WO2023039905A1 - Ai数据的传输方法、装置、设备及存储介质 - Google Patents

Ai数据的传输方法、装置、设备及存储介质 Download PDF

Info

Publication number
WO2023039905A1
WO2023039905A1 PCT/CN2021/119444 CN2021119444W WO2023039905A1 WO 2023039905 A1 WO2023039905 A1 WO 2023039905A1 CN 2021119444 W CN2021119444 W CN 2021119444W WO 2023039905 A1 WO2023039905 A1 WO 2023039905A1
Authority
WO
WIPO (PCT)
Prior art keywords
function
network element
communication device
data
functional
Prior art date
Application number
PCT/CN2021/119444
Other languages
English (en)
French (fr)
Inventor
尤心
Original Assignee
Oppo广东移动通信有限公司
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 Oppo广东移动通信有限公司 filed Critical Oppo广东移动通信有限公司
Priority to PCT/CN2021/119444 priority Critical patent/WO2023039905A1/zh
Priority to CN202180100126.4A priority patent/CN117597969A/zh
Publication of WO2023039905A1 publication Critical patent/WO2023039905A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W24/00Supervisory, monitoring or testing arrangements
    • H04W24/02Arrangements for optimising operational condition

Definitions

  • the present application relates to the field of communication technology, and in particular to a transmission method, device, equipment and storage medium for artificial intelligence (AI) data.
  • AI artificial intelligence
  • the introduction of an AI function is being considered, and the performance of the mobile communication system is optimized using the AI function.
  • Embodiments of the present application provide an AI data transmission method, device, device, and storage medium, which ensure the normal use of AI functions in a mobile communication system, so that the mobile communication system can improve system performance based on AI functions. Described technical scheme is as follows:
  • a method for transmitting AI data is provided, the method is executed by a first communication device, the first communication device includes: a first AI functional entity, and the method includes:
  • the first AI functional entity and the AI functional network element execute an interaction process for AI data transmission
  • the AI function network element is a network element used for AI function management in the mobile communication system, and the AI function management includes controlling and managing the AI data transmission, and the AI data is related to the AI enabling function data.
  • a method for transmitting AI data is provided, the method is performed by an AI functional network element, and the method includes:
  • the AI functional network element and the first AI functional entity in the first communication device execute an interaction process for AI data transmission
  • the AI function network element is a network element used for AI function management in the mobile communication system, and the AI function management includes controlling and managing the AI data transmission, and the AI data is related to the AI enabling function data.
  • an AI data transmission device comprising: a first AI functional entity module;
  • the first AI functional entity module is configured to execute an interaction process for AI data transmission with the AI functional network element module;
  • the device to which the AI function network element module belongs is a device for AI function management in the mobile communication system, and the AI function management includes controlling and managing the AI data transmission, and the AI data is related to the AI empowerment function-related data.
  • an AI data transmission device comprising: an AI functional network element module;
  • the AI functional network element module is configured to execute an interaction process for AI data transmission with the first AI functional entity module
  • the device is a device for AI function management in a mobile communication system
  • the AI function management includes controlling and managing the transmission of the AI data
  • the AI data is data related to the AI enabling function.
  • a communication device includes: a transceiver; wherein,
  • the transceiver is configured to perform an interaction process for AI data transmission with an AI functional network element
  • the AI function network element is a network element used for AI function management in the mobile communication system, and the AI function management includes controlling and managing the AI data transmission, and the AI data is related to the AI enabling function data.
  • a network element device includes: a transceiver; wherein,
  • the transceiver is configured to execute an interaction process for AI data transmission with a first AI functional entity in the first communication device;
  • the network element device is a network element used for AI function management in the mobile communication system
  • the AI function management includes controlling and managing the AI data transmission
  • the AI data is data related to the AI enabling function .
  • a computer-readable storage medium wherein executable instructions are stored in the readable storage medium, and the executable instructions are loaded and executed by a processor to realize the AI described in the above aspect The method of data transfer.
  • a chip is provided, the chip includes a programmable logic circuit and/or program instructions, and when the chip is run on a computer device, it is used to realize the AI data described in the above aspect. the transmission method.
  • a computer program product is provided.
  • the computer program product runs on a processor of a computer device, the computer device executes the AI data transmission method described in the above aspect.
  • a first AI functional entity to a first communication device in a mobile communication system, and introduce an AI functional network element for AI function management, the first AI functional entity and the AI functional network element in the first communication device, and execute It is used for the interactive process of AI data transmission, thereby ensuring the normal use of AI functions in the mobile communication system, so that the mobile communication system can improve system performance based on AI functions.
  • Fig. 1 is a schematic diagram of a neural network provided by an exemplary embodiment of the present application
  • Fig. 2 is a schematic diagram of a neural network with multiple hidden layers provided by an exemplary embodiment of the present application
  • Fig. 3 is a schematic diagram of a convolutional neural network provided by an exemplary embodiment of the present application.
  • FIG. 4 is a block diagram of a mobile communication system provided by an exemplary embodiment of the present application.
  • FIG. 5 is a flowchart of a method for transmitting AI data provided by an exemplary embodiment of the present application
  • FIG. 6 is a flowchart of a method for transmitting AI data provided by an exemplary embodiment of the present application.
  • FIG. 7 is a flowchart of a method for transmitting AI data provided by an exemplary embodiment of the present application.
  • FIG. 8 is a flowchart of a method for transmitting AI data provided by an exemplary embodiment of the present application.
  • Fig. 9 is a structural block diagram of an AI data transmission device provided by an exemplary embodiment of the present application.
  • Fig. 10 is a structural block diagram of an AI data transmission device provided by an exemplary embodiment of the present application.
  • Fig. 11 is a schematic structural diagram of a communication device provided by an exemplary embodiment of the present application.
  • the current mobile communication system provides greater flexibility, emphasizing the wide application of different scenarios and the full use of limited resources.
  • the basic principles of most of the current work are still based on theoretical modeling of the actual communication environment or simple parameter selection. down, gradually weakening.
  • FIG. 1 shows a schematic diagram of a neural network provided by an embodiment of the present application.
  • the basic structure of a simple neural network includes: input layer, hidden layer and output layer. Among them, the input layer is responsible for receiving data, the hidden layer is responsible for processing data, and the final result is generated in the output layer.
  • each node represents a processing unit, which can also be regarded as simulating a neuron. Multiple neurons form a layer of neural network, and multi-layer information transmission and processing constructs an overall neural network.
  • neural network deep learning algorithms have been proposed, as shown in Figure 2, more hidden layers are introduced, and feature learning is performed layer by layer through multi-hidden layer neural network training, which greatly improves the It improves the learning and processing capabilities of the neural network, and is widely used in pattern recognition, signal processing, optimization combination, anomaly detection, etc.
  • FIG. 3 shows a schematic diagram of a convolutional neural network provided by an embodiment of the present application.
  • a convolutional neural network its basic structure includes: an input layer, multiple convolutional layers, multiple pooling layers, a fully connected layer, and an output layer.
  • the introduction of convolutional layer and pooling layer effectively controls the sharp increase of network parameters, limits the number of parameters and excavates the characteristics of local structures, which improves the robustness of the algorithm.
  • the present application proposes an implementation method for enabling AI functions, thereby ensuring that the AI functions can be used normally in the mobile communication system.
  • FIG. 4 shows a block diagram of a mobile communication system provided by an exemplary embodiment of the present application.
  • the mobile communication system may include: a terminal device 10 , an access network element 20 and a core network element 30 .
  • the terminal equipment 10 may refer to a user equipment (User Equipment, UE), an access terminal, a subscriber unit, a subscriber station, a mobile station, a mobile station, a remote station, a remote terminal, a mobile device, a wireless communication device, a user agent or a user device.
  • UE User Equipment
  • the terminal device 10 may also be a cellular phone, a cordless phone, a Session Initiation Protocol (Session Initiation Protocol, SIP) phone, a Wireless Local Loop (Wireless Local Loop, WLL) station, a Personal Digital Assistant (PDA) ), handheld devices with wireless communication functions, computing devices or other processing devices connected to wireless modems, vehicle-mounted devices, wearable devices, terminal devices in the fifth generation mobile communication system (5th Generation System, 5GS) or future evolution
  • the terminal equipment in the Public Land Mobile Network (Public Land Mobile Network, PLMN), etc. is not limited in this embodiment of the present application.
  • the devices mentioned above are collectively referred to as terminal devices.
  • the terminal device 10 may include a first AI functional entity.
  • the access network element 20 is a network element device deployed in the access network to provide the terminal device 10 with a wireless communication function.
  • the access network element 20 may include various forms of macro base stations, micro base stations, relay stations, access points, and so on.
  • the names of devices with access network element functions may be different. For example, in 5G NR systems, they are called gNodeB or gNB. With the evolution of communication technology, the name "access network element" may change.
  • access network elements For convenience of description, in the embodiment of the present application, the above-mentioned devices that provide the wireless communication function for the terminal device 10 are collectively referred to as access network elements.
  • a communication relationship may be established between the terminal device 10 and the core network element 30 through the access network element 20 .
  • the access network element 20 may be one or more eNodeBs in EUTRAN (Evolved Universal Terrestrial Radio Access Network, Evolved Universal Terrestrial Radio Network) or EUTRAN ;
  • EUTRAN Evolved Universal Terrestrial Radio Access Network
  • EUTRAN EUTRAN
  • the access network element 20 may be a RAN (Radio Access Network, radio access network) or one or more gNBs in the RAN.
  • the access network element 20 may include a first AI functional entity.
  • the access network element 20 includes an AI functional network element.
  • the core network element 30 is a network element device deployed in the core network.
  • the functions of the core network element 30 are mainly to provide user connections, manage users, and carry out services, and provide an interface to the external network as a bearer network.
  • the core network element 30 includes an AI functional network element.
  • the "5G NR system" in the embodiment of the present application may also be called a 5G system or an NR system, but those skilled in the art can understand its meaning.
  • the technical solutions described in the embodiments of this application can be applied to LTE systems, 5G NR systems, and subsequent evolution systems of 5G NR systems, and can also be applied to systems such as NB-IoT (Narrow Band Internet of Things, narrowband Internet of Things) system and other communication systems, this application is not limited to this.
  • NB-IoT Near Band Internet of Things, narrowband Internet of Things
  • Fig. 5 shows a flowchart of an AI data transmission method provided by an exemplary embodiment of the present application. This method can be applied in the mobile communication system shown in Figure 4, and this method comprises:
  • Step 510 the first AI functional entity in the first communication device and the AI functional network element execute an interaction process for AI data transmission.
  • the first AI function entity in the first communication device interacts with the AI function network element, so that the AI data is transmitted between the first communication device and the AI function network element.
  • the first communication device is a communication device in a mobile communication system, and the first communication device supports an AI function.
  • the first communication device is a terminal device.
  • the first communication device is an access network element.
  • the first communication device includes a first AI functional entity, and the first AI functional entity is a functional entity related to an AI function in the first communication device.
  • the first AI functional entity is located in any of the following protocol layers, or between any two adjacent protocol layers in the following protocol layers; wherein,
  • the protocol layer includes at least one of the following: Non-Access Stratum (NAS) layer; Service Data Adaptation Protocol (Service Data Adaptation Protocol, SDAP) layer; Packet Data Convergence Protocol (Packet Data Convergence Protocol, PDCP) layer; radio link control (Radio Link Control, RLC) layer; media access control (Medium Access Control, MAC) layer; physical (PHY) layer.
  • NAS Non-Access Stratum
  • SDAP Service Data Adaptation Protocol
  • Packet Data Convergence Protocol Packet Data Convergence Protocol
  • RLC Radio Link Control
  • MAC Medium Access Control
  • PHY physical
  • the first AI functional entity on the terminal device side may be located at the NAS layer, or the SDAP layer, or the PDCP layer, or the RLC layer, or the MAC layer, or the PHY layer; it may also be located at a layer between the above protocol layers, namely Between any two adjacent protocol layers, such as: between NAS layer and SDAP layer, between SDAP layer and PDCP layer, between PDCP layer and RLC layer, between RLC layer and MAC layer, between MAC layer and PHY layer between.
  • the first AI functional entity is located in a centralized unit (Centralized Unit, CU), or located in a distributed unit (Distributed Unit, DU), or located in any of the following In a protocol layer, or between any two adjacent protocol layers in the following protocol layers; wherein, the protocol layer includes at least one of the following: NAS layer; SDAP layer; PDCP layer; RLC layer; MAC layer ; PHY layer.
  • the first AI functional entity on the network element side of the access network may be located in the CU or DU; it may be located in the NAS layer, or the SDAP layer, or the PDCP layer, or the RLC layer, or the MAC layer, or the PHY layer; it may also be located in the A layer between the above protocol layers, that is, between any two adjacent protocol layers, such as: between the NAS layer and the SDAP layer, between the SDAP layer and the PDCP layer, between the PDCP layer and the RLC layer, between the RLC layer and the Between the MAC layer, between the MAC layer and the PHY layer.
  • AI function network element is a network element used for AI function management in the mobile communication system.
  • AI function management includes control and management of AI data transmission, and AI data is data related to AI enabling functions.
  • the AI enabling function refers to endowing the mobile communication system with an AI function, so as to apply the AI function and optimize the performance of the mobile communication system.
  • AI data is data related to AI empowerment functions. It can be understood that AI data is data related to the application process of AI models in mobile communication systems. For example, AI data includes: capability information related to AI functions, AI model The model configuration information, the input data of the AI model, and the output data of the AI model; for example, the AI data does not include the data involved in the model training process of the AI model, such as: the training data of the AI model.
  • the AI function network element is any one of the existing network elements; or, the AI function network element is a new network element other than the existing network element; wherein, the existing network element includes: the access network element Meta, Network Data Analytics Function (NWDAF), Access and Mobility Management Function (Access and Mobility Management Function, AMF), Session Management Function (Session Management Function, SMF), Policy Control Function (Policy Control Function) , PCF), unified data management (Unified Data Management, UDM), user plane function (User Plane Function, UPF), network repository function (Network Repository Function, NRF) and network exposure function (Network Exposure Function, NEF).
  • NWDAF Network Data Analytics Function
  • AMF Access and Mobility Management Function
  • Session Management Function Session Management Function
  • Policy Control Function Policy Control Function
  • PCF Policy Control Function
  • UDM Unified Data Management
  • UDM User Plane Function
  • NRF Network Repository Function
  • NEF Network Exposure Function
  • the AI function network element introduced in the embodiment of the present application may be a network element defined in the current communication protocol standard, that is, an existing network element, and the existing network element includes an access network element and a core network element
  • the network elements of the access network include: gNB, eNB, and the network elements of the core network include: NWDAF, AMF, SMF, PCF, UDM, UPF, and NRFNEF;
  • NWDAF Access Mobility Management Function
  • AMF Access Management Function
  • SMF Serving Mobility Management Function
  • PCF PackeNB
  • UDM User Data Management Function
  • NRFNEF NRFNEF
  • a newly added network element other than the network element the newly added network element may be an access network element or a core network element.
  • the transmission of AI data between the first communication device and the AI functional network element is realized based on an existing interface, or, the transmission of AI data between the first communication device and the AI functional network element is based on a new dedicated AI data transmission interface to implement.
  • the method provided in this embodiment introduces the first AI function entity into the first communication device in the mobile communication system, and introduces the AI function network element for AI function management.
  • the first communication device in the first communication device An AI functional entity and an AI functional network element execute an interactive process for AI data transmission, thereby ensuring the normal use of the AI function in the mobile communication system, so that the mobile communication system can improve system performance based on the AI function.
  • the following interaction may be performed between the first functional entity in the first communication device and the AI functional network element:
  • the first AI functional entity and the AI functional network element execute an interaction process for performing AI capability negotiation.
  • the interaction process for AI capability negotiation refers to an interaction process for communicating capabilities related to the AI function of the first communication device between the first terminal device and the AI function network element.
  • the first AI functional entity and the AI functional network element execute an interaction process for configuring the AI model.
  • the interaction process for configuring the AI model refers to the interaction process in which the AI function network element configures the AI model for the first terminal device.
  • the first AI functional entity and the AI functional network element execute an interaction process for inputting an AI model.
  • the interaction process for inputting the AI model refers to the interaction process for the input data of the AI model to be exchanged between the first terminal device and the AI function network element.
  • the first AI functional entity and the AI functional network element execute an interaction process for performing AI capability negotiation.
  • Fig. 6 shows a flowchart of an AI data transmission method provided by an exemplary embodiment of the present application. This method can be applied in the mobile communication system shown in Figure 4, and this method comprises:
  • Step 610 The AI function network element sends a capability request to the first communication device, where the capability request is used to request the first communication device to report the capability related to the AI function.
  • the network element with the AI function sends a capability request to the first communication device, requesting the first communication device to report the capabilities related to the AI function.
  • Step 620 The first AI functional entity in the first communication device receives the capability request.
  • Step 630 The first AI function entity in the first communication device reports capability information to the AI function network element, where the capability information is used to provide capabilities related to the AI function.
  • the first AI functional entity reports capability information to the AI functional network element based on the capability request.
  • the first AI functional entity may also actively report to the The AI function network element reports the capability information, that is, step 630 is directly performed without the above steps 610 and 620 .
  • the first communication device reports its own capability information to the AI function network element.
  • the capability information includes at least one of the following:
  • bit used to indicate whether the AI function is supported in the capability information. If the bit is "0", it means that the first communication device supports the AI function. If the bit is "1”, it means that the first communication device supports the AI function. A communication device does not support the AI function.
  • the agreed types of AI functions in the mobile communication system include: type 1, type 2, type 3, etc., each type corresponds to a corresponding type identifier, and the type identifier is used to identify the type of the AI function.
  • the capability information includes a type identifier, and the first communication device notifies the AI function network element of the type of the AI function supported by the first communication device through the type identifier.
  • types of AI functions include: AI functions for improving prediction accuracy, AI functions for optimizing decisions, AI functions for improving calculation accuracy, etc., which are not limited in the present application.
  • the AI models stipulated in the mobile communication system include: AI model 1, AI model 2, AI model 3, etc., each AI model corresponds to a corresponding AI model identifier, and the AI model identifier is used to identify the AI model.
  • the capability information includes an AI model identifier, and the first communication device notifies the AI function network element of the AI model supported by the first communication device through the AI model identifier.
  • the agreed AI algorithms in the mobile communication system include: AI algorithm 1, AI algorithm 2, AI algorithm 3, etc., each AI algorithm corresponds to a corresponding AI algorithm identifier, and the AI algorithm identifier is used to identify the AI algorithm.
  • the capability information includes an AI algorithm identifier, and the first communication device notifies the AI function network element of the AI algorithm supported by the first communication device through the AI algorithm identifier.
  • AI algorithms include: linear regression algorithm, decision tree algorithm, random forest algorithm, logistic regression algorithm, neural network algorithm, Bayesian algorithm, K nearest neighbor algorithm, K-means algorithm and Markov algorithm, etc., the present application There is no restriction on this.
  • the capability information further includes computing power and storage capability of the first communication device to which the first AI functional entity belongs.
  • Step 640 The AI function network element receives capability information.
  • the AI function network element manages and controls the first communication device to perform the AI function based on the capability information of the first communication device, for example: the AI function network element sends the model configuration information of the AI model to the first communication device , so that the first communication device can determine the AI model based on the model configuration information, and use the AI model to improve system performance.
  • the method provided in this embodiment introduces the first AI function entity into the first communication device in the mobile communication system, and introduces the AI function network element for AI function management.
  • the first communication device in the first communication device An AI functional entity and an AI functional network element execute an interactive process for AI data transmission, thereby ensuring the normal use of the AI function in the mobile communication system, so that the mobile communication system can improve system performance based on the AI function.
  • the first AI functional entity in the first communication device and the AI functional network element execute an interaction process for AI capability negotiation, helping the AI functional network element to clarify the AI function of the first communication device related capabilities.
  • the first AI functional entity and the AI functional network element execute an interaction process for configuring the AI model.
  • Fig. 7 shows a flow chart of an AI data transmission method provided by an exemplary embodiment of the present application. This method can be applied in the mobile communication system shown in Figure 4, and this method comprises:
  • Step 710 The first AI function entity in the first communication device sends a service request to the AI function network element, and the service request includes a requirement for enabling the AI function.
  • the service request is used to request the network element with the AI function to meet the requirement of the first communication device for enabling the AI function.
  • the AI function can be implemented independently on the terminal device side.
  • the AI functional entity in the terminal device sends a service request to the AI functional network element, and the AI functional entity in the terminal device uses the corresponding model configuration information after obtaining the corresponding model configuration information. It can also be implemented on the side of the network element of the access network separately.
  • the AI function entity in the network element of the access network sends a service request to the network element of the AI function, and the AI function entity in the network element of the access network receives
  • the AI function can be enabled after the model configuration information is configured; it can also be implemented jointly on the terminal device side and the access network element side, for example: the AI function entity in the terminal device and the AI function entity in the access network element provide the AI function
  • the AI functional entity in the terminal device and the AI functional entity in the network element of the access network will enable the same type of AI function synchronously or asynchronously after obtaining the corresponding model configuration information.
  • the requirement for enabling the AI function includes at least one of the following:
  • CSI Channel-State Information
  • the terminal device determines the current CSI by measuring the reference signal configured by the network element of the access network, and feeds back the current CSI to the network element of the access network, so that the network element of the access network can determine the current CSI based on the CSI.
  • Channel conditions refers to that the first communication device predicts the CSI to be fed back based on the AI function, so as to determine the channel situation as early as possible.
  • the first communication device is an access network element, and the access network element performs CSI feedback prediction based on an AI function.
  • the first communication device is a terminal device, and the terminal device performs CSI feedback prediction based on an AI function.
  • the first communication device is an access network element and a terminal device, and the terminal device and the access network element jointly perform CSI feedback prediction based on an AI function.
  • Beam management enhancement refers to that the first communication device predicts beam failure information based on an AI function, and further determines a target beam that needs to be restored or switched.
  • the first communication device is a terminal device, and the terminal device performs beam management enhancement based on the AI function.
  • the terminal predicts the beam based on the AI model, and performs beam recovery or switching in advance based on the prediction result, avoiding failures caused by beam failure recovery. Data transfer interrupted.
  • the first communication device is an access network element, and the access network element performs beam management enhancement based on an AI function.
  • the first communication device is an access network element and a terminal device, and the terminal device and the access network element jointly perform beam management enhancement based on an AI function.
  • the handover decision enhancement refers to that the first communication device optimizes the decision of cell handover based on the AI function.
  • the first communication device is a terminal device, and the terminal device determines handover conditions, and/or information such as handover success rate and/or target cell identity based on the AI function, so as to enhance handover decision.
  • the first communication device is a network element of the access network, and the network element of the access network determines the handover condition based on the AI function, and/or the information such as the handover success rate and/or the identity of the target cell, so as to enhance the handover decision.
  • the first communication device is an access network element and a terminal device, and the terminal device and the access network element jointly perform handover decision enhancement based on an AI function.
  • the positioning accuracy enhancement refers to that the first communication device improves the accuracy of the positioning result based on the AI function.
  • the first communication device is a terminal device, and the terminal device optimizes positioning accuracy based on an AI algorithm to eliminate accuracy errors caused by non-line-of-sight (Non-Line of Sight, NLOS) scenarios, thereby enhancing positioning accuracy.
  • the first communication device is an access network element, and the access network element optimizes positioning accuracy based on an AI algorithm, and eliminates accuracy errors caused by NLOS scenarios, thereby enhancing positioning accuracy.
  • the first communication device is an access network element and a terminal device, and the terminal device and the access network element jointly perform positioning accuracy enhancement based on an AI function.
  • Random access enhancement means that the first communication device optimizes random access resource allocation based on the AI function.
  • the first communication device is an access network element, and the access network element performs random access enhancement based on an AI function.
  • the first communication device is a terminal device, and the terminal device performs random access enhancement based on an AI function.
  • the first communication device is an access network element and a terminal device, and the terminal device and the access network element jointly perform random access enhancement based on an AI function.
  • Resource scheduling enhancement means that the first communication device optimizes resource scheduling based on AI functions, and resource scheduling includes: dynamic resource scheduling and semi-static resource scheduling.
  • the first communication device is an access network element, and the access network element performs resource scheduling enhancement based on an AI function.
  • the first communication device is a terminal device, and the terminal device performs enhanced resource scheduling based on an AI function.
  • the first communication device is an access network element and a terminal device, and the terminal device and the access network element jointly perform resource scheduling enhancement based on an AI function.
  • the terminal path prediction refers to that the first communication device predicts the running trajectory of the terminal device based on the AI function, so as to help reserve corresponding resources and balance cell load.
  • the first communication device is an access network element, and the access network element performs terminal path prediction based on an AI function.
  • the first communication device is a terminal device, and the terminal device performs terminal path prediction based on an AI function.
  • the first communication device is an access network element and a terminal device, and the terminal device and the access network element jointly perform terminal path prediction based on the AI function.
  • the terminal service prediction refers to that the first communication device predicts the service type of the terminal device based on the AI function, thereby helping to optimize resource reservation and allocation.
  • the first communication device is an access network element, and the access network element performs terminal service prediction based on an AI function.
  • the first communication device is a terminal device, and the terminal device performs terminal service prediction based on an AI function.
  • the first communication device is an access network element and a terminal device, and the terminal device and the access network element jointly perform terminal service prediction based on an AI function.
  • the cell load prediction refers to that the first communication device predicts the load of the cell based on the AI function, so as to help optimize resource allocation.
  • the first communication device is an access network element, and the access network element performs cell load prediction based on an AI function.
  • the first communication device is a terminal device, and the terminal device performs cell load prediction based on an AI function.
  • the first communication device is an access network element and a terminal device, and the terminal device and the access network element jointly perform cell load prediction based on an AI function.
  • requirements for enabling the AI function may also include other types of requirements, which are not limited in the present application.
  • Step 720 The AI function network element receives the service request.
  • Step 730 The AI function network element sends model configuration information to the first communication device, and the model configuration information is used for the first communication device to determine the AI model corresponding to the requirement of enabling the AI function.
  • the AI function network element After the AI function network element receives the service request, it sends the first communication device the model configuration information indicating the AI model corresponding to the requirement of enabling the AI function, and the first communication device downloads the service request from the AI function network element. (download) model configuration information.
  • the AI functional entity has trained an AI model based on historical data, and determined model configuration information corresponding to the AI model.
  • Step 740 The first AI function entity in the first communication device receives the model configuration information.
  • the model configuration information includes at least one of the following: a hierarchical structure of the AI model; weight information of network parameters of the AI model.
  • the first AI functional entity updates the received model configuration information
  • the first AI functional entity sends the updated model configuration information to the AI functional network element
  • the AI functional network receives the updated model configuration information sent by the first AI functional entity.
  • the first communication device may further update the AI model based on its own real-time information as input, and upload (upload) the model configuration information corresponding to the updated AI model to the AI function NE.
  • the first AI functional entity receives real-time information sent by at least one other AI functional entity, and the real-time information is used to provide input data to the AI model in the first AI functional entity.
  • the first communication device is an access network element, and the first AI functional entity in the access network element performs resource scheduling enhancement based on the AI function.
  • the first AI functional entity receives the model configuration information and determines the corresponding AI model, in addition to using the real-time information on its own side as the input of the AI model, it also receives real-time information sent by other AI functional entities in at least one terminal device , and the received real-time information on the terminal device side is also used as the input of the AI model.
  • other AI functional entities receive indication information from the first AI functional entity or AI functional network element, and send real-time information to the first AI functional entity based on the indication information.
  • the AI functional network element sends the model configuration information to the first communication device, it sends instruction information to other AI functional entities, instructing other AI functional entities to send real-time information to the first AI functional entity.
  • the first AI functional entity sends indication information to other AI functional entities, instructing other AI functional entities to send real-time information to the first AI functional entity.
  • AI functions include: CSI feedback prediction, beam management enhancement, handover decision enhancement, positioning accuracy enhancement, random access enhancement, resource scheduling enhancement, terminal path prediction, terminal service prediction and cell load prediction, correspondingly, real-time information
  • the types include: the input type corresponding to CSI feedback prediction, the input type corresponding to beam management enhancement, the input type corresponding to handover decision enhancement, the input type corresponding to positioning accuracy enhancement, the input type corresponding to random access enhancement, and the input type corresponding to resource scheduling enhancement The input type, the input type corresponding to the terminal path prediction, the input type corresponding to the terminal service prediction, and the input type corresponding to the cell load prediction.
  • the first AI functional entity in the first communication device determines the AI module based on the model configuration information, inputs the model input data, and obtains the output of the AI model. Functional requirements, and then optimize the performance in the mobile communication system.
  • the method provided in this embodiment introduces the first AI function entity into the first communication device in the mobile communication system, and introduces the AI function network element for AI function management.
  • the first communication device in the first communication device An AI functional entity and an AI functional network element execute an interactive process for AI data transmission, thereby ensuring the normal use of the AI function in the mobile communication system, so that the mobile communication system can improve system performance based on the AI function.
  • the first AI functional entity and the AI functional network element in the first communication device execute an interaction process for configuring the AI model, and help the first communication device determine the AI model, thereby using the AI model Meet their own needs, and then optimize the performance in the mobile communication system.
  • the first AI functional entity and the AI functional network element execute an interaction process for inputting an AI model.
  • Fig. 8 shows a flow chart of an AI data transmission method provided by an exemplary embodiment of the present application. This method can be applied in the mobile communication system shown in Figure 4, and this method comprises:
  • Step 810 The first AI functional entity in the first communication device sends auxiliary data information to the AI functional network element, where the auxiliary data information is used to provide input data of the AI model.
  • the first communication device has a requirement for enabling the AI function, and the first communication device carries the data that needs to be processed to meet the requirement in the auxiliary data information, and sends it to the AI function network element, and the AI function network element converts the above The data is fed into the AI model for processing.
  • the first communication device is a network element of an access network
  • the network element of the access network has a requirement for resource scheduling enhancement based on an AI function
  • the first AI functional entity in the network element of the access network sends an auxiliary Data information
  • auxiliary data information carries data that affects resource scheduling decisions
  • the AI functional network element uses the above data as input data for the AI model to output resource scheduling decisions.
  • the first AI function entity in the first communication device sends a service request to the AI function network element, and the service request includes a requirement for enabling the AI function; the AI function network element receives the service request accordingly.
  • the first communications device actively reports the assistance data information. That is, the first communication device directly sends the auxiliary data information after reporting the service request.
  • the first communication device reports the auxiliary data information under the trigger of the AI function network element.
  • the AI functional network element sends an auxiliary data request to the first AI functional entity, and the auxiliary data request is used to request the first communication device to send auxiliary data information; the first AI functional entity responds The auxiliary data request sent by the AI functional network element is received, where the auxiliary data request is used to request the first communication device to send auxiliary data information. That is, after the first communication device reports the service request, the AI function network element first requests the first communication device to send the auxiliary data information, and then the first communication device sends the auxiliary data information accordingly.
  • Step 820 The AI function network element receives the auxiliary data information.
  • the AI function network element inputs the data carried in the auxiliary data information into the AI model, obtains the output of the AI model, and sends the output to the first communication device, and the above output can meet the requirement of the first communication device. It can meet the needs of AI functions, and then optimize the performance in the mobile communication system.
  • the first AI functional entity on the first communication device side exchanges auxiliary data information with the AI functional network element
  • the first AI functional entity on the first communication device side may also communicate with other The AI functional entity in the communication device exchanges the auxiliary data information, so as to realize the transmission of the auxiliary data information.
  • an AI model is stored in an AI functional entity in other communication devices, and the data in the auxiliary data information is used as input data of the AI model.
  • other communication devices transmit the auxiliary data information again, and finally transmit it to the AI functional network element for processing.
  • the first AI function entity exchanges assistance data information with the second AI function entity of the second communication device through the first interface
  • the second communication device is a communication device of a different type from the first communication device.
  • auxiliary data information is exchanged between the terminal equipment and the network elements of the access network.
  • the first interface includes: a Uu interface; or, an interface dedicated to the AI function. That is, the first interface may be an existing Uu interface, or a newly added interface dedicated to the AI function.
  • the auxiliary data information is carried in at least one of the following messages: a Radio Resource Control (Radio Resource Control, RRC) message, a MAC Control Element (Control Element, CE), a NAS message, and a message dedicated to the AI function.
  • RRC Radio Resource Control
  • CE MAC Control Element
  • NAS NAS message
  • the first AI functional entity exchanges assistance data information with a third AI functional entity of a third communication device, where the third communication device is a communication device of the same type as the first communication device.
  • auxiliary data information can be exchanged between communication devices of the same type.
  • auxiliary data information is exchanged between terminal equipment and terminal equipment, and auxiliary data information is exchanged between access network elements and access network elements.
  • the method provided in this embodiment introduces the first AI function entity into the first communication device in the mobile communication system, and introduces the AI function network element for AI function management.
  • the first communication device in the first communication device An AI functional entity and an AI functional network element execute an interactive process for AI data transmission, thereby ensuring the normal use of the AI function in the mobile communication system, so that the mobile communication system can improve system performance based on the AI function.
  • the first AI functional entity in the first communication device and the AI functional network element execute an interaction process for AI model input, and the first AI functional entity in the first communication device will
  • the auxiliary data information is reported to the AI function network element, and the AI function network element uses the AI model to process the data in the auxiliary data information, so as to meet the requirement of enabling the AI function of the first communication device.
  • the steps performed by the first communication device can be implemented separately as the AI data transmission method on the side of the first communication device, and the steps performed by the AI function network element can be separately implemented as an AI function network element- The transmission method of AI data on the side.
  • Fig. 9 shows a structural block diagram of an AI data transmission device provided by an exemplary embodiment of the present application.
  • the device can be implemented as a first communication device, or can be implemented as a part of the first communication device.
  • the device includes: the first communication device An AI functional entity module 901;
  • the first AI function entity module 901 is configured to execute an interaction process for AI data transmission with the AI function network element module;
  • the device to which the AI function network element module belongs is a device for AI function management in the mobile communication system, and the AI function management includes controlling and managing the AI data transmission, and the AI data is related to the AI empowerment function-related data.
  • the device to which the AI functional network element module belongs is any one of existing network elements; or, the device to which the AI functional network element module belongs is a network element other than the existing network element A newly added network element of ;
  • the existing network elements include: access network elements, NWDAF, AMF, SMF, PCF, UDM, UPF, NRF and NEF.
  • the first AI functional entity module 901 is located in any of the following protocol layers, or located in any two adjacent between protocol layers;
  • the protocol layer includes at least one of the following: NAS layer; SDAP layer; PDCP layer; RLC layer; MAC layer;
  • the first AI functional entity module 901 is located in a CU, or located in a DU, or located in any of the following protocol layers, or , located between any two adjacent protocol layers in the following protocol layers;
  • the protocol layer includes at least one of the following: NAS layer; SDAP layer; PDCP layer; RLC layer; MAC layer;
  • the first AI function entity module 901 is configured to execute an interaction process for AI capability negotiation with the AI function network element module;
  • the first AI functional entity module 901 is configured to execute an interaction process for configuring an AI model with the AI functional network element module;
  • the first AI function entity module 901 is configured to execute an interaction process for AI model input with the AI function network element module.
  • the first AI function entity module 901 is configured to receive a capability request sent by the AI function network element module, and the capability request is used to request the device to report capabilities related to the AI function ;
  • the first AI function entity module 901 is configured to report capability information to the AI function network element module, where the capability information is used to provide capabilities related to the AI function.
  • the capability information includes at least one of the following:
  • the first AI function entity module 901 is configured to send a service request to the AI function network element module, where the service request includes a requirement for enabling the AI function; the first The AI function entity module 901 is configured to receive model configuration information sent by the AI function network element module, where the model configuration information is used for the device to determine the AI model corresponding to the requirement for enabling the AI function.
  • the first AI function entity module 901 is configured to update the received model configuration information; the first AI function entity module 901 is configured to provide the AI function The network element module sends the updated model configuration information.
  • the model configuration information includes at least one of the following:
  • the weight information of the network parameters of the AI model is the weight information of the network parameters of the AI model.
  • the requirement for enabling the AI function includes at least one of the following:
  • the AI function performs terminal path prediction, terminal service prediction based on AI function, and cell load prediction based on AI function.
  • the first AI function entity module 901 is configured to send auxiliary data information to the AI function network element module, where the auxiliary data information is used to provide input data of the AI model.
  • the first AI function entity module 901 is configured to receive an assistance data request sent by the AI function network element module, and the assistance data request is used to request the device to send the assistance data request. Data information.
  • the first AI functional entity module 901 is configured to exchange the auxiliary data information with the second AI functional entity module through a first interface, and the device to which the second AI functional entity module belongs is a different type of device than the one described.
  • the first interface includes:
  • auxiliary data information is carried in at least one of the following messages:
  • RRC messages RRC messages, MAC CE, NAS messages and messages dedicated to AI functions.
  • the first AI function entity module 901 is configured to exchange the auxiliary data information with a third AI function entity module, and the device to which the third AI function entity module belongs is related to the Devices of the same type.
  • Fig. 10 shows a structural block diagram of an AI data transmission device provided by an exemplary embodiment of the present application.
  • the device can be realized as an AI function network element, or can be realized as a part of an AI function network element, and the device includes: AI Functional network element module 1001;
  • the AI functional network element module 1001 is configured to execute an interaction process for AI data transmission with the first AI functional entity module;
  • the device is a device for AI function management in a mobile communication system
  • the AI function management includes controlling and managing the transmission of the AI data
  • the AI data is data related to the AI enabling function.
  • the device is any one of existing network elements; or, the device is a new network element other than the existing network element;
  • the existing network elements include: access network elements, NWDAF, AMF, SMF, PCF, UDM, UPF, NRF and NEF.
  • the first AI functional entity module when the device to which the first AI functional entity module belongs is a terminal device, the first AI functional entity module is located in any of the following protocol layers, or located in the following protocol layer Between any two adjacent protocol layers in
  • the protocol layer includes at least one of the following: NAS layer; SDAP layer; PDCP layer; RLC layer; MAC layer;
  • the first AI functional entity module when the device to which the first AI functional entity module belongs is an access network element, the first AI functional entity module is located in a CU, or located in a DU, or located in the following In any protocol layer, or between any two adjacent protocol layers in the following protocol layers;
  • the protocol layer includes at least one of the following: NAS layer; SDAP layer; PDCP layer; RLC layer; MAC layer;
  • the AI functional network element module 1001 is configured to execute an interaction process for AI capability negotiation with the first AI functional entity module;
  • the AI functional network element module 1001 is configured to execute an interaction process for configuring an AI model with the first AI functional entity module;
  • the AI function network element module 1001 is configured to execute an interaction process for AI model input with the first AI function entity module.
  • the AI functional network element module 1001 is configured to send a capability request to the first AI functional entity module, where the capability request is used to request the The device reports capabilities related to AI functions; the AI function network element module 1001 is configured to receive capability information reported by the first AI function entity module, and the capability information is used to provide capabilities related to AI functions.
  • the capability information includes at least one of the following:
  • Types of AI functions supported by the device to which the first AI function entity module belongs
  • the AI algorithm supported by the device to which the first AI functional entity module belongs is the AI algorithm supported by the device to which the first AI functional entity module belongs.
  • the AI function network element module 1001 is configured to receive a service request sent by the first AI function entity module, where the service request includes a requirement for enabling the AI function; the AI The functional network element module 1001 is configured to send model configuration information to the first AI functional entity module, and the model configuration information is used for the device to which the first AI functional entity module belongs to determine the AI function-enabled The AI model corresponding to the requirement.
  • the AI function network element module 1001 is configured to receive the updated model configuration information sent by the first AI function entity module.
  • the model configuration information includes at least one of the following:
  • the weight information of the network parameters of the AI model is the weight information of the network parameters of the AI model.
  • the requirement for enabling the AI function includes at least one of the following:
  • the AI function performs terminal path prediction, terminal service prediction based on AI function, and cell load prediction based on AI function.
  • the AI functional network element module 1001 is configured to receive auxiliary data information sent by the first AI functional entity module, where the auxiliary data information is used to provide input data for the AI model .
  • the AI functional network element module 1001 is configured to send an auxiliary data request to the first AI functional entity module, and the auxiliary data request is used to request the first AI functional entity module The associated device sends the auxiliary data information.
  • Fig. 11 shows a schematic structural diagram of a communication device provided by an exemplary embodiment of the present application (the above-mentioned first communication device or AI function network element), the communication device includes: a processor 1101, a receiver 1102, a transmitter 1103 , memory 1104 and bus 1105 .
  • the processor 1101 includes one or more processing cores, and the processor 1101 executes various functional applications by running software programs and modules.
  • the receiver 1102 and the transmitter 1103 can be realized as a transceiver 1106, and the transceiver 1106 can be a communication chip.
  • the memory 1104 is connected to the processor 1101 through the bus 1105 .
  • the memory 1104 may be used to store a computer program, and the processor 1101 is used to execute the computer program, so as to implement various steps performed by the communication device in the foregoing method embodiments.
  • the memory 1104 can be realized by any type of volatile or non-volatile storage device or their combination, and the volatile or non-volatile storage device includes but not limited to: random-access memory (Random-Access Memory, RAM) And read-only memory (Read-Only Memory, ROM), erasable programmable read-only memory (Erasable Programmable Read-Only Memory, EPROM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), flash memory or other solid-state storage technologies, compact disc read-only memory (CD-ROM), high-density digital video disc (Digital Video Disc, DVD) or other optical storage, tape cartridges, tapes, disks storage or other magnetic storage devices.
  • RAM Random-Access Memory
  • ROM read-only memory
  • EPROM erasable programmable read-only memory
  • EPROM erasable programmable Read-Only Memory
  • EEPROM Electrically erasable programmable read-only memory
  • the processor and the transceiver in the communication device involved in the embodiment of the present application may execute any of the methods shown in FIG. 5 to FIG. 8 above, and the first communication device The steps to be executed will not be repeated here.
  • the transceiver 1106 is configured to execute an interaction process for AI data transmission with an AI functional network element
  • the AI function network element is a network element used for AI function management in the mobile communication system, and the AI function management includes controlling and managing the AI data transmission, and the AI data is related to the AI enabling function data.
  • the processor and the transceiver in the communication device involved in the embodiment of the present application can execute the methods shown in any one of the above-mentioned Figures 5 to 8, and the AI function network element The steps to be executed will not be repeated here.
  • the transceiver 1106 is configured to execute an interaction process for AI data transmission with the first AI functional entity in the first communication device;
  • the AI function network element is a network element used for AI function management in the mobile communication system, and the AI function management includes controlling and managing the AI data transmission, and the AI data is related to the AI enabling function data.
  • a computer-readable storage medium stores at least one instruction, at least one program, a code set or an instruction set, the at least one instruction, the At least one section of program, the code set or instruction set is loaded and executed by the processor to implement the AI data transmission method provided by the above method embodiments.
  • a chip is also provided, the chip includes a programmable logic circuit and/or program instructions, and when the chip is run on a computer device, it is used to implement the AI data described in the above aspects. transfer method.
  • a computer program product is also provided.
  • the computer program product is run on a processor of a computer device, the computer device is made to execute the method for transmitting AI data described in the above aspects.
  • the program can be stored in a computer-readable storage medium.
  • the above-mentioned The storage medium mentioned may be a read-only memory, a magnetic disk or an optical disk, and the like.

Landscapes

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

Abstract

本申请公开了一种AI数据的传输方法、装置、设备及存储介质,涉及通信技术领域。该方法由第一通信设备执行,所述第一通信设备包括:第一AI功能实体,所述方法包括:所述第一AI功能实体与AI功能网元,执行用于进行AI数据传输的交互流程;其中,所述AI功能网元是移动通信系统中用于AI功能管理的网元,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。本申请实施例所提供的方法,保障了AI功能在移动通信系统中的正常使用,使得移动通信系统可以基于AI功能提升系统性能。

Description

AI数据的传输方法、装置、设备及存储介质 技术领域
本申请涉及通信技术领域,特别涉及一种人工智能(Artificial Intelligence,AI)数据的传输方法、装置、设备及存储介质。
背景技术
在移动通信系统中,正在考虑引入AI功能,使用AI功能来优化移动通信系统的性能。
如何在移动通信系统中实现AI功能,相关技术尚未提供较好的解决方案。
发明内容
本申请实施例提供了一种AI数据的传输方法、装置、设备及存储介质,保障了AI功能在移动通信系统中的正常使用,使得移动通信系统可以基于AI功能提升系统性能。所述技术方案如下:
根据本申请的一个方面,提供了一种AI数据的传输方法,所述方法由第一通信设备执行,所述第一通信设备包括:第一AI功能实体,所述方法包括:
所述第一AI功能实体与AI功能网元,执行用于进行AI数据传输的交互流程;
其中,所述AI功能网元是移动通信系统中用于AI功能管理的网元,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
根据本申请的一个方面,提供了一种AI数据的传输方法,所述方法由AI功能网元执行,所述方法包括:
所述AI功能网元与第一通信设备中的第一AI功能实体,执行用于进行AI数据传输的交互流程;
其中,所述AI功能网元是移动通信系统中用于AI功能管理的网元,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
根据本申请的一个方面,提供了一种AI数据的传输装置,所述装置包括:第一AI功能实体模块;
所述第一AI功能实体模块,用于与AI功能网元模块,执行用于进行AI数据传输的交互流程;
其中,所述AI功能网元模块所属的装置是移动通信系统中用于AI功能管理的装置,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
根据本申请的一个方面,提供了一种AI数据的传输装置,所述装置包括:AI功能网元模块;
所述AI功能网元模块,用于与第一AI功能实体模块,执行用于进行AI数据传输的交互流程;
其中,所述装置是移动通信系统中用于AI功能管理的装置,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
根据本申请的一个方面,提供了一种通信设备,所述通信设备包括:收发器;其中,
所述收发器,用于与AI功能网元,执行用于进行AI数据传输的交互流程;
其中,所述AI功能网元是移动通信系统中用于AI功能管理的网元,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
根据本申请的一个方面,提供了一种网元设备,所述网元设备包括:收发器;其中,
所述收发器,用于与第一通信设备中的第一AI功能实体,执行用于进行AI数据传输的交互流程;
其中,所述网元设备是移动通信系统中用于AI功能管理的网元,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
根据本申请的一个方面,提供了一种计算机可读存储介质,所述可读存储介质中存储有可执行指令,所述可执行指令由处理器加载并执行以实现如上述方面所述的AI数据的传输方法。
根据本申请实施例的一个方面,提供了一种芯片,所述芯片包括可编程逻辑电路和/或程序指令,当所述芯片在计算机设备上运行时,用于实现上述方面所述的AI数据的传输方法。
根据本申请的一个方面,提供了一种计算机程序产品,该计算机程序产品在计算机设备的处理器上运行时,使得计算机设备执行上述方面所述的AI数据的传输方法。
本申请实施例提供的技术方案至少包括如下有益效果:
为移动通信系统中的第一通信设备引入第一AI功能实体,并引入用于进行AI功能管理的AI功能网元,第一通信设备中的第一AI功能实体与AI功能网元,执行用于进行AI数据传输的交互流程,从而保障了AI功能在移动通信系统中的正常使用,使得移动通信系统可以基于AI功能提升系统性能。
附图说明
为了更清楚地说明本申请实施例中的技术方案,下面将对实施例描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1是本申请一个示例性实施例提供的神经网络的示意图;
图2是本申请一个示例性实施例提供的多隐藏层的神经网络的示意图;
图3是本申请一个示例性实施例提供的卷积神经网络的示意图;
图4是本申请一个示例性实施例提供的移动通信系统的框图;
图5是本申请一个示例性实施例提供的AI数据的传输方法的流程图;
图6是本申请一个示例性实施例提供的AI数据的传输方法的流程图;
图7是本申请一个示例性实施例提供的AI数据的传输方法的流程图;
图8是本申请一个示例性实施例提供的AI数据的传输方法的流程图;
图9是本申请一个示例性实施例提供的AI数据的传输装置的结构框图;
图10是本申请一个示例性实施例提供的AI数据的传输装置的结构框图;
图11是本申请一个示例性实施例提供的通信设备的结构示意图。
具体实施方式
为使本申请的目的、技术方案和优点更加清楚,下面将结合附图对本申请实施方式作进一步地详细描述。
首先,对本申请实施例中涉及的名词进行简单介绍:
当前的移动通信系统较以往来说,提供了更大的灵活性,强调对不同场景的广泛适用,以及对有限资源的充分利用。但是,目前大部分工作的基础原理还是基于对实际通信环境的理论建模或者简单的参数选取来完成的,这种基本的工作方式所能带来的增益在多变的场景及复杂的通信环境下,逐渐弱化。针对这种情况,当前有必要采用新的方法和思路,与传统无线通信理论及系统相结合,从而打破性能瓶颈,进一步提升移动通信系统的性能。
人工智能(Artificial Intelligence,AI):
近年来,以神经网络为代表的人工智能研究在很多领域都取得了非常大的成果,其也将在未来很长的一段时间内,对人们的生产生活起到重要的影响。
请参考图1,其示出了本申请一个实施例提供的神经网络的示意图。如图1所示,一个简单的神经网络的基本结构包括:输入层、隐藏层和输出层。其中,输入层负责接收数据,隐藏层负责处理数据,而最后的结果在输出层产生。如图1所示,各个节点代表一个处理单元,也可以认为是模拟了一个神经元,多个神经元组成一层神经网络,多层的信息传递与处理构造出了一个整体的神经网络。
随着神经网络研究的不断发展,近年来,神经网络深度学习算法被提出,如图2所示,较多的隐藏层被引入,通过多隐藏层的神经网络逐层训练进行特征学习,极大地提升了神经网络的学习和处理能力,并在模式识别、信号处理、优化组合、异常探测等方面被广泛应用。
同时,随着深度学习的发展,卷积神经网络也被进一步地研究。请参考图3,其示出了本申请一个实施例提供的卷积神经网络的示意图。如图3所示,在一个卷积神经网络中,其基本结构包括:输入层、多个卷积层、多个池化层、全连接层及输出层。卷积层和池化层的引入,有效地控制了网络参数的剧增,限制了参数的个数并挖掘了局部结构的特点,提高了算法的鲁棒性。
在移动通信系统中,支持AI功能来优化系统性能是非常有前景的,但是,相关技术中所提供的第三代合作伙伴项目(the 3 rd Generation Partnership Project,3GPP)协议栈并不适用于AI相关模型,数据的传输。因此,本申请提出了一种用于赋能AI功能的实现方式,从而保障了AI功能可以在移动通信系统中正常使用。
图4示出了本申请一个示例性实施例提供的移动通信系统的框图,该移动通信系统可以包括:终端设备10、接入网网元20和核心网网元30。
终端设备10可以指用户设备(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、无线通信设备、用户代理或用户装置。可选地,终端设备10还可以是蜂窝电话、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、个人数字处理(Personal Digita1 Assistant,PDA)、具有无线通信功能的手持设备、计算设备或连接到无线调制解调器的其它处理设备、车载设备、可穿戴设备,第五代移动通信系统(5th Generation System,5GS)中的终端设备或者未来演进的公用陆地移动通信网络(Pub1ic Land Mobi1e Network,PLMN)中的终端设备等,本申请实施例对此并不限定。为方便描述,上面提到的设备统称为终端设备。终端设备10的数量通常为多个,每一个接入网网元20所管理的小区内可以分布一个或 多个终端设备10。在本申请实施例中,终端设备10中可以包括第一AI功能实体。
接入网网元20是一种部署在接入网中用以为终端设备10提供无线通信功能的网元设备。接入网网元20可以包括各种形式的宏基站,微基站,中继站,接入点等等。在采用不同的无线接入技术的系统中,具备接入网网元功能的设备的名称可能会有所不同,例如在5G NR系统中,称为gNodeB或者gNB。随着通信技术的演进,“接入网网元”这一名称可能会变化。为方便描述,本申请实施例中,上述为终端设备10提供无线通信功能的装置统称为接入网网元。可选地,通过接入网网元20,终端设备10和核心网网元30之间可以建立通信关系。示例性地,在LTE(Long Term Evolution,长期演进)系统中,接入网网元20可以是EUTRAN(Evolved Universal Terrestrial Radio Access Network,演进的通用陆地无线网)或者EUTRAN中的一个或者多个eNodeB;在5G NR系统中,接入网网元20可以是RAN(Radio Access Network,无线接入网)或者RAN中的一个或者多个gNB。在本申请实施例中,接入网网元20中可以包括第一AI功能实体。可选的,接入网网元20中包括AI功能网元。
核心网网元30是部署在核心网中的网元设备,核心网网元30的功能主要是提供用户连接、对用户的管理以及对业务完成承载,作为承载网络提供到外部网络的接口。可选的,核心网网元30中包括AI功能网元。
本申请实施例中的“5G NR系统”也可以称为5G系统或者NR系统,但本领域技术人员可以理解其含义。本申请实施例描述的技术方案可以适用于LTE系统,也可以适用于5G NR系统,也可以适用于5G NR系统后续的演进系统,还可以适用于诸如NB-IoT(Narrow Band Internet of Things,窄带物联网)系统等其他通信系统,本申请对此不作限定。
图5示出了本申请一个示例性实施例提供的AI数据的传输方法的流程图。该方法可以应用于如图4示出的移动通信系统中,该方法包括:
步骤510,第一通信设备中的第一AI功能实体与AI功能网元,执行用于进行AI数据传输的交互流程。
第一通信设备中的第一AI功能实体与AI功能网元之间进行交互,以使得第一通信设备和AI功能网元之间传输AI数据。
其中,第一通信设备是移动通信系统中的通信设备,第一通信设备支持AI功能。示例性的,第一通信设备为终端设备。示例性的,第一通信设备为接入网网元。在本申请实施例中,第一通信设备包括第一AI功能实体,第一AI功能实体是第一通信设备中与AI功能相关的功能实体。
可选的,在第一通信设备是终端设备的情况下,第一AI功能实体位于如下任意一个协议层中,或,位于如下协议层中的任意两个相邻的协议层之间;其中,协议层包括如下中的至少一种:非接入(Non-Access Stratum,NAS)层;服务数据适配协议(Service Data Adaptation Protocol,SDAP)层;分组数据汇聚协议(Packet Data Convergence Protocol,PDCP)层;无线链路控制(Radio Link Control,RLC)层;媒体接入控制(Medium Access Control,MAC)层;物理(PHY)层。
也即,终端设备侧的第一AI功能实体可以位于NAS层,或SDAP层,或PDCP层,或RLC层,或MAC层,或PHY层;也可以位于上述协议层之间的一层,即任意两个相邻的协议层之间,如:NAS层和SDAP层之间,SDAP层和PDCP层之间,PDCP层和RLC层之间,RLC层和MAC层之间,MAC层和PHY层之间。
可选的,在第一通信设备是接入网网元的情况下,第一AI功能实体位于集中单元(Centralized Unit,CU),或,位于分布单元(Distributed Unit,DU),或位于如下任意一个协议层中,或,位于如下协议层中的任意两个相邻的协议层之间;其中,协议层包括如下中的至少一种:NAS层;SDAP层;PDCP层;RLC层;MAC层;PHY层。
也即,接入网网元侧的第一AI功能实体可以位于CU,或DU;可以位于NAS层,或SDAP层,或PDCP层,或RLC层,或MAC层,或PHY层;也可以位于上述协议层之间的一层,即任意两个相邻的协议层之间,如:NAS层和SDAP层之间,SDAP层和PDCP层之间,PDCP层和RLC层之间,RLC层和MAC层之间,MAC层和PHY层之间。
其中,AI功能网元是移动通信系统中用于AI功能管理的网元。AI功能管理包括对AI数据传输进行控制管理,AI数据是与AI赋能功能有关的数据。
AI赋能功能指的是为移动通信系统赋予AI功能,从而应用AI功能,优化移动通信系统的性能。
AI数据是与AI赋能功能有关的数据,可以理解为AI数据是与AI模型在移动通信系统中的应用过程有关的数据,示例性的,AI数据包括:AI功能相关的能力信息、AI模型的模型配置信息、AI模型的输入数据、AI模型的输出数据;示例性的,AI数据不包括AI模型在模型训练过程中所涉及的数据,如:AI模型的训练数据。
可选的,AI功能网元是已有网元中的任意一个;或,AI功能网元是已有网元之外的一个新增网元;其中,已有网元包括:接入网网元、网络数据分析功能(Network Data Analytics Function,NWDAF)、接入和移动性管理功能(Access and Mobility Management Function,AMF)、会话管理功能(Session Management Function,SMF)、控制策略功能(Policy Control Function,PCF)、统一数据管理(Unified Data Management,UDM)、用户面功能(User Plane Function,UPF)、网络存储库功能(Network Repository Function,NRF)和网络公开功能(Network Exposure Function,NEF)。
也即,本申请实施例中引入的AI功能网元可以是当前的通信协议标准中已定义的网元,即已有网元,已有网元包括接入网网元和核心网网元,接入网网元包括:gNB、eNB,核心网网元包括:NWDAF、AMF、SMF、PCF、UDM、UPF、NRFNEF;本申请实施例中引入的AI功能网元也可以是除已有网元之外的一个新增网元,该新增网元可以是接入网网元,也可以是核心网网元。可以理解的是,上述已有网元仅进行示例性的说明,也可以包括未示出的其他网元,本申请实施例对此不加以限制。
可选的,第一通信设备和AI功能网元之间传输AI数据基于已有接口来实现,或,第一通信设备和AI功能网元之间传输AI数据基于新增的专用于AI数据传输的接口来实现。
综上所述,本实施例提供的方法,为移动通信系统中的第一通信设备引入第一AI功能实体,并引入用于进行AI功能管理的AI功能网元,第一通信设备中的第一AI功能实体与AI功能网元,执行用于进行AI数据传输的交互流程,从而保障了AI功能在移动通信系统中的正常使用,使得移动通信系统可以基于AI功能提升系统性能。
在基于图5的示意性实施例中,第一通信设备中的第一功能实体与AI功能网元之间,可以进行如下交互:
·第一AI功能实体与AI功能网元,执行用于进行AI能力协商的交互流程。
用于进行AI能力协商的交互流程指的是,第一终端设备与AI功能网元之间沟通第一通信设备的AI功能相关的能力的交互流程。
·第一AI功能实体与AI功能网元,执行用于进行AI模型配置的交互流程。
用于进行AI模型配置的交互流程指的是,AI功能网元为第一终端设备配置AI模型的交互流程。
·第一AI功能实体与AI功能网元,执行用于进行AI模型输入的交互流程。
用于进行AI模型输入的交互流程指的是,第一终端设备与AI功能网元之间交互AI模型的输入数据的交互流程。
可以理解的是,第一通信设备中的第一功能实体与AI功能网元之间,也可以进行其他与AI赋能功能相关的交互,本申请对此不加以限制。下面,对如上示出的交互流程进行进一步说明。
·第一AI功能实体与AI功能网元,执行用于进行AI能力协商的交互流程。
图6示出了本申请一个示例性实施例提供的AI数据的传输方法的流程图。该方法可以应用于如图4示出的移动通信系统中,该方法包括:
步骤610:AI功能网元向第一通信设备发送能力请求,能力请求用于请求第一通信设备上报AI功能相关的能力。
为了获知第一通信设备的AI功能相关的能力,AI功能网元向第一通信设备发送能力请求,请求第一通信设备上报AI功能相关的能力。
步骤620:第一通信设备中的第一AI功能实体接收能力请求。
步骤630:第一通信设备中的第一AI功能实体向AI功能网元上报能力信息,能力信息用于提供AI功能相关的能力。
可以理解的是,上述步骤610至步骤630中,第一AI功能实体基于能力请求,向AI功能网元上报能力信息,在另一种可能的实现方式中,第一AI功能实体也可以主动向AI功能网元上报能力信息,也即,不需要上述步骤610和步骤620,直接执行步骤630。
第一通信设备响应于能力请求,向AI功能网元上报自身的能力信息。
可选的,能力信息包括如下中的至少一种:
·第一通信设备是否支持AI功能。
示例性的,能力信息中存在用于指示是否支持AI功能的比特位,若该比特位为“0”,则表示第一通信设备支持AI功能,若该比特位为“1”,则表示第一通信设备不支持AI功能。
·第一通信设备支持的AI功能的类型。
示例性的,移动通信系统中约定好了AI功能的类型包括:类型1、类型2和类型3等等,每个类型对应有相应的类型标识,类型标识用于标识AI功能的类型。能力信息中包括类型标识,第一通信设备通过类型标识将第一通信设备支持的AI功能的类型告知AI功能网元。
示例性的,AI功能的类型包括:用于提高预测准确度的AI功能、用于优化决策的AI功能、用于提高计算精度的AI功能等等,本申请对此不加以限制。
·第一通信设备支持的AI模型。
示例性的,移动通信系统中约定好了AI模型包括:AI模型1、AI模型2和AI模型3等等,每个AI模型对应有相应的AI模型标识,AI模型标识用于标识AI模型。能力信息中包括AI模型标识,第一通信设备通过AI模型标识将第一通信设备支持的AI模型告知AI功能网元。
·第一通信设备支持的AI算法。
示例性的,移动通信系统中约定好了AI算法包括:AI算法1、AI算法2和AI算法3等 等,每个AI算法对应有相应的AI算法标识,AI算法标识用于标识AI算法。能力信息中包括AI算法标识,第一通信设备通过AI算法标识将第一通信设备支持的AI算法告知AI功能网元。
示例性的,AI算法包括:线性回归算法、决策树算法、随机森林算法、逻辑回归算法、神经网络算法、贝叶斯算法、K近邻算法、K均值算法和马尔科夫算法等等,本申请对此不加以限制。
示例性的,能力信息还包括第一AI功能实体所属的的第一通信设备的算力,存储能力等。
步骤640:AI功能网元接收能力信息。
可选的,在步骤640之后,AI功能网元基于第一通信设备的能力信息,管理控制第一通信设备执行AI功能,如:AI功能网元向第一通信设备发送AI模型的模型配置信息,使得第一通信设备可以基于模型配置信息确定AI模型,使用AI模型提升系统性能。
综上所述,本实施例提供的方法,为移动通信系统中的第一通信设备引入第一AI功能实体,并引入用于进行AI功能管理的AI功能网元,第一通信设备中的第一AI功能实体与AI功能网元,执行用于进行AI数据传输的交互流程,从而保障了AI功能在移动通信系统中的正常使用,使得移动通信系统可以基于AI功能提升系统性能。
同时,本实施例提供的方法,第一通信设备中的第一AI功能实体与AI功能网元,执行用于进行AI能力协商的交互流程,帮助AI功能网元明确第一通信设备的AI功能相关的能力。
·第一AI功能实体与AI功能网元,执行用于进行AI模型配置的交互流程。
图7示出了本申请一个示例性实施例提供的AI数据的传输方法的流程图。该方法可以应用于如图4示出的移动通信系统中,该方法包括:
步骤710:第一通信设备中的第一AI功能实体向AI功能网元发送业务请求,业务请求中包括使能AI功能的需求。
业务请求用于请求AI功能网元满足第一通信设备使能AI功能的需求。可以理解的是,AI功能可以单独在终端设备侧实现,如:终端设备中的AI功能实体向AI功能网元发送业务请求,终端设备中的AI功能实体在得到相应的模型配置信息后,使能AI功能;也可以单独在接入网网元侧实现,如:接入网网元中的AI功能实体向AI功能网元发送业务请求,接入网网元中的AI功能实体在得到相应的模型配置信息后,使能AI功能;也可以在终端设备侧和接入网网元侧联合实现,如:终端设备中的AI功能实体和接入网网元中的AI功能实体向AI功能网元发送同一类型的业务请求,终端设备中的AI功能实体和接入网网元中的AI功能实体在得到相应的模型配置信息后,同步或非同步使能同一类型的AI功能。
可选的,使能AI功能的需求包括如下中的至少一种:
·基于AI功能进行信道状态信息(Channel-State Information,CSI)反馈预测。
在移动通信系统中,终端设备通过对接入网网元配置的参考信号的测量,确定当前的CSI,并向接入网网元反馈当前的CSI,供接入网网元基于CSI确定当前的信道情况。CSI反馈预测指的是第一通信设备基于AI功能,对待反馈的CSI进行预测,从而尽早确定信道情况。
示例性的,第一通信设备为接入网网元,接入网网元基于AI功能进行CSI反馈预测。示例性的,第一通信设备为终端设备,终端设备基于AI功能进行CSI反馈预测。示例性的,第一通信设备为接入网网元和终端设备,终端设备和接入网网元基于AI功能联合进行CSI反馈预测。
·基于AI功能进行波束管理增强。
波束管理增强指的是第一通信设备基于AI功能,对波束失败信息进行预测,并且进一步确定需要请求恢复或切换的目标波束。
示例性的,第一通信设备为终端设备,终端设备基于AI功能进行波束管理增强,比如终端基于AI模型对于波束进行预测,基于预测结果提前进行波束恢复或切换,避免了由于波束失败恢复导致的数据传输中断。示例性的,第一通信设备为接入网网元,接入网网元基于AI功能进行波束管理增强。示例性的,第一通信设备为接入网网元和终端设备,终端设备和接入网网元基于AI功能联合进行波束管理增强。
·基于AI功能进行切换决策增强。
切换决策增强指的是第一通信设备基于AI功能,优化小区切换的决策。
示例性的,第一通信设备为终端设备,终端设备基于AI功能确定切换条件,和或切换成功率和或目标小区标识等信息,从而进行切换决策增强。示例性的,第一通信设备为接入网网元,接入网网元基于AI功能确定切换条件,和或切换成功率和或目标小区标识等信息,从而进行切换决策增强。示例性的,第一通信设备为接入网网元和终端设备,终端设备和接入网网元基于AI功能联合进行切换决策增强。
·基于AI功能进行定位精度增强。
定位精度增强指的是第一通信设备基于AI功能,提高定位结果的精度。
示例性的,第一通信设备为终端设备,终端设备基于AI算法优化定位精度,消除非视距(Non-Line of Sight,NLOS)场景带来的精度误差,从而进行定位精度增强。示例性的,第一通信设备为接入网网元,接入网网元基于AI算法优化定位精度,消除NLOS场景带来的精度误差,从而进行定位精度增强。示例性的,第一通信设备为接入网网元和终端设备,终端设备和接入网网元基于AI功能联合进行定位精度增强。
·基于AI功能进行随机接入增强。
随机接入增强指的是第一通信设备基于AI功能,优化随机接入资源分配。
示例性的,第一通信设备为接入网网元,接入网网元基于AI功能进行随机接入增强。示例性的,第一通信设备为终端设备,终端设备基于AI功能进行随机接入增强。示例性的,第一通信设备为接入网网元和终端设备,终端设备和接入网网元基于AI功能联合进行随机接入增强。
·基于AI功能进行资源调度增强。
资源调度增强指的是第一通信设备基于AI功能,优化资源调度,资源调度包括:动态资源调度以及半静态资源调度。
示例性的,第一通信设备为接入网网元,接入网网元基于AI功能进行资源调度增强。示例性的,第一通信设备为终端设备,终端设备基于AI功能进行资源调度增强。示例性的,第一通信设备为接入网网元和终端设备,终端设备和接入网网元基于AI功能联合进行资源调度增强。
·基于AI功能进行终端路径预测。
终端路径预测指的是第一通信设备基于AI功能,预测终端设备的运行轨迹,从而帮助预留相应资源以及均衡小区负载。
示例性的,第一通信设备为接入网网元,接入网网元基于AI功能进行终端路径预测。示例性的,第一通信设备为终端设备,终端设备基于AI功能进行终端路径预测。示例性的,第一通信设备为接入网网元和终端设备,终端设备和接入网网元基于AI功能联合进行终端路径 预测。
·基于AI功能进行终端业务预测。
终端业务预测指的是第一通信设备基于AI功能,预测终端设备的业务类型,从而帮助优化资源的预留与分配。
示例性的,第一通信设备为接入网网元,接入网网元基于AI功能进行终端业务预测。示例性的,第一通信设备为终端设备,终端设备基于AI功能进行终端业务预测。示例性的,第一通信设备为接入网网元和终端设备,终端设备和接入网网元基于AI功能联合进行终端业务预测。
·基于AI功能进行小区负载预测。
小区负载预测指的是是第一通信设备基于AI功能,预测小区的负载情况,从而帮助优化资源分配。
示例性的,第一通信设备为接入网网元,接入网网元基于AI功能进行小区负载预测。示例性的,第一通信设备为终端设备,终端设备基于AI功能进行小区负载预测。示例性的,第一通信设备为接入网网元和终端设备,终端设备和接入网网元基于AI功能联合进行小区负载预测。
可以理解的是,使能AI功能的需求也可以包括其他类型的需求,本申请对此不加以限制。
步骤720:AI功能网元接收业务请求。
步骤730:AI功能网元向第一通信设备发送模型配置信息,模型配置信息用于供第一通信设备确定与使能AI功能的需求所对应的AI模型。
也即,AI功能网元收到业务请求后,为第一通信设备发送用于指示与使能AI功能的需求所对应的AI模型的模型配置信息,第一通信设备从AI功能网元处下载(download)模型配置信息。
可选的,在步骤730之前,AI功能实体基于历史数据训练好AI模型,并确定AI模型对应的模型配置信息。
步骤740:第一通信设备中的第一AI功能实体接收模型配置信息。
可选的,模型配置信息包括如下中的至少一种:AI模型的层级结构;AI模型的网络参数的权重信息。
可选的,在步骤740之后,还包括如下步骤:第一AI功能实体对接收到的模型配置信息进行更新;第一AI功能实体向AI功能网元发送更新后的模型配置信息;AI功能网元接收第一AI功能实体发送的更新后的模型配置信息。
也即,第一通信设备收到模型配置信息,确定相应的AI模型后,可以基于自己的实时信息作为输入,进一步更新AI模型,并上传(upload)更新后的AI模型对应的模型配置信息给AI功能网元。
可选的,在步骤740之后,还包括如下步骤:第一AI功能实体接收至少一个其他AI功能实体发送的实时信息,该实时信息用于向第一AI功能实体中的AI模型提供输入数据。
示例性的,第一通信设备为接入网网元,接入网网元中的第一AI功能实体基于AI功能进行资源调度增强。第一AI功能实体在接收到模型配置信息,确定相应的AI模型后,除了将本侧的实时信息作为AI模型的输入之外,还接收至少一个终端设备中的其他AI功能实体发送的实时信息,并将接收到的终端设备侧的实时信息也作为AI模型的输入。
可选的,其他AI功能实体接收来自于第一AI功能实体或者AI功能网元的指示信息,基于指示信息向第一AI功能实体发送实时信息。示例性的,AI功能网元向第一通信设备发 送模型配置信息后,向其他AI功能实体发送指示信息,指示其他AI功能实体向第一AI功能实体发送实时信息。示例性的,第一AI功能实体在接收到模型配置信息后,向其他AI功能实体发送指示信息,指示其他AI功能实体向第一AI功能实体发送实时信息。
可选的,其他AI功能实体发送的实时信息的类型与第一功能实体中使能的AI功能存在对应关系,不同的AI功能对应于不同的实时信息的类型。示例性的,AI功能包括:CSI反馈预测、波束管理增强、切换决策增强、定位精度增强、随机接入增强、资源调度增强、终端路径预测、终端业务预测和小区负载预测,相应的,实时信息的类型包括:CSI反馈预测对应的输入类型、波束管理增强对应的输入类型、切换决策增强对应的输入类型、定位精度增强对应的输入类型、随机接入增强对应的输入类型、资源调度增强对应的输入类型、终端路径预测对应的输入类型、终端业务预测对应的输入类型和小区负载预测对应的输入类型。
可选的,在步骤740之后,第一通信设备中的第一AI功能实体基于模型配置信息确定AI模块,输入模型输入数据,得到AI模型的输出,上述输出能够满足第一通信设备使能AI功能的需求,进而优化移动通信系统中的性能。
综上所述,本实施例提供的方法,为移动通信系统中的第一通信设备引入第一AI功能实体,并引入用于进行AI功能管理的AI功能网元,第一通信设备中的第一AI功能实体与AI功能网元,执行用于进行AI数据传输的交互流程,从而保障了AI功能在移动通信系统中的正常使用,使得移动通信系统可以基于AI功能提升系统性能。
同时,本实施例提供的方法,第一通信设备中的第一AI功能实体与AI功能网元,执行用于进行AI模型配置的交互流程,帮助第一通信设备确定AI模型,从而使用AI模型满足自身需求,进而优化移动通信系统中的性能。
·第一AI功能实体与AI功能网元,执行用于进行AI模型输入的交互流程。
图8示出了本申请一个示例性实施例提供的AI数据的传输方法的流程图。该方法可以应用于如图4示出的移动通信系统中,该方法包括:
步骤810:第一通信设备中的第一AI功能实体向AI功能网元发送辅助数据信息,辅助数据信息用于提供AI模型的输入数据。
也即,第一通信设备具有使能AI功能的需求,第一通信设备将满足该需求所需要进行处理的数据携带在辅助数据信息中,发送给AI功能网元,由AI功能网元将上述数据输入AI模型进行处理。
示例性的,第一通信设备为接入网网元,接入网网元具有基于AI功能进行资源调度增强的需求,接入网网元中的第一AI功能实体向AI功能网元发送辅助数据信息,辅助数据信息中携带影响资源调度决策的数据,AI功能网元将上述数据作为AI模型的输入数据,输出资源调度决策。
可选的,在步骤810之前,第一通信设备中的第一AI功能实体向AI功能网元发送业务请求,业务请求中包括使能AI功能的需求;AI功能网元相应接收业务请求。
在一种可能的实现方式中,第一通信设备主动上报辅助数据信息。也即,第一通信设备在上报业务请求后,直接发送辅助数据信息。
在另一种可能的实现方式中,第一通信设备在AI功能网元的触发下,上报辅助数据信息。示例性的,在步骤810之前,还包括如下步骤:AI功能网元向第一AI功能实体发送辅助数据请求,辅助数据请求用于请求第一通信设备发送辅助数据信息;第一AI功能实体相应接收AI功能网元发送的辅助数据请求,辅助数据请求用于请求第一通信设备发送辅助数据信息。 也即,第一通信设备在上报业务请求后,AI功能网元先请求第一通信设备发送辅助数据信息,第一通信设备再相应发送辅助数据信息。
步骤820:AI功能网元接收辅助数据信息。
可选的,在步骤820之后,AI功能网元将辅助数据信息中携带的数据输入AI模型,得到AI模型的输出,并将输出发送给第一通信设备,上述输出能够满足第一通信设备使能AI功能的需求,进而优化移动通信系统中的性能。
可选的,除了在上述步骤810至步骤820中,第一通信设备侧第一AI功能实体与AI功能网元交互辅助数据信息之外,第一通信设备侧第一AI功能实体也可以与其他通信设备中的AI功能实体交互辅助数据信息,以实现辅助数据信息的传输。可选的,其他通信设备中的AI功能实体中存储有AI模型,将辅助数据信息中的数据作为AI模型的输入数据。可选的,其他通信设备将辅助数据信息再次进行传递,最终传递给AI功能网元进行处理。
在一种可能的实现方式中,第一AI功能实体通过第一接口与第二通信设备的第二AI功能实体交互辅助数据信息,第二通信设备是与第一通信设备不同类型的通信设备。
也即,不同类型的通信设备之间可以基于第一接口交互辅助数据信息。如:终端设备和接入网网元之间交互辅助数据信息。
可选的,第一接口包括:Uu接口;或,专用于AI功能的接口。也即,第一接口可以是现有的Uu接口,也可以是新增的专用于AI功能的接口。
可选的,辅助数据信息携带在如下消息中的至少一种:无线资源控制(Radio Resource Control,RRC)消息、MAC控制信元(Control Element,CE)、NAS消息和专用于AI功能的消息。
在另一种可能的实现方式中,第一AI功能实体与第三通信设备的第三AI功能实体交互辅助数据信息,第三通信设备是与第一通信设备相同类型的通信设备。
也即,相同类型的通信设备之间可以交互辅助数据信息。如:终端设备和终端设备之间交互辅助数据信息,接入网网元和接入网网元之间交互辅助数据信息。
综上所述,本实施例提供的方法,为移动通信系统中的第一通信设备引入第一AI功能实体,并引入用于进行AI功能管理的AI功能网元,第一通信设备中的第一AI功能实体与AI功能网元,执行用于进行AI数据传输的交互流程,从而保障了AI功能在移动通信系统中的正常使用,使得移动通信系统可以基于AI功能提升系统性能。
同时,本实施例提供的方法,第一通信设备中的第一AI功能实体与AI功能网元,执行用于进行AI模型输入的交互流程,由第一通信设备中的第一AI功能实体将辅助数据信息上报给AI功能网元,AI功能网元使用AI模型对辅助数据信息中的数据进行处理,以满足第一通信设备使能AI功能的需求。
需要说明的是,上述方法实施例可以分别单独实施,也可以组合实施,本申请对此不进行限制。
在上述各个实施例中,由第一通信设备执行的步骤可以单独实现成为第一通信设备设备一侧的AI数据的传输方法,由AI功能网元执行的步骤可以单独实现成为AI功能网元一侧的AI数据的传输方法。
图9示出了本申请一个示例性实施例提供的AI数据的传输装置的结构框图,该装置可以实现成为第一通信设备,或者,实现成为第一通信设备中的一部分,该装置包括:第一AI 功能实体模块901;
所述第一AI功能实体模块901,用于与AI功能网元模块,执行用于进行AI数据传输的交互流程;
其中,所述AI功能网元模块所属的装置是移动通信系统中用于AI功能管理的装置,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
在一个可选的实施例中,所述AI功能网元模块所属的装置是已有网元中的任意一个;或,所述AI功能网元模块所属的装置是所述已有网元之外的一个新增网元;
其中,所述已有网元包括:接入网网元、NWDAF、AMF、SMF、PCF、UDM、UPF、NRF和NEF。
在一个可选的实施例中,在所述装置是终端设备的情况下,所述第一AI功能实体模块901位于如下任意一个协议层中,或,位于如下协议层中的任意两个相邻的协议层之间;
其中,所述协议层包括如下中的至少一种:NAS层;SDAP层;PDCP层;RLC层;MAC层;PHY层。
在一个可选的实施例中,在所述装置是接入网网元的情况下,所述第一AI功能实体模块901位于CU,或,位于DU,或位于如下任意一个协议层中,或,位于如下协议层中的任意两个相邻的协议层之间;
其中,所述协议层包括如下中的至少一种:NAS层;SDAP层;PDCP层;RLC层;MAC层;PHY层。
在一个可选的实施例中,所述第一AI功能实体模块901,用于与所述AI功能网元模块,执行用于进行AI能力协商的交互流程;
或,
所述第一AI功能实体模块901,用于与所述AI功能网元模块,执行用于进行AI模型配置的交互流程;
或,
所述第一AI功能实体模块901,用于与所述AI功能网元模块,执行用于进行AI模型输入的交互流程。
在一个可选的实施例中,所述第一AI功能实体模块901,用于接收所述AI功能网元模块发送的能力请求,所述能力请求用于请求所述装置上报AI功能相关的能力;所述第一AI功能实体模块901,用于向所述AI功能网元模块上报能力信息,所述能力信息用于提供AI功能相关的能力。
在一个可选的实施例中,所述能力信息包括如下中的至少一种:
所述装置是否支持AI功能;
所述装置支持的AI功能的类型;
所述装置支持的AI模型;
所述装置支持的AI算法。
在一个可选的实施例中,所述第一AI功能实体模块901,用于向所述AI功能网元模块发送业务请求,所述业务请求中包括使能AI功能的需求;所述第一AI功能实体模块901,用于接收所述AI功能网元模块发送的模型配置信息,所述模型配置信息用于供所述装置装置确定与所述使能AI功能的需求所对应的AI模型。
在一个可选的实施例中,所述第一AI功能实体模块901,用于对接收到的所述模型配置 信息进行更新;所述第一AI功能实体模块901,用于向所述AI功能网元模块发送更新后的所述模型配置信息。
在一个可选的实施例中,所述模型配置信息包括如下中的至少一种:
AI模型的层级结构;
AI模型的网络参数的权重信息。
在一个可选的实施例中,所述使能AI功能的需求包括如下中的至少一种:
基于AI功能进行CSI反馈预测、基于AI功能进行波束管理增强、基于AI功能进行切换决策增强、基于AI功能进行定位精度增强、基于AI功能进行随机接入增强、基于AI功能进行资源调度增强、基于AI功能进行终端路径预测、基于AI功能进行终端业务预测和基于AI功能进行小区负载预测。
在一个可选的实施例中,所述第一AI功能实体模块901,用于向所述AI功能网元模块发送辅助数据信息,所述辅助数据信息用于提供所述AI模型的输入数据。
在一个可选的实施例中,所述第一AI功能实体模块901,用于接收所述AI功能网元模块发送的辅助数据请求,所述辅助数据请求用于请求所述装置发送所述辅助数据信息。
在一个可选的实施例中,所述第一AI功能实体模块901,用于通过第一接口与第二AI功能实体模块交互所述辅助数据信息,所述第二AI功能实体模块所属的装置是与所述装置不同类型的装置。
在一个可选的实施例中,所述第一接口包括:
Uu接口;
或,
专用于AI功能的接口。
在一个可选的实施例中,所述辅助数据信息携带在如下消息中的至少一种:
RRC消息、MAC CE、NAS消息和专用于AI功能的消息。
在一个可选的实施例中,所述第一AI功能实体模块901,用于与第三AI功能实体模块交互所述辅助数据信息,所述第三AI功能实体模块所属的装置是与所述装置相同类型的装置。
图10示出了本申请一个示例性实施例提供的AI数据的传输装置的结构框图,该装置可以实现成为AI功能网元,或者,实现成为AI功能网元中的一部分,该装置包括:AI功能网元模块1001;
所述AI功能网元模块1001,用于与第一AI功能实体模块,执行用于进行AI数据传输的交互流程;
其中,所述装置是移动通信系统中用于AI功能管理的装置,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
在一个可选的实施例中,所述装置是已有网元中的任意一个;或,所述装置是所述已有网元之外的一个新增网元;
其中,所述已有网元包括:接入网网元、NWDAF、AMF、SMF、PCF、UDM、UPF、NRF和NEF。
在一个可选的实施例中,在所述第一AI功能实体模块所属的装置是终端设备的情况下,所述第一AI功能实体模块位于如下任意一个协议层中,或,位于如下协议层中的任意两个相邻的协议层之间;
其中,所述协议层包括如下中的至少一种:NAS层;SDAP层;PDCP层;RLC层;MAC层;PHY层。
在一个可选的实施例中,在所述第一AI功能实体模块所属的装置是接入网网元的情况下,所述第一AI功能实体模块位于CU,或,位于DU,或位于如下任意一个协议层中,或,位于如下协议层中的任意两个相邻的协议层之间;
其中,所述协议层包括如下中的至少一种:NAS层;SDAP层;PDCP层;RLC层;MAC层;PHY层。
在一个可选的实施例中,所述AI功能网元模块1001,用于与所述第一AI功能实体模块,执行用于进行AI能力协商的交互流程;
或,
所述AI功能网元模块1001,用于与所述第一AI功能实体模块,执行用于进行AI模型配置的交互流程;
或,
所述AI功能网元模块1001,用于与所述第一AI功能实体模块,执行用于进行AI模型输入的交互流程。
在一个可选的实施例中,所述AI功能网元模块1001,用于向所述第一AI功能实体模块发送能力请求,所述能力请求用于请求所述第一AI功能实体模块所属的装置上报AI功能相关的能力;所述AI功能网元模块1001,用于接收所述第一AI功能实体模块上报的能力信息,所述能力信息用于提供AI功能相关的能力。
在一个可选的实施例中,所述能力信息包括如下中的至少一种:
所述第一AI功能实体模块所属的装置是否支持AI功能;
所述第一AI功能实体模块所属的装置支持的AI功能的类型;
所述第一AI功能实体模块所属的装置支持的AI模型;
所述第一AI功能实体模块所属的装置支持的AI算法。
在一个可选的实施例中,所述AI功能网元模块1001,用于接收所述第一AI功能实体模块发送的业务请求,所述业务请求中包括使能AI功能的需求;所述AI功能网元模块1001,用于向所述第一AI功能实体模块发送模型配置信息,所述模型配置信息用于供所述第一AI功能实体模块所属的装置确定与所述使能AI功能的需求所对应的AI模型。
在一个可选的实施例中,所述AI功能网元模块1001,用于接收所述第一AI功能实体模块发送的更新后的所述模型配置信息。
在一个可选的实施例中,所述模型配置信息包括如下中的至少一种:
AI模型的层级结构;
AI模型的网络参数的权重信息。
在一个可选的实施例中,所述使能AI功能的需求包括如下中的至少一种:
基于AI功能进行CSI反馈预测、基于AI功能进行波束管理增强、基于AI功能进行切换决策增强、基于AI功能进行定位精度增强、基于AI功能进行随机接入增强、基于AI功能进行资源调度增强、基于AI功能进行终端路径预测、基于AI功能进行终端业务预测和基于AI功能进行小区负载预测。
在一个可选的实施例中,所述AI功能网元模块1001,用于接收所述第一AI功能实体模块发送的辅助数据信息,所述辅助数据信息用于提供所述AI模型的输入数据。
在一个可选的实施例中,所述AI功能网元模块1001,用于向所述第一AI功能实体模块 发送辅助数据请求,所述辅助数据请求用于请求所述第一AI功能实体模块所属的装置发送所述辅助数据信息。
图11示出了本申请一个示例性实施例提供的通信设备(如上所述的第一通信设备或AI功能网元)的结构示意图,该通信设备包括:处理器1101、接收器1102、发射器1103、存储器1104和总线1105。
处理器1101包括一个或者一个以上处理核心,处理器1101通过运行软件程序以及模块,从而执行各种功能应用。
接收器1102和发射器1103可以实现为一个收发器1106,该收发器1106可以是一块通信芯片。
存储器1104通过总线1105与处理器1101相连。
存储器1104可用于存储计算机程序,处理器1101用于执行该计算机程序,以实现上述方法实施例中通信设备执行的各个步骤。
此外,存储器1104可以由任何类型的易失性或非易失性存储设备或者它们的组合实现,易失性或非易失性存储设备包括但不限于:随机存储器(Random-Access Memory,RAM)和只读存储器(Read-Only Memory,ROM)、可擦写可编程只读存储器(Erasable Programmable Read-Only Memory,EPROM)、电可擦写可编程只读存储器(Electrically Erasable Programmable Read-Only Memory,EEPROM)、闪存或其他固态存储其技术,只读光盘(Compact Disc Read-Only Memory,CD-ROM)、高密度数字视频光盘(Digital Video Disc,DVD)或其他光学存储、磁带盒、磁带、磁盘存储或其他磁性存储设备。
其中,当通信设备实现为第一通信设备时,本申请实施例涉及的通信设备中的处理器和收发器,可以执行上述图5至图8任一所示的方法中,由第一通信设备执行的步骤,此处不再赘述。
在一种可能的实现方式中,当通信设备实现为第一通信设备时,
所述收发器1106,用于与AI功能网元,执行用于进行AI数据传输的交互流程;
其中,所述AI功能网元是移动通信系统中用于AI功能管理的网元,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
其中,当通信设备实现为AI功能网元时,本申请实施例涉及的通信设备中的处理器和收发器,可以执行上述图5至图8任一所示的方法中,由AI功能网元执行的步骤,此处不再赘述。
在一种可能的实现方式中,当通信设备实现为AI功能网元时,
所述收发器1106,用于与第一通信设备中的第一AI功能实体,执行用于进行AI数据传输的交互流程;
其中,所述AI功能网元是移动通信系统中用于AI功能管理的网元,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
在示例性实施例中,还提供了一种计算机可读存储介质,所述计算机可读存储介质中存储有至少一条指令、至少一段程序、代码集或指令集,所述至少一条指令、所述至少一段程序、所述代码集或指令集由处理器加载并执行以实现上述各个方法实施例提供的AI数据的传输方法。
在示例性实施例中,还提供了一种芯片,所述芯片包括可编程逻辑电路和/或程序指令,当所述芯片在计算机设备上运行时,用于实现上述方面所述的AI数据的传输方法。
在示例性实施例中,还提供了一种计算机程序产品,该计算机程序产品在计算机设备的处理器上运行时,使得计算机设备执行上述方面所述的AI数据的传输方法。
本领域普通技术人员可以理解实现上述实施例的全部或部分步骤可以通过硬件来完成,也可以通过程序来指令相关的硬件完成,所述的程序可以存储于一种计算机可读存储介质中,上述提到的存储介质可以是只读存储器,磁盘或光盘等。
以上所述仅为本申请的可选实施例,并不用以限制本申请,凡在本申请的精神和原则之内,所作的任何修改、等同替换、改进等,均应包含在本申请的保护范围之内。

Claims (37)

  1. 一种人工智能AI数据的传输方法,其特征在于,所述方法由第一通信设备执行,所述第一通信设备包括:第一AI功能实体,所述方法包括:
    所述第一AI功能实体与AI功能网元,执行用于进行AI数据传输的交互流程;
    其中,所述AI功能网元是移动通信系统中用于AI功能管理的网元,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
  2. 根据权利要求1所述的方法,其特征在于,
    所述AI功能网元是已有网元中的任意一个;或,所述AI功能网元是所述已有网元之外的一个新增网元;
    其中,所述已有网元包括:接入网网元、网络数据分析功能NWDAF、接入和移动性管理功能AMF、会话管理功能SMF、控制策略功能PCF、统一数据管理UDM、用户面功能UPF、网络存储库功能NRF和网络公开功能NEF。
  3. 根据权利要求1或2所述的方法,其特征在于,
    在所述第一通信设备是终端设备的情况下,所述第一AI功能实体位于如下任意一个协议层中,或,位于如下协议层中的任意两个相邻的协议层之间;
    其中,所述协议层包括如下中的至少一种:非接入NAS层;服务数据适配协议SDAP层;分组数据汇聚协议PDCP层;无线链路控制RLC层;媒体接入控制MAC层;物理PHY层。
  4. 根据权利要求1或2所述的方法,其特征在于,
    在所述第一通信设备是接入网网元的情况下,所述第一AI功能实体位于集中单元CU,或,位于分布单元DU,或位于如下任意一个协议层中,或,位于如下协议层中的任意两个相邻的协议层之间;
    其中,所述协议层包括如下中的至少一种:NAS层;SDAP层;PDCP层;RLC层;MAC层;PHY层。
  5. 根据权利要求1至4任一所述的方法,其特征在于,所述第一AI功能实体与AI功能网元,执行用于进行AI数据传输的交互流程,包括:
    所述第一AI功能实体与所述AI功能网元,执行用于进行AI能力协商的交互流程;
    或,
    所述第一AI功能实体与所述AI功能网元,执行用于进行AI模型配置的交互流程;
    或,
    所述第一AI功能实体与所述AI功能网元,执行用于进行AI模型输入的交互流程。
  6. 根据权利要求5所述的方法,其特征在于,所述第一AI功能实体与所述AI功能网元,执行用于进行AI能力协商的交互流程,包括:
    所述第一AI功能实体接收所述AI功能网元发送的能力请求,所述能力请求用于请求所述第一通信设备上报AI功能相关的能力;
    所述第一AI功能实体向所述AI功能网元上报能力信息,所述能力信息用于提供AI功能相关的能力。
  7. 根据权利要求6所述的方法,其特征在于,所述能力信息包括如下中的至少一种:
    所述第一通信设备是否支持AI功能;
    所述第一通信设备支持的AI功能的类型;
    所述第一通信设备支持的AI模型;
    所述第一通信设备支持的AI算法。
  8. 根据权利要求5所述的方法,其特征在于,所述第一AI功能实体与所述AI功能网元,执行用于进行AI模型配置的交互流程,包括:
    所述第一AI功能实体向所述AI功能网元发送业务请求,所述业务请求中包括使能AI功能的需求;
    所述第一AI功能实体接收所述AI功能网元发送的模型配置信息,所述模型配置信息用于供所述第一通信设备确定与所述使能AI功能的需求所对应的AI模型。
  9. 根据权利要求8所述的方法,其特征在于,所述方法还包括:
    所述第一AI功能实体对接收到的所述模型配置信息进行更新;
    所述第一AI功能实体向所述AI功能网元发送更新后的所述模型配置信息。
  10. 根据权利要求8或9所述的方法,其特征在于,所述模型配置信息包括如下中的至少一种:
    AI模型的层级结构;
    AI模型的网络参数的权重信息。
  11. 根据权利要求8至10任一所述的方法,其特征在于,所述使能AI功能的需求包括如下中的至少一种:
    基于AI功能进行信道状态信息CSI反馈预测、基于AI功能进行波束管理增强、基于AI功能进行切换决策增强、基于AI功能进行定位精度增强、基于AI功能进行随机接入增强、基于AI功能进行资源调度增强、基于AI功能进行终端路径预测、基于AI功能进行终端业务预测和基于AI功能进行小区负载预测。
  12. 根据权利要求5所述的方法,其特征在于,所述第一AI功能实体与所述AI功能网元,执行用于进行AI模型输入的交互流程,包括:
    所述第一AI功能实体向所述AI功能网元发送辅助数据信息,所述辅助数据信息用于提供所述AI模型的输入数据。
  13. 根据权利要求12所述的方法,其特征在于,所述方法还包括:
    所述第一AI功能实体接收所述AI功能网元发送的辅助数据请求,所述辅助数据请求用于请求所述第一通信设备发送所述辅助数据信息。
  14. 根据权利要求12或13所述的方法,其特征在于,所述方法还包括:
    所述第一AI功能实体通过第一接口与第二通信设备的第二AI功能实体交互所述辅助数据信息,所述第二通信设备是与所述第一通信设备不同类型的通信设备。
  15. 根据权利要求14所述的方法,其特征在于,所述第一接口包括:
    Uu接口;
    或,
    专用于AI功能的接口。
  16. 根据权利要求14或15所述的方法,其特征在于,所述辅助数据信息携带在如下消息中的至少一种:
    无线资源控制RRC消息、媒体接入控制控制信元MAC CE、非接入层NAS消息和专用于AI功能的消息。
  17. 根据权利要求12至16任一所述的方法,其特征在于,所述方法还包括:
    所述第一AI功能实体与第三通信设备的第三AI功能实体交互所述辅助数据信息,所述第三通信设备是与所述第一通信设备相同类型的通信设备。
  18. 一种人工智能AI数据的传输方法,其特征在于,所述方法由AI功能网元执行,所述方法包括:
    所述AI功能网元与第一通信设备中的第一AI功能实体,执行用于进行AI数据传输的交互流程;
    其中,所述AI功能网元是移动通信系统中用于AI功能管理的网元,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
  19. 根据权利要求18所述的方法,其特征在于,
    所述AI功能网元是已有网元中的任意一个;或,所述AI功能网元是所述已有网元之外的一个新增网元;
    其中,所述已有网元包括:接入网网元、网络数据分析功能NWDAF、接入和移动性管理功能AMF、会话管理功能SMF、控制策略功能PCF、统一数据管理UDM、用户面功能UPF、网络存储库功能NRF和网络公开功能NEF。
  20. 根据权利要求18或19所述的方法,其特征在于,
    在所述第一通信设备是终端设备的情况下,所述第一AI功能实体位于如下任意一个协议层中,或,位于如下协议层中的任意两个相邻的协议层之间;
    其中,所述协议层包括如下中的至少一种:非接入NAS层;服务数据适配协议SDAP层;分组数据汇聚协议PDCP层;无线链路控制RLC层;媒体接入控制MAC层;物理PHY层。
  21. 根据权利要求18或19所述的方法,其特征在于,
    在所述第一通信设备是接入网网元的情况下,所述第一AI功能实体位于集中单元CU,或,位于分布单元DU,或位于如下任意一个协议层中,或,位于如下协议层中的任意两个相邻的协议层之间;
    其中,所述协议层包括如下中的至少一种:NAS层;SDAP层;PDCP层;RLC层;MAC层;PHY层。
  22. 根据权利要求18至21任一所述的方法,其特征在于,所述AI功能网元与第一通信设备中的第一AI功能实体,执行用于进行AI数据传输的交互流程,包括:
    所述AI功能网元与所述第一AI功能实体,执行用于进行AI能力协商的交互流程;
    或,
    所述AI功能网元与所述第一AI功能实体,执行用于进行AI模型配置的交互流程;
    或,
    所述AI功能网元与所述第一AI功能实体,执行用于进行AI模型输入的交互流程。
  23. 根据权利要求22所述的方法,其特征在于,所述AI功能网元与所述第一AI功能实体,执行用于进行AI能力协商的交互流程,包括:
    所述AI功能网元向所述第一AI功能实体发送能力请求,所述能力请求用于请求所述第一通信设备上报AI功能相关的能力;
    所述AI功能网元接收所述第一AI功能实体上报的能力信息,所述能力信息用于提供AI功能相关的能力。
  24. 根据权利要求23所述的方法,其特征在于,所述能力信息包括如下中的至少一种:
    所述第一通信设备是否支持AI功能;
    所述第一通信设备支持的AI功能的类型;
    所述第一通信设备支持的AI模型;
    所述第一通信设备支持的AI算法。
  25. 根据权利要求22所述的方法,其特征在于,所述AI功能网元与所述第一AI功能实体,执行用于进行AI模型配置的交互流程,包括:
    所述AI功能网元接收所述第一AI功能实体发送的业务请求,所述业务请求中包括使能AI功能的需求;
    所述AI功能网元向所述第一AI功能实体发送模型配置信息,所述模型配置信息用于供所述第一通信设备确定与所述使能AI功能的需求所对应的AI模型。
  26. 根据权利要求25所述的方法,其特征在于,所述方法还包括:
    所述AI功能网元接收所述第一AI功能实体发送的更新后的所述模型配置信息。
  27. 根据权利要求25或26所述的方法,其特征在于,所述模型配置信息包括如下中的至少一种:
    AI模型的层级结构;
    AI模型的网络参数的权重信息。
  28. 根据权利要求25至27任一所述的方法,其特征在于,所述使能AI功能的需求包括如下中的至少一种:
    基于AI功能进行信道状态信息CSI反馈预测、基于AI功能进行波束管理增强、基于AI功能进行切换决策增强、基于AI功能进行定位精度增强、基于AI功能进行随机接入增强、基于AI功能进行资源调度增强、基于AI功能进行终端路径预测、基于AI功能进行终端业务预测和基于AI功能进行小区负载预测。
  29. 根据权利要求22所述的方法,其特征在于,所述AI功能网元与所述第一AI功能实体,执行用于进行AI模型输入的交互流程,包括:
    所述AI功能网元接收所述第一AI功能实体发送的辅助数据信息,所述辅助数据信息用于提供所述AI模型的输入数据。
  30. 根据权利要求29所述的方法,其特征在于,所述方法还包括:
    所述AI功能网元向所述第一AI功能实体发送辅助数据请求,所述辅助数据请求用于请求所述第一通信设备发送所述辅助数据信息。
  31. 一种人工智能AI数据的传输装置,其特征在于,所述装置包括:第一AI功能实体模块;
    所述第一AI功能实体模块,用于与AI功能网元模块,执行用于进行AI数据传输的交互流程;
    其中,所述AI功能网元模块所属的装置是移动通信系统中用于AI功能管理的装置,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
  32. 一种人工智能AI数据的传输装置,其特征在于,所述装置包括:AI功能网元模块;
    所述AI功能网元模块,用于与第一AI功能实体模块,执行用于进行AI数据传输的交互流程;
    其中,所述装置是移动通信系统中用于AI功能管理的装置,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
  33. 一种通信设备,其特征在于,所述通信设备包括:收发器;其中,
    所述收发器,用于与AI功能网元,执行用于进行AI数据传输的交互流程;
    其中,所述AI功能网元是移动通信系统中用于AI功能管理的网元,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
  34. 一种网元设备,其特征在于,所述网元设备包括:收发器;其中,
    所述收发器,用于与第一通信设备中的第一AI功能实体,执行用于进行AI数据传输的交互流程;
    其中,所述网元设备是移动通信系统中用于AI功能管理的网元,所述AI功能管理包括对所述AI数据传输进行控制管理,所述AI数据是与AI赋能功能有关的数据。
  35. 一种计算机可读存储介质,其特征在于,所述可读存储介质中存储有可执行指令,所述可执行指令由处理器加载并执行以实现如权利要求1至30任一所述的AI数据的传输方法。
  36. 一种芯片,其中,所述芯片包括可编程逻辑电路和/或程序指令,当所述芯片运行时,用于实现如权利要求1至30任一所述的AI数据的传输方法。
  37. 一种计算机程序产品或计算机程序,其中,所述计算机程序产品或计算机程序包括计算机指令,所述计算机指令存储在计算机可读存储介质中,处理器从所述计算机可读存储介质读取并执行所述计算机指令,以实现如权利要求1至30任一所述的AI数据的传输方法。
PCT/CN2021/119444 2021-09-18 2021-09-18 Ai数据的传输方法、装置、设备及存储介质 WO2023039905A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2021/119444 WO2023039905A1 (zh) 2021-09-18 2021-09-18 Ai数据的传输方法、装置、设备及存储介质
CN202180100126.4A CN117597969A (zh) 2021-09-18 2021-09-18 Ai数据的传输方法、装置、设备及存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/119444 WO2023039905A1 (zh) 2021-09-18 2021-09-18 Ai数据的传输方法、装置、设备及存储介质

Publications (1)

Publication Number Publication Date
WO2023039905A1 true WO2023039905A1 (zh) 2023-03-23

Family

ID=85602359

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/119444 WO2023039905A1 (zh) 2021-09-18 2021-09-18 Ai数据的传输方法、装置、设备及存储介质

Country Status (2)

Country Link
CN (1) CN117597969A (zh)
WO (1) WO2023039905A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2024093151A1 (en) * 2023-04-14 2024-05-10 Lenovo (Beijing) Limited Terminal device, network device, and method for ai model transfer

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110139289A (zh) * 2018-02-09 2019-08-16 中兴通讯股份有限公司 一种调度方法及调度系统
CN111082960A (zh) * 2019-04-15 2020-04-28 中兴通讯股份有限公司 数据的处理方法及装置
US20200364571A1 (en) * 2018-02-06 2020-11-19 Huawei Technologies Co., Ltd. Machine learning-based data processing method and related device
WO2021163895A1 (zh) * 2020-02-18 2021-08-26 Oppo广东移动通信有限公司 网络模型的管理方法及建立或修改会话的方法、装置
CN113365287A (zh) * 2020-03-06 2021-09-07 华为技术有限公司 通信方法及装置
WO2022021421A1 (zh) * 2020-07-31 2022-02-03 Oppo广东移动通信有限公司 模型管理方法、系统、装置、通信设备及存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20200364571A1 (en) * 2018-02-06 2020-11-19 Huawei Technologies Co., Ltd. Machine learning-based data processing method and related device
CN110139289A (zh) * 2018-02-09 2019-08-16 中兴通讯股份有限公司 一种调度方法及调度系统
CN111082960A (zh) * 2019-04-15 2020-04-28 中兴通讯股份有限公司 数据的处理方法及装置
WO2021163895A1 (zh) * 2020-02-18 2021-08-26 Oppo广东移动通信有限公司 网络模型的管理方法及建立或修改会话的方法、装置
CN113365287A (zh) * 2020-03-06 2021-09-07 华为技术有限公司 通信方法及装置
WO2022021421A1 (zh) * 2020-07-31 2022-02-03 Oppo广东移动通信有限公司 模型管理方法、系统、装置、通信设备及存储介质

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2024093151A1 (en) * 2023-04-14 2024-05-10 Lenovo (Beijing) Limited Terminal device, network device, and method for ai model transfer

Also Published As

Publication number Publication date
CN117597969A (zh) 2024-02-23

Similar Documents

Publication Publication Date Title
US20230179490A1 (en) Artificial intelligence-based communication method and communication apparatus
US20230209390A1 (en) Intelligent Radio Access Network
WO2019206196A1 (zh) 一种模型更新方法、装置及系统
CN109479246B (zh) 功率余量报告的上报方法和装置
WO2019095922A1 (zh) 一种Pcell或PScell管理方法及装置
US20230262478A1 (en) Model configuration method and apparatus
US20220394727A1 (en) Scheduling method, scheduling algorithm training method, related system, and storage medium
WO2023039905A1 (zh) Ai数据的传输方法、装置、设备及存储介质
US20240015534A1 (en) Model processing method, communication apparatus, and system
US20230289615A1 (en) Training a machine learning model
US20240137783A1 (en) Signalling support for split ml-assistance between next generation random access networks and user equipment
WO2021146923A1 (zh) 一种无线参数调整方法和装置
WO2023097522A1 (zh) 用于无线信道处理的模型更新方法、装置、设备及介质
US20230319662A1 (en) Method and apparatus for programmable and customized intelligence for traffic steering in 5g networks using open ran architectures
EP4258730A1 (en) Method and apparatus for programmable and customized intelligence for traffic steering in 5g networks using open ran architectures
WO2023227135A1 (zh) 通信方法和通信装置
KR20190118067A (ko) 5g 네트워크 환경에서 lte 및 nr 액세스간 트래픽 경로 제어를 위한 기구 및 방법
US20230318794A1 (en) Optimizing physical cell id assignment in a wireless communication network
WO2023011423A1 (zh) 一种负载均衡策略的确定方法和装置
WO2024087119A1 (zh) 模型控制方法、装置、设备及介质
WO2024093503A1 (zh) 一种处理模型的方法和装置
WO2024099016A1 (zh) 一种通信方法及装置
CN114205059B (zh) 信息交互方法、装置及存储介质
WO2023070666A1 (en) Wireless communication method and apparatus of supporting artificial intelligence
WO2024094038A1 (zh) 切换或更新ai模型的方法以及通信装置

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: 21957182

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 202180100126.4

Country of ref document: CN

NENP Non-entry into the national phase

Ref country code: DE