WO2022048546A1 - 数据传输方法和装置 - Google Patents

数据传输方法和装置 Download PDF

Info

Publication number
WO2022048546A1
WO2022048546A1 PCT/CN2021/115798 CN2021115798W WO2022048546A1 WO 2022048546 A1 WO2022048546 A1 WO 2022048546A1 CN 2021115798 W CN2021115798 W CN 2021115798W WO 2022048546 A1 WO2022048546 A1 WO 2022048546A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
cell
task
information
data segment
Prior art date
Application number
PCT/CN2021/115798
Other languages
English (en)
French (fr)
Inventor
耿婷婷
胡星星
曾清海
Original Assignee
华为技术有限公司
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 华为技术有限公司 filed Critical 华为技术有限公司
Priority to EP21863610.8A priority Critical patent/EP4203542A4/en
Publication of WO2022048546A1 publication Critical patent/WO2022048546A1/zh
Priority to US18/177,432 priority patent/US20230262490A1/en

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W24/00Supervisory, monitoring or testing arrangements
    • H04W24/02Arrangements for optimising operational condition
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00Machine learning
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/08Learning methods
    • G06N3/098Distributed learning, e.g. federated learning
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00Computing arrangements using knowledge-based models
    • G06N5/04Inference or reasoning models

Definitions

  • the embodiments of the present application relate to the field of communications, and more particularly, to a data transmission method and apparatus.
  • the network needs to support different requirements such as ultra-high speed, ultra-low latency, ultra-high reliability and/or ultra-multiple connections, which makes network planning, network configuration and resource scheduling more and more difficult. getting more complicated.
  • These new requirements, new scenarios, and new features have brought unprecedented challenges to mobile network planning, O&M, and efficient operation.
  • Network planning, self-optimization of network configuration and resource scheduling through manual experience or simple algorithms have disadvantages such as time-consuming, high cost, and poor adaptability of self-optimization and scheduling algorithms, which cannot cope with these new challenges.
  • MIMO multiple input multiple output
  • the embodiments of the present application provide a data transmission method and device, which can implement the application of machine learning in a mobile network and improve network work efficiency.
  • a communication method is provided, and the method can be executed by a terminal device or a module (eg, a chip) configured in (or used for) the terminal device.
  • a module eg, a chip
  • the method includes: sending a first data segment and first information to a first network device, where the first data segment is one of one or more data segments corresponding to the first data, and the first information is used to indicate the first data segment The target device for the data segment.
  • the first data is data for machine learning.
  • it can be used for model training and/or inference, etc.
  • the terminal device sends the collected data segments for machine learning to the first network device, and provides corresponding auxiliary information (ie, first information) when sending the data segments to indicate the target device of the data segments, so that
  • the first network device determines the target device of the data segment according to the auxiliary information, realizes that the data collected by the terminal device for machine learning is accurately transmitted to the device performing machine learning in the network, and can realize the application of machine learning in the mobile network, so as to
  • the network provides better communication services based on the results of machine learning and improves network work efficiency.
  • the method further includes: receiving configuration information from the first network device or the second network device, where the configuration information is used to configure the terminal device to collect the first data .
  • the first network device or the second network device configures the terminal device to collect data for machine learning, that is, the terminal device collects the data for machine learning according to the configuration and can send it to the network configured to collect the data
  • the device (that is, the second network device) can also continue to transmit the collected data to the device performing machine learning in the network through the network device currently establishing the connection (that is, the first network device) after the movement or handover occurs.
  • the connected network device may determine the target device of the first data segment according to the auxiliary information. It can realize the application of machine learning in mobile network and improve network work efficiency.
  • the configuration information is used to indicate one or more of the following contents:
  • a first area in which the terminal device collects the first data the first area including at least one cell
  • a first identifier used to identify the device that performs the machine learning
  • the configuration information may indicate the cell or area for collecting the first data, so that the terminal device determines the range of collecting the data.
  • the configuration information may also include information related to the task to which the collected data belongs, so that the terminal device collects data required by the network according to the related information.
  • the configuration information may also include the identifier of the device performing machine learning, so that when the terminal device sends the first data to the network, the identifier is carried in the auxiliary information, so that the data can accurately reach the device performing machine learning.
  • the target device is a device that performs the machine learning.
  • the terminal device indicates the device performing machine learning through auxiliary information, so that the first network device can determine the target device of the data segment according to the auxiliary information.
  • the first network device receives Machine learning is performed after the first data.
  • the first network device forwards the first data segment to the device performing machine learning according to the auxiliary information.
  • the target device is a third network device.
  • the target device is a third network device
  • the third network device may be a device that configures the terminal device to perform machine learning (ie, the second network device), or other network devices
  • the auxiliary information indicates the third network device
  • the first information is used to indicate the target device of the first data segment, including: the first information is used to indicate one or more of the following contents:
  • the auxiliary information indicates the third network device through the identifier of the third network device.
  • the auxiliary information indicates a cell identity, so as to notify the third device to be a network device corresponding to the cell identity that manages the cell.
  • the auxiliary information indicates an area identifier, and the network device corresponding to the cell in the area is a third network device.
  • the first information is carried in a first message, and the first message further includes second information, where the second information is used to indicate one of the following contents or more;
  • An identifier of a first task wherein the first task is a task for collecting the first data
  • the first data segment is the last data segment in the one or more data segments
  • the second cell is a cell in a second area, where the second cell is managed by the first network device, and the second cell is a serving cell of the terminal device, and the second area includes at least one cell;
  • the first duration where the first duration is the duration between when the terminal device generates the first data segment and sends the first data segment.
  • the terminal device may also send the identifier of the first task, so that the device performing machine learning determines the task of the first data segment.
  • the terminal device may also send the sequence number of the data segment, so that the device performing machine learning can sort the data segment corresponding to the first data according to the sequence number.
  • the terminal device may also indicate whether the first data segment is the last data segment of the first data, so that the device performing machine learning determines whether all data segments of the first data are received.
  • the terminal device may also indicate whether the second cell is a cell in the second area, so that the first network device determines whether to deliver the data segment to the device performing machine learning to which the first network device is connected.
  • the terminal device may also indicate the first duration, so that the device performing machine learning determines whether the first data is valid.
  • the first data includes one or more of the following data:
  • a communication method is provided, and the method can be executed by a first network device or a module (such as a chip) configured in (or used for) the first network device, and the method is executed by the first network device as follows: example to illustrate.
  • the method includes: receiving a first data segment and first information from a terminal device, where the first data segment is one of one or more data segments corresponding to the first data, and the first information is used to indicate the first data
  • the target device of the segment send the first data segment to the target device.
  • the method further includes: sending configuration information to the terminal device, where the configuration information is used to configure the terminal device to collect the first data.
  • the configuration information is used to indicate one or more of the following contents:
  • a first area in which the terminal device collects the first data the first area including at least one cell
  • a first identifier used to identify the device that performs the machine learning
  • the target device is a device that performs the machine learning.
  • the target device is a third network device.
  • the first information is used to indicate one or more of the following contents:
  • the first information is carried in a first message, and the first message further includes second information, where the second information is used to indicate one or more of the following contents ;
  • An identifier of a first task wherein the first task is a task of collecting the first data
  • the first data segment is the last data segment in the one or more data segments
  • the second cell is a cell in a second area
  • the second cell is managed by the first network device, and the second cell is a serving cell of the terminal device, wherein the second area includes at least one cell;
  • the first duration where the first duration is the duration between when the terminal device generates the first data segment and sends the first data segment.
  • the method further includes:
  • the second information is sent to the target device.
  • the first data includes one or more of the following data:
  • a communication method is provided, and the method can be executed by a second network device or a module (such as a chip) configured in (or used for) the first network device, and the method is executed by the second network device as follows: example to illustrate.
  • the method includes: receiving a first data segment from a first network device, where the first data segment is one of one or more data segments corresponding to the first data; sending the first data segment to a device performing the machine learning .
  • the method further includes: receiving configuration information from the device performing machine learning, where the configuration information is used to configure the terminal device to collect the first data.
  • the method further includes: sending the configuration information to the terminal device.
  • the configuration information is used to indicate one or more of the following contents:
  • a first area in which the terminal device collects the first data the first area including at least one cell
  • a first identifier used to identify the device that performs the machine learning
  • the method further includes:
  • An identifier of a first task wherein the first task is a task of collecting the first data
  • the first duration where the first duration is the duration between when the terminal device generates the first data segment and sends the first data segment.
  • the method further includes: sending the second information to the device performing the machine learning.
  • the first data includes one or more of the following data:
  • a communication method is provided, and the method can be executed by a device performing machine learning or a module (such as a chip) configured in (or used for) the device performing machine learning, and the method is hereinafter performed by a device performing machine learning
  • a module such as a chip
  • the device implementation is used as an example to illustrate.
  • the method includes: sending configuration information to a first network device or a second network device, where the configuration information is used to configure a terminal device to collect the first data; receiving a first data segment from the first device, the first data segment being the first data segment One of one or more data segments corresponding to a piece of data, wherein the first device is a first network device or a third network device.
  • the configuration information is used to indicate one or more of the following contents:
  • a first area in which the terminal device collects the first data the first area including at least one cell
  • a first identifier used to identify the device that performs the machine learning
  • the method further includes: receiving second information from the first device, where the second information is used to indicate one or more of the following:
  • An identifier of a first task wherein the first task is a task of collecting the first data
  • the first data segment is the last data segment in the one or more data segments
  • the first duration where the first duration is the duration between when the terminal device generates the first data segment and sends the first data segment.
  • the first data includes one or more of the following data:
  • a fifth aspect provides a communication device.
  • the device may include modules that perform one-to-one correspondence with the methods/operations/steps/actions described in the first aspect.
  • the modules may be hardware circuits, or However, software can also be implemented in combination with hardware circuits and software.
  • the device includes: a processing unit for collecting first data; a transceiver unit for sending a first data segment and first information to a first network device, where the first data segment corresponds to the first data One of one or more data segments, and the first information is used to indicate the target device of the first data segment.
  • the transceiver unit is further configured to receive configuration information from the first network device or the second network device, where the configuration information is used to configure the terminal device to collect the first network device. data.
  • the configuration information is used to indicate one or more of the following contents:
  • a first area in which the terminal device collects the first data the first area including at least one cell
  • a first identifier used to identify the device that performs the machine learning
  • the target device is a device that performs the machine learning.
  • the target device is a third network device.
  • the first information is used to indicate the target device of the first data segment, including: the first information is used to indicate one or more of the following contents:
  • the first information is carried in a first message, and the first message further includes second information, where the second information is used to indicate one of the following contents or more;
  • An identifier of a first task wherein the first task is a task for collecting the first data
  • the first data segment is the last data segment in the one or more data segments
  • the second cell is a cell in a second area, where the second cell is managed by the first network device, and the second cell is a serving cell of the terminal device, and the second area includes at least one cell;
  • the first duration where the first duration is the duration between when the terminal device generates the first data segment and sends the first data segment.
  • the first data includes one or more of the following data:
  • a sixth aspect provides a communication device.
  • the device may include a one-to-one module for performing the method/operation/step/action described in the second aspect.
  • the module may be a hardware circuit, or However, software can also be implemented in combination with hardware circuits and software.
  • the apparatus includes: a transceiver unit for receiving a first data segment and first information from a terminal device, where the first data segment is one of one or more data segments corresponding to the first data, the The first information is used to indicate the target device of the first data segment; the processing unit is used to determine the target device according to the first information; the transceiver unit is further used to send the first data segment to the target device.
  • the transceiver unit is further configured to send configuration information to the terminal device, where the configuration information is used to configure the terminal device to collect the first data.
  • the configuration information is used to indicate one or more of the following contents:
  • a first area in which the terminal device collects the first data the first area including at least one cell
  • a first identifier used to identify the device that performs the machine learning
  • the first task is a task for collecting the first data
  • the target device is a device that performs the machine learning.
  • the target device is a third network device.
  • the first information is used to indicate one or more of the following contents:
  • the first information is carried in a first message, and the first message further includes second information, where the second information is used to indicate one or more of the following contents ;
  • the first data segment is the last data segment in the one or more data segments
  • the second cell is a cell in a second area
  • the second cell is managed by the first network device, and the second cell is a serving cell of the terminal device, wherein the second area includes at least one cell;
  • the first duration where the first duration is the duration between when the terminal device generates the first data segment and sends the first data segment.
  • the transceiver unit is further configured to send the second information to the target device.
  • the first data includes one or more of the following data:
  • a communication device may include modules that perform the methods/operations/steps/actions described in the third aspect one-to-one, and the modules may be hardware circuits, software, or hardware circuits combined with software.
  • the device includes: a transceiver unit for receiving a first data segment from a first network device, where the first data segment is one of one or more data segments corresponding to the first data; a processing unit, for determining the device for performing the machine learning; the transceiver unit is further configured to send the first data segment to the device for performing the machine learning.
  • the transceiver unit is further configured to receive configuration information from the device performing machine learning, where the configuration information is used to configure the terminal device to collect the first data.
  • the transceiver unit is further configured to send the configuration information to the terminal device.
  • the configuration information is used to indicate one or more of the following contents:
  • a first area in which the terminal device collects the first data the first area including at least one cell
  • a first identifier used to identify the device that performs the machine learning
  • the first task is a task for collecting the first data
  • the transceiver unit is further configured to receive second information from the first network device, where the second information is used to indicate one or more of the following contents :
  • An identifier of a first task wherein the first task is a task for collecting the first data
  • the first duration where the first duration is the duration between when the terminal device generates the first data segment and sends the first data segment.
  • the transceiver unit is further configured to send the second information to the device performing the machine learning.
  • the first data includes one or more of the following data:
  • a communication device may include modules that perform one-to-one correspondence with the methods/operations/steps/actions described in the fourth aspect.
  • the modules may be hardware circuits, or However, software can also be implemented in combination with hardware circuits and software.
  • the device includes: a processing unit for determining configuration information, the configuration information for configuring the terminal device to collect first data, the first data being data used for machine learning; a transceiver unit for sending the first data to the first A network device or a second network device sends the configuration information, and the configuration information is used to configure the terminal device to collect first data, where the first data is data used for machine learning; the transceiver unit is further configured to receive data from the first device.
  • a first data segment where the first data segment is one of one or more data segments corresponding to the first data, wherein the first device is a first network device or a third network device.
  • the configuration information is used to indicate one or more of the following contents:
  • a first area in which the terminal device collects the first data the first area including at least one cell
  • a first identifier used to identify the device that performs the machine learning
  • the identifier of the first task, the first task is a task for collecting the first data
  • the transceiver unit is further configured to receive second information from the first device, where the second information is used to indicate one or more of the following contents:
  • An identifier of a first task wherein the first task is a task for collecting the first data
  • the first data segment is the last data segment in the one or more data segments
  • the first duration where the first duration is the duration between when the terminal device generates the first data segment and sends the first data segment.
  • the first data includes one or more of the following data:
  • a communication apparatus including a processor.
  • the processor may implement the first aspect and the method in any possible implementation manner of the first aspect.
  • the communication device further includes a memory, and the processor is coupled to the memory and can be configured to execute instructions in the memory, so as to implement the first aspect and the method in any possible implementation manner of the first aspect.
  • the communication device further includes a communication interface, and the processor is coupled to the communication interface.
  • the communication interface may be a transceiver, a pin, a circuit, a bus, a module, or other types of communication interfaces, which are not limited.
  • the communication apparatus is a terminal device.
  • the communication interface may be a transceiver, or an input/output interface.
  • the communication device is a chip configured in the terminal device.
  • the communication interface may be an input/output interface.
  • the transceiver may be a transceiver circuit.
  • the input/output interface may be an input/output circuit.
  • a communication apparatus including a processor.
  • the processor may implement the method in the second aspect and any possible implementation manner of the second aspect.
  • the communication device further includes a memory, and the processor is coupled to the memory and can be configured to execute instructions in the memory, so as to implement the second aspect and the method in any possible implementation manner of the second aspect.
  • the communication device further includes a communication interface, and the processor is coupled to the communication interface.
  • the communication apparatus is a first network device.
  • the communication interface may be a transceiver, or an input/output interface.
  • the communication apparatus is a chip configured in the first network device.
  • the communication interface may be an input/output interface.
  • the transceiver may be a transceiver circuit.
  • the input/output interface may be an input/output circuit.
  • a communication apparatus including a processor.
  • the processor may implement the third aspect and the method in any possible implementation manner of the third aspect.
  • the communication device further includes a memory, and the processor is coupled to the memory and can be configured to execute instructions in the memory, so as to implement the third aspect and the method in any possible implementation manner of the third aspect.
  • the communication device further includes a communication interface, and the processor is coupled to the communication interface.
  • the communication apparatus is a second network device.
  • the communication interface may be a transceiver, or an input/output interface.
  • the communication apparatus is a chip configured in the second network device.
  • the communication interface may be an input/output interface.
  • the transceiver may be a transceiver circuit.
  • the input/output interface may be an input/output circuit.
  • a communication apparatus including a processor.
  • the processor may implement the fourth aspect and the method in any possible implementation manner of the fourth aspect.
  • the communication device further includes a memory, and the processor is coupled to the memory and can be configured to execute instructions in the memory, so as to implement the fourth aspect and the method in any possible implementation manner of the fourth aspect.
  • the communication device further includes a communication interface, and the processor is coupled to the communication interface.
  • the communication apparatus is a device that performs machine learning.
  • the communication interface may be a transceiver, or an input/output interface.
  • the communication device is a chip configured in a device that performs machine learning.
  • the communication interface may be an input/output interface.
  • the transceiver may be a transceiver circuit.
  • the input/output interface may be an input/output circuit.
  • a thirteenth aspect provides a processor, comprising: an input circuit, an output circuit, and a processing circuit.
  • the processing circuit is configured to receive a signal through the input circuit and transmit a signal through the output circuit, so that the processor performs the method of the first aspect to the fourth aspect and any one of possible implementations of the first aspect to the fourth aspect .
  • the above-mentioned processor may be one or more chips
  • the input circuit may be input pins
  • the output circuit may be output pins
  • the processing circuit may be transistors, gate circuits, flip-flops and various logic circuits, etc. .
  • the input signal received by the input circuit may be received and input by, for example, but not limited to, a receiver
  • the signal output by the output circuit may be, for example, but not limited to, output to and transmitted by a transmitter
  • the circuit can be the same circuit that acts as an input circuit and an output circuit at different times.
  • the embodiments of the present application do not limit the specific implementation manners of the processor and various circuits.
  • a fourteenth aspect provides a computer program product, the computer program product comprising: a computer program (also referred to as code, or instructions), when the computer program is executed, causes the computer to execute the above-mentioned first to fourth aspects Aspects and methods of any possible implementations of the first to fourth aspects.
  • a computer program also referred to as code, or instructions
  • a computer-readable storage medium stores a computer program (also referred to as code, or instruction), when it runs on a computer, causing the computer to execute the above-mentioned first aspect
  • a computer program also referred to as code, or instruction
  • a communication system including at least two of the aforementioned terminal device, first network device, second network device, and device for performing machine learning.
  • FIG. 1 is a schematic diagram of a wireless communication system 100 suitable for an embodiment of the present application
  • FIG. 2 is a schematic structural diagram of a network device suitable for an embodiment of the present application
  • FIG. 3 is another schematic structural diagram of a network device suitable for an embodiment of the present application.
  • FIG. 4 is a schematic flowchart of a data transmission method provided by an embodiment of the present application.
  • Fig. 5 is another schematic flowchart of the data transmission method provided by the embodiment of the present application.
  • FIG. 6 is a schematic block diagram of an example of a communication device of the present application.
  • FIG. 7 is a schematic structural diagram of a terminal device applicable to an embodiment of the present application.
  • FIG. 8 is a schematic structural diagram of a network device suitable for an embodiment of the present application.
  • FIG. 9 is a schematic structural diagram of an apparatus for performing machine learning applicable to an embodiment of the present application.
  • GSM global system for mobile communications
  • CDMA code division multiple access
  • WCDMA wideband code division multiple access
  • GPRS general packet radio service
  • LTE long term evolution
  • FDD frequency division duplex
  • TDD time division duplex
  • UMTS universal mobile telecommunication system
  • WiMAX worldwide interoperability for microwave access
  • 5G fifth generation
  • 5G new radio
  • NR new radio
  • V2X may include vehicle to network (V2N), vehicle to vehicle (V2V), vehicle to infrastructure (V2I), and/or vehicle to pedestrian (vehicle to pedestrian, V2P), etc.
  • V2X may include vehicle to network (V2N), vehicle to vehicle (V2V), vehicle to infrastructure (V2I), and/or vehicle to pedestrian (vehicle to pedestrian, V2P), etc.
  • FIG. 1 is a schematic diagram of a wireless communication system 100 suitable for an embodiment of the present application.
  • the wireless communication system 100 may include at least one network device, for example, the network device 110 shown in FIG. 1 .
  • One or more network devices included in the at least one network device are devices that perform machine learning.
  • the wireless communication system 100 may further include at least one terminal device, for example, the terminal device 120 shown in FIG. 1 .
  • the device performing machine learning in the system 100 configures the terminal device to collect data for machine learning.
  • the terminal device collects data for machine learning and sends it to the network and provides auxiliary information.
  • the device in the network can forward the collected terminal data to the device performing machine learning according to the auxiliary information provided by the terminal device, so that the device performing machine learning can perform machine learning based on the data from the terminal device.
  • This solution can realize the application of machine learning in the mobile network, so that the network can provide better communication services and realize the intelligence of the RAN.
  • the terminal equipment in the embodiments of the present application may also be referred to as user equipment (user equipment, UE), access terminal, subscriber unit, subscriber station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal , wireless communication device, user agent or user device.
  • user equipment user equipment
  • UE user equipment
  • access terminal subscriber unit, subscriber station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal , wireless communication device, user agent or user device.
  • the terminal device in the embodiment of the present application may be a mobile phone (mobile phone), a tablet computer (pad), a computer with a wireless transceiver function, a virtual reality (virtual reality, VR) terminal device, an augmented reality (augmented reality, AR) terminal equipment, wireless terminals in industrial control, wireless terminals in self driving, wireless terminals in remote medical, wireless terminals in smart grid, transportation security ( wireless terminals in transportation safety), wireless terminals in smart cities, wireless terminals in smart homes, cellular phones, cordless phones, session initiation protocol (SIP) phones, wireless local Wireless local loop (WLL) station, personal digital assistant (PDA), handheld device with wireless communication function, computing device or other processing device, in-vehicle device, wearable device, terminal device in 5G network Or the terminal equipment in the public land mobile network (Public Land Mobile Network, PLMN) that evolves in the future.
  • a virtual reality (virtual reality, VR) terminal device an augmented reality (augmented reality, AR) terminal equipment
  • wireless terminals in industrial control wireless terminals
  • wearable devices can also be called wearable smart devices, which is a general term for the intelligent design of daily wear and the development of wearable devices using wearable technology, such as glasses, gloves, watches, clothing and shoes.
  • a wearable device is a portable device that is worn directly on the body or integrated into the user's clothing or accessories.
  • Wearable device is not only a hardware device, but also realizes powerful functions through software support, data interaction, and cloud interaction.
  • wearable smart devices include full-featured, large-scale, complete or partial functions without relying on smart phones, such as smart watches or smart glasses, and only focus on a certain type of application function, which needs to cooperate with other devices such as smart phones. Use, such as all kinds of smart bracelets, smart jewelry, etc. for physical sign monitoring.
  • the terminal device may also be a terminal device in an internet of things (Internet of things, IoT) system.
  • IoT Internet of things
  • Its main technical feature is to connect items to the network through communication technology, so as to realize the intelligent network of human-machine interconnection and interconnection of things.
  • the network device in this embodiment of the present application may be a device with a wireless transceiver function.
  • the equipment includes but is not limited to: base station, evolved node B (evolved node B, eNB), radio network controller (radio network controller, RNC), node B (node B, NB), base station controller (base station controller, BSC), base transceiver station (BTS), home base station (for example, home evolved nodeB, or home node B, HNB), baseband unit (BBU), wireless fidelity (wireless fidelity, WIFI) system
  • the access point (AP), wireless relay node, wireless backhaul node, transmission point (TP) or transmission and reception point (TRP), etc. can also be 5G (such as Access network (RAN) equipment in NR) systems, such as next-generation base stations (generation node B, gNB), TRP or TP, or, one or a group of base stations (including multiple antennas) in 5G systems panel) antenna panel.
  • 5G such as Access
  • the gNB may include a centralized unit (CU) and a distributed unit (DU), as shown in FIG. 2 .
  • the gNB may also include an active antenna unit (active antenna unit, AAU for short).
  • the CU implements some functions of the gNB, and the DU implements some functions of the gNB.
  • the CU may be responsible for processing non-real-time protocols and services, for example, it may implement a radio resource control (RRC) layer, a service data adaptation protocol (SDAP) layer, and/or a packet data convergence layer protocol (packet data convergence layer protocol).
  • RRC radio resource control
  • SDAP service data adaptation protocol
  • packet data convergence layer protocol packet data convergence layer protocol
  • the DU is responsible for handling physical layer protocols and real-time services.
  • radio link control radio link control
  • MAC media access control
  • PHY physical (physical, PHY) layer
  • One DU can be connected to only one CU or to multiple CUs, and one CU can be connected to multiple DUs, and communication between CUs and DUs can be performed through the F1 interface.
  • AAU implements some physical layer processing functions, radio frequency processing and related functions of active antennas.
  • the network device may be a device including one or more of a CU node, a DU node, a CU-CP node, a CU-UP node, and an AAU node.
  • the CU can be divided into network devices in an access network (radio access network, RAN), and the CU can also be divided into network devices in a core network (core network, CN), which is not limited in this application.
  • the CU may further include a CU control plane (CU-control plane, CU-CP) node and a CU user plane (CU-user plane, CU-UP) node, as shown in FIG. 3 .
  • the CU-CP may be responsible for control plane functions, such as implementing RRC layer and PDCP layer control plane functions (PDCP-C).
  • the CU-UP may be responsible for user plane functions, such as implementing SDAP layer and PDCP layer user plane functions (PDCP-U).
  • the CU-CP and CU-UP communicate via E1 interface.
  • the CU-CP can communicate with the core network through the NG interface on behalf of the gNB, and communicate with the DU through the F1-C interface.
  • the CU-UP can communicate with the DU through the F1-U interface.
  • the CU-UP may implement the function of PDCP-C, but the present application is not limited to this.
  • the network device in this embodiment of the present application may also be a CU node or a DU node that constitutes a gNB, or the network device may also be a CU-CP node or a CU-UP node that constitutes a CU, but the present application is not limited thereto.
  • the network device provides services for the cell, and the terminal device communicates with the network device in the cell through transmission resources (eg, frequency domain resources, or spectrum resources) allocated by the network device.
  • the cell may belong to a macro base station (for example, a macro eNB or a macro gNB, etc.), or may belong to a base station corresponding to a small cell (small cell).
  • the small cells here may include: urban cells (metro cells), micro cells (micro cells), pico cells (pico cells), or femto cells (femto cells). These small cells have the characteristics of small coverage and low transmit power, and are suitable for providing high-speed data transmission services.
  • Artificial intelligence can simulate nonlinear models, which can effectively adapt to the actual environment and approach the performance limit.
  • the embodiments of the present application propose to apply artificial intelligence (such as machine learning (ML)) to mobile networks, which can greatly improve the efficiency of network planning, network configuration, and resource scheduling, and realize network intelligence.
  • artificial intelligence such as machine learning
  • the terminal device can send massive data for machine learning to the network side, and the device performing machine learning in the network performs machine learning after receiving the data, so that the network can Based on the results of machine learning, it provides better communication services for terminal devices and realizes RAN intelligence.
  • AI Artificial intelligence
  • Machine learning is an implementation of artificial intelligence that empowers machines to learn to perform functions that cannot be done directly by programming.
  • machine learning is a method of using data, training a model, and then using the model to predict outcomes.
  • Reinforcement learning is a field in machine learning that emphasizes how to act based on the environment to maximize the desired benefit.
  • Transfer learning is another area in machine learning. It focuses on storing solved models of existing problems and leveraging them on other different but related problems.
  • Training refers to the process in which the model learns to perform a specific task by optimizing parameters, such as weights, in an AI model or ML model.
  • the embodiments of the present application are applicable to, but not limited to, one or more of the following training methods: supervised learning, unsupervised learning, reinforcement learning, transfer learning, and the like.
  • Supervised learning is trained with a set of training samples that have been correctly labeled (correctly labeled means that each sample has an expected output value).
  • unsupervised learning refers to a method that automatically classifies or groups incoming data without given pre-labeled training samples.
  • Inference refers to the use of trained AI models or ML models to perform tasks. Input the actual data into the AI model or ML model for processing to obtain the corresponding prediction results.
  • the prediction result may also be referred to as an inference result or a decision result.
  • a distributed AI training method which puts the training process of AI algorithms on multiple devices instead of aggregating them on one server, which can solve the problem of time-consuming and large communication overhead caused by data collection during centralized AI training.
  • the central node sends the AI model to multiple participating nodes, the participating nodes train the AI model based on their own data, and report the AI model trained by themselves to the central node in a gradient manner.
  • the central node averages or performs other operations on the gradient information fed back by multiple participating nodes to obtain a new AI model.
  • the central node can send the updated AI model to multiple participating nodes so that the participating nodes can retrain the AI model.
  • the participating nodes selected by the central node may be the same or different, and there is no restriction.
  • the embodiments of the present application can be applied to federated learning.
  • both the network device and the terminal device participate in the training of the machine learning AI model, and it can also be applied to centralized learning.
  • the network device can Conduct centralized AI model training, but this application is not limited to this.
  • the solution of this application can also be applied to other model training methods or AI algorithms.
  • “/” may indicate that the objects associated before and after are an “or” relationship, for example, A/B may indicate A or B; “and/or” may be used to describe that there are three types of associated objects A relationship, for example, A and/or B, can mean that A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural.
  • words such as “first” and “second” may be used to distinguish technical features with the same or similar functions. The words “first”, “second” and the like do not limit the quantity and execution order, and the words “first”, “second” and the like do not limit the difference.
  • words such as “exemplary” or “for example” are used to represent examples, illustrations or illustrations, and any embodiment or design solution described as “exemplary” or “for example” should not be construed are preferred or advantageous over other embodiments or designs.
  • the use of words such as “exemplary” or “such as” is intended to present the relevant concepts in a specific manner to facilitate understanding.
  • At least one (species) may also be described as one (species) or multiple (species), and the multiple (species) may be two (species), three (species), four (species) ) or more (species), which is not limited in this application.
  • the terminal device and/or the network device and/or the machine learning device may perform some or all of the steps in the embodiments of the present application, these steps or operations are only examples, and the embodiments of the present application may also be executed Other operations or variations of various operations.
  • various steps may be performed in different orders presented in the embodiments of the present application, and may not be required to perform all the operations in the embodiments of the present application.
  • FIG. 4 is a schematic flowchart of a data transmission method provided by an embodiment of the present application.
  • a device for performing machine learning (hereinafter referred to as a machine learning device) is configured on the network side, and the machine learning device can configure a terminal device to collect data for machine learning, and perform machine learning according to the data collected by the terminal device.
  • the machine learning device may be a network device that establishes a wireless connection with the terminal device (such as the first network device in FIG. 4 ), and in another embodiment, the machine learning device may be independent of the first network equipment.
  • the machine learning device independent of the first network device may be referred to as artificial intelligence control/controller (AIC), radio intelligence control/controller (RIC) or other names. limit.
  • the machine learning device when the machine learning device is independent of the first network device, the machine learning device performs S410.
  • the machine learning device sends configuration information A to the network device A, where the configuration information A is used to configure the terminal device to collect the first data.
  • the network device A receives the configuration information A from the machine learning device.
  • the first data is data used for machine learning.
  • the first data may be training data, model parameter gradients (information), or inference results.
  • the training data is used for model training in machine learning.
  • the training data can be the original data collected or measured by the terminal device, the data after processing the original data (such as normalization, which can also be normalized or standardized), and the feature engineering data (used to represent the original data) at least one of the characteristic data).
  • the model parameter gradient may be the gradient of the updated AI model parameters obtained by the participating nodes in the machine learning relative to the AI model parameters before the update.
  • the updated AI model is collected (or measured) by the participating nodes based on the AI model before the update and by itself.
  • the training data is obtained after training the AI model.
  • the inference result can be the inference result obtained after inputting the actual data into the AI model for inference in the application.
  • the configuration information A may indicate a cell in which the terminal device collects the first data or a first area in which the first data is collected.
  • the first area includes at least one cell.
  • the configuration information A includes an identifier of a cell, and is used to instruct the terminal device to collect the cell of the first data.
  • the machine learning device notifies the network device A of the cell where the terminal device needs to collect the first data through the configuration information A, and the network device A forwards the cell to the terminal device.
  • the configuration information A is transparently transmitted by the network device A to the terminal device, so that the terminal device determines the cell for collecting the first data according to the identification of the cell in the configuration information A, and the terminal device is in the cell corresponding to the identification of the cell.
  • the first data is collected in, but the application is not limited to this.
  • the identity of the cell may be at least one of the following:
  • CGI Cell global identifier
  • PCI physical cell identifier
  • NTN ID non-public network identifier
  • NTN ID non-terrestrial network identifier
  • the configuration information A includes an identifier of a first area, and the first area includes one or more cells. After acquiring the identifier of the first area, the terminal device collects the first data when the terminal device is within the coverage of a cell in the first area, but the present application is not limited to this. Optionally, when the configuration information does not include the area identifier, the terminal device may consider that the first data is collected in the serving cell.
  • the identifier of the first area is an identifier of at least one cell, and/or the identifier of the first area is a machine learning area identifier, a tracking area code (TAC), a radio access network notification At least one of an area code (radio access network notification area code, RANAC), a slice identifier, a service identifier, or other identification information that can identify an area.
  • TAC tracking area code
  • RANAC radio access network notification area code
  • the configuration information A includes indication information A, where the indication information A is used to indicate whether collecting the first data is a cell-level task or an area-level task.
  • the terminal device may consider that the network configures the terminal device to collect the first data in the serving cell that receives the configuration information A; or, the configuration information A includes the identifier of the cell, The terminal device collects the first data in the cell corresponding to the identity of the cell; when the indication information A indicates that collecting the first data is a regional task, the terminal device may consider that the network configures the terminal device in the serving cell where the configuration information A is received.
  • the first data is collected in the area of or; or, the configuration information A includes an identifier of the first area in which the terminal device collects the first data, but the present application is not limited to this.
  • the configuration information A indicates a first identifier, where the first identifier is an identifier of a machine learning device.
  • the configuration information A includes the identifier of the machine learning device to identify the machine learning device that configures the terminal device to collect the first data.
  • the configuration information A includes information about one or more tasks, including information about a first task, where the first task is a task for collecting first data.
  • the relevant information of the first task may include one or more of the following:
  • the identifier of the first task the name of the first task, the type of the first task, and the action type of the first task.
  • the identifier of the first task is used to identify the first task, or in other words, the identifier of the first task is used to indicate the task of collecting the first data.
  • the type of the first task may be: collecting raw data (or referred to as raw data collection), model publishing, model training, inference, and inference result publishing.
  • the type of the first task is to collect raw data, indicating that the first data is raw data, and the first task is a task for collecting raw data.
  • the type of the first task is model release, indicating that the first data is model parameters (eg, gradient information, etc.), and the first task is a task for collecting model parameter information, but the present application is not limited thereto.
  • the type of the first task in the configuration information A can be replaced with the type of the first data, and the type of the first data can be used to indicate the type of data that the terminal device needs to collect, or in other words, the type of the first data.
  • the type indicates that the first task is a task for collecting data of that type.
  • the type of the first data is used to indicate that the first data is one or more of training data, model parameter gradients (information), or inference results.
  • the action type of the first task may include one of various actions.
  • the multiple actions include one or more of the following actions: start, pause, resume, activate, and deactivate.
  • the action type of the first task may also be referred to as the state of the first task.
  • the configuration information A may include conditions for reporting the first data, data parameters or data sets to be collected.
  • the configuration information A indicates the conditions for reporting the gradient of the model parameters, and/or the parameters of the neural network for model training.
  • the configuration information A indicates at least one group of measurement types, and optionally the configuration information A also indicates a measurement reporting manner.
  • the measurement type can also be a new measurement type based on the requirements when applying machine learning to mobile networks.
  • the measurement reporting method may be periodic, one-time, event-triggered, or may be buffered and reported.
  • the cache reporting may refer to a mechanism in which a terminal device in a disconnected state collects the first data and then stores the first data, and sends the first data to a network device or a machine learning device when the terminal device transitions to a connected state.
  • the specific manner of the measurement report can be configured through the configuration information A, which is not limited in this application.
  • the network device A sends configuration information B to the terminal device, where the configuration information B is used to configure the terminal device to collect the first data.
  • the terminal device receives the configuration information B from the network device A.
  • the configuration information B indicates one or more of the following:
  • the cell where the terminal device collects the first data the first area where the first data is collected, the identity of the machine learning device, the related information of the first task, the conditions for reporting the first data, the reference parameters for training, and the reference for training data set.
  • the related information of the first task indicates one or more pieces of information among the identifier of the first task, the name of the first task, the type of the first task, and the action type.
  • the network device A is a device capable of performing machine learning.
  • the network device A sends the configuration information B to the terminal device according to the requirements of the machine learning task, so as to notify the terminal device to collect the first data.
  • the network device A determines, according to the configuration information A received from the machine learning device in S410, that the machine learning device requires the terminal device to collect the first data.
  • the configuration information B is generated according to the configuration information A and sent to the terminal device.
  • the network device A does not process the configuration information A after receiving it, and transparently transmits the configuration information A to the terminal device, and the configuration information B is the configuration information A.
  • the network device A does not read the configuration information of the machine learning device for the terminal device. Specific parameter content.
  • the terminal device collects the first data.
  • the terminal device receives the configuration information B from the network device A in S420, and determines, according to the configuration information B, that the device performing machine learning configures the terminal device to collect the first data, or determines that the terminal device needs to report to the network device A according to the configuration information B. first data. The terminal device collects the first data according to the configuration information B.
  • the configuration information B includes the identifier of the first area, and the first area includes a plurality of cells.
  • the configuration information B instructs the terminal device to collect and report the original data used for training, and also configures the measurement type and the measurement reporting method using periodic reporting. Then, when the terminal device is within the coverage of the first area, it collects the original data for training (that is, an example of the first data) according to the configured measurement type, and periodically reports the collected first data according to the indicated measurement reporting method. a data.
  • the terminal device sends the data segment A (ie, an example of the first data segment) and the first information to the network device A.
  • the data segment A ie, an example of the first data segment
  • the network device A receives the data segment A and the first information from the terminal device.
  • the data segment A is one of one or more data segments corresponding to the first data.
  • the terminal device may send the first data as a data segment to the network device A; or, the terminal device may send the first data in segments, for example, the first data is divided into multiple data segments, and each data segment includes the first data segment. part.
  • the data segment may be referred to as a data block, a data segment or a data packet, which is not limited in this application.
  • the terminal device sends the first data in segments.
  • the device performing machine learning configures the terminal device to periodically report the first data, and the terminal device collects the data segment A of the first data within a period and reports it to the network device.
  • the first information is used to indicate a target device of the data segment A, where the target device is a device that performs machine learning.
  • the apparatus for performing machine learning may perform model training and/or inference based on the first data.
  • the target device is the network device A.
  • the device performing machine learning is not the network device A, as shown in FIG. 2 , when the machine learning device performs machine learning, the target device is the machine learning device.
  • the first information includes an identification of the first task and/or a first identification, where the first identification is used to identify a device that performs machine learning.
  • the first information includes the identifier of the first task, so that the network device A can determine the task corresponding to the first data after receiving the first information, such as determining that the first data belongs to a machine learning device and is a terminal device according to the identifier of the first task
  • the configured machine learning task is forwarded to the machine learning device.
  • the network device A is a device performing machine learning, and the network device A determines according to the task identifier in the first information that the first data belongs to the data of the first task configured by the network device A for the terminal device, but the present application is not limited to this.
  • the first information includes the first identifier, so that the network device A determines the target device of the first data according to the first identifier. If the network device A is a device that performs machine learning, the first identifier is the identifier of the network device A. If the device that performs machine learning is not the network device A but the machine learning device in FIG. 4 , then the network device A determines the first identifier. The data needs to be forwarded to the machine learning device, but this application is not limited to this.
  • the first data and the first information are carried in a first message sent by the terminal device to the network device A, where the first message further includes the second information.
  • the first information and the second information may also be carried in different messages.
  • the second information is used to indicate one or more of the following:
  • the data segment A is the last data segment in one or more data segments corresponding to the first data
  • the first duration where the first duration is the duration between when the terminal device generates or stores the data segment A and sends the data segment A.
  • the second information includes the serial number of the data segment, so that the device performing machine learning can sort the multiple data segments of the first data according to the serial number after receiving the data segment A, so as to correctly read the first data .
  • the second information indicates whether the data segment A is the last data segment corresponding to the first data, so that the device performing machine learning can determine whether all data segments corresponding to the first data are received, so as to decode or analyze the first data.
  • the second information may also indicate the first duration, so that after receiving the first duration, the device performing machine learning can determine whether the first data is valid.
  • the second information further indicates whether the cell A is a cell in the second area, where the cell A is managed by the network device A, and the cell A is the current serving cell of the terminal device.
  • network device A determines that the first data segment can be forwarded to a device corresponding to network device A (or establishing a connection) performing machine learning.
  • the first message is a radio resource control (radio resource control, RRC) message. That is, the first data is control plane data.
  • RRC radio resource control
  • the first duration may specifically be the duration between the time when the terminal device generates or stores the data segment A (for example, the time when the data segment A is stored in the memory of the terminal device) and the time at which the RRC message is sent, or the first duration may specifically be is the time period between when the terminal device receives the configuration information B and sends the first message, but the present application is not limited to this.
  • the network device A is a gNB, and after the gNB receives the RRC message (ie, an example of the first message) from the terminal device, the RRC protocol layer of the gNB processes the first message.
  • the RRC message ie, an example of the first message
  • the terminal device After collecting the first data, the terminal device segments the first data and generates an RRC message, where the RRC message includes the first data segment and the first information.
  • the RRC protocol layer processes the first message, for example, reads the first information in the first message to determine the target device and so on.
  • the gNB is a device that performs machine learning. After the gNB receives the RRC message from the terminal device, the RRC protocol layer determines whether it has received all the data segments of the first data according to the first message. After all data segments of the data, the gNB performs machine learning according to the first data, but the present application is not limited thereto.
  • the network device A is a CU node or CU-CP node that constitutes a gNB, and the CU node or CU-CP node processes the first message by the RRC protocol layer after receiving the RRC message from the terminal device .
  • the network device A is a CU-CP node.
  • the RRC protocol layer processes the first message, for example, reads the first information in the first message to determine the target device, etc. .
  • the CU-CP node is a device that performs machine learning. After the CU-CP node receives the RRC message from the terminal device, the RRC protocol layer determines whether all data of the first data is received according to the first message. segment, after receiving all data segments of the first data, the CU-CP node performs machine learning according to the first data, but the present application is not limited to this.
  • the first message is layer 2 control signaling.
  • the layer 2 control signaling may be at least one of a PDCP layer control protocol data unit (protocol data unit, PDU), an RLC layer control PDU, and a MAC control element (control element, CE).
  • PDU PDCP layer control protocol data unit
  • RLC layer control PDU RLC layer control PDU
  • CE MAC control element
  • the network device A is a DU node, or a CU node and a DU node that constitute a gNB.
  • the DU node receives the layer 2 control signaling from the terminal device, the corresponding layer 2 protocol layer processes the DU node. Layer 2 signaling messages.
  • the network device A is a DU node.
  • the corresponding layer 2 protocol layer (such as RLC or MAC) processes the first message, for example, reads the first message in the first message. information to determine the target device, etc.
  • the DU node is a device that performs machine learning. After the DU node receives the layer 2 signaling message from the terminal device, the corresponding layer 2 protocol layer determines whether the first data is received according to the layer 2 signaling message. After receiving all the data segments of the first data, the DU node performs machine learning according to the first data, but the present application is not limited to this.
  • the network device A When the network device A is a device that performs machine learning, after receiving the first data through the above steps, the network device A uses the first data to perform machine learning.
  • the network device A configures the terminal device to collect training data
  • the network device performs model training according to the acquired first data to obtain a training model, so as to provide better network services through the trained model, but this application Not limited to this.
  • the first data is an inference result obtained by the terminal device according to the trained model, and the network device A judges, according to the inference result, whether the parameters of the model used by the terminal device need to be adjusted or the training model needs to be further optimized. Not limited to this.
  • the network device A When the network device A is not a device that performs machine learning, the network device A performs S450.
  • the network device A sends the data segment A and the second information to the machine learning device.
  • the machine learning device arranges the received data segments according to the serial number of the data segment A in the second information.
  • the machine learning device determines whether all data segments corresponding to the first data are received according to the second information.
  • the machine learning device receives the data segment A and the second information from the network device A. After the machine learning device receives all the data segments of the first data, it performs machine learning according to the first data.
  • the network configures the terminal device to collect data for machine learning, the terminal device performs data collection according to the network configuration, and the terminal device reports the collected data to the network device with auxiliary information to indicate the target device corresponding to the data. , so that the network device can determine the target device of the data according to the auxiliary information.
  • the terminal device can report the data segment to the network, and the auxiliary information can also indicate the serial number of the data segment and/or whether it is the last data segment, so that the machine can execute the data in the network.
  • the learning device can sort the data segments and determine whether to receive all the data segments of the first data. It can realize the accurate transmission of massive data to the equipment performing machine learning in the network, so that the network can provide better communication services according to the results of machine learning, and realize RAN intelligence.
  • the embodiment of the present application also provides a data transmission method, which considers how the terminal device reports data used for machine learning to the device performing machine learning in the event of a cell handover, so as to avoid wasting the collected data and improve communication efficiency .
  • FIG. 5 is another schematic flowchart provided by an embodiment of the present application.
  • the terminal device is moved from the coverage of the network device A to the coverage of the network device B.
  • the terminal device may move to the coverage of network device B after collecting the first data or after collecting part of the first data, or by performing cell selection, cell reselection or handover by the network device.
  • B provides access services.
  • the terminal equipment establishes a wireless connection with the network equipment B, and the cell B managed by the network equipment B provides services for the terminal equipment, that is, the cell B managed by the network equipment B is the current serving cell of the terminal equipment.
  • S510 may occur after S430 and before S440, that is, the terminal device moves after collecting the first data or after collecting part of the first data.
  • S510 may also occur after S440, for example, the terminal device moves to the cell B after sending part of the data segment of the first data to the network device A, but the present application is not limited to this.
  • the terminal device sends the data segment B (ie, another example of the first data segment) and the first information to the network device B.
  • the data segment B ie, another example of the first data segment
  • the network device B receives the data segment B and the first information from the terminal device.
  • the data segment B is one of one or more data segments corresponding to the first data.
  • the first information is used to indicate the target device of the data segment B.
  • the first data and the first information are carried in a first message sent by the terminal device to the network device B, where the first message further includes second information, where the second information is used to indicate one or more of the following:
  • the identifier of the first task, the first task is the task of collecting the first data
  • the data segment B is the last data segment in one or more data segments corresponding to the first data
  • the first duration where the first duration is the duration between when the terminal device generates or stores the data segment B and sends the data segment B.
  • the first message is a radio resource control (radio resource control, RRC) message.
  • RRC radio resource control
  • the terminal device sends the second information, and after receiving the second information, the device performing machine learning can determine the task corresponding to the data segment B according to the identifier of the first task in the second information.
  • the device performing machine learning may also sort the data segments according to the sequence number, and/or determine whether all data segments of the first data are received according to whether data segment B is the last data segment, thereby decoding or analyzing the first data , for machine learning.
  • the second information includes the first duration, it is determined whether the first data is valid according to the first duration.
  • the implementation manner in which the first information indicates the target device may include, but is not limited to, the following manners:
  • the target device is a machine learning device.
  • the first information is indicative of a machine learning device.
  • the first information includes the identification of the machine learning device
  • the terminal device indicates through the first information that the target device of the data segment B is the machine learning device
  • the network device B receives the first information according to the identification of the machine learning device. It is determined to forward the data segment B to the machine learning device corresponding to the identifier, but the present application is not limited to this.
  • the network device B After the network device B determines the machine learning device according to the first information, it executes S530 corresponding to the first mode in FIG. 5 .
  • the network device B sends the data segment B and the second information to the machine learning device.
  • the machine learning device receives the data segment B and the second information from the network device B.
  • the machine learning device may sort the data segments according to the sequence number of the data segment B in the second information.
  • the second information indicates whether the data segment B is the last data segment of the first data, it is determined whether all data segments of the first data are received, so as to determine whether to decode the first data and perform machine learning according to the first data.
  • the machine learning device may determine whether the first data is valid according to the first duration.
  • the target device is the network device C.
  • the first information indicates the network device C.
  • the network device C may be the network device A that sends the configuration information B to the terminal device, or may be another network device that establishes an interface with the device performing machine learning, or the network device C is the device performing machine learning.
  • the first information is used to indicate one or more of the following content:
  • Identification of network device C ie, an example of a third network device
  • the identifier of the first cell may be at least one of the following:
  • the identifier of the second area is the identifier of at least one cell, or the identifier of the second area is machine learning area identifier, TAC, RANAC, slice identifier, service identifier, or other identifier information that can identify the area. at least one of.
  • the first information includes the CGI and TAC of the first cell.
  • the network device B determines the network device C corresponding to the first cell according to the CGI of the first cell, so as to forward the data segment B to the network. device C.
  • network device B determines the corresponding network device C according to the CGI of the first cell, it can send the data segment B to the network device C through the interface between the network devices; if there is no direct connection between the network device B and the network device C
  • the interface between network devices, network device B can send data segment B and the CGI and TAC to network device C through the core network device, and the core network device can determine a specific forwarding path according to some or all of the CGI and TAC information.
  • the core network device may send the network identification information of the network device B and/or the network identification information of the network device C to the network device C. Further, the core network device may also send the TAC of the network device B and/or the TAC of the network device C to the network device C.
  • the network device C can be made to know the relevant information of the network device that sends the first information, so that the network device C can optimize the data forwarding path.
  • the network device C can use the network identification information of the network device B or the network The identification information and the TAC establish a direct interface between network devices with the network device B, so that the network device B can send the data segment B to the network device C through the interface.
  • the core network device may send the relevant information of the network device C to the network device B, and the network device B establishes an interface between the network devices with the network device C according to the relevant information of the network device C, thereby sending data to the network device C through the interface.
  • Section B the specific process can refer to the above description, which is not repeated here.
  • the network device B may also send the second information to the network device C. It should be understood that the present application is not limited to the above-mentioned forwarding mechanism.
  • the first information includes an identifier of the second area
  • the network device B forwards the data segment B to one or more network devices corresponding to the cells in the second area.
  • network device B can determine that there is a direct interface between network devices between it and a network device in the second area, such as network device C
  • network device B can forward data segment B to the network through the interface.
  • Device C if network device B cannot send data segment B to network devices in the second area through the interface between network devices, network device B can send data segment B to the second area through its connected core network device or machine learning node
  • the network device within sends the data segment B, but the present application is not limited to this.
  • the network device B may also send the second information to the network devices in the second area.
  • the first cell is a cell that collects the first data, that is, a cell that the terminal device indicated by the configuration information B collects the first data.
  • the second area is the first area, that is, the second area is an area where the terminal device indicated by the configuration information B collects the first data.
  • the network device B After the network device B determines the network device C according to the first information, it executes S530 corresponding to the second mode in FIG. 5 .
  • the network device B sends the data segment B and the second information to the network device C.
  • the network device C receives the data segment B and the second information from the network device B.
  • the network device C sends the data segment B and the second information to the machine learning device.
  • the network device C determines that the data segment B is data used for machine learning, and the network device C forwards the data segment B and the second information C to the relevant network device C establishes a connected device for performing machine learning, but the present application is not limited thereto.
  • the network device C may determine that the data segment B is data for machine learning according to the identifier of the first task in the second information.
  • the first information sent by the terminal device to the network device B indicates both the forwarding device, that is, the network device C, and the target device, that is, the machine learning device.
  • the network device B also indicates the target device of the data segment B to the network device C, that is, the machine learning device.
  • the network device C After receiving the data segment B and the second information, the network device C sends the data segment B and the second information to the machine learning device in S540 according to the target device indicated by the network device B, but the present application is not limited to this.
  • the machine learning device receives the data segment B and the second information from the network device C at S540.
  • the machine learning device performs machine learning according to the first data after receiving all the data segments of the first data.
  • the machine learning device configures the terminal device to collect data for training
  • the machine learning device performs model training according to the acquired first data to obtain a training model, so as to provide better network services through the trained model
  • the present application is not limited to this.
  • the first data is an inference result obtained by the terminal device according to the trained model, and the machine learning device judges according to the inference result whether the parameters of the model used by the terminal device need to be adjusted or the training model needs to be further optimized. Not limited to this.
  • the network configures the terminal device to collect data for machine learning, the terminal device performs data collection according to the network configuration, the terminal device reports the collected data to the network device, and carries auxiliary information, and moves the terminal device to other devices in time.
  • the cell When the cell is within the coverage area, it can also notify the moving network device B of the target device of the first data through the auxiliary information, so that the network device B can forward the first data to the device performing machine learning, which can avoid the collected data. be wasted and improve communication efficiency. It can realize the accurate transmission of massive data to the equipment performing machine learning in the network, so that the network can provide better communication services according to the results of machine learning, and realize the intelligentization of RAN.
  • each network element may include a hardware structure and/or a software module, and implement the above functions in the form of a hardware structure, a software module, or a hardware structure plus a software module. Whether one of the above functions is performed in the form of a hardware structure, a software module, or a hardware structure plus a software module depends on the specific application and design constraints of the technical solution.
  • FIG. 6 is a schematic block diagram of a communication apparatus provided by an embodiment of the present application.
  • the communication apparatus 600 may include a processing unit 610 and a transceiver unit 620 .
  • the communication apparatus 600 may correspond to the terminal equipment in the above method embodiments, or a chip configured (or used in) the terminal equipment, or other apparatuses capable of implementing the methods of the terminal equipment, Modules, circuits or units, etc.
  • the communication apparatus 600 may correspond to the terminal equipment in the methods 400 and 500 according to the embodiments of the present application, and the communication apparatus 600 may include the method for executing the terminal equipment in the methods 400 and 500 in FIG. 4 and FIG. 5 . method unit.
  • each unit in the communication device 600 and the above-mentioned other operations and/or functions are to implement the corresponding processes of the methods 400 and 500 in FIG. 4 and FIG. 5 , respectively.
  • the transceiver unit 620 in the communication apparatus 600 may be an input/output interface or circuit of the chip, and the processing in the communication apparatus 600 Unit 610 may be a processor in a chip.
  • the communication apparatus 600 may further include a processing unit 610, and the processing unit 610 may be configured to process instructions or data to implement corresponding operations.
  • the communication device 600 may further include a storage unit 630, the storage unit 630 may be used to store instructions or data, and the processing unit 610 may execute the instructions or data stored in the storage unit, so as to enable the communication device to implement corresponding operations .
  • the transceiver unit 620 in the communication device 600 in the communication device 600 may correspond to the transceiver 710 in the terminal device 700 shown in FIG. 7
  • the storage unit 630 may correspond to the terminal device 700 shown in FIG. 7 . of memory.
  • the transceiver unit 620 in the communication apparatus 600 may be implemented through a communication interface (such as a transceiver or an input/output interface), for example, it may correspond to the terminal shown in FIG. 7 .
  • the transceiver 710 in the device 700, the processing unit 610 in the communication device 600 may be implemented by at least one processor, for example, may correspond to the processor 720 in the terminal device 700 shown in FIG.
  • the processing unit 610 may be implemented by at least one logic circuit.
  • the communication apparatus 600 may correspond to the network device in the above method embodiments, for example, or a chip configured (or used in) the network device, or other methods capable of implementing the network device device, module, circuit or unit, etc.
  • the communication apparatus 600 may correspond to a network device in the methods 400 and 500 according to the embodiments of the present application, such as network device A, network device B, or network device C.
  • the communication apparatus 600 may include means for performing the method performed by the network device in the methods 400 and 500 in FIG. 4 and FIG. 5 .
  • each unit in the communication device 600 and the above-mentioned other operations and/or functions are to implement the corresponding processes of the methods 400 and 500 in FIG. 4 and FIG. 5 , respectively.
  • the transceiver unit in the communication device 600 is an input/output interface or circuit in the chip
  • the processing unit in the communication device 600 610 may be a processor in a chip.
  • the communication apparatus 600 may further include a processing unit 610, and the processing unit 610 may be configured to process instructions or data to implement corresponding operations.
  • the communication apparatus 600 may further include a storage unit 630, which may be used to store instructions or data, and the processing unit may execute the instructions or data stored in the storage unit 630 to enable the communication apparatus to implement corresponding operations.
  • the storage unit 630 in the communication apparatus 600 may correspond to the memory in the network device 800 shown in FIG. 8 .
  • the transceiver unit 620 in the communication apparatus 600 may be implemented through a communication interface (such as a transceiver or an input/output interface), for example, may correspond to the network shown in FIG. 8 .
  • the transceiver 810 in the device 800, the processing unit 610 in the communication device 600 may be implemented by at least one processor, for example, may correspond to the processor 820 in the network device 800 shown in FIG.
  • the processing unit 610 may be implemented by at least one logic circuit.
  • the communication apparatus 600 may correspond to the apparatus for performing machine learning in the above method embodiments, for example, or a chip configured (or used for) in the apparatus for performing machine learning, or other A device, module, circuit or unit, etc. capable of performing a method of a machine device.
  • the communication apparatus 600 may correspond to a device for performing machine learning in the methods 400 and 500 according to the embodiments of the present application, and the communication apparatus 600 may include a device for performing the methods 400 and 500 in FIG. 4 and FIG. 5 .
  • a unit of a method performed by a machine learning device A unit of a method performed by a machine learning device.
  • each unit in the communication device 600 and the above-mentioned other operations and/or functions are to implement the corresponding processes of the methods 400 and 500 in FIG. 4 and FIG. 5 , respectively.
  • the transceiver unit in the communication device 600 is an input/output interface or circuit in the chip, and the communication device 600
  • the processing unit 610 may be a processor in a chip.
  • the communication apparatus 600 may further include a processing unit 610, and the processing unit 610 may be configured to process instructions or data to implement corresponding operations.
  • the communication apparatus 600 may further include a storage unit 630, which may be used to store instructions or data, and the processing unit may execute the instructions or data stored in the storage unit 630 to enable the communication apparatus to implement corresponding operations.
  • the storage unit 630 in the communication apparatus 600 may correspond to the memory in the apparatus 900 for performing machine learning shown in FIG. 9 .
  • the transceiver unit 620 in the communication apparatus 600 may be implemented through a communication interface (such as a transceiver or an input/output interface), for example, it may correspond to the one shown in FIG. 9 .
  • the transceiver 910 in the apparatus 900 for performing machine learning shown in FIG. 9 the processing unit 610 in the communication apparatus 600 may be implemented by at least one processor, for example, may correspond to the processing in the apparatus 900 for performing machine learning shown in FIG. 9 .
  • the processor 920, the processing unit 610 in the communication device 600 may be implemented by at least one logic circuit.
  • FIG. 7 is a schematic structural diagram of a terminal device 600 provided by an embodiment of the present application.
  • the terminal device 700 can be applied to the system shown in FIG. 1 to perform the functions of the terminal device in the foregoing method embodiments.
  • the terminal device 700 includes a processor 720 and a transceiver 710 .
  • the terminal device 700 further includes a memory.
  • the processor 720, the transceiver 710 and the memory can communicate with each other through an internal connection path to transmit control and/or data signals, the memory is used to store computer programs, and the processor 720 is used to execute the computer in the memory. program to control the transceiver 710 to send and receive signals.
  • the above-mentioned processor 720 and the memory can be combined into a processing device, and the processor 720 is configured to execute the program codes stored in the memory to realize the above-mentioned functions.
  • the memory can also be integrated in the processor 720, or be independent of the processor 720.
  • the processor 720 may correspond to the processing unit in FIG. 6 .
  • the above transceiver 710 may correspond to the transceiver unit in FIG. 6 .
  • the transceiver 710 may include a receiver (or receiver, receiving circuit) and a transmitter (or transmitter, transmitting circuit). Among them, the receiver is used for receiving signals, and the transmitter is used for transmitting signals.
  • the terminal device 700 shown in FIG. 7 can implement the processes involving the terminal device in the method embodiments shown in FIG. 4 and FIG. 5 .
  • the operations and/or functions of each module in the terminal device 700 are respectively to implement the corresponding processes in the foregoing method embodiments.
  • the above-mentioned processor 720 may be used to perform the actions implemented by the terminal device described in the foregoing method embodiments, and the transceiver 710 may be used to perform the operations described in the foregoing method embodiments that the terminal device sends to or receives from the network device. action.
  • the transceiver 710 may be used to perform the operations described in the foregoing method embodiments that the terminal device sends to or receives from the network device. action.
  • the above-mentioned terminal device 700 may further include a power supply for providing power to various devices or circuits in the terminal device.
  • the terminal device 700 may further include one or more of an input unit, a display unit, an audio circuit, a camera, a sensor, etc., and the audio circuit may also include a speaker, microphone, etc.
  • FIG. 8 is a schematic structural diagram of a network device provided by an embodiment of the present application.
  • the network device 800 may be applied to the system shown in FIG. 1 to perform the functions of the network device in the foregoing method embodiments.
  • it may be a schematic diagram of a related structure of a network device.
  • the network device 800 shown in FIG. 8 can implement various processes involving a network device (eg, network device A, network device B, or network device C) in the method embodiments shown in FIG. 4 and FIG. 5 .
  • the operations and/or functions of each module in the network device 800 are respectively to implement the corresponding processes in the foregoing method embodiments.
  • the network device 800 shown in FIG. 8 may be an eNB or a gNB.
  • the network device includes network devices such as CU, DU, and AAU.
  • CU may be specifically divided into CU-CP and CU-CP. UP. This application does not limit the specific architecture of the network device.
  • the network device 800 shown in FIG. 8 may be a CU node or a CU-CP node.
  • An embodiment of the present application further provides a processing apparatus, including a processor and a (communication) interface; the processor is configured to execute the method in any of the above method embodiments.
  • the above-mentioned processing device may be one or more chips.
  • the processing device may be a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on chip (SoC), or a It is a central processing unit (CPU), a network processor (NP), a digital signal processing circuit (DSP), or a microcontroller (microcontroller unit). , MCU), it can also be a programmable logic device (PLD) or other integrated chips.
  • FPGA field programmable gate array
  • ASIC application specific integrated circuit
  • SoC system on chip
  • MCU microcontroller unit
  • MCU programmable logic device
  • PLD programmable logic device
  • the present application also provides a computer program product, the computer program product includes: computer program code, when the computer program code is executed by one or more processors, the computer program code including the processor is executed.
  • the apparatus executes the methods in the embodiments shown in FIG. 4 and FIG. 5 .
  • the technical solutions provided in the embodiments of the present application may be implemented in whole or in part by software, hardware, firmware, or any combination thereof.
  • software When implemented in software, it can be implemented in whole or in part in the form of a computer program product.
  • the computer program product includes one or more computer instructions.
  • the computer may be a general purpose computer, a special purpose computer, a computer network, network equipment, terminal equipment, core network equipment, machine learning equipment, or other programmable devices.
  • the computer instructions may be stored in or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be downloaded from a website site, computer, server, or data center Transmission to another website site, computer, server, or data center by wire (eg, coaxial cable, optical fiber, digital subscriber line, DSL) or wireless (eg, infrared, wireless, microwave, etc.).
  • the computer-readable storage medium can be any available media that can be accessed by a computer, or a data storage device such as a server, data center, etc. that includes one or more available media integrated.
  • the usable media may be magnetic media (eg, floppy disks, hard disks, magnetic tapes), optical media (eg, digital video discs (DVDs)), or semiconductor media, and the like.
  • the present application further provides a computer-readable storage medium, where the computer-readable storage medium stores program codes, and when the program codes are executed by one or more processors, the processing includes the processing
  • the device of the controller executes the method in the embodiment shown in FIG. 4 and FIG. 5 .
  • the present application further provides a system, which includes the aforementioned one or more network devices.
  • the system may further include one or more of the aforementioned terminal devices.
  • the network equipment in each of the above apparatus embodiments completely corresponds to the terminal equipment and the network equipment or terminal equipment in the method embodiments, and corresponding steps are performed by corresponding modules or units.
  • a processing unit processor
  • processor For functions of specific units, reference may be made to corresponding method embodiments.
  • the number of processors may be one or more.
  • the disclosed system, apparatus and method may be implemented in other manners.
  • the apparatus embodiments described above are only illustrative.
  • the division of the units is only a logical function division. In actual implementation, there may be other division methods.
  • multiple units or components may be combined or Can be integrated into another system, or some features can be ignored, or not implemented.
  • the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of devices or units, and may be in electrical, mechanical or other forms.
  • the units described as separate components may or may not be physically separated, and components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.
  • the embodiments may refer to each other.
  • the methods and/or terms between the method embodiments may refer to each other, such as the functions and/or the device embodiments.
  • terms may refer to each other, eg, functions and/or terms between an apparatus embodiment and a method embodiment may refer to each other.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Computational Linguistics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Mobile Radio Communication Systems (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本申请实施例提供了一种数据传输方法和装置,该方法包括:终端设备向第一网络设备发送第一数据段和辅助信息,该第一数据段是用于机器学习的数据对应的一个或多个数据段中的一个,辅助信息指示该第一数据段的目标设备。通过该方法,第一网络设备能够根据辅助信息确定目标设备,实现终端设备收集到的用于机器学习的海量数据准确地传递至网络中执行机器学习的设备,以便网络根据机器学习的结果提供更好的通信服务,提升网络工作效率。

Description

数据传输方法和装置
本申请要求于2020年09月04日提交中国国家知识产权局、申请号为202010923904.7、申请名称为“数据传输方法和装置”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请实施例涉及通信领域,更具体地,涉及一种数据传输方法和装置。
背景技术
当前的移动网络,由于支持的业务越来越多样,网络需要支持超高速率、超低时延、超高可靠和/或超多连接等不同需求,这使得网络规划、网络配置和资源调度越来越复杂。这些新需求、新场景和新特性给移动网络规划、运维和高效运营带来了前所未有的挑战。通过人工经验或简单的算法进行网络规划、网络配置自优化和资源调度存在耗时长、成本高、自优化和调度算法适应性差等弊端,无法应对这些新的挑战。比如,目前常用的多输入多输出(multiple input multiple output,MIMO)场景中,大量的数据处理通常都是代数或矩阵的线性运算,或者,基于高斯分布等的假设进行数据处理,但实际环境是复杂多变的,这些算法很难达到性能理论上限,且适用环境有限。
发明内容
本申请实施例提供了一种数据传输方法和装置,能够实现机器学习在移动网络中的应用,提升网络工作效率。
第一方面,提供了一种通信方法,该方法可以由终端设备或配置于(或用于)终端设备的模块(如芯片)执行,以下以该方法由终端设备执行为例进行说明。
该方法包括:向第一网络设备发送第一数据段和第一信息,该第一数据段是第一数据对应的一个或多个数据段中的一个,该第一信息用于指示该第一数据段的目标设备。
可选地,第一数据为用于机器学习的数据。例如可以用于进行模型训练和/或推理等。
根据上述方案,终端设备将收集到的用于机器学习的数据分段发送给第一网络设备,并在发送数据段时提供相应的辅助信息(即第一信息)指示数据段的目标设备,以便第一网络设备根据辅助信息确定数据段的目标设备,实现终端设备收集到的用于机器学习的数据准确地传递至网络中执行机器学习的设备,能够实现机器学习在移动网络中的应用,以便网络根据机器学习的结果提供更好的通信服务,提升网络工作效率。
结合第一方面,在第一方面的某些实现方式中,该方法还包括:接收来自该第一网络设备或第二网络设备的配置信息,该配置信息用于配置终端设备收集该第一数据。
根据上述方案,由第一网络设备或第二网络设备配置终端设备收集用于机器学习的数据,也就是说,终端设备根据配置收集用于机器学习的数据后可以发送给配置其收集数据的网络设备(即第二网络设备),也可以在发生移动或切换后,继续通过当前建立连接的网络设备(即第一网络设备)将收集到的数据传递至网络中执行机器学习的设备,当前建立连接的网络设备可以根据辅助信息确定第一数据段的目标设备。能够实现机器学习在移动网络中的应用, 提升网络工作效率。
结合第一方面,在第一方面的某些实现方式中,该配置信息用于指示以下一种或多种内容:
该终端设备收集该第一数据的小区;
该终端设备收集该第一数据的第一区域,该第一区域包括至少一个小区;
第一标识,用于标识执行该机器学习的设备;
第一任务的标识,其中,该第一任务是收集该第一数据的任务;和,
该第一任务的类型。
根据上述方案,配置信息可以指示收集第一数据的小区或区域,使得终端设备确定收集数据的范围。配置信息还可以包括收集数据所属任务相关信息,使得终端设备根据该相关信息收集网络需要的数据。配置信息还可以包括执行机器学习的设备的标识,使得终端设备向网络发送第一数据时,在辅助信息中携带该标识,以实现数据能够准确地到达执行机器学习的设备。
结合第一方面,在第一方面的某些实现方式中,该目标设备为执行该机器学习的设备。
根据上述方案,终端设备通过辅助信息指示执行机器学习的设备,使得第一网络设备能够根据辅助信息确定数据段的目标设备,当执行机器学习的设备为第一网络设备时,第一网络设备接收到第一数据后执行机器学习。当执行机器学习的设备不是第一网络设备时,第一网络设备根据辅助信息将第一数据段转发给执行机器学习的设备。
结合第一方面,在第一方面的某些实现方式中,该目标设备为第三网络设备。
根据上述方案,目标设备是第三网络设备,该第三网络设备可以是配置终端设备执行机器学习的设备(即第二网络设备),也可以是其他网络设备,辅助信息指示第三网络设备,以便第一网络设备根据辅助信息将第一数据段转发至第三网络设备,最终实现终端设备收集到的数据准确地到达执行机器学习的设备。
结合第一方面,在第一方面的某些实现方式中,该第一信息用于指示该第一数据段的目标设备,包括:该第一信息用于指示以下一种或多种内容:
该第三网络设备的标识;
第一小区的标识,其中,该第一小区为该第三网络设备管理的小区;和,
第二区域的标识,其中,该第二区域中包括至少一个小区,且该至少一个小区中包括该第三网络设备管理的第一小区。
根据上述方案,辅助信息通过第三网络设备的标识指示第三网络设备。或者辅助信息指示小区标识,以通知第三设备为小区标识对应的管理该小区的网络设备。或者,辅助信息指示区域标识,该区域中的小区对应的网络设备为第三网络设备。
结合第一方面,在第一方面的某些实现方式中,该第一信息承载在第一消息中,该第一消息还包括第二信息,该第二信息用于指示以下内容中的一种或多种;
第一任务的标识,其中,该第一任务是用于收集该第一数据的任务;
该第一数据段在该一个或多个数据段中的序列号;
该第一数据段是否为该一个或多个数据段中的最后一个数据段;
第二小区是否为第二区域中的一个小区,其中,该第二小区是由该第一网络设备管理的,且该第二小区为终端设备的服务小区,该第二区域包括至少一个小区;和,
第一时长,其中,该第一时长为终端设备生成该第一数据段到发送该第一数据段之间的时间长度。
根据上述方案,终端设备还可以发送第一任务的标识,以便执行机器学习的设备确定该第一数据段该的任务。终端设备还可以发送该数据段的序列号,以便执行机器学习的设备能够根据序列号对第一数据对应的数据段进行排序。终端设备还可以指示该第一数据段是否为第一数据的最后一个数据段,以便执行机器学习的设备确定是否接收到所有第一数据的数据段。终端设备还可以指示第二小区是否为第二区域中的一个小区,以便第一网络设备确定是否将该数据段传递至第一网络设备所连接的执行机器学习的设备。终端设备还可以指示第一时长,以便执行机器学习的设备确定第一数据是否有效。
结合第一方面,在第一方面的某些实现方式中,该第一数据包括以下一种或多种数据:
训练数据、模型参数梯度和推理结果。
第二方面,提供了一种通信方法,该方法可以由第一网络设备或配置于(或用于)第一网络设备的模块(如芯片)执行,以下以该方法由第一网络设备执行为例进行说明。
该方法包括:接收来自终端设备的第一数据段和第一信息,该第一数据段是第一数据对应的一个或多个数据段中的一个,该第一信息用于指示该第一数据段的目标设备;向该目标设备发送该第一数据段。
关于第一数据的具体介绍参见第一方面,不再赘述。
结合第二方面,在第二方面的某些实现方式中,该方法还包括:向该终端设备发送配置信息,该配置信息用于配置终端设备收集该第一数据。
结合第二方面,在第二方面的某些实现方式中,该配置信息用于指示以下一种或多种内容:
该终端设备收集该第一数据的小区;
该终端设备收集该第一数据的第一区域,该第一区域包括至少一个小区;
第一标识,用于标识执行该机器学习的设备;
第一任务的标识,其中,该第一任务是收集该第一数据的任务;和,
该第一任务的类型。
结合第二方面,在第二方面的某些实现方式中,该目标设备为执行该机器学习的设备。
结合第二方面,在第二方面的某些实现方式中,该目标设备为第三网络设备。
结合第二方面,在第二方面的某些实现方式中,该第一信息用于指示以下一种或多种内容:
该第三网络设备的标识;
第一小区的标识,其中,该第一小区为该第三网络设备管理的小区;和,
第二区域的标识,其中,该第二区域包括至少一个小区,且该至少一个小区中包括该第三网络设备管理的第一小区。
结合第二方面,在第二方面的某些实现方式中,该第一信息承载在第一消息中,第一消息还包括第二信息,该第二信息用于指示以下一种或多种内容;
第一任务的标识,其中,该第一任务是收集该第一数据的任务;
该第一数据段在该一个或多个数据段中的序列号;
该第一数据段是否为该一个或多个数据段中的最后一个数据段;
第二小区是否为第二区域中的一个小区,该第二小区是由第一网络设备管理的,且该第二小区为终端设备的服务小区,其中,该第二区域包括至少一个小区;和,
第一时长,该第一时长为该终端设备生成该第一数据段到发送该第一数据段之间的时间长度。
结合第二方面,在第二方面的某些实现方式中,该方法还包括:
向该目标设备发送该第二信息。
结合第二方面,在第二方面的某些实现方式中,该第一数据包括以下一种或多种数据:
训练数据、模型参数梯度和推理结果。
第三方面,提供了一种通信方法,该方法可以由第二网络设备或配置于(或用于)第一网络设备的模块(如芯片)执行,以下以该方法由第二网络设备执行为例进行说明。
该方法包括:接收来自第一网络设备的第一数据段,该第一数据段是第一数据对应的一个或多个数据段中的一个;向执行该机器学习的设备发送该第一数据段。
关于第一数据的介绍请参考第一方面,这里不再赘述。
结合第三方面,在第三方面的某些实现方式中,该方法还包括:接收来自该执行机器学习的设备的配置信息,该配置信息用于配置终端设备收集该第一数据。
结合第三方面,在第三方面的某些实现方式中,该方法还包括:向该终端设备发送该配置信息。
结合第三方面,在第三方面的某些实现方式中,该配置信息用于指示以下一种或多种内容:
该终端设备收集该第一数据的小区;
该终端设备收集该第一数据的第一区域,该第一区域包括至少一个小区;
第一标识,用于标识执行该机器学习的设备;
第一任务的标识,其中,该第一任务是收集该第一数据的任务;和,
该第一任务的类型。
结合第三方面,在第三方面的某些实现方式中,该方法还包括:
接收来自该第一网络设备的第二信息,其中,该第二信息用于指示以下一种或多种内容:
第一任务的标识,其中,该第一任务是收集该第一数据的任务;
该第一数据段在该一个或多个数据段中的序列号;
该第一数据段是否为该一个或多个数据段中的最后一个数据段;和,
第一时长,该第一时长为该终端设备生成该第一数据段到发送该第一数据段之间的时间长度。
结合第三方面,在第三方面的某些实现方式中,该方法还包括:向该执行该机器学习的设备发送该第二信息。
结合第三方面,在第三方面的某些实现方式中,该第一数据包括以下一种或多种数据:
训练数据、模型参数梯度和推理结果。
第四方面,提供了一种通信方法,该方法可以由执行机器学习的设备或配置于(或用于)执行机器学习的设备的模块(如芯片)执行,以下以该方法由执行机器学习的设备执行为例进行说明。
该方法包括:向第一网络设备或第二网络设备发送配置信息,该配置信息用于配置终端设备收集该第一数据;接收来自第一设备的第一数据段,该第一数据段是第一数据对应的一个或多个数据段中的一个,其中,该第一设备为第一网络设备或第三网络设备。
关于第一数据的介绍请参考第一方面,这里不再赘述。
结合第四方面,在第四方面的某些实现方式中,该配置信息用于指示以下一种或多种内容:
该终端设备收集该第一数据的小区;
该终端设备收集该第一数据的第一区域,该第一区域包括至少一个小区;
第一标识,用于标识执行该机器学习的设备;
第一任务的标识,该第一任务是收集该第一数据的任务;和
该第一任务的类型。
结合第四方面,在第四方面的某些实现方式中,该方法还包括:接收来自该第一设备的第二信息,其中,该第二信息用于指示以下一种或多种内容:
第一任务的标识,其中,该第一任务是收集该第一数据的任务;
该第一数据段在该一个或多个数据段中的序列号;
该第一数据段是否为该一个或多个数据段中的最后一个数据段;
第一时长,该第一时长为该终端设备生成该第一数据段到发送该第一数据段之间的时间长度。
结合第四方面,在第四方面的某些实现方式中,该第一数据包括以下一种或多种数据:
训练数据、模型参数梯度和推理结果。
第五方面,提供了一种通信装置,一种设计中,该装置可以包括执行第一方面中所描述的方法/操作/步骤/动作所一一对应的模块,该模块可以是硬件电路,也可是软件,也可以是硬件电路结合软件实现。一种设计中,该装置包括:处理单元,用于收集第一数据;收发单元,用于向第一网络设备发送第一数据段和第一信息,该第一数据段是第一数据对应的一个或多个数据段中的一个,该第一信息用于指示该第一数据段的目标设备。
关于第一数据的介绍请参考第一方面,这里不再赘述。
结合第五方面,在第五方面的某些实现方式中,该收发单元还用于接收来自该第一网络设备或第二网络设备的配置信息,该配置信息用于配置终端设备收集该第一数据。
结合第五方面,在第五方面的某些实现方式中,该配置信息用于指示以下一种或多种内容:
该终端设备收集该第一数据的小区;
该终端设备收集该第一数据的第一区域,该第一区域包括至少一个小区;
第一标识,用于标识执行该机器学习的设备;
第一任务的标识,其中,该第一任务是收集该第一数据的任务;和,
该第一任务的类型。
结合第五方面,在第五方面的某些实现方式中,该目标设备为执行该机器学习的设备。
结合第五方面,在第五方面的某些实现方式中,该目标设备为第三网络设备。
结合第五方面,在第五方面的某些实现方式中,该第一信息用于指示该第一数据段的目标设备,包括:该第一信息用于指示以下一种或多种内容:
该第三网络设备的标识;
第一小区的标识,其中,该第一小区为该第三网络设备管理的小区;和,
第二区域的标识,其中,该第二区域中包括至少一个小区,且该至少一个小区中包括该第二网络设备管理的第一小区。
结合第五方面,在第五方面的某些实现方式中,该第一信息承载在第一消息中,该第一消息还包括第二信息,该第二信息用于指示以下内容中的一种或多种;
第一任务的标识,其中,该第一任务是用于收集该第一数据的任务;
该第一数据段在该一个或多个数据段中的序列号;
该第一数据段是否为该一个或多个数据段中的最后一个数据段;
第二小区是否为第二区域中的一个小区,其中,该第二小区是由该第一网络设备管理的,且该第二小区为终端设备的服务小区,该第二区域包括至少一个小区;和,
第一时长,其中,该第一时长为终端设备生成该第一数据段到发送该第一数据段之间的时间长度。
结合第五方面,在第五方面的某些实现方式中,该第一数据包括以下一种或多种数据:
训练数据、模型参数梯度和推理结果。
第六方面,提供了一种通信装置,一种设计中,该装置可以包括执行第二方面中所描述的方法/操作/步骤/动作所一一对应的模块,该模块可以是硬件电路,也可是软件,也可以是硬件电路结合软件实现。一种设计中,该装置包括:收发单元,用于接收来自终端设备的第一数据段和第一信息,该第一数据段是第一数据对应的一个或多个数据段中的一个,该第一信息用于指示该第一数据段的目标设备;处理单元,用于根据该第一信息确定目标设备;该收发单元还用于向该目标设备发送该第一数据段。
关于第一数据的介绍请参考第一方面,这里不再赘述。
结合第六方面,在第六方面的某些实现方式中,该收发单元还用于向该终端设备发送配置信息,该配置信息用于配置终端设备收集该第一数据。
结合第六方面,在第六方面的某些实现方式中,该配置信息用于指示以下一种或多种内容:
该终端设备收集该第一数据的小区;
该终端设备收集该第一数据的第一区域,该第一区域包括至少一个小区;
第一标识,用于标识执行该机器学习的设备;
第一任务的标识,其中,该第一任务是用于收集该第一数据的任务;和,
该第一任务的类型。
结合第六方面,在第六方面的某些实现方式中,该目标设备为执行该机器学习的设备。
结合第六方面,在第六方面的某些实现方式中,该目标设备为第三网络设备。
结合第六方面,在第六方面的某些实现方式中,该第一信息用于指示以下一种或多种内容:
该第三网络设备的标识;
第一小区的标识,其中,该第一小区为该第三网络设备管理的小区;和,
第二区域的标识,其中,该第二区域包括至少一个小区,且该至少一个小区中包括该第三网络设备管理的第一小区。
结合第六方面,在第六方面的某些实现方式中,该第一信息承载在第一消息中,第一消息还包括第二信息,该第二信息用于指示以下一种或多种内容;
第一任务的标识,其中,该第一任务用于是收集该第一数据的任务;
该第一数据段在该一个或多个数据段中的序列号;
该第一数据段是否为该一个或多个数据段中的最后一个数据段;
第二小区是否为第二区域中的一个小区,该第二小区是由第一网络设备管理的,且该第二小区为终端设备的服务小区,其中,该第二区域包括至少一个小区;和,
第一时长,该第一时长为该终端设备生成该第一数据段到发送该第一数据段之间的时间长度。
结合第六方面,在第六方面的某些实现方式中,该收发单元还用于向该目标设备发送该第二信息。
结合第六方面,在第六方面的某些实现方式中,该第一数据包括以下一种或多种数据:
训练数据、模型参数梯度和推理结果。
第七方面,提供了一种通信装置。一种设计中,该装置可以包括执行第三方面中所描述的方法/操作/步骤/动作所一一对应的模块,该模块可以是硬件电路,也可是软件,也可以是硬件电路结合软件实现。一种设计中,该装置包括:收发单元,用于接收来自第一网络设备的第一数据段,该第一数据段是第一数据对应的一个或多个数据段中的一个;处理单元,用于确定执行该机器学习的设备;该收发单元还用于向该执行机器学习的设备发送该第一数据段。
关于第一数据的介绍请参考第一方面,这里不再赘述。
结合第七方面,在第七方面的某些实现方式中,该收发单元还用于接收来自该执行机器学习的设备的配置信息,该配置信息用于配置终端设备收集该第一数据。
结合第七方面,在第七方面的某些实现方式中,该收发单元还用于向该终端设备发送该配置信息。
结合第七方面,在第七方面的某些实现方式中,该配置信息用于指示以下一种或多种内容:
该终端设备收集该第一数据的小区;
该终端设备收集该第一数据的第一区域,该第一区域包括至少一个小区;
第一标识,用于标识执行该机器学习的设备;
第一任务的标识,其中,该第一任务是用于收集该第一数据的任务;和,
该第一任务的类型。
结合第七方面,在第七方面的某些实现方式中,该收发单元还用于接收来自该第一网络设备的第二信息,其中,该第二信息用于指示以下一种或多种内容:
第一任务的标识,其中,该第一任务是用于收集该第一数据的任务;
该第一数据段在该一个或多个数据段中的序列号;
该第一数据段是否为该一个或多个数据段中的最后一个数据段;和,
第一时长,该第一时长为该终端设备生成该第一数据段到发送该第一数据段之间的时间长度。
结合第七方面,在第七方面的某些实现方式中,该收发单元还用于向该执行该机器学习的设备发送该第二信息。
结合第七方面,在第七方面的某些实现方式中,该第一数据包括以下一种或多种数据:
训练数据、模型参数梯度和推理结果。
第八方面,提供了一种通信装置,一种设计中,该装置可以包括执行第四方面中所描述的方法/操作/步骤/动作所一一对应的模块,该模块可以是硬件电路,也可是软件,也可以是硬件电路结合软件实现。一种设计中,该装置包括:处理单元,用于确定配置信息,该配置信息用于配置终端设备收集第一数据,该第一数据为用于机器学习的数据;收发单元,用于向第一网络设备或第二网络设备发送该配置信息,该配置信息用于配置终端设备收集第一数据,该第一数据为用于机器学习的数据;该收发单元还用于接收来自第一设备的第一数据段,该第一数据段是第一数据对应的一个或多个数据段中的一个,其中,该第一设备为第一网络设备或第三网络设备。
结合第八方面,在第八方面的某些实现方式中,该配置信息用于指示以下一种或多种内容:
该终端设备收集该第一数据的小区;
该终端设备收集该第一数据的第一区域,该第一区域包括至少一个小区;
第一标识,用于标识执行该机器学习的设备;
第一任务的标识,该第一任务是用于收集该第一数据的任务;
该第一任务的类型。
结合第八方面,在第八方面的某些实现方式中,该收发单元还用于接收来自该第一设备的第二信息,其中,该第二信息用于指示以下一种或多种内容:
第一任务的标识,其中,该第一任务是用于收集该第一数据的任务;
该第一数据段在该一个或多个数据段中的序列号;
该第一数据段是否为该一个或多个数据段中的最后一个数据段;
第一时长,该第一时长为该终端设备生成该第一数据段到发送该第一数据段之间的时间长度。
结合第八方面,在第八方面的某些实现方式中,该第一数据包括以下一种或多种数据:
训练数据、模型参数梯度和推理结果。
第九方面,提供了一种通信装置,包括处理器。该处理器可以实现上述第一方面以及第一方面中任一种可能实现方式中的方法。可选地,该通信装置还包括存储器,该处理器与该存储器耦合,可用于执行存储器中的指令,以实现上述第一方面以及第一方面中任一种可能实现方式中的方法。可选地,该通信装置还包括通信接口,处理器与通信接口耦合。本申请实施例中,通信接口可以是收发器、管脚、电路、总线、模块或其它类型的通信接口,不予限制。
在一种实现方式中,该通信装置为终端设备。当该通信装置为终端设备时,该通信接口可以是收发器,或,输入/输出接口。
在另一种实现方式中,该通信装置为配置于终端设备中的芯片。当该通信装置为配置于终端设备中的芯片时,该通信接口可以是输入/输出接口。
可选地,该收发器可以为收发电路。可选地,该输入/输出接口可以为输入/输出电路。
第十方面,提供了一种通信装置,包括处理器。该处理器可以实现上述第二方面以及第二方面中任一种可能实现方式中的方法。可选地,该通信装置还包括存储器,该处理器与该存储器耦合,可用于执行存储器中的指令,以实现上述第二方面以及第二方面中任一种可能实现方式中的方法。可选地,该通信装置还包括通信接口,处理器与通信接口耦合。
在一种实现方式中,该通信装置为第一网络设备。当该通信装置为第一网络设备时,该通信接口可以是收发器,或,输入/输出接口。
在另一种实现方式中,该通信装置为配置于第一网络设备中的芯片。当该通信装置为配置于第一网络设备中的芯片时,该通信接口可以是输入/输出接口。
可选地,该收发器可以为收发电路。可选地,该输入/输出接口可以为输入/输出电路。
第十一方面,提供了一种通信装置,包括处理器。该处理器可以实现上述第三方面以及第三方面中任一种可能实现方式中的方法。可选地,该通信装置还包括存储器,该处理器与该存储器耦合,可用于执行存储器中的指令,以实现上述第三方面以及第三方面中任一种可能实现方式中的方法。可选地,该通信装置还包括通信接口,处理器与通信接口耦合。
在一种实现方式中,该通信装置为第二网络设备。当该通信装置为第二网络设备时,该通信接口可以是收发器,或,输入/输出接口。
在另一种实现方式中,该通信装置为配置于第二网络设备中的芯片。当该通信装置为配 置于第二网络设备中的芯片时,该通信接口可以是输入/输出接口。
可选地,该收发器可以为收发电路。可选地,该输入/输出接口可以为输入/输出电路。
第十二方面,提供了一种通信装置,包括处理器。该处理器可以实现上述第四方面以及第四方面中任一种可能实现方式中的方法。可选地,该通信装置还包括存储器,该处理器与该存储器耦合,可用于执行存储器中的指令,以实现上述第四方面以及第四方面中任一种可能实现方式中的方法。可选地,该通信装置还包括通信接口,处理器与通信接口耦合。
在一种实现方式中,该通信装置为执行机器学习的设备。当该通信装置为执行机器学习的设备时,该通信接口可以是收发器,或,输入/输出接口。
在另一种实现方式中,该通信装置为配置于执行机器学习的设备中的芯片。当该通信装置为配置于执行机器学习的设备中的芯片时,该通信接口可以是输入/输出接口。
可选地,该收发器可以为收发电路。可选地,该输入/输出接口可以为输入/输出电路。
第十三方面,提供了一种处理器,包括:输入电路、输出电路和处理电路。该处理电路用于通过该输入电路接收信号,并通过该输出电路发射信号,使得该处理器执行第一方面至第四方面以及第一方面至第四方面中任一种可能实现方式中的方法。
在具体实现过程中,上述处理器可以为一个或多个芯片,输入电路可以为输入管脚,输出电路可以为输出管脚,处理电路可以为晶体管、门电路、触发器和各种逻辑电路等。输入电路所接收的输入的信号可以是由例如但不限于接收器接收并输入的,输出电路所输出的信号可以是例如但不限于输出给发射器并由发射器发射的,且输入电路和输出电路可以是同一电路,该电路在不同的时刻分别用作输入电路和输出电路。本申请实施例对处理器及各种电路的具体实现方式不做限定。
第十四方面,提供了一种计算机程序产品,该计算机程序产品包括:计算机程序(也可以称为代码,或指令),当该计算机程序被运行时,使得计算机执行上述第一方面至第四方面以及第一方面至第四方面中任一种可能实现方式中的方法。
第十五方面,提供了一种计算机可读存储介质,该计算机可读存储介质存储有计算机程序(也可以称为代码,或指令)当其在计算机上运行时,使得计算机执行上述第一方面至第四方面以及第一方面至第四方面中任一种可能实现方式中的方法。
第十六方面,提供了一种通信系统,包括前述的终端设备、第一网络设备、第二网络设备和执行机器学习的设备中的至少两种设备。
附图说明
图1是适用于本申请实施例的无线通信系统100的示意图;
图2是适用于本申请实施例的网络设备的一个示意性架构图;
图3是适用于本申请实施例的网络设备另一个示意性架构图;
图4是本申请实施例提供的数据传输方法的一个示意性流程图;
图5是本申请实施例提供的数据传输方法的另一个示意性流程图;
图6是本申请的通信装置的一例的示意性框图;
图7是适用于本申请实施例的终端设备的一个示意性结构图;
图8是适用于本申请实施例的网络设备的一个示意性结构图;
图9是适用于本申请实施例的执行机器学习的设备的一个的示意性结构图。
具体实施方式
本申请实施例的技术方案可以应用于各种通信系统,例如:全球移动通信(global system formobile communications,GSM)系统、码分多址(code division multiple access,CDMA)系统、宽带码分多址(wideband code division multiple access,WCDMA)系统、通用分组无线业务(general packet radio service,GPRS)、长期演进(long term evolution,LTE)系统、LTE频分双工(frequency division duplex,FDD)系统、LTE时分双工(time division duplex,TDD)、通用移动通信系统(universal mobile telecommunication system,UMTS)、全球互联微波接入(worldwide interoperability for microwave access,WiMAX)通信系统、第五代(5th generation,5G)通信系统、新无线(new radio,NR)接入技术、车到其它设备(vehicle-to-x V2X)通信、车联网、机器类通信(machine type communication,MTC)、或物联网(internet of things,IoT)等。其中V2X可以包括车到互联网(vehicle to network,V2N)、车到车(vehicle to-vehicle,V2V)、车到基础设施(vehicle to infrastructure,V2I)、和/或车到行人(vehicle to pedestrian,V2P)等。
图1是适用于本申请实施例的无线通信系统100的示意图。
如图1所示,该无线通信系统100可以包括至少一个网络设备,例如图1所示的网络设备110。该至少一个网络设备中包括一个或多个网络设备为执行机器学习的设备。该无线通信系统100还可以包括至少一个终端设备,例如图1所示的终端设备120。根据本申请实施例提供的方案,该系统100中执行机器学习的设备配置终端设备收集用于机器学习的数据。终端设备根据该配置收集用于机器学习的数据发送给网络并提供辅助信息。网络中的设备可以根据终端设备提供的辅助信息将终端数据收集到数据转发至执行机器学习的设备,使得执行机器学习的设备根据来自终端设备的数据进行机器学习。该方案可以实现机器学习在移动网络中的应用,使网络能够提供更好的通信服务,实现RAN智能化。
本申请实施例中的终端设备也可以称为用户设备(user equipment,UE)、接入终端、用户单元、用户站、移动站、移动台、远方站、远程终端、移动设备、用户终端、终端、无线通信设备、用户代理或用户装置。本申请的实施例中的终端设备可以是手机(mobile phone)、平板电脑(pad)、带无线收发功能的电脑、虚拟现实(virtual reality,VR)终端设备、增强现实(augmented reality,AR)终端设备、工业控制(industrial control)中的无线终端、无人驾驶(self driving)中的无线终端、远程医疗(remote medical)中的无线终端、智能电网(smart grid)中的无线终端、运输安全(transportation safety)中的无线终端、智慧城市(smart city)中的无线终端、智慧家庭(smart home)中的无线终端、蜂窝电话、无绳电话、会话启动协议(session initiation protocol,SIP)电话、无线本地环路(wireless local loop,WLL)站、个人数字助理(personal digital assistant,PDA)、具有无线通信功能的手持设备、计算设备或其它处理设备、车载设备、可穿戴设备,5G网络中的终端设备或者未来演进的公用陆地移动通信网络(public land mobile network,PLMN)中的终端设备等。
其中,可穿戴设备也可以称为穿戴式智能设备,是应用穿戴式技术对日常穿戴进行智能化设计、开发出可以穿戴的设备的总称,如眼镜、手套、手表、服饰及鞋等。可穿戴设备即直接穿在身上,或是整合到用户的衣服或配件的一种便携式设备。可穿戴设备不仅仅是一种硬件设备,更是通过软件支持以及数据交互、云端交互来实现强大的功能。广义穿戴式智能设备包括功能全、尺寸大、可不依赖智能手机实现完整或者部分的功能,例如:智能手表或智能眼镜等,以及只专注于某一类应用功能,需要和其它设备如智能手机配合使用,如各类进行体征监测的智能手环、智能首饰等。
此外,终端设备还可以是物联网(internet of things,IoT)系统中的终端设备。IoT是未 来信息技术发展的重要组成部分,其主要技术特点是将物品通过通信技术与网络连接,从而实现人机互连,物物互连的智能化网络。
应理解,本申请实施例对于终端设备的具体形式不作限定。
本申请实施例中的网络设备可以是一种具有无线收发功能的设备。该设备包括但不限于:基站、演进型节点B(evolved node B,eNB)、无线网络控制器(radio network controller,RNC)、节点B(node B,NB)、基站控制器(base station controller,BSC)、基站收发台(base transceiver station,BTS)、家庭基站(例如,home evolved nodeB,或home node B,HNB)、基带单元(baseband unit,BBU)、无线保真(wireless fidelity,WIFI)系统中的接入点(access point,AP)、无线中继节点、无线回传节点、传输点(transmission point,TP)或者发送接收点(transmission and reception point,TRP)等,还可以为5G(如NR)系统中的接入网(radio access network,RAN)设备,如下一代基站(generation node B,gNB)、TRP或TP,或者,5G系统中的基站的一个或一组(包括多个天线面板)天线面板。
在一些部署中,gNB可以包括集中式单元(centralized unit,CU)和分布式单元(distributed unit,DU),如图2所示。gNB还可以包括有源天线单元(active antenna unit,简称AAU)。CU实现gNB的部分功能,DU实现gNB的部分功能。CU可以负责处理非实时协议和服务,例如可以实现无线资源控制(radio resource control,RRC)层,业务数据自适应协议(service data adaptation protocol,SDAP)层,和/或分组数据汇聚层协议(packet data convergence protocol,PDCP)层的功能。DU负责可以处理物理层协议和实时服务。例如可以实现无线链路控制(radio link control,RLC)层、媒体接入控制(media access control,MAC)层和物理(physical,PHY)层的功能。一个DU可以仅连接到一个CU或者连接到多个CU,而一个CU可以连接到多个DU,CU与DU之间可以通过F1接口进行通信。AAU实现部分物理层处理功能、射频处理及有源天线的相关功能。由于RRC层的信息最终会被递交至PHY层从而变成PHY层的信息,或者,由PHY层的信息转变而来,因而,在这种架构下,高层信令,如RRC层信令,也可以认为是由DU发送的,或者,由DU+AAU发送的。可以理解的是,网络设备可以为包括CU节点、DU节点、CU-CP节点、CU-UP节点的、AAU节点中一项或多项的设备。此外,可以将CU划分为接入网(radio access network,RAN)中的网络设备,也可以将CU划分为核心网(core network,CN)中的网络设备,本申请对此不做限定。其中,CU可以进一步的包括CU控制面(CU-control plane,CU-CP)节点和CU用户面(CU-user plane,CU-UP)节点,如图3所示。CU-CP可以负责控制面功能,例如实现RRC层和PDCP层控制面功能(PDCP-C)。CU-UP可以负责用户面功能,例如实现SDAP层和PDCP层用户面功能(PDCP-U)。CU-CP和CU-UP通过E1接口进行通信。CU-CP可以代表gNB通过NG接口与核心网通信,通过F1-C接口与DU进行通信。CU-UP可以通过F1-U接口与DU进行通信。另一实现方式中,CU-UP可以实现PDCP-C的功能,但本申请不限于此。
本申请实施例中的网络设备还可以是构成gNB的CU节点或DU节点,或者,网络设备还可以是构成CU的CU-CP节点或CU-UP节点,但本申请不限于此。
网络设备为小区提供服务,终端设备通过网络设备分配的传输资源(例如,频域资源,或者说,频谱资源)在小区中与网络设备进行通信。该小区可以属于宏基站(例如,宏eNB或宏gNB等),也可以属于小小区(small cell)对应的基站。这里的小小区可以包括:城市小区(metro cell)、微小区(micro cell)、微微小区(pico cell)、或毫微微小区(femto cell)等。这些小小区具有覆盖范围小、发射功率低的特点,适用于提供高速率的数据传输服务。
人工智能(artificial intelligence,AI)可以模拟非线性模型,从而能够有效适应实际环境, 逼近性能极限。本申请实施例提出将人工智能(例如机器学习(mechine learning,ML))运用于移动网络,可以大大提升网络规划、网络配置和资源调度的效率,实现网络智能化。而人工智能(例如机器学习)需要获取大量数据,通过机器学习算法对获取的数据进行模型训练和/或决策推理,输出AI模型和/或决策结果。在移动网络中,通过本申请实施例提供的方法,可以实现终端设备将用于机器学习的海量数据发送给网络侧,网络中的执行机器学习的设备接收到该数据后执行机器学习,以便网络根据机器学习的结果,为终端设备提供更好的通信服务,实现RAN智能化。
下面对本申请实施例中涉及到的定义进行说明。
1、人工智能
人工智能AI是让机器具有学习能力,能够积累经验,从而能够解决人类通过经验可以解决的诸如自然语言理解、图像识别和下棋等问题。
2、机器学习
机器学习是人工智能的一种实现方式,能够赋予机器学习能力,以此让机器完成直接编程无法完成的功能的方法。从实践的意义上来说,机器学习是一种通过利用数据,训练出模型,然后使用模型预测结果的一种方法。强化学习是机器学习中的一个领域,强调如何基于环境而行动,以取得最大化的预期利益。迁移学习是机器学习中的另一个领域。它专注于存储已有问题的解决模型,并将其利用在其他不同但相关问题上。
3、训练(training)或学习
训练是指一个处理过程,在该处理过程中通过优化一个AI模型或ML模型中的参数,如加权值,使模型学会执行某项特定的任务。本申请实施例适用于但不限于以下一种或多种训练方法:监督学习、无监督学习、强化学习、和迁移学习等。有监督学习是利用一组具有已经打好正确标签的训练样本来训练(已经打好正确标签是指每个样本有一个期望的输出值)。与有监督学习不同,无监督学习是指一种方法,该方法没有给定事先标记过的训练样本,自动对输入的数据进行分类或分群。
4、推理(inference)
推理是指利用训练之后的AI模型或ML模型执行任务。将实际数据输入AI模型或ML模型进行处理,得到对应的预测结果。该预测结果还可以称为推理结果或决策结果。
5、联邦学习(federated learning)
一种分布式AI训练方法,将AI算法的训练过程放在多个设备上进行,而不是聚合到一个服务器上,能够解决集中式AI训练时收集数据导致的耗时和大量通信开销问题。同时,由于不用将设备数据发送到服务器,也能够减少隐私安全问题。具体过程如下:中心节点向多个参与节点发送AI模型,参与节点基于自己的数据进行AI模型训练,并将自己训练的AI模型以梯度的方式上报给中心节点。中心节点对多个参与节点反馈的梯度信息进行平均或其他运算,得到新的AI模型。可选地,中心节点可以将更新后的AI模型发送给多个参与节点让参与节点重新进行AI模型训练。不同次联邦学习过程中,中心节点选择的参与节点可能相同,也可能不同,不予限制。
可以理解的,本申请实施例可以适用于联邦学习,例如,网络设备和终端设备均参与机器学习AI模型的训练,也可以适用于集中式学习,如终端设备上报收集到的数据后由网络设备进行集中式AI模型训练,但本申请不限于此。本申请的方案也可以适用于其他模型训练方式或AI算法。
在本申请实施例中,“/”可以表示前后关联的对象是一种“或”的关系,例如,A/B可 以表示A或B;“和/或”可以用于描述关联对象存在三种关系,例如,A和/或B,可以表示:单独存在A,同时存在A和B,单独存在B这三种情况,其中A,B可以是单数或者复数。为了便于描述本申请实施例的技术方案,在本申请实施例中,可以采用“第一”、“第二”等字样对功能相同或相似的技术特征进行区分。该“第一”、“第二”等字样并不对数量和执行次序进行限定,并且“第一”、“第二”等字样也并不限定一定不同。在本申请实施例中,“示例性的”或者“例如”等词用于表示例子、例证或说明,被描述为“示例性的”或者“例如”的任何实施例或设计方案不应被解释为比其它实施例或设计方案更优选或更具优势。使用“示例性的”或者“例如”等词旨在以具体方式呈现相关概念,便于理解。
在本申请实施例中,至少一个(种)还可以描述为一个(种)或多个(种),多个(种)可以是两个(种)、三个(种)、四个(种)或者更多个(种),本申请不做限制。
下面结合附图对本申请实施例提供的数据传输方法进行详细说明。
应理解,本申请实施例中,终端设备和/或网络设备和/或机器学习设备可以执行本申请实施例中的部分或全部步骤,这些步骤或操作仅是示例,本申请实施例还可以执行其它操作或者各种操作的变形。此外,各个步骤可以按照本申请实施例呈现的不同的顺序来执行,并且有可能并非要执行本申请实施例中的全部操作。
图4是本申请实施例提供的数据传输方法的一个示意性流程图。
在本申请实施例中,网络侧配置有执行机器学习的设备(下文称为机器学习设备),机器学习设备可以配置终端设备收集用于机器学习的数据,并根据终端设备收集的数据进行机器学习优化网络服务。在一种实施方式中,机器学习设备可以是与终端设备建立无线连接的网络设备(如图4中的第一网络设备),在另一种实施方式中,机器学习设备可以独立于第一网络设备。独立于第一网络设备的机器学习设备可以称为人工智能控制/控制器(artificial intelligence control/controller,AIC)、无线智能控制/控制器(radio intelligence control/controller,RIC)或其它名称,不予限制。
可选地,当机器学习设备独立于第一网络设备时,机器学习设备执行S410。
S410,机器学习设备向网络设备A发送配置信息A,该配置信息A用于配置终端设备收集第一数据。
相应地,该网络设备A接收来自该机器学习设备的该配置信息A。其中,第一数据为用于机器学习的数据。
作为示例非限定,第一数据可以是训练数据、模型参数梯度(信息)或推理结果。
其中,训练数据用于进行机器学习中的模型训练。例如,训练数据可以是终端设备收集或测量的原始数据、对原始数据进行处理(如归一化(normalization,也可以成为正规化、标准化))后的数据和特征工程数据(用来表示原始数据的特征数据)中的至少一种。模型参数梯度可以是机器学习的参与节点得到的更新的AI模型参数相对于更新前的AI模型参数的梯度,该更新的AI模型是参与节点基于更新前的AI模型和自己收集(或测量)的训练数据进行AI模型训练后得到的。推理结果可以是在应用中将实际数据输入AI模型进行推理后得到的推理结果。
可选地,该配置信息A可以指示终端设备在其中收集该第一数据的小区或收集该第一数据的第一区域。其中,第一区域包括至少一个小区。
例如,配置信息A包括小区的标识,用于指示终端设备收集第一数据的小区。机器学习设备通过该配置信息A向网络设备A通知终端设备需要收集第一数据的小区,由网络设备A转发至终端设备。或者,由网络设备A将该配置信息A透传至终端设备,以便终端设备根据 该配置信息A中的小区的标识确定用于收集第一数据的小区,终端设备在该小区的标识对应的小区中收集该第一数据,但本申请不限于此。
作为示例非限定,该小区的标识可以为以下至少一种:
小区全局标识(cell global identifier,CGI)、物理小区标识(physical cell identifier,PCI)及频点、小区标识(cell identifier)、非公网标识(non-public network identifier,NPN ID)和非陆地网络标识(non-terrestrial network identifier,NTN ID)。
再例如,配置信息A包括第一区域的标识,第一区域包括一个或多个小区。终端设备获取到该第一区域的标识后,当终端设备在第一区域中的小区的覆盖范围内时,收集第一数据,但本申请不限于此。可选地,当配置信息不包括区域标识的情况下,终端设备可以认为在服务小区内收集该第一数据。
作为示例非限定,该第一区域的标识为至少一个小区的标识,和/或,该第一区域的标识为机器学习区域标识、跟踪区码(tracking area code,TAC)、无线接入网通知区域码(radio access network notification area code,RANAC)、切片标识、业务标识或者其他可以标识区域的标识信息中的至少一种。
可选地,该配置信息A中包括指示信息A,该指示信息A用于指示收集第一数据为小区级任务还是区域级任务。
例如,当指示信息A指示收集第一数据为小区级任务时,终端设备可以认为网络配置终端设备在接收到该配置信息A的服务小区收集第一数据;或者,配置信息A包括小区的标识,终端设备在该小区的标识对应的小区中收集第一数据;当指示信息A指示收集第一数据为区域级任务时,终端设备可以认为网络配置终端设备在接收到该配置信息A的服务小区所在的区域收集第一数据;或者,配置信息A包括第一区域的标识,终端设备在该第一区域中收集第一数据,但本申请不限于此。
可选地,该配置信息A指示第一标识,该第一标识为机器学习设备的标识。
例如,配置信息A中包括该机器学习设备的标识,以标识配置终端设备收集第一数据的机器学习设备。
可选地,该配置信息A包括一个或多个任务的相关信息,其中包括第一任务的相关信息,该第一任务是用于收集第一数据的任务。
其中,第一任务的相关信息可以包括以下一种或多种:
第一任务的标识、第一任务的名称、第一任务的类型、和第一任务的动作类型。
其中,第一任务的标识用于标识该第一任务,或者说,该第一任务的标识用于指示收集第一数据的任务。
第一任务的类型可以是:收集原始数据(或称为原始数据收集)、模型发布、模型训练、推理和推理结果发布。
例如,第一任务的类型为收集原始数据,表示第一数据为原始数据,第一任务是用于收集原始数据的任务。或者,第一任务的类型为模型发布,表示第一数据为模型参数(例如,梯度信息等),第一任务是用于收集模型参数信息的任务,但本申请不限于此。
可选地,配置信息A中的第一任务的类型可以替换为第一数据的类型,该第一数据的类型可以用于指示终端设备需要收集的数据的类型,或者说,该第一数据的类型指示第一任务是用于收集该类型的数据的任务。可选地,第一数据的类型用于指示第一数据为训练数据、模型参数梯度(信息)或推理结果中的一种或多种。
第一任务的动作类型可以包括多种动作中的一种。其中,该多种动作中包括以下一种或 多种动作:开始、暂停、继续、激活和去激活。第一任务的动作类型还可以称为第一任务的状态。
可选地,配置信息A可以包括上报第一数据的条件、需要收集的数据参数或数据集。
例如,配置信息A指示上报模型参数梯度的条件,和/或进行模型训练的神经网络的参数等。
可选地,配置信息A指示至少一组组测量类型,以及可选地配置信息A还指示测量报告方式。
其中,测量类型可以参考第三代合作伙伴计划(3rd generation partnership project,3GPP)技术协议37.320中的测量类型,但本申请不限于此。测量类型也可以是根据为机器学习应用于移动网络时的需求新增的测量类型。测量报告方式可以是周期性地、一次性地、事件触发的或者可以缓存上报。其中,缓存上报可以指处于非连接态的终端设备收集第一数据后存储第一数据,当该终端设备转入连接态后向网络设备或机器学习设备发送该第一数据的机制。测量报告的具体方式可以通过配置信息A进行配置,本申请对此不做限定。
S420,网络设备A向终端设备发送配置信息B,配置信息B用于配置终端设备收集第一数据。
相应地,该终端设备接收来自该网络设备A的该配置信息B。
该配置信息B指示以下一种或多种内容:
终端设备用于收集该第一数据的小区、收集该第一数据的第一区域、机器学习设备的标识、第一任务的相关信息、上报第一数据的条件、训练的参考参数和训练的参考数据集。
其中,第一任务的相关信息指示第一任务的标识、第一任务的名称、第一任务的类型和动作类型中的一种或多种信息。
一种实施方式中,网络设备A为能够执行机器学习的设备。
网络设备A根据机器学习任务需求,向终端设备发送该配置信息B,以通知终端设备收集第一数据。
另一种实施方式,网络设备A根据S410中接收到的来自机器学习设备的配置信息A,确定机器学习设备需要终端设备收集第一数据。
可选地,网络设备A处理配置信息A后,根据配置信息A生成配置信息B并发送给终端设备。或者网络设备A接收到配置信息A后不做处理,将配置信息A透传至终端设备,配置信息B即为配置信息A,或者说,网络设备A不读取机器学习设备为终端设备配置的具体参数内容。
S430,终端设备收集第一数据。
终端设备在S420中接收到来自网络设备A的配置信息B,根据该配置信息B确定执行机器学习的设备配置终端设备收集第一数据,或者根据该配置信息B确定终端设备需要向网络设备A上报第一数据。终端设备根据该配置信息B收集第一数据。
例如,配置信息B包括第一区域的标识,第一区域包括多个小区。配置信息B指示终端设备收集并上报用于训练的原始数据,且还配置了测量类型以及采用周期性上报的测量上报方式。则终端设备在第一区域的覆盖范围内时,根据配置的测量类型收集用于训练的原始数据(即第一数据的一个示例),并根据指示的测量上报方式周期性地上报收集到的第一数据。
S440,终端设备向网络设备A发送数据段A(即第一数据段的一个示例)和第一信息。
相应地,网络设备A接收来自终端设备的该数据段A和第一信息。该数据段A是第一数据对应的一个或多个数据段中的一个。终端设备可以将第一数据作为一个数据段发送给网 络设备A;或者,终端设备将第一数据分段发送,比如第一数据被分为多个数据段,每个数据段包括第一数据的一部分。其中,数据段可以称为数据块、数据分段或数据包,本申请对此不做限定。
例如,用于机器学习的该第一数据的数据量较大,无法一次发送给网络设备A,则终端设备将第一数据分段发送。或者,执行机器学习的设备配置终端设备周期性地上报第一数据,终端设备在一个周期内收集到该第一数据的数据段A,并上报给网络设备。
该第一信息用于指示数据段A的目标设备,其中该目标设备为执行机器学习的设备。该执行机器学习的设备可以根据第一数据执行模型训练和/或推理。
当该网络设备A执行机器学习时,该目标设备为网络设备A。当执行机器学习的设备不是网络设备A,如图2中机器学习设备执行机器学习时,该目标设备为该机器学习设备。
可选地,第一信息包括第一任务的标识和/或第一标识,该第一标识用于标识执行机器学习的设备。
例如,第一信息包括第一任务的标识,以便网络设备A接收到该第一信息后确定第一数据对应的任务,如根据第一任务的标识确定该第一数据属于机器学习设备为终端设备配置的机器学习任务,并将该第一数据转发给机器学习设备。或者,网络设备A为执行机器学习的设备,网络设备A根据第一信息中的任务标识确定该第一数据属于网络设备A为终端设备配置的第一任务的数据,但本申请不限于此。
再例如,第一信息包括第一标识,以便网络设备A根据该第一标识确定第一数据的目标设备。如网络设备A为执行机器学习的设备,则该第一标识为网络设备A的标识,如执行机器学习的设备不是网络设备A而是图4中机器学习设备,则网络设备A确定该第一数据需要转发至机器学习设备,但本申请不限于此。
可选地,第一数据和第一信息承载在终端设备向网络设备A发送的第一消息中,该第一消息还包括第二信息。可选地,第一信息和第二信息还可以承载于不同的消息中。其中,该第二信息用于指示以下一项或多项:
该数据段A在第一数据对应的一个或多个数据段中的序列号;和,
该数据段A是否为第一数据对应的一个或多个数据段中的最后一个数据段;
第一时长,该第一时长为终端设备生成或存储数据段A到发送该数据段A之间的时间长度。
根据上述方案,第二信息包括数据段的序列号,可以使执行机器学习的设备接收到该数据段A后根据序列号对第一数据的多个数据段进行排序,以便正确读取第一数据。第二信息指示数据段A是否为第一数据对应的最后一个数据段,以便执行机器学习的设备可以确定是否接收到了第一数据对应的所有数据段,以对第一数据进行解码或分析。第二信息还可以指示第一时长,使得执行机器学习的设备接收到该第一时长后,可以确定该第一数据是否有效。
可选地,第二信息还指示小区A是否为第二区域中的一个小区,其中,该小区A是由该网络设备A管理的,且该小区A为终端设备的当前服务小区。
例如,当终端设备指示小区A是第二区域中的小区时,网络设备A确定该第一数据段可以转发至与网络设备A对应的(或者说建立连接的)执行机器学习的设备。
作为示例非限定,该第一消息为无线资源控制(radio resource control,RRC)消息。也就是说,该第一数据为控制面数据。
该第一时长具体可以是终端设备生成或存储该数据段A的时刻(例如,存储到终端设备的内存的时刻)到发送该RRC消息的时刻之间的时长,或者,该第一时长具体可以是终端设 备收到配置信息B到发送该第一消息之间的时长,但本申请不限于此。
一种实施方式中,该网络设备A是gNB,该gNB接收到来自终端设备的该RRC消息(即第一消息的一个示例)后,由gNB的RRC协议层处理该第一消息。
例如,终端设备收集到第一数据后,对第一数据分段并生成RRC消息,该RRC消息包括第一数据段和第一信息。gNB接收到该RRC消息后,由RRC协议层处理该第一消息,例如,读取第一消息中的第一信息确定目标设备等。
再例如,该gNB为执行机器学习的设备,gNB接收到来自终端设备的该RRC消息后,由RRC协议层根据该第一消息确定是否接收到该第一数据的所有数据段,收到第一数据的所有数据段后,该gNB根据第一数据执行机器学习,但本申请不限于此。
另一种实施方式中,该网络设备A是构成gNB的CU节点或CU-CP节点,该CU节点或CU-CP节点接收到来自终端设备的该RRC消息后由RRC协议层处理该第一消息。
例如,该网络设备A是CU-CP节点,该CU-CP节点接收到该第一消息后由RRC协议层处理该第一消息,例如,读取第一消息中的第一信息确定目标设备等。
再例如,该CU-CP节点为执行机器学习的设备,CU-CP节点接收到来自终端设备的该RRC消息后,由RRC协议层根据该第一消息确定是否接收到该第一数据的所有数据段,收到第一数据的所有数据段后,该CU-CP节点根据第一数据执行机器学习,但本申请不限于此。
作为示例非限定,该第一消息为层2控制信令。该层2控制信令可以为PDCP层控制协议数据单元(protocol data unit,PDU),RLC层控制PDU,MAC控制元素(control element,CE)的至少一种。
一种实施方式中,该网络设备A是构成gNB的DU节点,或CU节点和DU节点,该DU节点接收到来自终端设备的该层2控制信令后,由对应的层2协议层处理该层2信令消息。
例如,该网络设备A是DU节点,该DU节点接收到该第一消息后由对应的层2协议层(如RLC或MAC)处理该第一消息,例如,读取第一消息中的第一信息确定目标设备等。
再例如,该DU节点为执行机器学习的设备,DU节点接收到来自终端设备的层2信令消息后,由对应的层2协议层根据该层2信令消息确定是否接收到该第一数据的所有数据段,收到第一数据的所有数据段后,该DU节点根据第一数据执行机器学习,但本申请不限于此。
当网络设备A为执行机器学习的设备时,网络设备A通过上述步骤接收到第一数据后,利用第一数据进行机器学习。
例如,网络设备A配置终端设备收集训练数据的情况下,网络设备根据获取到的该第一数据,进行模型训练以得到训练模型,以便通过训练后的模型提供更好的网络服务,但本申请不限于此。
再例如,该第一数据为终端设备根据训练后的模型得到的推理结果,网络设备A根据该推理结果判断终端设备使用的模型的参数是否需要调整或者进一步对训练模型进行优化等,但本申请不限于此。
当网络设备A不是执行机器学习的设备时,网络设备A执行S450。
S450,网络设备A向机器学习设备发送数据段A和第二信息。
可选地,机器学习设备根据第二信息中的数据段A的序列号对接收到的数据段进行排列。在第二信息指示数据段A是否为第一数据对应的最后一个数据段的情况下,机器学习设备根据第二信息确定是否接收到第一数据对应的所有数据段。
相应地,机器学习设备接收来自网络设备A的数据段A和第二信息。在机器学习设备接 收到第一数据的所有数据段后,根据该第一数据执行机器学习。
根据上述方案,网络配置终端设备收集用于机器学习的数据,终端设备根据网络的配置执行数据收集,终端设备将收集到的数据上报给网络设备,并携带辅助信息,指示该数据对应的目标设备,以便网络设备根据辅助信息可以确定数据的目标设备。在用于机器学习的数据量较大的情况下,终端设备可以将数据分段上报给网络,辅助信息还可以指示数据段的序列号和/或是否为最后一个数据段,以便网络中执行机器学习的设备收到辅助信息后能够对数据段进行排序并确定是否接收到第一数据的全部数据段。能够实现海量数据准确地传递至网络中的执行机器学习的设备,使得网络能够根据机器学习的结果提供更好的通信服务,实现RAN智能化。
本申请实施例还提供了一种数据传输方法,考虑终端设备在发生小区切换的情况下如何将用于机器学习的数据上报至执行机器学习的设备,避免收集到的数据被浪费,提高通信效率。
图5为本申请实施例提供的另一个示意性流程图。
需要说明的是,在未另行定义或说明的情况下,图5实施例与图4实施例中相同或相似的部分可以参考图4实施例中的描述,为了简要,在此不再赘述。
S510,终端设备由网络设备A的覆盖范围移动至网络设备B的覆盖范围。
终端设备可以在图4所示的实施例中收集第一数据后或收集到部分第一数据后,移动至网络设备B的覆盖范围,或通过执行小区选择、小区重选或者切换后由网络设备B提供接入服务。终端设备与网络设备B建立无线连接,由网络设备B管理的小区B为该终端设备提供服务,即网络设备B管理的小区B为终端设备的当前服务小区。
例如,S510可能发生在S430之后S440之前,即终端设备收集第一数据后或收集到部分第一数据后发生了移动。或者,S510也可能发生在S440之后,如终端设备向网络设备A发送了部分第一数据的数据段后移动至小区B,但本申请不限于此。
S520,终端设备向网络设备B发送数据段B(即第一数据段的另一个示例)和第一信息。
相应地,网络设备B接收来自终端设备的该数据段B和该第一信息。该数据段B为第一数据对应的一个或多个数据段中的一个。该第一信息用于指示数据段B的目标设备。
可选地,第一数据和第一信息承载在终端设备向网络设备B发送的第一消息中,该第一消息还包括第二信息,该第二信息用于指示以下一项或多项:
第一任务的标识,该第一任务为收集第一数据的任务;
该数据段B在第一数据对应的一个或多个数据段中的序列号;
该数据段B是否为第一数据对应的一个或多个数据段中的最后一个数据段;和,
第一时长,该第一时长为终端设备生成或存储数据段B到发送该数据段B之间的时间长度。
作为示例非限定,该第一消息为无线资源控制(radio resource control,RRC)消息。
根据上述方案,终端设备发送该第二信息,执行机器学习的设备接收到该第二信息后,可以根据第二信息中的第一任务的标识确定数据段B对应的任务。执行机器学习的设备还可以根据序列号对数据段进行排序,和/或,根据数据段B是否是最后一个数据段确定是否接收到该第一数据的所有数据段,从而解码或分析第一数据,进行机器学习。以及,当第二信息包括第一时长时,根据第一时长确定第一数据是否有效。
第一信息指示目标设备的实施方式可以包括但不限于以下方式:
方式一,该目标设备为机器学习设备。该第一信息指示机器学习设备。
例如,第一信息中包括机器学习设备的标识,终端设备通过第一信息指示该数据段B的目标设备为该机器学习设备,网络设备B接收到该第一信息后根据该机器学习设备的标识确定将数据段B转发至该标识对应的机器学习设备,但本申请不限于此。
网络设备B根据第一信息确定机器学习设备后,执行图5中方式一对应的S530。
S530,网络设备B向机器学习设备发送数据段B和第二信息。
相应第,机器学习设备接收来自网络设备B的数据段B和第二信息。
机器学习设备接收到数据段B后,当第二信息指示数据段B的序列号时,机器学习设备可以根据第二信息中的数据段B的序列号对数据段进行排序。当第二信息指示数据段B是否为第一数据的最后一个数据段时,确定是否接收到第一数据的所有数据段,从而判断是否解码第一数据并根据第一数据进行机器学习。当第二信息指示第一时长时,机器学习设备可以根据该第一时长确定该第一数据是否有效。
方式二,该目标设备为网络设备C。该第一信息指示该网络设备C。
可选地,该网络设备C可以是向终端设备发送配置信息B的网络设备A,也可以是与执行机器学习的设备建立接口的其他网络设备,或者网络设备C为执行机器学习的设备。
可选地,该第一信息用于指示以下一种或多种内容:
网络设备C(即第三网络设备的一个示例)的标识;
第一小区的标识,该第一小区为网络设备C管理的小区;和,
第二区域的标识,第二区域包括至少一个小区,且第二区域中包括网络设备C管理的第一小区。
作为示例非限定,该第一小区的标识可以为以下至少一种:
CGI、PCI及频点、小区标识(cell identifier)、NPN ID和NTN ID。
作为示例非限定,该第二区域的标识为至少一个小区的标识,或者,该第二区域的标识为机器学习区域标识、TAC、RANAC、切片标识、业务标识或者其他可以标识区域的标识信息中的至少一种。
例如,第一信息中包括第一小区的CGI和TAC,网络设备B接收到该第一信息后根据第一小区的CGI确定该第一小区对应的网络设备C,从而将数据段B转发至网络设备C。具体地,若网络设备B和网络设备C之间有直接的网络设备间的接口(比如,F1,X2,Xn,E1,F1-C,F1-U或者其他网络设备间的接口中的至少一种),网络设备B根据第一小区的CGI确定对应的网络设备C后,可以将数据段B通过网络设备间的接口发送给网络设备C;若网络设备B和网络设备C之间没有直接的网络设备间的接口,网络设备B可以通过核心网设备向网络设备C发送数据段B以及该CGI和TAC,核心网设备可以根据CGI和TAC的部分或全部信息确定具体的转发路径。可选地,核心网设备可以向网络设备C发送网络设备B的网络标识信息和/或网络设备C的网络标识信息。进一步的,核心网设备还可以向网络设备C发送网络设备B的TAC和/或网络设备C的TAC。通过该机制,可以使得网络设备C获知发送第一信息的网络设备的相关信息,以便网络设备C优化数据的转发路径,比如网络设备C可以根据网络设备B的网络标识信息或者网络设备B的网络标识信息和TAC,与网络设备B建立直接的网络设备间的接口,以便网络设备B可以通过该接口向该网络设备C发送数据段B。或者,核心网设备可以向网络设备B发送网络设备C的相关信息,由网络设备B根据网络设备C的相关信息与网络设备C建立网络设备间的接口,从而通过该接口向网络设备C发送数据段B,具体过程可以参考上述描述,在此不再赘述。
可选地,网络设备B还可以向网络设备C发送第二信息。应理解,本申请不限于上述转 发机制。
或者,该第一信息包括第二区域的标识,网络设备B接收到该第一信息后,将数据段B转发至第二区域中的小区对应的一个或多个网络设备。具体地,若网络设备B可以确定其和第二区域内的网络设备,如网络设备C,之间有直接的网络设备间的接口,网络设备B可以将数据段B通过该接口转发给该网络设备C;若网络设备B无法通过网络设备间的接口向第二区域内的网络设备发送数据段B,网络设备B可以通过其连接的核心网设备或者机器学习节点发送数据段B向第二区域内的网络设备发送数据段B,但本申请不限于此。可选地,网络设备B还可以向第二区域内的网络设备发送第二信息。
可选地,第一小区为收集第一数据的小区,即配置信息B指示的终端设备收集第一数据的小区。
可选地,第二区域为第一区域,即第二区域为配置信息B指示的终端设备收集第一数据的区域。
网络设备B根据第一信息确定网络设备C后,执行图5中方式二对应的S530。
S530,网络设备B向网络设备C发送数据段B和第二信息。
相应地,网络设备C接收来自网络设备B的该数据段B和该第二信息。
S540,网络设备C向机器学习设备发送数据段B和第二信息。
例如,网络设备C在S530中接收到数据段B和第二信息后,确定该数据段B为用于机器学习的数据,网络设备C将该数据段B和第二信息C转发给与网络设备C建立连接的执行机器学习的设备,但本申请不限于此。可选地,网络设备C可以根据第二信息中第一任务的标识确定该数据段B为用于机器学习的数据。
再例如,在S520中,终端设备向网络设备B发送的第一信息中既指示了转发设备即网络设备C又指示了目标设备即机器学习设备。网络设备B在S530中还向网络设备C指示了数据段B的目标设备,即机器学习设备。网络设备C接收到数据段B和第二信息后根据网络设备B指示的目标设备,将该数据段B和第二信息在S540中发送给该机器学习设备,但本申请不限于此。
相应地,机器学习设备在S540接收来自网络设备C的该数据段B和第二信息。机器学习设备在接收到第一数据的所有数据段后根据第一数据进行机器学习。
例如,机器学习设备配置终端设备收集用于训练的数据的情况下,机器学习设备根据获取到的该第一数据进行模型训练以得到训练模型,以便通过训练后的模型提供更好的网络服务,但本申请不限于此。
再例如,该第一数据为终端设备根据训练后的模型得到的推理结果,机器学习设备根据该推理结果判断终端设备使用的模型的参数是否需要调整或者进一步对训练模型进行优化等,但本申请不限于此。
根据上述方案,网络配置终端设备收集用于机器学习的数据,终端设备根据网络的配置执行数据收集,终端设备将收集到的数据上报给网络设备,并携带辅助信息,及时在终端设备移动至其他小区的覆盖范围内时,也能够通过辅助信息通知移动至的网络设备B该第一数据的目标设备,以实现网络设备B将第一数据转发至执行机器学习的设备,能够避免收集到的数据被浪费,提高通信效率。能够实现海量数据准确地传递至网络中执行机器学习的设备,使得网络能够根据机器学习的结果提供更好的通信服务,实现RAN智能化。
以上,结合图2至图5详细说明了本申请实施例提供的方法。以下,结合图6至图8详细说明本申请实施例提供的装置。为了实现上述本申请实施例提供的方法中的各功能,各网 元可以包括硬件结构和/或软件模块,以硬件结构、软件模块、或硬件结构加软件模块的形式来实现上述各功能。上述各功能中的某个功能以硬件结构、软件模块、还是硬件结构加软件模块的方式来执行,取决于技术方案的特定应用和设计约束条件。
图6是本申请实施例提供的通信装置的示意性框图。如图6所示,该通信装置600可以包括处理单元610和收发单元620。
在一种可能的设计中,该通信装置600可对应于上文方法实施例中的终端设备,或者配置于(或用于)终端设备中的芯片,或者其他能够实现终端设备的方法的装置、模块、电路或单元等。
应理解,该通信装置600可对应于根据本申请实施例的方法400、500中的终端设备,该通信装置600可以包括用于执行图4、图5中的方法400、500中终端设备执行的方法的单元。并且,该通信装置600中的各单元和上述其他操作和/或功能分别为了实现图4、图5中的方法400、500的相应流程。
还应理解,该通信装置600为配置于(或用于)终端设备中的芯片时,该通信装置600中的收发单元620可以为芯片的输入/输出接口或电路,该通信装置600中的处理单元610可以为芯片中的处理器。
可选地,通信装置600还可以包括处理单元610,该处理单元610可以用于处理指令或者数据,以实现相应的操作。
可选地,通信装置600还可以包括存储单元630,该存储单元630可以用于存储指令或者数据,处理单元610可以执行该存储单元中存储的指令或者数据,以使该通信装置实现相应的操作。该通信装置600中的该通信装置600中的收发单元620为可对应于图7中示出的终端设备700中的收发器710,存储单元630可对应于图7中示出的终端设备700中的存储器。
应理解,各单元执行上述相应步骤的具体过程在上述方法实施例中已经详细说明,为了简洁,在此不再赘述。
还应理解,该通信装置600为终端设备时,该通信装置600中的收发单元620为可通过通信接口(如收发器或输入/输出接口)实现,例如可对应于图7中示出的终端设备700中的收发器710,该通信装置600中的处理单元610可通过至少一个处理器实现,例如可对应于图7中示出的终端设备700中的处理器720,该通信装置600中的处理单元610可通过至少一个逻辑电路实现。
在另一种可能的设计中,该通信装置600可对应于上文方法实施例中的网络设备,例如,或者配置于(或用于)网络设备中的芯片,或者其他能够实现网络设备的方法的装置、模块、电路或单元等。
应理解,该通信装置600可对应于根据本申请实施例的方法400、500中的网络设备,如网络设备A、网络设备B或网路设备C。该通信装置600可以包括用于执行图4、图5中的方法400、500中网络设备执行的方法的单元。并且,该通信装置600中的各单元和上述其他操作和/或功能分别为了实现图4、图5中的方法400、500的相应流程。
还应理解,该通信装置600为配置于(或用于)网络设备中的芯片时,该通信装置600中的收发单元为芯片中的输入/输出接口或电路,该通信装置600中的处理单元610可为芯片中的处理器。
可选地,通信装置600还可以包括处理单元610,该处理单元610可以用于处理指令或者数据,以实现相应的操作。
可选地,通信装置600还可以包括存储单元630,该存储单元可以用于存储指令或者数据,处理单元可以执行该存储单元630中存储的指令或者数据,以使该通信装置实现相应的操作。该通信装置600中的存储单元630为可对应于图8中示出的网络设备800中的存储器。
应理解,各单元执行上述相应步骤的具体过程在上述方法实施例中已经详细说明,为了简洁,在此不再赘述。
还应理解,该通信装置600为网络设备时,该通信装置600中的收发单元620为可通过通信接口(如收发器或输入/输出接口)实现,例如可对应于图8中示出的网络设备800中的收发器810,该通信装置600中的处理单元610可通过至少一个处理器实现,例如可对应于图8中示出的网络设备800中的处理器820,该通信装置600中的处理单元610可通过至少一个逻辑电路实现。
在另一种可能的设计中,该通信装置600可对应于上文方法实施例中的执行机器学习的设备,例如,或者配置于(或用于)执行机器学习的设备中的芯片,或者其他能够执行机器设备的方法的装置、模块、电路或单元等。
应理解,该通信装置600可对应于根据本申请实施例的方法400、500中的执行机器学习的设备,该通信装置600可以包括用于执行图4、图5中的方法400、500中执行机器学习的设备执行的方法的单元。并且,该通信装置600中的各单元和上述其他操作和/或功能分别为了实现图4、图5中的方法400、500的相应流程。
还应理解,该通信装置600为配置于(或用于)执行机器学习的设备中的芯片时,该通信装置600中的收发单元为芯片中的输入/输出接口或电路,该通信装置600中的处理单元610可为芯片中的处理器。
可选地,通信装置600还可以包括处理单元610,该处理单元610可以用于处理指令或者数据,以实现相应的操作。
可选地,通信装置600还可以包括存储单元630,该存储单元可以用于存储指令或者数据,处理单元可以执行该存储单元630中存储的指令或者数据,以使该通信装置实现相应的操作。该通信装置600中的存储单元630为可对应于图9中示出的执行机器学习的设备900中的存储器。
应理解,各单元执行上述相应步骤的具体过程在上述方法实施例中已经详细说明,为了简洁,在此不再赘述。
还应理解,该通信装置600为执行机器学习的设备时,该通信装置600中的收发单元620为可通过通信接口(如收发器或输入/输出接口)实现,例如可对应于图9中示出的执行机器学习的设备900中的收发器910,该通信装置600中的处理单元610可通过至少一个处理器实现,例如可对应于图9中示出的执行机器学习的设备900中的处理器920,该通信装置600中的处理单元610可通过至少一个逻辑电路实现。
图7是本申请实施例提供的终端设备600的结构示意图。该终端设备700可应用于如图1所示的系统中,执行上述方法实施例中终端设备的功能。如图所示,该终端设备700包括处理器720和收发器710。可选地,该终端设备700还包括存储器。其中,处理器720、收发器710和存储器之间可以通过内部连接通路互相通信,传递控制和/或数据信号,该存储器用于存储计算机程序,该处理器720用于执行该存储器中的该计算机程序,以控制该收发器710收发信号。
上述处理器720可以和存储器可以合成一个处理装置,处理器720用于执行存储器中存储的程序代码来实现上述功能。具体实现时,该存储器也可以集成在处理器720中,或者独 立于处理器720。该处理器720可以与图6中的处理单元对应。
上述收发器710可以与图6中的收发单元对应。收发器710可以包括接收器(或称接收机、接收电路)和发射器(或称发射机、发射电路)。其中,接收器用于接收信号,发射器用于发射信号。
应理解,图7所示的终端设备700能够实现图4、图5所示方法实施例中涉及终端设备的过程。终端设备700中的各个模块的操作和/或功能,分别为了实现上述方法实施例中的相应流程。具体可参见上述方法实施例中的描述,为避免重复,此处适当省略详细描述。
上述处理器720可以用于执行前面方法实施例中描述的由终端设备内部实现的动作,而收发器710可以用于执行前面方法实施例中描述的终端设备向网络设备发送或从网络设备接收的动作。具体请见前面方法实施例中的描述,此处不再赘述。
可选地,上述终端设备700还可以包括电源,用于给终端设备中的各种器件或电路提供电源。
除此之外,为了使得终端设备的功能更加完善,该终端设备700还可以包括输入单元、显示单元、音频电路、摄像头和传感器等中的一个或多个,所述音频电路还可以包括扬声器、麦克风等。
图8是本申请实施例提供的网络设备的结构示意图,该网络设备800可应用于如图1所示的系统中,执行上述方法实施例中网络设备的功能。例如可以为网络设备的相关结构的示意图。
应理解,图8所示的网络设备800能够实现图4、图5所示方法实施例中涉及网络设备(如网络设备A、网络设备B或网络设备C)的各个过程。网络设备800中的各个模块的操作和/或功能,分别为了实现上述方法实施例中的相应流程。具体可参见上述方法实施例中的描述,为避免重复,此处适当省略详细描述。
应理解,图8所示出的网络设备800可以是eNB或gNB,可选地,网络设备包含CU、DU和AAU的网络设备等,可选地,CU可以具体分为CU-CP和CU-UP。本申请对于网络设备的具体架构不作限定。
应理解,图8所示出的网络设备800可以是CU节点或CU-CP节点。
本申请实施例还提供了一种处理装置,包括处理器和(通信)接口;所述处理器用于执行上述任一方法实施例中的方法。
应理解,上述处理装置可以是一个或多个芯片。例如,该处理装置可以是现场可编程门阵列(field programmable gate array,FPGA),可以是专用集成芯片(application specific integrated circuit,ASIC),还可以是系统芯片(system on chip,SoC),还可以是中央处理器(central processor unit,CPU),还可以是网络处理器(network processor,NP),还可以是数字信号处理电路(digital signal processor,DSP),还可以是微控制器(micro controller unit,MCU),还可以是可编程控制器(programmable logic device,PLD)或其他集成芯片。
根据本申请实施例提供的方法,本申请还提供一种计算机程序产品,该计算机程序产品包括:计算机程序代码,当该计算机程序代码由一个或多个处理器执行时,使得包括该处理器的装置执行图4、图5所示实施例中的方法。
本申请实施例提供的技术方案可以全部或部分地通过软件、硬件、固件或者其任意组合来实现。当使用软件实现时,可以全部或部分地以计算机程序产品的形式实现。所述计算机程序产品包括一个或多个计算机指令。在计算机上加载和执行所述计算机程序指令时,全部或部分地产生按照本发明实施例所述的流程或功能。所述计算机可以是通用计算机、专用计 算机、计算机网络、网络设备、终端设备、核心网设备、机器学习设备或者其他可编程装置。所述计算机指令可以存储在计算机可读存储介质中,或者从一个计算机可读存储介质向另一个计算机可读存储介质传输,例如,所述计算机指令可以从一个网站站点、计算机、服务器或数据中心通过有线(例如同轴电缆、光纤、数字用户线(digital subscriber line,DSL))或无线(例如红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。所述计算机可读存储介质可以是计算机可以存取的任何可用介质或者是包含一个或多个可用介质集成的服务器、数据中心等数据存储设备。所述可用介质可以是磁性介质(例如,软盘、硬盘、磁带)、光介质(例如,数字视频光盘(digital video disc,DVD))、或者半导体介质等。
根据本申请实施例提供的方法,本申请还提供一种计算机可读存储介质,该计算机可读存储介质存储有程序代码,当该程序代码由一个或多个处理器运行时,使得包括该处理器的装置执行图4、图5所示实施例中的方法。
根据本申请实施例提供的方法,本申请还提供一种系统,其包括前述的一个或多个网络设备。还系统还可以进一步包括前述的一个或多个终端设备。
上述各个装置实施例中网络设备与终端设备和方法实施例中的网络设备或终端设备完全对应,由相应的模块或单元执行相应的步骤,例如通信单元(收发器)执行方法实施例中接收或发送的步骤,除发送、接收外的其它步骤可以由处理单元(处理器)执行。具体单元的功能可以参考相应的方法实施例。其中,处理器可以为一个或多个。
本领域普通技术人员可以意识到,结合本文中所公开的实施例描述的各示例的单元及算法步骤,能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行,取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能,但是这种实现不应认为超出本申请的范围。
所属领域的技术人员可以清楚地了解到,为描述的方便和简洁,上述描述的系统、装置和单元的具体工作过程,可以参考前述方法实施例中的对应过程,在此不再赘述。
在本申请所提供的几个实施例中,应该理解到,所揭露的系统、装置和方法,可以通过其它的方式实现。例如,以上所描述的装置实施例仅仅是示意性的,例如,所述单元的划分,仅仅为一种逻辑功能划分,实际实现时可以有另外的划分方式,例如多个单元或组件可以结合或者可以集成到另一个系统,或一些特征可以忽略,或不执行。另一点,所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口,装置或单元的间接耦合或通信连接,可以是电性,机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的,作为单元显示的部件可以是或者也可以不是物理单元,即可以位于一个地方,或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
在本申请实施例中,在无逻辑矛盾的前提下,各实施例之间可以相互引用,例如方法实施例之间的方法和/或术语可以相互引用,例如装置实施例之间的功能和/或术语可以相互引用,例如装置实施例和方法实施例之间的功能和/或术语可以相互引用。
以上所述,仅为本申请的具体实施方式,但本申请的保护范围并不局限于此,任何熟悉本技术领域的技术人员在本申请揭露的技术范围内,可轻易想到变化或替换,都应涵盖在本申请的保护范围之内。因此,本申请的保护范围应以所述权利要求的保护范围为准。

Claims (60)

  1. 一种数据传输方法,其特征在于,包括:
    向第一网络设备发送第一数据段和第一信息,所述第一数据段是第一数据对应的一个或多个数据段中的一个,所述第一数据为用于机器学习的数据,所述第一信息用于指示所述第一数据段的目标设备。
  2. 根据权利要求1所述的方法,其特征在于,所述方法还包括:
    接收来自所述第一网络设备或第二网络设备的配置信息,所述配置信息用于配置终端设备收集所述第一数据。
  3. 根据权利要求2所述的方法,其特征在于,所述配置信息用于指示以下一种或多种内容:
    所述终端设备收集所述第一数据的小区;
    所述终端设备收集所述第一数据的第一区域,其中,所述第一区域包括至少一个小区;
    第一标识,用于标识执行所述机器学习的设备;
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;和,
    所述第一任务的类型。
  4. 根据权利要求1至3中任一项所述的方法,其特征在于,所述目标设备为执行所述机器学习的设备。
  5. 根据权利要求1至3中任一项所述的方法,其特征在于,所述目标设备为第三网络设备。
  6. 根据权利要求5所述的方法,其特征在于,所述第一信息用于指示所述第一数据段的目标设备,包括:所述第一信息用于指示以下一种或多种内容:
    所述第三网络设备的标识;
    第一小区的标识,其中,所述第一小区为所述第三网络设备管理的小区;和,
    第二区域的标识,其中,所述第二区域中包括至少一个小区,且所述至少一个小区中包括所述第三网络设备管理的第一小区。
  7. 根据权利要求1至6中任一项所述的方法,其特征在于,所述第一信息承载在第一消息中,所述第一消息还包括第二信息,所述第二信息用于指示以下内容中的一种或多种:
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;
    所述第一数据段在所述一个或多个数据段中的序列号;
    所述第一数据段是否为所述一个或多个数据段中的最后一个数据段;
    第二小区是否为第二区域中的一个小区,其中,所述第二小区是由所述第一网络设备管理的,且所述第二小区为终端设备的服务小区,所述第二区域包括至少一个小区;和,
    第一时长,其中,所述第一时长为所述终端设备生成所述第一数据段到发送所述第一数据段之间的时间长度。
  8. 根据权利要求1至7中任一项所述的方法,其特征在于,所述第一数据包括以下一种或多种数据:
    训练数据、模型参数梯度和推理结果。
  9. 一种数据传输方法,其特征在于,包括:
    接收来自终端设备的第一数据段和第一信息,所述第一数据段是第一数据对应的一个或多个数据段中的一个,所述第一数据为用于机器学习的数据,所述第一信息用于指示所述第一数据段的目标设备;
    向所述目标设备发送所述第一数据段。
  10. 根据权利要求9所述的方法,其特征在于,所述方法还包括:
    向所述终端设备发送配置信息,所述配置信息用于配置所述终端设备收集所述第一数据。
  11. 根据权利要求10所述的方法,其特征在于,所述配置信息用于指示以下一种或多种内容:
    所述终端设备收集所述第一数据的小区;
    所述终端设备收集所述第一数据的第一区域,其中,所述第一区域包括至少一个小区;
    第一标识,用于标识执行所述机器学习的设备;
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;和,
    所述第一任务的类型。
  12. 根据权利要求9至11中任一项所述的方法,其特征在于,所述目标设备为执行所述机器学习的设备。
  13. 根据权利要求9至11中任一项所述的方法,其特征在于,所述目标设备为第三网络设备。
  14. 根据权利要求13所述的方法,其特征在于,所述第一信息用于指示以下一种或多种内容:
    所述第三网络设备的标识;
    第一小区的标识,其中,所述第一小区为所述第三网络设备管理的小区;和,
    第二区域的标识,其中,所述第二区域包括至少一个小区,且所述至少一个小区中包括所述第三网络设备管理的第一小区。
  15. 根据权利要求9至14中任一项所述的方法,其特征在于,所述第一信息承载在第一消息中,第一消息还包括第二信息,所述第二信息用于指示以下一种或多种内容:
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;
    所述第一数据段在所述一个或多个数据段中的序列号;
    所述第一数据段是否为所述一个或多个数据段中的最后一个数据段;
    第二小区是否为第二区域中的一个小区,所述第二小区是由第一网络设备管理的,且所述第二小区为所述终端设备的服务小区,其中,所述第二区域包括至少一个小区;和,
    第一时长,所述第一时长为所述终端设备生成所述第一数据段到发送所述第一数据段之间的时间长度。
  16. 根据权利要求15所述的方法,其特征在于,所述方法还包括:
    向所述目标设备发送所述第二信息。
  17. 根据权利要求9至16中任一项所述的方法,其特征在于,所述第一数据包括以下一种或多种数据:
    训练数据、模型参数梯度和推理结果。
  18. 一种数据传输方法,其特征在于,包括:
    接收来自第一网络设备的第一数据段,所述第一数据段是第一数据对应的一个或多个数据段中的一个,所述第一数据为用于机器学习的数据;
    向执行所述机器学习的设备发送所述第一数据段。
  19. 根据权利要求18所述的方法,其特征在于,所述方法还包括:
    接收来自所述执行机器学习的设备的配置信息,所述配置信息用于配置终端设备收集所述第一数据。
  20. 根据权利要求19所述的方法,其特征在于,所述方法还包括:
    向所述终端设备发送所述配置信息。
  21. 根据权利要求19或20所述的方法,其特征在于,所述配置信息用于指示以下一种或多种内容:
    所述终端设备收集所述第一数据的小区;
    所述终端设备收集所述第一数据的第一区域,其中,所述第一区域包括至少一个小区;
    第一标识,用于标识执行所述机器学习的设备;
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;和,
    所述第一任务的类型。
  22. 根据权利要求18至21中任一项所述的方法,其特征在于,所述方法还包括:
    接收来自所述第一网络设备的第二信息,其中,所述第二信息用于指示以下一种或多种内容:
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;
    所述第一数据段在所述一个或多个数据段中的序列号;
    所述第一数据段是否为所述一个或多个数据段中的最后一个数据段;和,
    第一时长,所述第一时长为终端设备生成所述第一数据段到发送所述第一数据段之间的时间长度。
  23. 根据权利要求22所述的方法,其特征在于,所述方法还包括:
    向所述执行所述机器学习的设备发送所述第二信息。
  24. 根据权利要求18至23中任一项所述的方法,其特征在于,所述第一数据包括以下一种或多种数据:
    训练数据、模型参数梯度和推理结果。
  25. 一种数据传输方法,其特征在于,包括:
    向第一网络设备或第二网络设备发送配置信息,所述配置信息用于配置终端设备收集第一数据,所述第一数据为用于机器学习的数据;
    接收来自第一设备的第一数据段,所述第一数据段是第一数据对应的一个或多个数据段中的一个,
    其中,所述第一设备为第一网络设备或第三网络设备。
  26. 根据权利要求25所述的方法,其特征在于,所述配置信息用于指示以下一种或多种内容:
    所述终端设备收集所述第一数据的小区;
    所述终端设备收集所述第一数据的第一区域,其中,所述第一区域包括至少一个小区;
    第一标识,用于标识执行所述机器学习的设备;
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;和,
    所述第一任务的类型。
  27. 根据权利要求25或26所述的方法,其特征在于,所述方法还包括:
    接收来自所述第一设备的第二信息,其中,所述第二信息用于指示以下一种或多种内容:
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;
    所述第一数据段在所述一个或多个数据段中的序列号;
    所述第一数据段是否为所述一个或多个数据段中的最后一个数据段;
    第一时长,所述第一时长为所述终端设备生成所述第一数据段到发送所述第一数据段之间的时间长度。
  28. 一种数据传输装置,其特征在于,包括:
    处理单元,用于收集第一数据;
    收发单元,用于向第一网络设备发送第一数据段和第一信息,所述第一数据段是第一数据对应的一个或多个数据段中的一个,所述第一数据为用于机器学习的数据,所述第一信息用于指示所述第一数据段的目标设备。
  29. 根据权利要求28所述的装置,其特征在于,
    所述收发单元还用于接收来自所述第一网络设备或第二网络设备的配置信息,所述配置信息用于配置终端设备收集所述第一数据。
  30. 根据权利要求29所述的装置,其特征在于,所述配置信息用于指示以下一种或多种内容:
    所述终端设备收集所述第一数据的小区;
    所述终端设备收集所述第一数据的第一区域,其中,所述第一区域包括至少一个小区;
    第一标识,用于标识执行所述机器学习的设备;
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;和,
    所述第一任务的类型。
  31. 根据权利要求28至30中任一项所述的装置,其特征在于,所述目标设备为执行所述机器学习的设备。
  32. 根据权利要求28至30中任一项所述的装置,其特征在于,所述目标设备为第三网络设备。
  33. 根据权利要求32所述的装置,其特征在于,所述第一信息用于指示所述第一数据段的目标设备,包括:所述第一信息用于指示以下一种或多种内容:
    所述第三网络设备的标识;
    第一小区的标识,其中,所述第一小区为所述第三网络设备管理的小区;和,
    第二区域的标识,其中,所述第二区域中包括至少一个小区,且所述至少一个小区中包括所述第三网络设备管理的第一小区。
  34. 根据权利要求28至33中任一项所述的装置,其特征在于,所述第一信息承载在第一消息中,所述第一消息还包括第二信息,所述第二信息用于指示以下内容中的一种或多种:
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;
    所述第一数据段在所述一个或多个数据段中的序列号;
    所述第一数据段是否为所述一个或多个数据段中的最后一个数据段;
    第二小区是否为第二区域中的一个小区,其中,所述第二小区是由所述第一网络设备管理的,且所述第二小区为终端设备的服务小区,所述第二区域包括至少一个小区;和,
    第一时长,其中,所述第一时长为所述终端设备生成所述第一数据段到发送所述第一数据段之间的时间长度。
  35. 根据权利要求28至34中任一项所述的装置,其特征在于,所述第一数据包括以下一种或多种数据:
    训练数据、模型参数梯度和推理结果。
  36. 一种数据传输装置,其特征在于,包括:
    收发单元,用于接收来自终端设备的第一数据段和第一信息,所述第一数据段是第一数据对应的一个或多个数据段中的一个,所述第一数据为用于机器学习的数据,所述第一信息用于指示所述第一数据段的目标设备;
    处理单元,用于根据所述第一信息确定目标设备;
    所述收发单元还用于向所述目标设备发送所述第一数据段。
  37. 根据权利要求36所述的装置,其特征在于,
    所述收发单元还用于向所述终端设备发送配置信息,所述配置信息用于配置所述终端设备收集所述第一数据。
  38. 根据权利要求37所述的装置,其特征在于,所述配置信息用于指示以下一种或多种内容:
    所述终端设备收集所述第一数据的小区;
    所述终端设备收集所述第一数据的第一区域,其中,所述第一区域包括至少一个小区;
    第一标识,用于标识执行所述机器学习的设备;
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;和,
    所述第一任务的类型。
  39. 根据权利要求36至38中任一项所述的装置,其特征在于,所述目标设备为执行所述机器学习的设备。
  40. 根据权利要求36至38中任一项所述的装置,其特征在于,所述目标设备为第三网络设备。
  41. 根据权利要求40所述的装置,其特征在于,所述第一信息用于指示以下一种或多种内容:
    所述第三网络设备的标识;
    第一小区的标识,其中,所述第一小区为所述第三网络设备管理的小区;和,
    第二区域的标识,其中,所述第二区域包括至少一个小区,且所述至少一个小区中包括所述第三网络设备管理的第一小区。
  42. 根据权利要求36至41中任一项所述的装置,其特征在于,所述第一信息承载在第一消息中,第一消息还包括第二信息,所述第二信息用于指示以下一种或多种内容:
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;
    所述第一数据段在所述一个或多个数据段中的序列号;
    所述第一数据段是否为所述一个或多个数据段中的最后一个数据段;
    第二小区是否为第二区域中的一个小区,所述第二小区是由第一网络设备管理的,且所述第二小区为所述终端设备的服务小区,其中,所述第二区域包括至少一个小区;和,
    第一时长,所述第一时长为所述终端设备生成所述第一数据段到发送所述第一数据段之间的时间长度。
  43. 根据权利要求42所述的装置,其特征在于,
    所述收发单元还用于向所述目标设备发送所述第二信息。
  44. 根据权利要求36至43中任一项所述的装置,其特征在于,所述第一数据包括以下一种或多种数据:
    训练数据、模型参数梯度和推理结果。
  45. 一种数据传输装置,其特征在于,包括:
    收发单元,用于接收来自第一网络设备的第一数据段,所述第一数据段是第一数据对应的一个或多个数据段中的一个,所述第一数据为用于机器学习的数据;
    处理单元,用于确定执行所述机器学习的设备;
    所述收发单元还用于向执行所述机器学习的设备发送所述第一数据段。
  46. 根据权利要求45所述的装置,其特征在于,
    所述收发单元还用于接收来自所述执行机器学习的设备的配置信息,所述配置信息用于配置终端设备收集所述第一数据。
  47. 根据权利要求46所述的装置,其特征在于,
    所述收发单元还用于向所述终端设备发送所述配置信息。
  48. 根据权利要求46或47所述的装置,其特征在于,所述配置信息用于指示以下一种或多种内容:
    所述终端设备收集所述第一数据的小区;
    所述终端设备收集所述第一数据的第一区域,其中,所述第一区域包括至少一个小区;
    第一标识,用于标识执行所述机器学习的设备;
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;和,
    所述第一任务的类型。
  49. 根据权利要求45至48中任一项所述的装置,其特征在于,
    所述收发单元还用于接收来自所述第一网络设备的第二信息,其中,所述第二信息用于指示以下一种或多种内容:
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;
    所述第一数据段在所述一个或多个数据段中的序列号;
    所述第一数据段是否为所述一个或多个数据段中的最后一个数据段;和,
    第一时长,所述第一时长为终端设备生成所述第一数据段到发送所述第一数据段之间的时间长度。
  50. 根据权利要求49所述的装置,其特征在于,
    所述收发单元还用于向所述执行所述机器学习的设备发送所述第二信息。
  51. 根据权利要求45至50中任一项所述的装置,其特征在于,所述第一数据包括以下一种或多种数据:
    训练数据、模型参数梯度和推理结果。
  52. 一种数据传输装置,其特征在于,包括:
    处理单元,用于确定配置信息,所述配置信息用于配置终端设备收集第一数据,所述第一数据为用于机器学习的数据;
    收发单元,用于向第一网络设备或第二网络设备发送所述配置信息;
    所述收发单元还用于接收来自第一设备的第一数据段,所述第一数据段是第一数据对应的一个或多个数据段中的一个,
    其中,所述第一设备为第一网络设备或第三网络设备。
  53. 根据权利要求52所述的装置,其特征在于,所述配置信息用于指示以下一种或多种内容:
    所述终端设备收集所述第一数据的小区;
    所述终端设备收集所述第一数据的第一区域,其中,所述第一区域包括至少一个小区;
    第一标识,用于标识执行所述机器学习的设备;
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;和,
    所述第一任务的类型。
  54. 根据权利要求52或53所述的装置,其特征在于,
    所述收发单元还用于接收来自所述第一设备的第二信息,其中,所述第二信息用于指示 以下一种或多种内容:
    第一任务的标识,其中,所述第一任务是用于收集所述第一数据的任务;
    所述第一数据段在所述一个或多个数据段中的序列号;
    所述第一数据段是否为所述一个或多个数据段中的最后一个数据段;
    第一时长,所述第一时长为所述终端设备生成所述第一数据段到发送所述第一数据段之间的时间长度。
  55. 一种通信装置,其特征在于,用于实现如权利要求1至27中任一项所述的方法。
  56. 一种通信装置,其特征在于,包括处理器和存储器,所述存储器和所述处理器耦合,所述处理器用于执行权利要求1至27中任一项所述的方法。
  57. 一种通信装置,其特征在于,包括处理器和通信接口,所述处理器利用所述通信接口,执行权利要求1至27中任一项所述的方法。
  58. 一种计算机可读存储介质,其特征在于,所述计算机可读存储介质上存储有指令,当所述指令在计算机上运行时,使得计算机执行权利要求1至27中任一项所述的方法。
  59. 一种计算机程序产品,其特征在于,包括指令,当所述指令在计算机上运行时,使得计算机执行权利要求1至27中任一项所述的方法。
  60. 一种通信系统,其特征在于,包括以下至少两个通信装置:
    第一通信装置,用于实现如权利要求1至8中任一项所述的方法,
    第二通信装置,用于实现如权利要求9至17中任一项所述的方法,
    第三通信装置,用于实现如权利要求18至24中任一项所述的方法,或,
    第四通信装置,用于实现如权利要求25至27中任一项所述的方法。
PCT/CN2021/115798 2020-09-04 2021-08-31 数据传输方法和装置 WO2022048546A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP21863610.8A EP4203542A4 (en) 2020-09-04 2021-08-31 DATA TRANSMISSION METHOD AND DEVICE
US18/177,432 US20230262490A1 (en) 2020-09-04 2023-03-02 Data transmission method and apparatus

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010923904.7A CN114143802A (zh) 2020-09-04 2020-09-04 数据传输方法和装置
CN202010923904.7 2020-09-04

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/177,432 Continuation US20230262490A1 (en) 2020-09-04 2023-03-02 Data transmission method and apparatus

Publications (1)

Publication Number Publication Date
WO2022048546A1 true WO2022048546A1 (zh) 2022-03-10

Family

ID=80438422

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/115798 WO2022048546A1 (zh) 2020-09-04 2021-08-31 数据传输方法和装置

Country Status (4)

Country Link
US (1) US20230262490A1 (zh)
EP (1) EP4203542A4 (zh)
CN (1) CN114143802A (zh)
WO (1) WO2022048546A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2024020747A1 (zh) * 2022-07-25 2024-02-01 北京小米移动软件有限公司 一种模型的生成方法及装置
WO2024035981A1 (en) * 2022-08-10 2024-02-15 Qualcomm Incorporated Machine learning (ml) data input configuration and reporting
WO2024130522A1 (zh) * 2022-12-19 2024-06-27 北京小米移动软件有限公司 数据处理方法、装置
WO2024169522A1 (zh) * 2023-02-15 2024-08-22 华为技术有限公司 通信方法和装置

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20230239239A1 (en) * 2022-01-25 2023-07-27 Qualcomm Incorporated Upper analog media access control (mac-a) layer functions for analog transmission protocol stack
WO2024011474A1 (zh) * 2022-07-14 2024-01-18 华为技术有限公司 一种通信方法和装置
WO2024138611A1 (zh) * 2022-12-30 2024-07-04 Oppo广东移动通信有限公司 通信方法和通信设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160323768A1 (en) * 2014-01-30 2016-11-03 Lg Electronics Inc. Method and apparatus for supporting minimization of drive tests for radio access network sharing in wireless communication system
CN110119808A (zh) * 2018-02-06 2019-08-13 华为技术有限公司 一种基于机器学习的数据处理方法以及相关设备
CN110121180A (zh) * 2018-02-05 2019-08-13 华为技术有限公司 一种数据分析装置、系统及方法
CN111444848A (zh) * 2020-03-27 2020-07-24 广州英码信息科技有限公司 一种基于联邦学习的特定场景模型升级方法和系统

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3903244A1 (en) * 2018-12-28 2021-11-03 Telefonaktiebolaget LM Ericsson (publ) A wireless device, a network node and methods therein for training of a machine learning model
CA3126601A1 (en) * 2019-01-13 2020-07-16 Strong Force Iot Portfolio 2016, Llc Methods, systems, kits and apparatuses for monitoring and managing industrial settings

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160323768A1 (en) * 2014-01-30 2016-11-03 Lg Electronics Inc. Method and apparatus for supporting minimization of drive tests for radio access network sharing in wireless communication system
CN110121180A (zh) * 2018-02-05 2019-08-13 华为技术有限公司 一种数据分析装置、系统及方法
CN110119808A (zh) * 2018-02-06 2019-08-13 华为技术有限公司 一种基于机器学习的数据处理方法以及相关设备
CN111444848A (zh) * 2020-03-27 2020-07-24 广州英码信息科技有限公司 一种基于联邦学习的特定场景模型升级方法和系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP4203542A4

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2024020747A1 (zh) * 2022-07-25 2024-02-01 北京小米移动软件有限公司 一种模型的生成方法及装置
WO2024035981A1 (en) * 2022-08-10 2024-02-15 Qualcomm Incorporated Machine learning (ml) data input configuration and reporting
WO2024130522A1 (zh) * 2022-12-19 2024-06-27 北京小米移动软件有限公司 数据处理方法、装置
WO2024169522A1 (zh) * 2023-02-15 2024-08-22 华为技术有限公司 通信方法和装置

Also Published As

Publication number Publication date
EP4203542A4 (en) 2024-02-28
US20230262490A1 (en) 2023-08-17
EP4203542A1 (en) 2023-06-28
CN114143802A (zh) 2022-03-04

Similar Documents

Publication Publication Date Title
WO2022048546A1 (zh) 数据传输方法和装置
US20230209390A1 (en) Intelligent Radio Access Network
EP4181556A1 (en) Artificial intelligence-based communication method and communication device
US20230300686A1 (en) Communication system for machine learning metadata
JP7495524B2 (ja) ローカルエリア機械学習無線リソース管理のための測定構成
WO2020200034A1 (zh) 一种网络接入的方法和装置
CN113966592A (zh) 更新在应用功能与核心网络之间协商的后台数据传输策略的方法、策略控制功能、以及应用功能
US20230023603A1 (en) Measurement object indication method and apparatus
CN115150847A (zh) 模型处理的方法、通信装置和系统
WO2023143267A1 (zh) 一种模型配置方法及装置
WO2022165846A1 (zh) 用于获取数据的方法和装置
CN117044361A (zh) 无线电接入网中确定性业务和非确定性业务的优化
WO2021027907A1 (zh) 通信方法和通信装置
WO2022209234A1 (ja) Ranノード、ue及び方法
WO2023280143A1 (zh) Ai任务的控制方法、终端、基站及存储介质
WO2022213897A1 (zh) 通信方法及通信装置
WO2021179165A1 (zh) 模型协调方法及装置
EP3833107A1 (en) Information transmission method and device
WO2023173981A1 (zh) 任务配置方法及装置
WO2024169522A1 (zh) 通信方法和装置
WO2022082608A1 (zh) 通信方法和通信装置
WO2024169329A1 (zh) 通信方法、通信装置及通信系统
WO2024169600A1 (zh) 一种通信方法及装置
WO2024148528A1 (zh) 通信方法以及通信设备
WO2024152877A1 (zh) 数据处理方法及装置

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: 21863610

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2021863610

Country of ref document: EP

Effective date: 20230324

NENP Non-entry into the national phase

Ref country code: DE