  • the Mobility MO classes are NRCellRelation MO for gNB NR technology, and EUtranCellRelation MO for eNB LTE.
  • the step of computing 930 comprises generating 1120 one or more topologic vector representations 1420 for each of the one or more cells 1020.
  • the step 1120 is provided.
  • the step of generating 1120 comprises training 1310 a second ML model 1410 to infer one or more topologic vector representations 1420 for each of the one or more cells 1020.
  • the step of training 1310 comprises training on a dataset 1415 comprising satellite imagery data.
  • the satellite imagery data comprises images of Earth collected by imaging satellites that revolve in the space around Earth. Satellite imagery data may be provided by various third-party data providers, such as Sentinel- 2.
  • One or more satellite imagery data 1430 may be associated with each of the one or more cells 1020.
  • the second ML model 1410 may correspond to a deep learning model.
  • the second ML model 1410 may correspond to a neural network model.
  • the second ML model 1410 may correspond to ResNet, as defined for example in He, Kaiming, et al. "Deep residual learning for image recognition.” Proceedings of the IEEE conference on computer vision and pattern recognition. 2016.
  • the second ML model 1410 may correspond to EfficientNet, as defined for example in Tan, Mingxing, and Quoc Le. "Efficientnet: Rethinking model scaling for convolutional neural networks.” International conference on machine learning. PMLR, 2019.
  • the second ML model 1410 may correspond to Vision Transformer, as defined for example in Dosovitskiy, Alexey, et al. "An image is worth 16x16 words: Transformers for image recognition at scale.” arXiv preprint arXiv:2010.11929 (2020).
  • the step of generating 1120 comprises running 1320 the second ML model 1410 with one or more satellite imagery data 1420 associated to the one or more cells 1020 as input to the second ML model 1410, and the one or more topologic vector representations 1420 for each of the one or more satellite imagery data 1430 as output of the second ML model 1410.
  • the one or more topologic vector representations 1420 may comprise one or more environment attributes.
  • the one or more environment attributes may comprise: land features of the one or more cells, environmental attributes of the one or more cells (e.g., parks, buildings).
  • the one or more vector representations 1420 may comprise one or more topological features.
  • the one or more topological features may comprise: infrastructure present in the one or more cells (e.g., roads, walkways, greenspace, detailed building footprint).
  • the step of generating 1120 comprises determining 1330, for each of the one or more cells 1020, one or more similar topologic vector representation 1420. Determining 1330 comprises comparing the one or more topologic vector representations 1420 by using a cosine similarity measurement between the one or more topologic vector representations 1420.
  • the step of computing 930 comprise generating 1130 one or more function vector representations 1620 for each of the one or more cells 1020.
  • the step 1130 is provided.
  • the step of generating 1130 comprises extracting geographical data 350 associated to each one of the one or more cells 1020.
  • the extracted geographical data 350 comprises geographical data related to Point of Interest in the cell (e.g., stadium, school, restaurant, ).
  • the extracted geographical data 350 is encoded into one or more functional vector representations 1620.
  • the encoding may be one-hot vectorizing.
  • the step of computing 930 comprises generating 1140 one or more semantic vector representations 1610 for each of the one or more cells 1020.
  • the step 1140 is provided.
  • the step of generating 1140 comprises extracting 1610 one or more historical traffic data 360 associated to each one of the one or more cells 1020.
  • the step of generating 1130 comprises training 1630 a third ML model 1710 to infer one or more semantic vector representations 1720 for each of the one or more cells 1020.
  • the step of training 1620 comprises training on a dataset 1715 comprising historical traffic data.
  • the historical traffic data is arranged in time series representing traffic.
  • the third ML model 1710 may correspond to a deep learning model.
  • the third model may correspond to a neural network model.
  • the step of generating 1120 comprises running 1630 the third ML model 1110 with the one or more historical traffic data 360 associated to the one or more cells 1020 as input to the third ML model 1710, and the one or more semantic vector representations 1720 as output of the third ML model.
  • the step of generating 1120 comprises determining 1640, for each of the one or more cells 1020, one or more similar semantic vector representation 1720. Determining 1640 comprises comparing the one or more semantic vector representations 1720 by using a cosine similarity measurement between the one or more semantic vector representations 1720.
  • the step of computing 930 comprises running 1150 a temporal neural network model 1810 with a network graph 1820 as input of the temporal neural network model 1910, and one or more predicted key performance indicators 1830 for each of the one or more cells 1020 as output of the temporal neural network model 1910.
  • the network graph 1820 is based on the one or more structural links, the one or more topological vector representations 1420, the one or more functional vector representations and the one or more semantic vector representations 1720.
  • an embodiment of the step 1150 is provided.
  • the graph G consists of nodes V.
  • the nodes V 1910 of the graph G corresponds to the one or more cells 1020.
  • node 1911 corresponds to cell 411
  • node 1912 correspond to cell 412.
  • only two nodes are represented (e.g., node 1911 and node 1912), however the skilled person would understand that it is possible to have more nodes.
  • edge 1920 correspond to a relation between the node 1911 (i.e., cell 411 ) and the node 1912 (i.e., cell 412).
  • the nodes V 1910 of the network graph 1820 are represented by the data 230.
  • the data 230 may be encoded into a one-hot vector.
  • the edges E 1920 of the network graph 1820 comprises edge features based on the generated one or more structural links in the step 1110, the generated one or more topologic vector representations 1420 in the step 1120, the generated one or more functional vector representations in the of the step 1130, and the generated one or more semantic vector representations 1720 in the step 1140.
  • Running 1150 the temporal graph neural network model 1810 comprises obtaining the one or more predicted KPIs 1830 for each node of the nodes V 1910 of the network graph 1820.
  • Each of the one or more predicted KPIs 1830 is a prediction of how each of the node of the nodes V 1910 would perform if the one or more recommended network updates 250 are deployed to each of the nodes V 1910.
  • the one or more predicted KPIs 1830 may relate to one or more of: accessibility of the cell 1020, integrity of the cell 1020, and mobility of the cell 1020.
  • the method 100 comprises updating 140 the network node 210 based on the one or more recommended updates 250.
  • an embodiment of the step 140 is provided.
  • the step of updating 140 comprises comparing 2010 the predicted one or more KPIs 1830 to the data 230, and triggering 2020 an update of the network node 210 if the one or more predicted KPIs 1830 show an improvement compared to the data 230.
  • a predicted signal strength measurement for a cell is higher than the signal strength measurement of the same cell, thus the predicted signal strength improvement shows an improvement compared to the signal strength measurement, and the recommended update is beneficial.
  • the apparatus 200 comprises a receiving unit 2110.
  • the apparatus 200 comprises a first computing unit 2120.
  • the apparatus 200 comprises a first determining unit 2135.
  • the apparatus 200 comprises an updating unit 2152.
  • the receiving unit 2110 is configured to cause the apparatus 200 to perform the step 110 of the method 100 as described above.
  • the first computing unit 2120 is configured to cause the apparatus 200 to perform the step 120 of the method 100 as described above.
  • the apparatus 200 may comprise a dividing unit 2121 .
  • the dividing unit 2121 is configured to cause the apparatus 200 to perform the step of dividing 510 as described above.
  • the apparatus 200 may comprise an associating unit 2122.
  • the associating unit 2122 is configured to cause the apparatus 200 to perform the step of associating 520 as described above.
  • the apparatus 200 may comprise a processing unit 2123.
  • the processing unit 2123 is configured to cause the apparatus 200 to perform the step of processing 530 as described above.
  • the apparatus 200 may further comprise a first extracting unit 2124.
  • the first extracting unit 2124 is configured to cause the apparatus 200 to perform the step of extracting 610 as described above.
  • the apparatus 200 may further comprise a first calculating unit
  • the first calculating unit 2125 is configured to cause the apparatus 200 to perform the step of calculating 620 as described above.
  • the apparatus 200 may further comprise a second calculating unit
  • the second calculating unit 2126 is configured to cause the apparatus 200 to perform the step of calculating 630 as described above.
  • the apparatus 200 may further comprise a third calculating unit
  • the third calculating unit 2127 is configured to cause the apparatus 200 to perform the step of calculating 640 as described above.
  • the apparatus 200 may further comprise an arranging unit 2128.
  • the arranging unit 2128 is configured to cause the apparatus 200 to perform the step of arranging 650 as described above.
  • the apparatus 200 may further comprise an encoding unit 2129.
  • the encoding unit 2129 is configured to cause the apparatus 200 to perform the step of encoding 660 as described above.
  • the apparatus 200 may further comprise a splitting unit 2130.
  • the splitting unit 2130 is configured to cause the apparatus 200 to perform the step of splitting 670 as described above.
  • the apparatus 200 may comprise a first training unit 2131 .
  • the first training unit 2131 is configured to cause the apparatus 200 to perform the step of training 540 as described above.
  • the apparatus 200 may further comprise a second training unit 2132.
  • the second training unit 2132 is configured to cause the apparatus 200 to perform the step of training 720 as described above.
  • the apparatus 200 may further comprise an evaluating unit 2133.
  • the evaluating unit 2133 is configured to cause the apparatus 200 to perform the step of evaluating 730 as described above.
  • the apparatus 200 may comprise a first generating unit 2134.
  • the first generating unit 2134 is configured to cause the apparatus 200 to perform the step of generating 550 as described above.
  • the first determining unit 2135 is configured to cause the apparatus 200 to perform the step 130 of the method 100 as described above.
  • the apparatus 200 may comprise a second determining unit 2136.
  • the second determining unit 2136 is configured to cause the apparatus 200 to perform the step of determining 910 as described above.
  • the apparatus 200 may further comprises a third determining unit
  • the third determining unit 2137 is configured to cause the apparatus 200 to perform the step of determining 920 as described above.
  • the apparatus 200 may further comprise a second computing unit
  • the second computing unit 2138 is configured to cause the apparatus 200 to perform the step of computing 930 as described above.
  • the apparatus 200 may further comprise a second generating unit 2139, a third generating unit 2140, a fourth generating unit 2141 , a fifth generating unit 2142 and a first running unit 2143.
  • the second generating unit 2139, the third generating unit 2140, the fourth generating unit 2141 , the fifth generating unit 2142 and the first running unit 2143 are configured to cause the apparatus 200 to perform the step of generating 1110, the step of generating 1120, the step a generating 1130, the step of generating 1140, and the step of running 1150, respectively, as described above.
  • the apparatus 200 further comprises a third training unit 2144, a second running unit 2145, and a fourth determining unit 2146.
  • the third training unit 2144, the second running unit 2145, and the fourth determining unit 2146 are configured to cause the apparatus 200 to perform the step of training 1310, the step of running 1320, and the step of determining 1330, respectively, as described above.
  • the apparatus 200 further comprises a second extracting unit 2147.
  • the second extracting unit 2147 is configured to cause the apparatus 200 to perform the step of extracting 1510 as described above.
  • the apparatus 200 further comprises a third extracting unit 2148, a fourth training unit 2149, a second running unit 2150, and a fifth determining unit 2151 .
  • the third extracting unit 2148, the fourth training unit 2149, the second running unit 2150, and the fifth determining unit 2151 are configured to cause the apparatus 200 to perform the step of extracting 1610, the step of training 1620, the step of running 1630, and the step of determining 1640, respectively, as described above.
  • the updating unit 2152 is configured to cause the apparatus 200 to perform the step 140 of the method 100 as described above.
  • the first computing unit 2135, the second computing unit 2138 are a same unit.
  • the first determining unit 2135, the second determining unit 2136, the third determining unit 2137, the fourth determining unit 2146, and the fifth determining unit 2151 are a same determining unit.
  • the first training unit 2131 , the second training unit 2132, the third training unit 2144, and the fourth training unit 2149 are a same training unit.
  • the first generating unit 2134, the second generating unit 2140, the third generating unit 2141 , the fourth generating unit 2142, and the fifth generating unit 2143 are a same generating unit.
  • the first extracting unit 2124, the second extracting unit 2147, and the third extracting unit 2148 are a same extracting unit.
  • the first calculating unit 2125, the second calculating unit 2126, and the third calculating unit 2127 are a same calculating unit.
  • the first running unit 2143, and the second running unit 2150 are a same running unit.
  • the apparatus 200 comprises a processor 2210, and a computer readable storage medium 2220 in the form of a memory 2225.
  • the memory 2225 contains a computer program 2230 comprising instruction executable by the processor 2210 whereby the apparatus 200 is operative to perform the steps of the method 100.
  • the (non-transitory) computer readable storage media mentioned above may be an Electrically Erasable Programmable Read-Only Memory (EEPROM), a flash memory, Field Programmable Gate Array, and a hard drive.
  • EEPROM Electrically Erasable Programmable Read-Only Memory
  • the processor 2210 of Figure 22 may be a single Central Processing Unit (CPU), but could also comprise two or more processing units.
  • the processor 2210 of Figure 22 may include general purpose microprocessors; instructions set processors and/or related chips sets and/or special purpose microprocessors such as Application Specific Circuits (ASICs).
  • ASICs Application Specific Circuits
  • the processor 2210 of Figure 22 may also comprise board memory for caching purposes.
  • the computer program 2230 of Figure 22 may be carried by a computer program product connected to the processor 2210 of Figure 22.
  • the computer program product may be or comprise a non-transitory computer readable storage medium on which the computer program 2230 of Figure 22 is stored.
  • the computer program product may be a flash memory, a Random-Access memory (RAM), a Read-Only Memory (ROM), or an EEPROM, and the computer programs described above could in alternative embodiments be distributed on different computer program products in the form of memories.
  • first”, “second”, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another.
  • first ML mode could be termed the second ML model, and similarly, the second ML model could be termed the first ML model.
  • the term “and/or” includes any and all combinations of one or more of the associated listed terms.
  • the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limited of example embodiments.
  • the single forms “a”, “an”, and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Environmental & Geological Engineering (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

L'invention divulgue un appareil (200), un procédé (100) et un programme informatique. L'appareil (200) est destiné à mettre à jour un nœud de réseau (210) sur la base d'une ou plusieurs mises à jour de réseau recommandées (250), le nœud de réseau (210) recouvrant une ou plusieurs zones de couverture (220, 221, 222), l'appareil (200) étant configuré pour : recevoir des données (230) en provenance du nœud de réseau (210), les données (230) concernant la ou les zones de couverture (220, 221, 222) ; calculer une performance de réseau actuelle (240) sur la base des données reçues (230) ; déterminer une ou plusieurs mises à jour de réseau recommandées (250) du nœud de réseau (210), la ou les mises à jour de réseau recommandées (250) étant basées sur la performance de réseau actuelle calculée (240) ; et mettre à jour le nœud de réseau (210) sur la base de la ou des mises à jour de réseau recommandées (250).
PCT/SE2023/051059 2023-10-26 2023-10-26 Appareil et procédé de mise à jour d'un nœud de réseau Pending WO2025089995A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/SE2023/051059 WO2025089995A1 (fr) 2023-10-26 2023-10-26 Appareil et procédé de mise à jour d'un nœud de réseau

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/SE2023/051059 WO2025089995A1 (fr) 2023-10-26 2023-10-26 Appareil et procédé de mise à jour d'un nœud de réseau

Publications (1)

Publication Number Publication Date
WO2025089995A1 true WO2025089995A1 (fr) 2025-05-01

Family

ID=95516192

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/SE2023/051059 Pending WO2025089995A1 (fr) 2023-10-26 2023-10-26 Appareil et procédé de mise à jour d'un nœud de réseau

Country Status (1)

Country Link
WO (1) WO2025089995A1 (fr)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018164618A1 (fr) * 2017-03-10 2018-09-13 Telefonaktiebolaget Lm Ericsson (Publ) Prédiction d'un indicateur de performance
US20190268779A1 (en) * 2018-02-23 2019-08-29 Google Llc Detecting Radio Coverage Problems
US20190306023A1 (en) * 2018-04-02 2019-10-03 Cisco Technology, Inc. Network configuration change analysis using machine learning
WO2020121084A1 (fr) * 2018-12-11 2020-06-18 Telefonaktiebolaget Lm Ericsson (Publ) Système et procédé d'amélioration de performance de modèle d'apprentissage automatique dans un réseau de communication
US20210227402A1 (en) * 2020-01-16 2021-07-22 T-Mobile Usa, Inc. Systems and methods for maintaining network quality
WO2022123292A1 (fr) * 2020-12-09 2022-06-16 Telefonaktiebolaget Lm Ericsson (Publ) Apprentissage de renforcement coordonné décentralisé pour optimiser des réseaux d'accès radio
US20230037893A1 (en) * 2021-08-02 2023-02-09 Samsung Electronics Co., Ltd. Method and network apparatus for generating real-time radio coverage map in wireless network
US11792096B1 (en) * 2022-09-23 2023-10-17 Verizon Patent And Licensing Inc. Method and system for predictive and feedback management of network performance

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018164618A1 (fr) * 2017-03-10 2018-09-13 Telefonaktiebolaget Lm Ericsson (Publ) Prédiction d'un indicateur de performance
US20190268779A1 (en) * 2018-02-23 2019-08-29 Google Llc Detecting Radio Coverage Problems
US20190306023A1 (en) * 2018-04-02 2019-10-03 Cisco Technology, Inc. Network configuration change analysis using machine learning
WO2020121084A1 (fr) * 2018-12-11 2020-06-18 Telefonaktiebolaget Lm Ericsson (Publ) Système et procédé d'amélioration de performance de modèle d'apprentissage automatique dans un réseau de communication
US20210227402A1 (en) * 2020-01-16 2021-07-22 T-Mobile Usa, Inc. Systems and methods for maintaining network quality
WO2022123292A1 (fr) * 2020-12-09 2022-06-16 Telefonaktiebolaget Lm Ericsson (Publ) Apprentissage de renforcement coordonné décentralisé pour optimiser des réseaux d'accès radio
US20230037893A1 (en) * 2021-08-02 2023-02-09 Samsung Electronics Co., Ltd. Method and network apparatus for generating real-time radio coverage map in wireless network
US11792096B1 (en) * 2022-09-23 2023-10-17 Verizon Patent And Licensing Inc. Method and system for predictive and feedback management of network performance

Similar Documents

Publication Publication Date Title
CN111246488B (zh) 基于测量的无线通信网络设计
US20240049003A1 (en) Managing a wireless device that is operable to connect to a communication network
CN106804046B (zh) 基于测量报告的移动台定位方法及装置
US9426044B2 (en) Radio access network geographic information system with multiple format
Braham et al. Fixed rank kriging for cellular coverage analysis
US10582332B2 (en) Autonomous localization in wireless networks
US20210219384A1 (en) Procedure for optimization of self-organizing network
US20220053285A1 (en) Positioning in Wireless Communication Networks
CN111368384B (zh) 预测天线工程参数的方法及设备
US20210083737A1 (en) Efficient data generation for beam pattern optimization
US12342189B2 (en) Cellular network indoor traffic auto-detection
CN112867147B (zh) 定位方法和定位装置
CN108293193A (zh) 一种用于完成路测数据的高效计算方法
CN115022795A (zh) Rf指纹识别映射更新
Qureshi et al. Enhanced MDT-based performance estimation for AI driven optimization in future cellular networks
Aguilar-Garcia et al. Location-aware self-organizing methods in femtocell networks
Alves et al. A novel approach for user equipment indoor/outdoor classification in mobile networks
Fang et al. An accurate and real-time commercial indoor localization system in LTE networks
WO2024074189A1 (fr) Systèmes et procédés pour programmer et déployer des mises à niveau de réseau dans des environnements urbains
WO2017020273A1 (fr) Procédé et dispositif de positionnement et procédé de génération d'une base de données de positionnement dans un réseau multiporteuse
WO2025089995A1 (fr) Appareil et procédé de mise à jour d'un nœud de réseau
US20230413312A1 (en) Network parameter for cellular network based on safety
EP4418617A1 (fr) Mécanisme de commande de rendement énergétique
WO2024079737A1 (fr) Premier nœud et procédés exécutés par ceux-ci pour gérer l'emplacement d'un nœud de réseau dans une zone géographique pour un fonctionnement dans un système de communication
US20240230839A1 (en) Landscape sensing using radio signals

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23956937

Country of ref document: EP

Kind code of ref document: A1

DPE1 Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101)