Communication method and system, first device and second device

By performing signaling interaction between devices in distributed model applications, the problem of difficult distributed model management and coordination is solved, the life cycle management of distributed model is realized, and the model training and update efficiency is improved.

CN119095177APending Publication Date: 2024-12-06CHINA TELECOM CORP LTD BEIJING RESEARCH INSTITUTE +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202310656904.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-06-05
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The lack of signaling processes suitable for distributed models in the prior art leads to problems of management and coordination difficulties in distributed model applications.

Method used

The first configuration information is sent to the second device through the first device, for life cycle management of the distributed model, and receive the second configuration information sent by the second device, determine and transmit the relevant management information and configuration information of the distributed model, Implement life cycle management decisions for distributed models.

Benefits of technology

Through signaling interaction between devices, relevant management information and configuration information of the distributed model are obtained, which facilitates subsequent life cycle management of the distributed model, solves the coordination and management problems in distributed model applications, and improves model training and update efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119095177A_ABST
    Figure CN119095177A_ABST
Patent Text Reader

Abstract

The invention relates to a communication method and system, first equipment and second equipment, and relates to the technical field of communication. The communication method is executed by a first device and comprises the steps that first configuration information is sent to a second device, and the first configuration information is used for life cycle management of a distributed model; second configuration information sent by the second device is received, and the second configuration information comprises configuration information of the second device about the distributed model.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the field of communication technology, and in particular to a communication method, system, first device, and second device. Background Art

[0002] 3GPP (3rd Generation Partnership Project) Rel-18 (version 18) studies key technologies for wireless AI (Artificial Intelligence) air interface.

[0003] As data security and privacy receive more and more attention, the security and privacy issues brought about by data centralization in existing AI / ML (Machine Learning) solutions can be solved through distributed frameworks such as federated learning. Summary of the invention

[0004] Currently, no relevant signaling process has been established in the standard regarding the application of the distributed model.

[0005] A technical problem to be solved by the present disclosure is to propose a communication solution suitable for applications of distributed models.

[0006] According to some embodiments of the present disclosure, a communication method is provided, which is executed by a first device and includes: sending first configuration information to a second device, wherein the first configuration information is used for lifecycle management of a distributed model; and receiving second configuration information sent by the second device, wherein the second configuration information includes configuration information of the second device regarding the distributed model.

[0007] In some embodiments, the method further includes: determining decision information for lifecycle management of the distributed model based on the second configuration information; and sending the decision information for lifecycle management to the second device.

[0008] In some embodiments, sending first configuration information to a second device includes: sending a distributed model training request to the second device based on the network coverage environment and / or historical information, wherein the training request includes: the first configuration information; receiving second configuration information sent by the second device includes: receiving a distributed model training response sent by the second device, wherein the training response includes: the second configuration information.

[0009] In some embodiments, the first configuration information includes at least one of the following: capability information of the first device, used to indicate the distributed model support capability and / or the type of supported distributed model; configuration information of the distributed model; relevant identification of the distributed model; information of the central node of the distributed model; and positioning configuration information.

[0010] In some embodiments, the positioning configuration information includes at least one of the following: support information for a downlink positioning reference signal, supported positioning types, recommended positioning types, and relevant information of a positioning data set.

[0011] In some embodiments, the relevant information of the positioning data set includes at least one of the following: data set identification, training data set configuration information, test data set configuration information, verification data set configuration information, measurement data type, measurement data dimension, and quantization type.

[0012] In some embodiments, when the distributed model includes a federated learning model, the configuration information of the distributed model includes at least one of the following: the type of the federated learning model, the aggregation configuration information of the federated learning model, the sending resource configuration information of the federated learning model, and the security encryption method adopted.

[0013] In some embodiments, the configuration information of the distributed model includes at least one of the following: structure, parameters, hyperparameters, loss function, objectives, and initial values ​​of the distributed model.

[0014] In some embodiments, the relevant identifier of the distributed model includes at least one of the following: an identifier of the distributed model, a function identifier of the distributed model, a training identifier of the distributed model, and a configuration identifier of the distributed model.

[0015] In some embodiments, the information of the central node of the distributed model includes at least one of the following: the type of the central node, and the identifier of the central node.

[0016] In some embodiments, the configuration information of the second device regarding the distributed model includes at least one of the following: capability information, positioning-related information, and service requirement information of the second device.

[0017] In some embodiments, the capability information of the second device includes at least one of the following: support information for distributed models, supported distributed model sizes, supported distributed model storage space, supported distributed model complexity levels, and supported distributed model categories; and / or positioning-related information includes at least one of the following: support information for downlink positioning reference signals, supported positioning types, support information for vertical dimension positioning, and support information for motion sensing functions; and / or business requirement information includes: at least one of positioning service quality QoS requirement information and positioning dimension requirement information.

[0018] In some embodiments, the decision of lifecycle management includes: at least one item of the first configuration information.

[0019] In some embodiments, the lifecycle management of the distributed model includes at least one of the following: configuration, data collection, data processing, training, updating, monitoring, and reasoning process of the distributed model.

[0020] In some embodiments, the distributed model includes a federated learning model; and / or the distributed model is used for positioning.

[0021] In some embodiments, the first configuration information is transmitted through system messages, radio resource control RRC signaling, media access control MAC control information element CE, downlink control information DCI or non-access layer NAS signaling; and / or the second configuration information is transmitted through system messages, radio resource control RRC signaling, media access control MAC control information element CE, downlink control information DCI or non-access layer NAS signaling; and / or the lifecycle management decision is transmitted through system messages, radio resource control RRC signaling, media access control MAC control information element CE, downlink control information DCI or non-access layer NAS signaling.

[0022] In some embodiments, the first device is a first base station, the second device is a terminal, the first base station serves as a central node of a distributed model, and the terminal is a distributed node. When the terminal is a terminal served by the first base station, the first base station communicates with the terminal through a Uu interface; or when the terminal is a terminal served by the second base station, the first base station communicates with the second base station through an Xn interface, and the second base station communicates with the terminal through a Uu interface, wherein the second base station and the first base station are in the same tracking area or cell group.

[0023] In some embodiments, the first device is a base station, the second device is a terminal, and the communication method further includes: determining a central node and distributed nodes of a distributed model from multiple terminals according to second configuration information of the multiple terminals, wherein the central node and the distributed nodes communicate through a PC5 interface.

[0024] According to some other embodiments of the present disclosure, a communication method is provided, including: receiving first configuration information sent by a first device, wherein the first configuration information is used for lifecycle management of a distributed model; and sending second configuration information to the first device, wherein the second configuration information includes configuration information of the second device regarding the distributed model.

[0025] In some embodiments, the method further includes: receiving a lifecycle management decision of the distributed model sent by the first device, wherein the lifecycle management decision is determined by the first device according to the second configuration information.

[0026] In some embodiments, receiving first configuration information sent by the first device includes: receiving a training request for a distributed model sent by the first device, wherein the training request includes: first configuration information; sending second configuration information to the first device includes: sending a training response for the distributed model to the first device, wherein the training response includes: second configuration information.

[0027] According to some further embodiments of the present disclosure, a first device is provided, comprising: a transmitter, configured to send first configuration information to a second device, wherein the first configuration information is used for lifecycle management of a distributed model; and a receiver, configured to receive second configuration information sent by the second device, wherein the second configuration information includes configuration information of the second device regarding the distributed model.

[0028] According to some further embodiments of the present disclosure, a second device is provided, comprising: a receiver for receiving first configuration information sent by a first device, wherein the first configuration information is used for lifecycle management of a distributed model; and a transmitter for sending second configuration information to the first device, wherein the second configuration information includes configuration information of the second device regarding the distributed model.

[0029] According to some further embodiments of the present disclosure, a communication device is provided, comprising: a processor; and a memory coupled to the processor, for storing instructions, wherein when the instructions are executed by the processor, the processor executes the communication method as described in any of the aforementioned embodiments.

[0030] According to some further embodiments of the present disclosure, a non-transitory computer-readable storage medium is provided, on which a computer program is stored, wherein the program, when executed by a processor, implements the communication method of any of the aforementioned embodiments.

[0031] According to some further embodiments of the present disclosure, a communication system is provided, including: a first device and a second device of any of the aforementioned embodiments.

[0032] In the present disclosure, the first device sends first configuration information to the second device, the first configuration information is used for life cycle management of the distributed model, and the first device receives second configuration information sent by the second device, the second configuration information includes configuration information of the second device about the distributed model. Through the signaling interaction between the first device and the second device, the devices can be informed of the relevant management information and configuration information of the distributed model, which is convenient for the subsequent life cycle management of the distributed model.

[0033] Other features and advantages of the present disclosure will become apparent from the following detailed description of exemplary embodiments of the present disclosure with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0035] Figure 1 A schematic flow chart showing a communication method according to some embodiments of the present disclosure.

[0036] Figure 2 A schematic flow chart showing communication methods according to other embodiments of the present disclosure.

[0037] Figure 3 A schematic flow chart showing communication methods according to some other embodiments of the present disclosure is provided.

[0038] Figure 4 A schematic structural diagram of a first device according to some embodiments of the present disclosure is shown.

[0039] Figure 5 A schematic structural diagram of a second device according to some embodiments of the present disclosure is shown.

[0040] Figure 6 A schematic diagram showing the structure of a communication device according to some embodiments of the present disclosure.

[0041] Figure 7 A schematic diagram showing the structure of a communication device according to some other embodiments of the present disclosure.

[0042] Figure 8 A schematic diagram showing the structure of a communication system according to some embodiments of the present disclosure. DETAILED DESCRIPTION

[0043] The technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all of the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present disclosure and its application or use. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0044] The technical solutions of various embodiments of the present disclosure can be applied to various communication systems according to the access standard, such as the fifth generation (5th generation, 5G) or new wireless (New Radio, NR) system.

[0045] The technical solution provided by the present disclosure can also be applied to future communication systems, such as 5G-Advanced communication systems and sixth-generation mobile communication systems. The technical solution provided by the present disclosure can also be applied to device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, machine-to-machine (M2M) communication, machine type communication (MTC), and Internet of Things (IoT) communication systems or other communication systems.

[0046] Terminal equipment, also known as user equipment (UE), terminal, etc. Terminal equipment is a device with wireless transceiver function, which can communicate with one or more core networks (CN) via the access network equipment in the (Radio) Access Network (R) AN). It can be deployed on land, including indoors or outdoors, handheld, wearable or vehicle-mounted; it can also be deployed on the water surface, such as on a ship; it can also be deployed in the air, for example, on an airplane, balloon or satellite. Terminal equipment can be a mobile phone, a tablet computer, a computer with wireless transceiver function, a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal in industrial control, a wireless terminal in self-driving, a wireless terminal in remote medical care, a wireless terminal in smart grid, a wireless terminal in transportation safety, a wireless terminal in smart city, a wireless terminal in smart home, etc.

[0047] (Radio) access network, used to provide network access for authorized user equipment in a specific area, and can use transmission tunnels of different qualities according to the level of user equipment, business requirements, etc. For example, (R)AN can manage wireless resources, provide access services for user equipment, and then complete the forwarding of control information and / or data information between user equipment and the core network.

[0048] (R)AN equipment can provide authorized users in a specific area with the function of accessing the communication network, and can specifically include wireless network equipment in the 3rd generation partnership project (3GPP) network and access points in non-3GPP networks.

[0049] The access network equipment may include: the next-generation Node Basestation (gNB) in the 5G system, the evolved Node B (eNB) in the long term evolution (LTE), the radio network controller (RNC), the Node B (NB), the base station controller (BSC), the base transceiver station (BTS), the home base station (for example, the Home Evolved Node B, or the Home Node B, HNB), the base band unit (BBU), the transmission point (TRP), the transmission point (TP), the small base station equipment (Pico), the mobile switching center, or the network equipment in the future network, etc.

[0050] The present disclosure provides a communication method, Figure 1 Give a description.

[0051] Figure 1 Flowcharts of some embodiments of the communication method disclosed herein. Figure 1 As shown, the method of this embodiment includes: steps S102 to S104.

[0052] In step S102, the first device sends first configuration information to the second device, and correspondingly, the second device receives the first configuration information sent by the first device.

[0053] In some embodiments, the first device is a base station or a core network element, the second device is a terminal, and the core network element is a LMF (Location Management Function), which is not limited to the examples given.

[0054] The first configuration information is AI / MI related configuration information. The first configuration information is used for life cycle management of distributed models. The distributed model performs distributed machine learning by multiple distributed nodes. For example, the distributed model includes a federated learning model. In some embodiments, the life cycle management of the distributed model includes at least one of the following: configuration, data collection, data processing, training, updating, monitoring, and reasoning process of the distributed model.

[0055] The first configuration information may also be used for positioning, where the positioning includes at least one of direct positioning based on AI / ML (eg, a distributed model) or indirect positioning based on AI / ML.

[0056] In some embodiments, the first configuration information includes at least one of the following: capability information of the first device, used to indicate the distributed model support capability and / or the type of supported distributed model; configuration information of the distributed model; relevant identification of the distributed model; information of the central node of the distributed model; and positioning configuration information.

[0057] In some embodiments, the first indication information is used to indicate the distributed model support capability, that is, whether the distributed model lifecycle management is supported, for example, whether the distributed model training is supported or whether federated learning is supported, etc. The form of the first indication information is not limited.

[0058] For example, the types of distributed models supported include: types of supported federated learning. For example, different values ​​can be used to represent different types of federated learning, and the types of federated learning can include horizontal federated learning, vertical federated learning, and federated transfer learning.

[0059] In some embodiments, the configuration information of the distributed model includes at least one of the following: AI / ML model configuration information, federated learning configuration information.

[0060] In some embodiments, the AI / ML model configuration information includes at least one of the following: structure, parameters, hyperparameters, loss function (optimization function), target (optimization target), and initial value of the distributed model.

[0061] In some embodiments, the federated learning configuration information includes at least one of the following: the type of federated learning model (or the type of federated learning), the aggregation configuration information of the federated learning model, the sending resource configuration information of the federated learning model, and the security encryption method used.

[0062] For example, types of federated learning models include: horizontal federated learning, vertical federated learning, and federated transfer learning.

[0063] For example, the aggregation configuration information of the federated learning model includes: at least one of the sending cycle and channel configuration information; the sending resource configuration information of the federated learning model includes: at least one of the receiving time window and sending channel configuration information; and the security encryption method used can be represented by the identifier of the security encryption method.

[0064] In some embodiments, the related identifiers of the distributed model include at least one of the following: an identifier of the distributed model, a function identifier of the distributed model, a training identifier of the distributed model, and a configuration identifier of the distributed model. The training identifier of the distributed model can be used to indicate the type of training method adopted by the distributed model. The distributed model can adopt different configuration methods, which can be represented by the configuration identifier of the distributed model.

[0065] In some embodiments, the information of the central node of the distributed model includes at least one of the following: the type of the central node, the identification (ID) of the central node. The type of the central node can be used to indicate that the central node is a base station, a core network element (such as LMF) or a terminal.

[0066] In some embodiments, the positioning configuration information includes at least one of the following: support information for a downlink positioning reference signal (PRS), supported positioning types, recommended positioning types, and related information of a positioning data set.

[0067] For example, the support information for downlink PRS is used to indicate whether the sending of downlink PRS is supported; the supported positioning types or recommended positioning types include: direct positioning or indirect positioning; the relevant information of the positioning data set includes at least one of the following: data set identifier (ID), training data set configuration information, test data set configuration information, verification data set configuration information, measurement data type, measurement data dimension, and quantization type.

[0068] The measurement data types include, for example, CIR (Channel Impulse Response), TDOA (Time Difference of Arrival), TOA (Time of Arriving), RSRP (Reference Signal Receiving Power). The measurement data dimensions include, for example, one-dimensional and two-dimensional. The quantization types include, for example, Float32, Float16, etc.

[0069] In step S104, the first device receives the second configuration information sent by the second device, and correspondingly, the second device sends the second configuration information to the first device.

[0070] For example, the second configuration information includes: configuration information of the second device regarding the distributed model.

[0071] In some embodiments, the second configuration information includes at least one of the following: capability information of the second device, positioning-related information, and service requirement information.

[0072] In some embodiments, the capability information of the second device includes at least one of the following: support information for distributed models, size of supported distributed models, storage space of supported distributed models, complexity level of supported distributed models, and category of supported distributed models.

[0073] For example, support information for distributed models can be used to indicate whether distributed model training or federated learning is supported, etc. The types of supported distributed models can include the types of supported federated learning (e.g., horizontal federation, vertical federation, federated transfer learning).

[0074] In some embodiments, the positioning-related information includes at least one of the following: support information for a downlink positioning reference signal (PRS), supported positioning types, support information for vertical dimension positioning, and support information for a motion sensing function.

[0075] For example, the support information for downlink PRS is used to indicate whether the reception of downlink PRS is supported, the supported positioning types include, for example, direct positioning, indirect positioning, etc., the support information for vertical dimension positioning is used to indicate whether vertical dimension positioning is supported, and the support information for motion sensing function is used to indicate whether the motion sensing function is supported.

[0076] In some embodiments, the service requirement information includes at least one of: positioning QoS (Quality of Service) requirement information and positioning dimension requirement information. Positioning QoS requirement information can be used to indicate positioning accuracy requirements, and positioning dimension requirement information includes, for example, two-dimensional positioning, three-dimensional positioning, and the like.

[0077] In some embodiments, the first configuration information is transmitted through system messages, RRC (Radio Resource Control) signaling, MAC (Media Access Control) CE (Control Element), DCI (Downlink Control Information) or NAS (Non-Access Stratum) signaling; and / or, the second configuration information is transmitted through system messages, RRC signaling, MAC CE, DCI or NAS signaling.

[0078] The order of the above steps S102 and S104 can be exchanged, and only step S104 is executed without executing step S102.

[0079] In the method of the above embodiment, the first device sends first configuration information to the second device, the first configuration information is used for life cycle management of the distributed model, and the first device receives second configuration information sent by the second device, the second configuration information includes configuration information of the second device about the distributed model. Through the signaling interaction between the first device and the second device, the devices can be informed of the relevant management information and configuration information of the distributed model, which is convenient for the subsequent life cycle management of the distributed model.

[0080] Combine the following Figure 2 Other embodiments of the communication method disclosed herein are described.

[0081] Figure 2 Flowcharts of other embodiments of the communication method disclosed herein. Figure 2 As shown, the method of this embodiment includes: steps S202 to S208.

[0082] In step S202, the first device sends first configuration information to the second device.

[0083] In step S204, the second device sends second configuration information to the first device.

[0084] In step S206, the first device determines decision information for lifecycle management of the distributed model according to the second configuration information.

[0085] In some embodiments, the first device sends first configuration information to multiple second devices and receives second configuration information sent by multiple second devices, and the first device determines decision information for lifecycle management of the distributed model based on the multiple second configuration information.

[0086] In some embodiments, the decision information for lifecycle management of the distributed model includes at least one item of information in the first configuration information. The first configuration information can refer to the aforementioned embodiment and will not be described in detail here.

[0087] The first device sends the first configuration information to the second device mainly to trigger the lifecycle management process of the distributed model. The decision information of the lifecycle management is mainly used to represent the configuration information finally determined by the first device about the distributed model, which is the configuration information that needs to be actually executed by the second device. Therefore, the contents of the first configuration information and the decision information of the lifecycle management may be different or not completely consistent.

[0088] In step S208, the first device sends decision information of lifecycle management of the distributed model to the second device.

[0089] The first device may send decision information of lifecycle management of the distributed model to multiple second devices.

[0090] For example, lifecycle management decisions are transmitted via system messages, RRC signaling, MAC CE, DCI, or NAS signaling.

[0091] In the method of the above embodiment, the first device and the second device interact through signaling to obtain configuration information of the distributed model from both parties. The first device can make decisions to determine the decision information of lifecycle management and send it to the second device, thereby realizing the complete configuration process of the distributed model.

[0092] Combine the following Figure 3 Taking the training of a distributed model as an example, some further embodiments of the communication method disclosed herein are described.

[0093] Figure 3 Flowcharts of some other embodiments of the communication method disclosed herein. Figure 3 As shown, the method of this embodiment includes: steps S302 to S308.

[0094] In step S302, the first device sends a distributed model training request to the second device based on the network coverage environment and / or historical information. Correspondingly, the second device receives the distributed model training request sent by the first device, wherein the training request includes: first configuration information.

[0095] In some embodiments, the second device is in an RRC connected state or an RRC_INACTIVE state.

[0096] For example, the first device may determine multiple second devices for distributed model training based on at least one of the network coverage environment and historical information, and send a distributed model training request to each second device.

[0097] For example, the base station sends different distributed model training requests to terminals covered by different beam directions according to the coverage environment. For example, the base station sends training requests for the same distributed model to terminals within the same beam coverage range or within the coverage range of a similar preset beam according to the coverage of the beam. For another example, the base station selects terminals whose channel environment characteristics are within a preset range based on the channel environment characteristics of terminals within different beam coverage ranges, and sends training requests for the same distributed model.

[0098] For example, the base station selects the terminal as a distributed node of the same distributed model according to the training status of the historical distributed model and sends a distributed model training request. For example, the training status of the historical distributed model is evaluated, and multiple distributed nodes of the distributed model with good training effect are still used as nodes for joint training and send training requests, while multiple distributed nodes of the distributed model with poor training effect may no longer send training requests.

[0099] For example, when a terminal initiates a positioning request, the base station first selects other terminals to form a federated learning group with the terminal based on at least one of the coverage beam, geographic location, and channel environment characteristics, and scores the training of the historical distributed model of each terminal. According to the scoring results, the terminals with scores below the preset value are deleted. If the number of remaining terminals meets the training requirements, a training request is sent to each terminal. If the training requirements are not met, a new terminal is added to the federated learning group based on at least one of the coverage beam, geographic location, and channel environment characteristics, and the above process is repeated until the training requirements are met.

[0100] Because different beam directions sent by the base station cover different geographical ranges, some geographical environment features are similar and suitable for the devices in the environment to jointly perform distributed model training of small models. Therefore, a distributed model training request is sent to the devices covered by the single or several beam directions. When it is known that the coverage environment within the geographical range covered by different beams has significantly different channel environment characteristics (for example, direct path, non-direct path, multipath reflection environment), that is, the devices in this area are not suitable for joint distributed model training, or the performance of the joint distributed model training of the devices in this area in the historical information is poor, the distributed model training request is not sent. When it is known that the coverage environment within the geographical range covered by different beams has different channel environment characteristics, but the joint distributed model training of the devices in this area in the historical information can significantly improve the generalization performance of the model, the distributed model training request is not sent, and the request is to train a large model suitable for a larger geographical range.

[0101] The content of the first configuration information can refer to the aforementioned embodiment and will not be repeated here.

[0102] In step S304, the first device receives a training response of the distributed model sent by the second device, and correspondingly, the second device sends a training response of the distributed model to the first device, wherein the training response includes: second configuration information.

[0103] The content of the second configuration information can refer to the aforementioned embodiment and will not be repeated here.

[0104] In step S306, the first device makes a decision based on the training response, and sends the training decision of the distributed model to the second device, wherein the training decision includes at least one item of information in the first configuration information.

[0105] The training decisions of different second devices may be different, for example, different training data sets, different initial values, etc.

[0106] In step S308, the second device performs local training of the distributed model according to the training decision.

[0107] After each round of training on the second device, the training results can be sent to the central node for aggregation and updating of model parameters. The training process of federated learning can be referred to and will not be repeated here.

[0108] The central node of the distributed model can be a base station or a core network element, or a terminal, and the distributed node can be a terminal. The following describes different situations.

[0109] In some embodiments, the first device is a first base station, the second device is a terminal, the first base station serves as a central node of a distributed model, and the terminal is a distributed node. When the terminal is a terminal served by the first base station, the first base station communicates with the terminal through a Uu interface; or when the terminal is a terminal served by the second base station, the first base station communicates with the second base station through an Xn interface, and the second base station communicates with the terminal through a Uu interface, wherein the second base station and the first base station are in the same tracking area (TA) or cell group (Cell Group).

[0110] In the case where the central node is the first base station, the first base station communicates with multiple terminals served by the first base station through the Uu interface. For example, during the training of the distributed model, the first base station and multiple terminals served by the first base station transmit model parameters through the Uu interface. For example, the central gNB communicates with multiple UEs served by the central gNB through the Uu interface.

[0111] The first base station and the second base station are located in the same TA or cell group. For the terminal served by the second base station, the first base station and the terminal can communicate through the Xn interface between the first base station and the second base station and the Uu interface between the second base station and the terminal. For example, when the central node gNB is in a TA area or gNB group, and the distributed node UE includes the service users of the central gNB and the service users of other second gNBs in the TA area or gNB group where the central gNB is located, the above signaling interaction between the central gNB and the UE is transmitted in the following ways: through the Uu interface between the second gNB and the UE; through the Xn interface between the central gNB and the second gNB and through the Uu interface between the second gNB and the UE.

[0112] In some embodiments, the first device is a base station, the second device is a terminal, and the base station determines the central node and distributed nodes of the distributed model from multiple terminals based on the second configuration information of the multiple terminals, wherein the central node and the distributed nodes communicate through the PC5 interface.

[0113] For example, based on the terminal capabilities in the second configuration information sent by multiple terminals to the base station, the base station selects a terminal with stronger terminal capabilities (memory, computing power, power, etc.) as the central node to aggregate and report the parameters of the distributed nodes.

[0114] For example, the gNB selects a central node UE according to the second configuration information reported by the UE; the distributed node UE and the central node UE perform distributed / federated learning model training through the PC5 interface; the distributed model training response information sent by the gNB to the distributed node UE and the central node UE may be the same or different.

[0115] The method of the above embodiment is applicable to the federated learning positioning scenario, and the establishment of a federated learning group can be achieved based on the above steps.

[0116] The solution disclosed in the present invention can be applied to the scenario of positioning based on the distributed positioning framework of federated learning, which can reduce the transmission of user-location-related data between the air interface and the network element interface, protect the security and privacy of user positioning-related data, and reduce the transmission overhead related to the air interface and model training. The solution disclosed in the present invention supports the central node as LMF or base station or adopts a decentralized distributed structure with high-capability terminals as central nodes and terminals as distributed nodes, and is also applicable to various solutions for distributed deployment or centralized deployment of computing resources on the RAN side in the future.

[0117] The method for establishing a federated learning group is designed in the scheme of the present invention. Since a high-precision positioning map of the geographical location is established through federated learning, which is related to the actual network deployment environment and the mobility range of the terminal, the signaling process, transmission interface, and configuration information that must be included in the signaling required for the establishment of the federated learning group are designed according to the network coverage environment, which usually includes the geographical range covered by a single or multiple base stations as the geographical range of distributed training. This is a basic solution for realizing distributed positioning based on federated learning on the RAN side.

[0118] The disclosed solution is used in federated learning positioning scenarios, reduces air interface data interaction, protects data security and privacy, reduces signaling overhead, and increases the generalization of the model. It is suitable for 6G, 5G-Advanced, wireless AI, high frequency / millimeter wave / terahertz, indoor / industrial Internet, and heavy NLOS scenarios.

[0119] The present disclosure also provides a first device, which may be a base station or a core network element (such as LMF). Figure 4 Give a description.

[0120] Figure 4 FIG. 1 is a structural diagram of some embodiments of the first device of the present disclosure. Figure 4 As shown, the first device 40 of this embodiment includes: a transmitter 410 and a receiver 420 .

[0121] The transmitter 410 is used to send first configuration information to the second device, wherein the first configuration information is used for lifecycle management of the distributed model.

[0122] In some embodiments, the first device 40 further includes a controller 430 for determining decision information for lifecycle management of the distributed model based on the second configuration information; the transmitter 410 is further used to send the decision information for lifecycle management to the second device.

[0123] The receiver 420 is used to receive second configuration information sent by the second device, wherein the second configuration information includes configuration information of the second device regarding the distributed model.

[0124] In some embodiments, the controller 430 is also used to determine the second device based on the network coverage environment and / or historical information, and the transmitter 410 is also used to send a training request for a distributed model to the second device, wherein the training request includes: first configuration information; the receiver 420 is also used to receive a training response for the distributed model sent by the second device, wherein the training response includes: second configuration information.

[0125] The specific contents of the first configuration information and the second configuration information can be referred to the aforementioned embodiment and will not be described in detail here.

[0126] The present disclosure also provides a second device, Figure 5 Give a description.

[0127] Figure 5 FIG. 1 is a structural diagram of some embodiments of the second device of the present disclosure. Figure 5 As shown, the second device 50 of this embodiment includes: a receiver 510 and a transmitter 520 .

[0128] The receiver 510 is used to receive first configuration information sent by the first device, wherein the first configuration information is used for lifecycle management of the distributed model.

[0129] The transmitter 520 is used to send second configuration information to the first device, wherein the second configuration information includes configuration information of the second device regarding the distributed model.

[0130] In some embodiments, the receiver 510 is further used to receive a lifecycle management decision of the distributed model sent by the first device, wherein the lifecycle management decision is determined by the first device according to the second configuration information.

[0131] In some embodiments, the receiver 510 is used to receive a training request for a distributed model sent by a first device, wherein the training request includes: first configuration information; the transmitter 520 is used to send a training response for a distributed model to the first device, wherein the training response includes: second configuration information.

[0132] The specific contents of the first configuration information and the second configuration information can be referred to the aforementioned embodiment and will not be described in detail here.

[0133] The communication device (first device or second device) in the embodiments of the present disclosure can be implemented by various computing devices or computer systems. Figure 6 as well as Figure 5 Give a description.

[0134] Figure 6 FIG. 1 is a structural diagram of some embodiments of the communication device disclosed in the present invention. Figure 6 As shown, the device 60 of this embodiment includes: a memory 610 and a processor 620 coupled to the memory 610 , and the processor 620 is configured to execute the communication method in any of some embodiments of the present disclosure based on the instructions stored in the memory 610 .

[0135] The memory 610 may include, for example, a system memory, a fixed non-volatile storage medium, etc. The system memory may store, for example, an operating system, an application program, a boot loader, a database, and other programs.

[0136] Figure 7 FIG. 1 is a structural diagram of some other embodiments of the communication device disclosed in the present invention. Figure 7 As shown, the device 70 of this embodiment includes: a memory 710 and a processor 720, which are similar to the memory 610 and the processor 620 respectively. It can also include an input and output interface 730, a network interface 740, a storage interface 750, etc. These interfaces 730, 740, 750 and the memory 710 and the processor 720 can be connected, for example, through a bus 760. Among them, the input and output interface 730 provides a connection interface for input and output devices such as a display, a mouse, a keyboard, and a touch screen. The network interface 740 provides a connection interface for various networked devices, for example, it can be connected to a database server or a cloud storage server. The storage interface 750 provides a connection interface for external storage devices such as SD cards and USB flash drives.

[0137] The present disclosure also provides a communication system.

[0138] Figure 8 FIG. 1 is a structural diagram of some embodiments of the communication system disclosed in the present invention. Figure 8 As shown, the system 8 of this embodiment includes: a first device 40 and a second device 50 .

[0139] Those skilled in the art will appreciate that the embodiments of the present disclosure may be provided as methods, systems, or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present disclosure may take the form of a computer program product implemented on one or more computer-usable non-transient storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0140] The present disclosure is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present disclosure. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0141] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0142] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0143] The above description is only a preferred embodiment of the present disclosure and is not intended to limit the present disclosure. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present disclosure should be included in the protection scope of the present disclosure.

Claims

1. A communication method, performed by a first device, comprising: Sending first configuration information to the second device, wherein the first configuration information is used for lifecycle management of the distributed model; Receive second configuration information sent by the second device, wherein the second configuration information includes configuration information of the second device regarding the distributed model.

2. The communication method according to claim 1, further comprising: Determining decision information for lifecycle management of the distributed model according to the second configuration information; Send the lifecycle management decision information to the second device.

3. The communication method according to claim 1, wherein: The sending the first configuration information to the second device includes: Sending a training request for the distributed model to the second device according to the network coverage environment and / or historical information, wherein the training request includes: the first configuration information; The receiving the second configuration information sent by the second device includes: Receive a training response of the distributed model sent by the second device, wherein the training response includes: the second configuration information.

4. The communication method according to claim 1, wherein: The first configuration information includes at least one of the following: The capability information of the first device is used to indicate the capability of supporting the distributed model and / or the type of the distributed model supported; Configuration information of the distributed model; The relevant identification of the distributed model; Information of the central node of the distributed model; Positioning configuration information.

5. The communication method according to claim 4, wherein: The positioning configuration information includes at least one of the following: support information for a downlink positioning reference signal, supported positioning types, recommended positioning types, and related information of a positioning data set.

6. The communication method according to claim 5, wherein: The relevant information of the positioning data set includes at least one of the following: data set identification, training data set configuration information, test data set configuration information, verification data set configuration information, measurement data type, measurement data dimension, and quantization type.

7. The communication method according to claim 4, wherein: In the case where the distributed model includes a federated learning model, the configuration information of the distributed model includes at least one of the following: the type of the federated learning model, the aggregation configuration information of the federated learning model, the sending resource configuration information of the federated learning model, and the adopted security encryption method.

8. The communication method according to claim 4, wherein: The configuration information of the distributed model includes at least one of the following: the structure, parameters, hyperparameters, loss function, target, and initial value of the distributed model.

9. The communication method according to claim 4, wherein: The related identifiers of the distributed model include at least one of the following: an identifier of the distributed model, a function identifier of the distributed model, a training identifier of the distributed model, and a configuration identifier of the distributed model.

10. The communication method according to claim 4, wherein: The information of the central node of the distributed model includes at least one of the following: the type of the central node and the identifier of the central node.

11. The communication method according to claim 1, wherein: The configuration information of the second device about the distributed model includes at least one of the following: capability information, positioning-related information, and service requirement information of the second device.

12. The communication method according to claim 11, wherein: The capability information of the second device includes at least one of the following: support information for the distributed model, supported size of the distributed model, supported storage space of the distributed model, supported complexity level of the distributed model, supported category of the distributed model; and / or The positioning related information includes at least one of the following: support information for downlink positioning reference signals, supported positioning types, support information for vertical dimension positioning, and support information for motion sensing functions; and / or The business demand information includes: at least one of positioning service quality QoS demand information and positioning dimension demand information.

13. The communication method according to claim 1, wherein: The decision of the lifecycle management includes: at least one item of information in the first configuration information.

14. The communication method according to any one of claims 1 to 13, wherein: The lifecycle management of the distributed model includes at least one of the following: configuration, data collection, data processing, training, updating, monitoring, and reasoning process of the distributed model.

15. The communication method according to any one of claims 1 to 13, wherein: The distributed model includes a federated learning model; and / or The distributed model is used for positioning.

16. The communication method according to any one of claims 1 to 13, wherein: The first configuration information is transmitted via system message, radio resource control RRC signaling, media access control MAC control element CE, downlink control information DCI or non-access layer NAS signaling; and / or The second configuration information is transmitted via system message, radio resource control RRC signaling, media access control MAC control element CE, downlink control information DCI or non-access layer NAS signaling; and / or The lifecycle management decision is transmitted through system messages, radio resource control RRC signaling, media access control MAC control element CE, downlink control information DCI or non-access layer NAS signaling.

17. The communication method according to any one of claims 1 to 13, wherein: The first device is a first base station, the second device is a terminal, the first base station serves as a central node of the distributed model, and the terminal is a distributed node. When the terminal is a terminal served by the first base station, the first base station communicates with the terminal through a Uu interface; or In the case where the terminal is a terminal served by a second base station, the first base station communicates with the second base station via an Xn interface, and the second base station communicates with the terminal via a Uu interface, wherein the second base station and the first base station are in the same tracking area or cell group.

18. The communication method according to any one of claims 1 to 13, wherein: The first device is a base station, the second device is a terminal, and the communication method further includes: According to the second configuration information of the plurality of terminals, a central node and distributed nodes of the distributed model are determined from the plurality of terminals, wherein the central node and the distributed nodes communicate with each other through a PC5 interface.

19. A communication method, performed by a second device, comprising: Receive first configuration information sent by a first device, wherein the first configuration information is used for lifecycle management of a distributed model; Sending second configuration information to the first device, wherein the second configuration information includes configuration information of the second device regarding the distributed model.

20. The communication method according to claim 19, further comprising: Receive a lifecycle management decision of the distributed model sent by the first device, wherein the lifecycle management decision is determined by the first device according to the second configuration information.

21. The communication method according to claim 19, wherein: The receiving the first configuration information sent by the first device includes: receiving a training request for the distributed model sent by the first device, wherein the training request includes: the first configuration information; The sending second configuration information to the first device includes: Sending a training response of the distributed model to the first device, wherein the training response includes: the second configuration information.

22. A first device, comprising: A transmitter, configured to send first configuration information to a second device, wherein the first configuration information is used for lifecycle management of a distributed model; A receiver is used to receive second configuration information sent by the second device, wherein the second configuration information includes configuration information of the second device regarding the distributed model.

23. A second device, comprising: A receiver, configured to receive first configuration information sent by a first device, wherein the first configuration information is used for lifecycle management of a distributed model; A transmitter is used to send second configuration information to the first device, wherein the second configuration information includes configuration information of the second device regarding the distributed model.

24. A communication device, comprising: processor; as well as A memory coupled to the processor, for storing instructions, wherein when the instructions are executed by the processor, the processor executes the communication method according to any one of claims 1 to 21.

25. A non-transitory computer-readable storage medium having a computer program stored thereon, wherein: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 21 are implemented.

26. A communication system comprising: The first device of claim 22 and the second device of claim 23.