Communication method, communication node, medium, and program product

Abstract

The invention discloses a communication method, a communication node, a medium and a program product. Comprising the following steps: receiving a training duration request sent by a second communication node; and performing training duration estimation according to the training duration request, determining an estimated training duration, and feeding back the estimated training duration to the second communication node. Before the longitudinal federated learning training is carried out, the longitudinal federated learning server estimates the time required by the training, and timely communicates and matches the estimated training duration obtained by estimation with the network element demand of a consumer. Therefore, the first communication node can carry out longitudinal federated learning training only under the condition that the user demand can be satisfied, and invalid consumption of training resources is reduced.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a communication method, communication node, medium, and program product. Background Technology

[0002] In the 5th Generation Mobile Communication (5G) system, the Network Data Analytics Function (NWDAF) is a 5G Core Network Function (5GC NF) located in the control plane, which can perform statistical data and machine learning related tasks in the 5G system.

[0003] Current 5G systems support multiple NWDAFs and / or Application Functions (AFs) as participants for federated learning training and / or inference, and also support multiple triggering processes for federated learning training. However, the training and / or inference time for vertical federated learning is relatively long, which may lead to training waste if consumer network elements are unwilling to wait for a long time. Summary of the Invention

[0004] This application provides a communication method, communication node, medium, and program product to solve the problem of untimely negotiation of waiting time between consumer network elements and federated learning server during federated learning. It enables timely communication of consumer network element demand time and federated learning training time during the process of triggering federated learning training, and conducts vertical federated learning training only when it can be determined that the user actually needs it, thereby reducing the ineffective consumption of training resources.

[0005] To achieve the above objectives, embodiments of this application provide a communication method applied to a first communication node, comprising:

[0006] Receive training duration request sent by the second communication node;

[0007] Based on the training duration request, the training duration is estimated, the estimated training duration is determined, and the estimated training duration is fed back to the second communication node.

[0008] To achieve the above objectives, embodiments of this application provide a communication method applied to a second communication node, comprising:

[0009] Receive a first model subscription request sent by a third communication node, and send a training duration request to the first communication node according to the first model subscription request; wherein, the first model subscription request includes the model required time;

[0010] Receive the estimated training duration sent by the first communication node;

[0011] The estimated training time is compared with the model's required time, and a vertical federated learning training instruction is sent to the first communication node based on the comparison result.

[0012] To achieve the above objectives, embodiments of this application provide a communication method applied to a first communication node, comprising:

[0013] Receive the first request time sent by the second communication node;

[0014] Perform training duration estimation to determine the estimated training duration;

[0015] The estimated training time is compared with the time required for the first requirement, and vertical federated learning training is initiated based on the comparison results.

[0016] To achieve the above objectives, embodiments of this application provide a communication method applied to a second communication node, comprising:

[0017] Receive a first model subscription request sent by a third communication node; wherein the first model subscription request includes the model requirement time;

[0018] Send the first required time to the first communication node.

[0019] To achieve the above objectives, embodiments of this application provide a communication method applied to a third communication node, comprising:

[0020] Receive a first instruction; wherein the first instruction includes a first identifier;

[0021] Send the first analysis request to the first communication node;

[0022] The first analysis request includes a first identifier.

[0023] The communication method provided in this application embodiment receives a training duration request sent by a second communication node; estimates the training duration based on the request, determines the estimated training duration, and feeds back the estimated training duration to the second communication node. By adopting the above technical solution, before conducting vertical federated learning training, the vertical federated learning server first estimates the training time required and promptly communicates and matches the estimated training duration with the needs of the consumer network element. This allows the first communication node to conduct vertical federated learning training only when it is determined that the user's needs can be met, reducing the ineffective consumption of training resources. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of a 5G core network architecture provided in the prior art;

[0025] Figure 2 A timing diagram illustrating a vertical federated learning triggering process provided in the prior art;

[0026] Figure 3 A flowchart illustrating a communication method provided in an embodiment of this application;

[0027] Figure 4 A flowchart illustrating a communication method provided in an embodiment of this application;

[0028] Figure 5 A flowchart illustrating a communication method provided in an embodiment of this application;

[0029] Figure 6 A flowchart illustrating a communication method provided in an embodiment of this application;

[0030] Figure 7 A flowchart illustrating a communication method provided in an embodiment of this application;

[0031] Figure 8 A timing example diagram of a vertical federated learning triggering process provided for an embodiment of this application;

[0032] Figure 9 A timing example diagram of a vertical federated learning triggering process provided for an embodiment of this application;

[0033] Figure 10 A timing example diagram of a vertical federated learning triggering process provided for an embodiment of this application;

[0034] Figure 11 A timing example diagram of a vertical federated learning triggering process provided for an embodiment of this application;

[0035] Figure 12 This is a schematic diagram of the structure of a communication device provided in an embodiment of this application;

[0036] Figure 13 This is a schematic diagram of the structure of a communication device provided in an embodiment of this application;

[0037] Figure 14 This is a schematic diagram of the structure of a communication device provided in an embodiment of this application;

[0038] Figure 15 This is a schematic diagram of the structure of a communication device provided in an embodiment of this application;

[0039] Figure 16 This is a schematic diagram of the structure of a communication device provided in an embodiment of this application;

[0040] Figure 17 This is a schematic diagram of the structure of a communication node provided in an embodiment of this application. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in detail below with reference to the accompanying drawings. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be arbitrarily combined with each other.

[0042] The steps illustrated in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases the steps shown or described may be performed in a different order than that presented here.

[0043] The communication method provided in this application can be used in 5G systems to trigger federated learning training for multiple NWDAFs and / or AFs that support federated learning training and / or inference. To clearly describe federated learning in 5G systems, a brief introduction is given here of the 5G core network architecture and the NWDAFs that perform federated learning as 5GC NFs based on the 5G core network architecture.

[0044] Figure 1 This is a schematic diagram of a 5G core network architecture provided in the prior art, which has the following functions:

[0045] 1) User Equipment (UE).

[0046] 2) Radio Access Network (RAN). The RAN manages radio resources, transmits user data received through the N3 interface to the UE, and transmits user data from the UE through the N3 interface. The RAN maps between Quality of Service (QoS) traffic in Dedicated Radio Bearer (DRB) and Protocol Data Unit (PDU) sessions.

[0047] 3) Access and Mobility Management Function (AMF). This function includes registration management, connection management, reachability management, and mobility management. It also performs access authentication and authorization. AMF is a Network Attached Storage (NAS) security endpoint used to forward SM NAS data between the UE and the Session Management Function (SMF).

[0048] 4) SMF. This function includes the following: session establishment, modification, and release; UE IP address allocation and management (including optional authorization functions); selection and control of User Plane (UP) functions and downlink data notification, etc. The SMF controls the User Plane Function (UPF) through the N4 interface. The SMF provides the UPF with Packet Detection Rules (PDR) to indicate how to detect user data traffic; provides Forwarding and Routing Control Association Rules (FAR); and provides QoS enforcement rules and Usage Reporting Rules (URR) to indicate how the UPF performs user data traffic forwarding, QoS processing, and usage reporting on user data traffic detected using PDR.

[0049] 5) UPF. This function includes the following: serving as an anchor point for intra / inter-radio access type mobility, packet routing and forwarding, traffic usage reporting, QoS processing for the UP, downlink packet buffering, and downlink data notification triggering. The General Packet Radio Service (GPRS) Tunneling Protocol for User Plane (GTP-U) tunnel is used for the N3 interface between the RAN and UPF. The GTP-U tunnel operates on a per-PDU session basis. For downlink traffic, the UPF binds the downlink traffic to the QoS traffic within the PDU session's GTP-U tunnel using the FAR received from the SMF. For uplink traffic, the RAN transmits user plane traffic to the QoS stream identified by the UE.

[0050] 6) Policy Control Function (PCF). The PCF provides QoS policy rules to control plane functions for enforcement. The PCF translates AF requests into PCC rules applicable to PDU sessions.

[0051] 7) Unified Data Management (UDM). The UDM performs 3GPP AKA authentication credential generation, access authorization based on subscription data, UE service NF registration management (e.g., storing AMF for UE storage services, SMF for UE PDU session storage services), and subscription management. The UDM accesses the UDR to retrieve UE subscription data and stores the UE context in the UDR. The UDM and UDR can be deployed together.

[0052] Based on the 5G core network architecture, the NWDAF is a 5GC NF located in the control plane, performing data statistics and machine learning-related tasks in the 5G system. The NWDAF can interact with different entities for various purposes:

[0053] Data is collected based on event subscriptions provided by AMF, SMF, UPF, PCF, UDM, NSCAF, AF (directly or through NEF) and Operations, Administration and Maintenance (OAM);

[0054] [Optional] Use the Data Collection Coordination Function (DCCF) for analysis and data collection;

[0055] Retrieve information from data repositories (e.g., retrieve UDRs related to users via UDM or retrieve PFD information via NEF (PFDF));

[0056] Collect location information data from the LCS system;

[0057] [Optional] Store and retrieve information from the Analytics Data Storage Function (ADRF);

[0058] [Optional] Analyze and collect data from the Messaging Framework Adaptor Function (MFAF);

[0059] Retrieve information about NF (e.g., retrieve NF-related information from NRF);

[0060] Provide analytics to consumers on demand.

[0061] Provides batch data related to the analysis ID.

[0062] Provides information on the accuracy of the analysis ID.

[0063] Provides information on the accuracy of machine learning (ML) models or indications of ML model accuracy degradation.

[0064] In some examples, a single instance or multiple instances of NWDAF can be deployed in a public terrestrial mobile network. NWDAF may contain the following logical functions:

[0065] Analysis Logic Function (AnLF): A logic function in NWDAF used to perform inference, derive analytical information (i.e., derive statistical data and / or predictions based on analytical consumer requests) and expose analytical services (i.e., Nnwdaf_AnalyticsSubscription or Nnwdaf_AnalyticsInfo).

[0066] Model Training Logic Function (MTLF): A logic function in NWDAF used to train ML models and expose new training services (such as providing pre-trained ML models).

[0067] An NWDAF can contain a Model Training Logical Function (MTLF) or an Analytics Logical Function (AnLF), or both.

[0068] The Data Collection Coordination and Function (DCCF) is also an NF on the 5G core network control plane. The DCCF is responsible for coordinating the collection and distribution of data requested by NF consumers. It prevents data sources from processing multiple subscriptions to the same data and prevents multiple notifications containing the same information from being sent due to incoordination of data consumer requests.

[0069] DCCF is applicable to:

[0070] NWDAF requests data from a data source (such as for computational analysis).

[0071] NF consumers are analyzed from the NWDAF data source.

[0072] An NF consumer that requests data from an ADRF data source.

[0073] ADRF that receives data from NF data sources.

[0074] To clearly describe the differences between the vertical federated learning triggering process in this embodiment and the existing vertical federated learning triggering process, a brief introduction to the existing vertical federated learning triggering process is given first. Figure 2 This is a timing diagram of a vertical federated learning triggering process provided in the prior art.

[0075] The training triggering process includes the following steps:

[0076] 1) If an NWDAF containing AnLF wants to perform inference but has no model, it will discover an NWDAF containing MTLF from the NRF and send a model subscription request to the NWDAF containing MTLF using Nnwdaf_MLModelProvision_Subscribe.

[0077] 2) If an NWDAF containing MTLF is discovered and decides to use Vertical Federated Learning (VFL), and cannot act as a VFL server itself, but wants an AF to act as a VFL server, then the NWDAF containing MTLF discovers an AF as a VFL server through NRF, and may use Naf_Training_Subscribe to send a Vertical Federated Learning training request to the VFL server AF, which includes the analysis ID and notification target address.

[0078] 3) NWDAF containing MTLF returns the model as unavailable and the reason for unavailability to AnLF.

[0079] One possible reason for unavailability is that training is in progress.

[0080] 4) AF initiates VFL training as a VFL server.

[0081] 5) The AF notification indicates that NWDAF training containing MTLF is complete.

[0082] 6) MTLF notifies AnLF: VFL model is available or VFL training is complete.

[0083] The reasoning triggering process includes the following steps:

[0084] 1) The NWDAF containing AnLF receives an Analytics Request from a consumer network element.

[0085] 2) The NWDAF containing AnLF sends an analysis request to the AF that acts as the VFL server.

[0086] 3) The VFL server initiates vertical federated learning inference.

[0087] 4) The VFL server returns the inference results (also known as the analysis results) to the NWDAF containing AnLF.

[0088] However, although current 5G systems support various triggering procedures for vertical federated learning training, including the aforementioned vertical federated learning triggering process, the long duration of vertical federated learning and the lack of negotiation on the waiting time between the vertical federated learning server and AnLF before triggering training can lead to wasted training if the consumer network element is unwilling to wait for an extended period. To address this issue, this application provides a communication method that can be implemented between a first communication node, a second communication node, and a third communication node. The first, second, and third communication nodes can be respectively... Figure 2 The participants in the vertical federated learning shown act as VFL servers, including NWDAFs containing MTLF and NWDAFs containing AnLF. The first communication node, second communication node, and third communication node are generally electronic devices with certain computing capabilities. In this embodiment, the triggering process for vertical federated learning training is mainly described by the communication method between the first and second communication nodes. In some possible implementations, the communication method can be implemented by the processor calling computer-executable instructions stored in memory.

[0089] In one exemplary implementation Figure 3 This application provides a flowchart of a communication method. This method is applicable to the longitudinal federated learning process in a 5G system, where training of communication nodes participating in the longitudinal federated learning is triggered. The method can be executed by a communication device, which can be implemented in software and / or hardware and integrated on the communication node. This method can be applied to a first communication node. In this application embodiment, the first communication node can be as follows: Figure 2 The AF that acts as the VFL server among the participants in the vertical federated learning shown is not limited in this embodiment.

[0090] like Figure 3 As shown, the communication method provided in this application embodiment specifically includes the following steps:

[0091] S101, Receive the training duration request sent by the second communication node.

[0092] In this embodiment, the training duration request can be specifically understood as a request instructing the first communication node to estimate the duration required for the vertical federated learning training it will perform.

[0093] In this embodiment, the second communication node may be as follows: Figure 2 The NWDAF containing MTLF is shown as an intermediate node among the participants in the vertical federated learning.

[0094] In a specific example, before triggering vertical federated learning, the second communication node will first send a training duration request to the first communication node to trigger the first communication node to estimate the training duration of the vertical federated learning to be triggered, so as to obtain the possible time for training the vertical federated learning before actually triggering the vertical federated learning.

[0095] S102. Estimate the training duration based on the training duration request, determine the estimated training duration, and feed back the estimated training duration to the second communication node.

[0096] In this embodiment, the estimated training duration can be understood as the estimated time length obtained by the first communication node for the duration that the longitudinal federated learning training to be triggered will consume, or the estimated time to complete the longitudinal federated learning training.

[0097] In a specific example, the first communication node, in response to the received training duration request, determines the type of model to be trained using longitudinal federated learning, estimates the time required to complete the training, and feeds back the estimated training duration to the second communication node.

[0098] The communication method provided in this application embodiment receives a training duration request sent by a second communication node; estimates the training duration based on the request, determines the estimated training duration, and feeds back the estimated training duration to the second communication node. By adopting the above technical solution, before conducting vertical federated learning training, the vertical federated learning server first estimates the training time required and promptly communicates and matches the estimated training duration with the needs of the consumer network element. This allows the first communication node to conduct vertical federated learning training only when it is determined that the user's needs can be met, reducing the ineffective consumption of training resources.

[0099] In one embodiment, estimating the training duration includes at least one of the following:

[0100] Training duration for vertical federated learning;

[0101] The time required to complete the longitudinal federated learning training.

[0102] In one embodiment, the training duration request includes an analysis identifier.

[0103] In this embodiment, the analysis identifier can be specifically understood as an identifier used to distinguish the purpose of a model in 5G system federated learning.

[0104] In one embodiment, estimating the training duration based on the training duration request includes:

[0105] Training time is estimated based on the analysis labels.

[0106] In a specific example, upon receiving a training duration request, the first communication node can determine the vertical federated learning model functionality required by the consumer network element based on the analysis identifier included in the training duration request, and thus determine the model type that needs to be trained using VFL. Knowing the model type that needs VFL training, the first communication node can estimate the VFL training duration for that model type and obtain the corresponding estimated training duration.

[0107] In one embodiment, the first communication node is a vertical federated training server; the second communication node is a model training logic function.

[0108] In one exemplary implementation Figure 4 This application provides a flowchart of a communication method. This method is applicable to the longitudinal federated learning process in a 5G system, where training of communication nodes participating in the longitudinal federated learning is triggered. The method can be executed by a communication device, which can be implemented in software and / or hardware and integrated on the communication node. This method can be applied to a second communication node. In this application embodiment, the second communication node can be, for example,... Figure 2 The NWDAF containing MTLF that serves as a transit node among the participants in the vertical federated learning shown is not limited in this embodiment.

[0109] like Figure 4 As shown, the communication method provided in this application embodiment specifically includes the following steps:

[0110] S201. Receive the first model subscription request sent by the third communication node, and send a training duration request to the first communication node according to the first model subscription request.

[0111] The first model subscription request includes the model requirement time.

[0112] In this embodiment, the third communication node may be as follows: Figure 2 The NWDAF, which includes AnLF, is shown as a participant in the vertical federated learning process that acts as a node interacting with consumer network elements.

[0113] In this embodiment, the first model subscription request can be specifically understood as a request from a third communication node, which, upon receiving a request from a consumer network element to perform inference but finding that it does not have a model, requests a second communication node discovered by the Network Repository Function (NRF) to subscribe to the model required by the consumer network element.

[0114] In this embodiment, the model demand time can be specifically understood as the latest time that the consumer expects to receive the model.

[0115] In a specific example, after receiving a request from a consumer network element, the third communication node sends a first model subscription request to the second communication node based on the consumer network element's requirements for the model. Since the second communication node does not contain a model that the third communication node can use in the vertical federated learning scenario of the 5G system, the second communication node can make the decision. If the decision requires using VFL for model training, the second communication node will send a training duration request to the first communication node that can initiate vertical federated learning training based on the first model subscription request, so as to instruct the first communication node to estimate the time required for vertical federated learning training.

[0116] S202, Receive the estimated training duration sent by the first communication node.

[0117] S203. Compare the estimated training time with the model's required time, and send a vertical federated learning training instruction to the first communication node based on the comparison result.

[0118] In this embodiment, the vertical federated learning training instruction can be specifically understood as information instructing the first communication node to initiate vertical federated learning training.

[0119] In a specific example, the second communication node compares the estimated training duration received by the first communication node with the model requirement time received by the third communication node to determine whether the vertical federated learning training initiated by the first communication node can be completed before the latest time that the consumer expects to receive the model. Based on the comparison result, it can determine whether the first communication node needs to perform vertical federated learning training, and if it is determined that the first communication node needs to perform vertical federated learning training, it sends a vertical federated learning training instruction to the first communication node.

[0120] In one embodiment, estimating the training duration includes at least one of the following:

[0121] Training duration for vertical federated learning;

[0122] The time required to complete the longitudinal federated learning training.

[0123] In one embodiment, sending a longitudinal federated learning training instruction to a first communication node based on the comparison result includes:

[0124] If the comparison result shows that the estimated training time is no later than the model's required time, a longitudinal federated learning training instruction is sent to the first communication node.

[0125] In a specific example, if the comparison result shows that the estimated training time is no later than the model's required time, that is, the latest time for the first communication node to complete the vertical federated learning training is no later than the model's required time, it can be considered that the time required for the first communication node to conduct the vertical federated learning training can meet the consumer's expected needs. In other words, it can be considered that the consumer can use the vertical federated learning model trained by the first communication node for inference. At this time, the second communication node can send a vertical federated learning training instruction to the first communication node to trigger the first communication node to initiate the vertical federated learning training.

[0126] In one embodiment, sending a longitudinal federated learning training instruction to a first communication node based on the comparison result includes:

[0127] If the comparison result shows that the estimated training time is later than the model's required time, the first requirement instruction is fed back to the third communication node.

[0128] Receive a second model subscription request sent by a third communication node; wherein the second model subscription request includes the adjusted model requirement time;

[0129] The estimated training time is compared with the adjusted model requirement time, and if the comparison result shows that the estimated training time is not later than the adjusted model requirement time, a vertical federated learning training instruction is sent to the first communication node.

[0130] In this embodiment, the first demand indication can be specifically understood as an indication used to inform the third communication node that the time for the first communication node to perform vertical federated learning training is insufficient to meet the consumer's demand.

[0131] In this embodiment, the second model subscription request can be specifically understood as a request from the third communication node to subscribe to the model required by the consumer network element when the third communication node decides that it still needs the model trained by the first communication node. The request may carry the adjusted latest time when the consumer expects to receive the model.

[0132] In a specific example, if the comparison result shows that the estimated training time is later than the required model time (i.e., the latest time for the first communication node to complete the vertical federated learning training is later than the required model time), it can be considered that the time required for the first communication node to conduct the vertical federated learning training cannot meet the consumer's expectations. In this case, the second communication node will send a first demand indication to the third communication node, indicating that the time required for the first communication node to conduct the vertical federated learning training is insufficient to meet the consumer's needs. However, if the third communication node decides to continue using the model trained by the first communication node, the second communication node will receive a second model subscription request from the third communication node, containing the adjusted required model time. The second communication node will compare the estimated training time and the adjusted required model time in the same way as described above. If the comparison result shows that the estimated training time is not later than the adjusted required model time, it is determined that the model trained by the first communication node can meet the consumer's expectations. At this point, the second communication node will send a vertical federated learning training indication to the first communication node to trigger the first communication node to initiate the vertical federated learning training.

[0133] In one embodiment, sending a longitudinal federated learning training instruction to a first communication node based on the comparison result includes:

[0134] If the comparison result shows that the estimated training time is later than the model's required time, the first requirement instruction is fed back to the third communication node.

[0135] If a second model subscription request is not received from a third communication node, no longitudinal federated learning training instruction will be sent to the first communication node.

[0136] In a specific example, if the comparison result shows that the estimated training time is later than the model's required time—that is, the latest time for the first communication node to complete the vertical federated learning training is later than the model's required time—it can be assumed that the time required for the first communication node to perform vertical federated learning training cannot meet the consumer's expected demand. In this case, the second communication node will provide feedback to the third communication node indicating that the time required for the first communication node to perform vertical federated learning training is insufficient to meet the consumer's demand. However, if the third communication node decides that the model trained by the first communication node is unnecessary, the third communication node will no longer provide feedback to the second communication node; that is, the second communication node will not receive the second model subscription request sent by the third communication node. In this case, the second communication node can determine that vertical federated learning training is unnecessary for the first communication node, and therefore will not send a vertical federated learning training instruction to the first communication node.

[0137] In one embodiment, the first demand indication includes at least one of the following:

[0138] The model's required timeframe cannot meet the indication;

[0139] Estimated training duration;

[0140] The model's time requirements cannot meet the indicated and estimated training duration.

[0141] In this embodiment, the indication that the model training time cannot be met can be understood as indicating that the third communication node cannot meet the indication that the second communication node it requested can provide the model training time, and thus cannot meet the latest time that the consumer network element expects to receive the model.

[0142] In one embodiment, the method includes at least one of the following:

[0143] The first model subscription request also includes an analytics identifier;

[0144] Training duration requests include analytics identifiers;

[0145] The training instructions for longitudinal federated learning include analysis identifiers.

[0146] In some examples, the first model subscription request also includes an analytics identifier, enabling the second communication node to understand the intended use of the model requested by the third communication node, thus identifying the first communication node that can be used to train that model. Simultaneously, the analytics identifier can be carried in a training duration request that instructs the first communication node to estimate the training duration, allowing the first communication node to estimate the training duration of the longitudinal federated learning model corresponding to the analytics identifier based on the analytics identifier included in the training duration request.

[0147] In some examples, the longitudinal federated learning training instruction includes an analysis identifier, enabling the first communication node receiving the longitudinal federated learning training instruction to perform longitudinal federated learning training on a model that is specifically suited to the model application corresponding to the analysis identifier.

[0148] In one embodiment, the first model subscription request includes at least one of the following:

[0149] The initial model subscription request sent by the third communication node;

[0150] The third communication node resends the model subscription request; prior to resending, the third communication node has received the first request indication sent by the other second communication nodes.

[0151] In some examples, after receiving the first demand indication sent by the second communication node, the third communication node can rediscover a new second communication node via NRF and initiate a model subscription request. That is, the first model subscription request received by the second communication node in this embodiment can be a model subscription request containing an analysis identifier and model demand time, resent by the third communication node after receiving the first demand indication sent by other communication nodes besides the second communication node in this embodiment.

[0152] In one embodiment, the first communication node is a vertical federated training server; the second communication node is a model training logic function; and the third communication node is an analysis logic function.

[0153] In one exemplary implementation Figure 5 This application provides a flowchart of a communication method. This method is applicable to the longitudinal federated learning process in a 5G system, where training of communication nodes participating in the longitudinal federated learning is triggered. The method can be executed by a communication device, which can be implemented in software and / or hardware and integrated on the communication node. This method can be applied to a first communication node. In this application embodiment, the first communication node can be as follows: Figure 2 The AF that acts as the VFL server among the participants in the vertical federated learning shown is not limited in this embodiment.

[0154] like Figure 5 As shown, the communication method provided in this application embodiment specifically includes the following steps:

[0155] S301, Receive the first demand time sent by the second communication node.

[0156] In this embodiment, the second communication node may be as follows: Figure 2 The NWDAF containing MTLF is shown as an intermediate node among the participants in the vertical federated learning.

[0157] In this embodiment, the first demand time can be specifically understood as information related to the consumer network element's expected model time, which is communicated by the second communication node to the first communication node.

[0158] S302. Estimate the training duration and determine the estimated training duration.

[0159] S303. Compare the estimated training time with the first required time, and initiate vertical federated learning training based on the comparison results.

[0160] In a specific example, after estimating the training duration and determining the estimated training duration, the first communication node can compare it with the received first demand time to determine whether the vertical federated learning training initiated by the first communication node can be completed before the latest time that the consumer expects to receive the model. Then, based on the comparison result, it can be determined whether the first communication node needs to perform vertical federated learning training. If it is determined that the first communication node needs to perform vertical federated learning training, the first communication node can directly initiate the vertical federated learning training.

[0161] In one embodiment, the first required time includes at least one of the following:

[0162] Model time requirement;

[0163] Training completion time required;

[0164] Training duration requirements.

[0165] In this embodiment, the model requirement time is included in the model subscription request from the third communication node to the second communication node, representing the latest time the consumer expects to receive the model. The training completion requirement time can be specifically understood as the time determined by the second communication node based on the model requirement time in the received model subscription request, and sent to the first communication node to inform it of the latest time the consumer expects to receive the model. The training duration requirement can be specifically understood as the acceptable time spent on model training given by the consumer network element.

[0166] In one embodiment, estimating the training duration includes at least one of the following:

[0167] Training duration for vertical federated learning;

[0168] The time required to complete the longitudinal federated learning training.

[0169] In one embodiment, before performing training duration estimation, the method further includes receiving an analysis identifier.

[0170] In some examples, when the second communication node receives the model subscription request from the third communication node, it can also obtain the analytics identifier required by the consumer from the model subscription request. In this case, the second communication node can send the analytics identifier to the first communication node so that the first communication node can receive the analytics identifier before performing training duration estimation and perform training duration estimation specifically for the longitudinal federated learning model corresponding to the analytics identifier.

[0171] In one embodiment, training duration estimation includes:

[0172] Training time is estimated based on the analysis labels.

[0173] In one embodiment, initiating longitudinal federated learning training based on the comparison results includes:

[0174] If the comparison results show that the estimated training time is no later than the first required time, then initiate vertical federated learning training.

[0175] In a specific example, if the comparison result shows that the estimated training time is no later than the first required time, that is, the latest time for the first communication node to complete the vertical federated learning training is no later than the first required time, it can be considered that the time required for the first communication node to carry out the vertical federated learning training can meet the consumer's expected needs. In other words, it can be considered that the consumer can use the vertical federated learning model trained by the first communication node for inference. At this time, the first communication node will directly initiate the vertical federated learning training.

[0176] In one embodiment, initiating longitudinal federated learning training based on the comparison results includes:

[0177] If the comparison result shows that the estimated training time is later than the first requirement time, a second requirement instruction is sent to the second communication node.

[0178] Receive the new first demand time sent by the second communication node;

[0179] The estimated training duration is compared with the new first demand time, and if the comparison result shows that the estimated training duration is no later than the new first demand time, vertical federated learning training is initiated.

[0180] In this embodiment, the second demand indication can be specifically understood as an indication used to inform the second communication node that the time for the first communication node to perform vertical federated learning training is insufficient to meet the consumer's demand.

[0181] In a specific example, if the comparison result shows that the estimated training time is later than the first required time, meaning the latest time for the first communication node to complete the vertical federated learning training is later than the model's required time, it can be considered that the time required for the first communication node to perform vertical federated learning training cannot meet the consumer's expectations. In this case, the first communication node will send a second requirement indication to the second communication node, indicating that the time for the first communication node to perform vertical federated learning training is insufficient to meet the consumer's needs. However, if the third communication node determines that the model still needs to be trained by the first communication node, it will send a new first required time to the first communication node through the second communication node. The first communication node will then compare the estimated training time with the new first required time in the same way as described above. If the comparison result shows that the estimated training time is not later than the new first required time, it will determine that the model trained by the first communication node can meet the consumer's expectations. At this point, the first communication node will directly initiate vertical federated learning training.

[0182] In one embodiment, initiating longitudinal federated learning training based on the comparison results includes:

[0183] If the comparison result shows that the estimated training time is later than the first requirement time, a second requirement instruction is sent to the second communication node.

[0184] If no new first demand time is received from the second communication node, vertical federated learning training will not be initiated.

[0185] In a specific example, if the comparison result shows that the estimated training time is later than the first required time, meaning the latest time for the first communication node to complete the vertical federated learning training is later than the model's required time, it can be considered that the time required for the first communication node to perform vertical federated learning training cannot meet the consumer's expected demand. In this case, the first communication node will send a second demand indication to the second communication node, indicating that the time required for the first communication node to perform vertical federated learning training is insufficient to meet the consumer's demand. If the third communication node decides that the model trained by the first communication node is not needed, the third communication node will no longer send a new first required time to the first communication node through the second communication node. At this point, the first communication node can determine that the process is terminated and will not initiate vertical federated learning training.

[0186] In one embodiment, the second demand indication includes at least one of the following:

[0187] The first requirement cannot be met within the specified timeframe;

[0188] Estimated training duration;

[0189] The first required time cannot meet the instructions and the estimated training duration.

[0190] In this embodiment, the indication that the first required time cannot be met can be understood as indicating that the training time of the model requested by the second communication node cannot meet the indication that the consumer network element expects to receive the model in the shortest possible time.

[0191] In one embodiment, the first communication node is a vertical federated training server; the second communication node is a model training logic function.

[0192] In one exemplary implementation Figure 6 This application provides a flowchart of a communication method. This method is applicable to the longitudinal federated learning process in a 5G system, where training of communication nodes participating in the longitudinal federated learning is triggered. The method can be executed by a communication device, which can be implemented in software and / or hardware and integrated on the communication node. This method can be applied to a second communication node. In this application embodiment, the second communication node can be, for example,... Figure 2The NWDAF containing MTLF that serves as a transit node among the participants in the vertical federated learning shown is not limited in this embodiment.

[0193] like Figure 6 As shown, the communication method provided in this application embodiment specifically includes the following steps:

[0194] S401, Receive the first model subscription request sent by the third communication node.

[0195] The first model subscription request includes the model requirement time.

[0196] S402, Send the first required time to the first communication node.

[0197] In a specific example, after receiving the first subscription request sent by the third communication node, the second communication node can determine the first demand time to represent the latest time that the consumer expects to receive the model, and send the first demand time to the first communication node.

[0198] In one embodiment, the first required time includes at least one of the following:

[0199] Model time requirement;

[0200] Training duration requirements;

[0201] Training completion time required.

[0202] In one embodiment, the method includes at least one of the following:

[0203] The first model subscription request also includes an analytics identifier;

[0204] Send the analysis identifier to the first communication node.

[0205] In one embodiment, the method further includes:

[0206] Receive the second demand indication sent by the first communication node, and send the second demand indication to the third communication node;

[0207] Receive a second model subscription request sent by a third communication node; wherein the second model subscription request includes the adjusted model requirement time;

[0208] The new first demand time is sent to the first communication node according to the second model subscription request.

[0209] In a specific example, if the first communication node determines that its estimated training time cannot meet the first required time, it will send a second required time instruction to the second communication node, which will then forward this instruction to the third communication node. If the third communication node decides that the model still needs to be trained by the first communication node, the second communication node will receive a second model subscription request from the third communication node, containing the adjusted model required time. The second communication node will then determine a new first required time based on the second model subscription request, using the same method as described above, and send this new first required time to the first communication node.

[0210] In one embodiment, the second demand indication includes at least one of the following:

[0211] The first requirement cannot be met within the specified timeframe;

[0212] Estimated training duration;

[0213] The first requirement time cannot meet the instructions and estimated training duration.

[0214] In one embodiment, estimating the training duration includes at least one of the following:

[0215] Training duration for vertical federated learning;

[0216] The time required to complete the longitudinal federated learning training.

[0217] In one embodiment, the first model subscription request includes at least one of the following:

[0218] The model subscription request sent initially by the third communication node;

[0219] The third communication node retransmits the model subscription request; wherein, prior to the retransmission, the third communication node has received a second demand indication sent by another second communication node.

[0220] In one embodiment, the first communication node is a vertical federated training server; the second communication node is a model training logic function; and the third communication node is an analysis logic function.

[0221] In one exemplary implementation Figure 7 This application provides a flowchart of a communication method. This method is applicable to the longitudinal federated learning process in a 5G system, where longitudinal federated learning inference is triggered on communication nodes participating in the learning. The method can be executed by a communication device, which can be implemented in software and / or hardware and integrated on the communication node. This method can be applied to a third communication node. In this application embodiment, the third communication node can be, for example,... Figure 2The NWDAF containing AnLF that acts as a node interacting with consumer network elements in the vertical federated learning participants shown is not limited in this embodiment.

[0222] like Figure 7 As shown, the communication method provided in this application embodiment specifically includes the following steps:

[0223] S501, Receive first instruction.

[0224] The first instruction includes a first identifier.

[0225] In this embodiment, the first indication can be specifically understood as an indication that the third communication node has completed training of the VFL model it requested.

[0226] In this embodiment, the first identifier can be specifically understood as an identifier used to indicate the identity of a VFL model that has completed training.

[0227] In a specific example, after the third communication node requests the model training from the second communication node, the first communication node will receive a first instruction containing a first identifier of the trained VFL model.

[0228] S502, Send the first analysis request to the first communication node.

[0229] The first analysis request includes a first identifier.

[0230] In this embodiment, the first analysis request can be specifically understood as a request to the first communication node to perform vertical federated learning inference.

[0231] In a specific example, when the third communication node determines that analysis and reasoning are required, it will select a model based on the content that needs to be analyzed and reasoned, and include a first identifier associated with the model identity in the first analysis request and send it to the first communication node, so that the first communication node can initiate federated learning reasoning based on the first analysis request.

[0232] In one embodiment, the first instruction includes at least one of the following:

[0233] Training completion instructions;

[0234] The model can be indicated.

[0235] In this embodiment, the training completion indication can be specifically understood as an indication that the VFL model training is complete.

[0236] In this embodiment, the model availability indicator can be specifically understood as an indicator that the trained VFL model is ready for use.

[0237] In some examples, the training completion indication can be an indication sent directly from the first communication node to the third communication node after the VFL model has been trained.

[0238] In some examples, the model availability indicator can be a second communication node notifying a third communication node that the requested model is available after the first communication node has completed training its VFL model.

[0239] In one embodiment, receiving a first instruction includes at least one of the following:

[0240] Receive the first instruction sent by the first communication node;

[0241] Receive the first instruction sent by the second communication node.

[0242] In one embodiment, before sending the first analysis request to the first communication node, the method further includes:

[0243] Received the second analysis request.

[0244] In this embodiment, the second analysis request can be specifically understood as a request information sent by the consumer network element to the third communication node, containing the analysis needs of the consumer network element.

[0245] In one embodiment, sending a first analysis request to a first communication node includes:

[0246] The first analysis request is sent to the first communication node according to the second analysis request.

[0247] In some specific examples, after receiving the second analysis request from the consumer network element, the third communication node can extract the specific demand information of the consumer network element, construct a first analysis request based on the specific demand information, and send the first analysis request to the corresponding first communication node.

[0248] In one embodiment, the second analysis request includes an analysis identifier.

[0249] In one embodiment, sending a first analysis request to a first communication node according to a second analysis request includes:

[0250] The first analysis request is sent to the first communication node based on the analysis identifier.

[0251] In a specific example, since the second analysis request received by the third communication node contains an analysis identifier, and the analysis identifier can be used to characterize the model functions required by the consumer network element, the third communication node can determine the first communication node that needs to request inference based on the analysis identifier in the second analysis request, and send the first analysis request to the determined first communication node.

[0252] In one embodiment, the first identifier includes at least one of the following:

[0253] Analytics ID;

[0254] Vertical Federated Learning Identifier (VFL ID);

[0255] Vertical federated learning correlation ID (VFL correlation ID);

[0256] Vertical federated learning model ID (VFL model ID);

[0257] Machine Learning Model ID (ML ID).

[0258] In one embodiment, when the first identifier includes an analysis identifier, the third communication node stores the mapping relationship between the analysis identifier and the vertical federated learning relationship identifier.

[0259] In one embodiment, after sending the first analysis request to the first communication node, the method further includes:

[0260] Receive the reasoning or analysis results fed back by the first communication node.

[0261] In one embodiment, the first communication node is a vertical federated training server; the second communication node is a model training logic function; and the third communication node is an analysis logic function.

[0262] The communication method of this application is illustrated below through some exemplary schemes. In the following schemes, AF (VFL server) refers to the first communication node, NWDAF containing MTLF refers to the second communication node, and NWDAF containing AnLF refers to the third communication node.

[0263] Solution 1: A training and inference triggering example is given, in which the MTLF is used to compare the estimated training time and the model required time during the longitudinal federated learning process, and the comparison result shows that the estimated training time is no later than the model required time. Figure 8 A timing example diagram of a vertical federated learning triggering process provided for embodiments of this application is shown below. Figure 8 As shown, the specific steps may include the following:

[0264] The following are the training trigger steps:

[0265] 1) If an NWDAF containing AnLF wants to perform inference but has no model, it will discover an NWDAF containing MTLF from the NRF and send a model subscription request to the NWDAF containing MTLF using Nnwdaf_MLModelProvision_Subscribe. AnLF may carry the following two parameters in the Nnwdaf_MLModelProvision_Subscribe service:

[0266] Analytics ID: Used to distinguish the purpose of the model;

[0267] Model Requirement Time: Indicates the latest time that consumers expect to receive the model.

[0268] 2) If the MTLF decision requires model training using VFL and received the model requirement time parameter sent by AnLF in step 1), then in the request sent to AF, it queries the training duration of VFL (or queries the time to complete VFL training).

[0269] In some examples, the Analytics ID parameter may be included in the request sent to AF.

[0270] 3) AF estimates the VFL training duration or the estimated time to complete VFL training based on the received Analytics ID, and returns the estimated VFL training duration or the estimated time to complete VFL training as the VFL server.

[0271] 4) MTLF compares the "estimated VFL training duration or estimated VFL training completion time" with the "model required time". If the VFL training completion time is no later than the model required time, it sends a VFL training instruction to AF, which includes the Analytics ID.

[0272] 5) MTLF returns an indication to AnLF that model training is in progress.

[0273] 6) AF initiates VFL training as a VFL server.

[0274] 7) The AF notifies the MTLF training to be complete, and the notification includes the VFL correlation ID, analysis ID, or other ID that can identify the VFL process, VFL model, or purpose of the VFL model.

[0275] 8) MTLF notifies AnLF: VFL model is available or VFL training is complete; this notification contains VFL correlationID or analysis ID or other ID that can identify the VFL process or VFL model, as well as the address information of the VFL server.

[0276] In some examples, the address information of the VFL server can be the AF ID.

[0277] Understandably, if the notification returned by the AF contains an analysis ID, the AF needs to internally store the mapping relationship between the analysis ID and the VFLcorrelation ID for subsequent VFL inference.

[0278] The following are the steps to trigger the reasoning:

[0279] 1) AnLF receives an Analytics Request from a consumer network element, which contains an Analytics ID.

[0280] 2) AnLF sends an analysis request to the AF, which acts as the VFL server, containing a VFL correlation ID, analysis ID, or other ID that can identify the VFL process or VFL model.

[0281] 3) The VFL server initiates vertical federated learning inference.

[0282] 4) The VFL server returns the inference or analysis results to AnLF.

[0283] Solution 2: A training and inference triggering example is given, in which the estimated training time and the model required time are compared by MTLF during the longitudinal federated learning process, and the comparison result shows that the estimated training time is later than the model required time. Figure 9 A timing example diagram of a vertical federated learning triggering process provided for embodiments of this application is shown below. Figure 9 As shown, the specific steps may include the following:

[0284] The following are the training trigger steps:

[0285] 1) If an NWDAF containing AnLF wants to perform inference but has no model, it will discover an NWDAF containing MTLF from the NRF and send a model subscription request to the NWDAF containing MTLF using Nnwdaf_MLModelProvision_Subscribe. AnLF may carry the following two parameters in the Nnwdaf_MLModelProvision_Subscribe service:

[0286] Analytics ID: Used to distinguish the purpose of the model;

[0287] Model Requirement Time: Indicates the latest time that consumers expect to receive the model.

[0288] 2) If the MTLF decision requires model training using VFL and received the model requirement time parameter sent by AnLF in step 1), then in the request sent to AF, it queries the training duration of VFL (or queries the time to complete VFL training).

[0289] In some examples, the Analytics ID parameter may be included in the request sent to AF.

[0290] 3) AF estimates the VFL training duration or the estimated time to complete VFL training based on the received Analytics ID, and returns the estimated VFL training duration or the estimated time to complete VFL training as the VFL server.

[0291] 4) MTLF compares the "estimated VFL training duration or estimated VFL training completion time" with the "model required time". If the VFL training completion time is later than the model required time, it returns an indication to AnLF that the VFL model training cannot be completed within the required time, and can also return the estimated VFL training completion time or VFL training duration at the same time.

[0292] 5a) If AnLF decides to continue needing the model, it will send a model subscription request to the NWDAF containing the MTLF using Nnwdaf_MLModelProvision_Subscribe, which includes the adjusted time for needing the model: representing the latest time the consumer expects to receive the ML model.

[0293] 5b) AnLF decides that it no longer needs the model, and discovers another MTLF through NRF and initiates a model subscription request, that is, performs the above steps 1) and subsequent operations on the other MTLF.

[0294] 6) MTLF compares the "estimated VFL training duration or estimated VFL training completion time" with the "adjusted model requirement time". If the VFL training completion time is no later than the adjusted model requirement time, it sends a VFL training instruction to AF, which includes the Analytics ID.

[0295] 7) MTLF returns to AnLF that the model is unavailable and the reason for the unavailability (e.g., VFL training is in progress).

[0296] 8) The AF, acting as the VFL server, initiates VFL training.

[0297] 9) The AF notifies the MTLF training to be complete, and the notification includes the VFL correlation ID, analysis ID, or other ID that can identify the VFL process, VFL model, or purpose of the VFL model.

[0298] 10) MTLF notifies AnLF: VFL model is available or VFL training is complete; the notification contains VFL correlationID or analysis ID or other ID that can identify the VFL process or VFL model, as well as the address information of the VFL server.

[0299] It is understood that the subsequent reasoning triggering process is the same as that of Scheme 1 above, and this application embodiment will not describe it in detail.

[0300] Solution 3: A training and inference triggering example is given, which compares the estimated training time with the model requirement time by AF during the longitudinal federated learning process. Figure 10 A timing example diagram of a vertical federated learning triggering process provided for embodiments of this application is shown below. Figure 10 As shown, the specific steps may include the following:

[0301] The following are the training trigger steps:

[0302] 1) If an NWDAF containing AnLF wants to perform inference but has no model, it will discover an NWDAF containing MTLF from the NRF and send a model subscription request to the NWDAF containing MTLF using Nnwdaf_MLModelProvision_Subscribe. AnLF may carry the following two parameters in the Nnwdaf_MLModelProvision_Subscribe service:

[0303] Analytics ID: Used to distinguish the purpose of the model;

[0304] Model Requirement Time: Indicates the latest time that consumers expect to receive the model.

[0305] 2) If the MTLF decision requires model training using VFL and received the model requirement time parameter sent by AnLF in step 1), then when initiating the VFL training request to AF, it should carry the VFL training duration requirement or the time to complete VFL training.

[0306] 3) AF estimates the VFL training duration requirement or the time to complete VFL training. If VFL training can be completed within the specified time or the VFL training duration requirement can be met, skip steps 3-6 and proceed directly to step 7b) and subsequent steps.

[0307] If VFL training cannot be completed within the specified time or the VFL training duration requirement cannot be met, return to MTLF that the VFL training duration requirement cannot be met or the VFL training cannot be completed within the specified time, and return the estimated time to complete VFL training or the duration of VFL training.

[0308] 4) MTLF returns to AnLF that it cannot meet the VFL training duration requirement or cannot complete VFL training within the specified time. It can also return the estimated time to complete VFL training or the duration of VFL training.

[0309] 5a) If AnLF decides to continue needing the model, it will send a model subscription request to the NWDAF containing the MTLF using Nnwdaf_MLModelProvision_Subscribe, which includes the adjusted time for needing the model: indicating the latest time the consumer expects to receive the ML model, and then proceed to step 6).

[0310] 5b) AnLF decides that it no longer needs the model, and discovers another MTLF through NRF and initiates a model subscription request, that is, performs the above steps 1) and subsequent operations on the other MTLF.

[0311] 6) When MTLF sends another VFL training request to AF, it carries the adjusted VFL training duration requirement or the time to complete VFL training received in step 5a), and executes step 7a).

[0312] 7a) AF estimates the VFL training duration requirement or the time to complete VFL training. If VFL training can be completed within the specified time or the adjusted VFL training duration requirement can be met, then proceed to step 7b).

[0313] 7b) AF initiates VFL training.

[0314] 8) MTLF returns to AnLF that the model is unavailable and the reason for the unavailability (e.g., VFL training is in progress).

[0315] 9) The AF notifies the MTLF training to be complete, and the notification includes the VFL correlation ID, analysis ID, or other ID that can identify the VFL process, VFL model, or purpose of the VFL model.

[0316] 10) MTLF notifies AnLF: VFL model is available or VFL training is complete; the notification contains VFL correlationID or analysis ID or other ID that can identify the VFL process or VFL model, as well as the address information of the VFL server.

[0317] It is understood that the subsequent reasoning triggering process is the same as that of Scheme 1 above, and this application embodiment will not describe it in detail.

[0318] Option 4: Provides an example of training end and inference triggering in a longitudinal federated learning process that only involves AnLF and AF. Figure 11 A timing example diagram of a vertical federated learning triggering process provided for embodiments of this application is shown below. Figure 11 As shown, the specific steps include the following:

[0319] 1) AF sends a VFL training completion indication, which includes a VFL correlation ID, analysis ID, or other ID that can identify the VFL process, VFL model, or the purpose of the VFL model.

[0320] Understandably, if the notification returned by the AF contains an analysis ID, the AF needs to internally store the mapping relationship between the analysis ID and the VFLcorrelation ID for subsequent VFL inference.

[0321] 2) AnLF receives an Analytics Request from a consumer network element, which contains an Analytics ID.

[0322] 3) AnLF sends an analysis request to the AF, which acts as the VFL server, containing a VFL correlation ID, analysis ID, or other ID that can identify the VFL process or VFL model.

[0323] 4) The VFL server initiates vertical federated learning inference.

[0324] 5) The VFL server returns the inference or analysis results to AnLF.

[0325] In one exemplary implementation Figure 12 This is a schematic diagram of a communication device provided in an embodiment of this application. The communication device is applied to a first communication node, such as... Figure 12 As shown, the device includes:

[0326] The first request receiving module 610 is configured to receive training duration requests sent by the second communication node.

[0327] The first duration estimation module 620 is configured to estimate the training duration based on the training duration request, determine the estimated training duration, and feed back the estimated training duration to the second communication node.

[0328] In one embodiment, estimating the training duration includes at least one of the following:

[0329] Training duration for vertical federated learning;

[0330] The time required to complete the longitudinal federated learning training.

[0331] In one embodiment, the training duration request includes an analysis identifier.

[0332] In one embodiment, estimating the training duration based on the training duration request includes:

[0333] Training time is estimated based on the analysis labels.

[0334] In one embodiment, the first communication node is a vertical federated training server; the second communication node is a model training logic function.

[0335] In one exemplary implementation Figure 13 This is a schematic diagram of a communication device provided in an embodiment of this application. The communication device is applied to a second communication node, such as... Figure 13 As shown, the device includes:

[0336] The first request sending module 710 is configured to receive a first model subscription request sent by a third communication node, and send a training duration request to the first communication node according to the first model subscription request. The first model subscription request includes the required model time.

[0337] The first duration receiving module 720 is configured to receive the estimated training duration sent by the first communication node.

[0338] The first instruction sending module 730 is configured to compare the estimated training time with the model's required time, and send a vertical federated learning training instruction to the first communication node based on the comparison result.

[0339] In one embodiment, estimating the training duration includes at least one of the following:

[0340] Training duration for vertical federated learning;

[0341] The time required to complete the longitudinal federated learning training.

[0342] In one embodiment, sending a longitudinal federated learning training instruction to a first communication node based on the comparison result includes:

[0343] If the comparison result shows that the estimated training time is no later than the model's required time, a longitudinal federated learning training instruction is sent to the first communication node.

[0344] In one embodiment, sending a longitudinal federated learning training instruction to a first communication node based on the comparison result includes:

[0345] If the comparison result shows that the estimated training time is later than the model's required time, the first requirement instruction is fed back to the third communication node.

[0346] Receive a second model subscription request sent by a third communication node; wherein the second model subscription request includes the adjusted model requirement time;

[0347] The estimated training time is compared with the adjusted model requirement time, and if the comparison result shows that the estimated training time is not later than the adjusted model requirement time, a vertical federated learning training instruction is sent to the first communication node.

[0348] In one embodiment, sending a longitudinal federated learning training instruction to a first communication node based on the comparison result includes:

[0349] If the comparison result shows that the estimated training time is later than the model's required time, the first requirement instruction is fed back to the third communication node.

[0350] If a second model subscription request is not received from a third communication node, no longitudinal federated learning training instruction will be sent to the first communication node.

[0351] In one embodiment, the first demand indication includes at least one of the following:

[0352] The model's required timeframe cannot meet the indication;

[0353] Estimated training duration;

[0354] The model's time requirements cannot meet the indicated and estimated training duration.

[0355] In one embodiment, the method includes at least one of the following:

[0356] The first model subscription request also includes an analytics identifier;

[0357] Training duration requests include analytics identifiers;

[0358] The training instructions for longitudinal federated learning include analysis identifiers.

[0359] In one embodiment, the first model subscription request includes at least one of the following:

[0360] The initial model subscription request sent by the third communication node;

[0361] The third communication node resends the model subscription request; prior to resending, the third communication node has received the first request indication sent by the other second communication nodes.

[0362] In one embodiment, the first communication node is a vertical federated training server; the second communication node is a model training logic function; and the third communication node is an analysis logic function.

[0363] In one exemplary implementation Figure 14 This is a schematic diagram of a communication device provided in an embodiment of this application. The communication device is applied to a first communication node, such as... Figure 14 As shown, the device includes:

[0364] The second time receiving module 810 is configured to receive the first required time sent by the second communication node.

[0365] The second duration estimation module 820 is configured to perform training duration estimation and determine the estimated training duration.

[0366] The training initiation module 830 is configured to compare the estimated training duration with the first required time, and initiate vertical federated learning training based on the comparison result.

[0367] In one embodiment, the first required time includes at least one of the following:

[0368] Model time requirement;

[0369] Training completion time required;

[0370] Training duration requirements.

[0371] In one embodiment, estimating the training duration includes at least one of the following:

[0372] Training duration for vertical federated learning;

[0373] The time required to complete the longitudinal federated learning training.

[0374] In one embodiment, before performing training duration estimation, the method further includes receiving an analysis identifier.

[0375] In one embodiment, training duration estimation includes:

[0376] Training time is estimated based on the analysis labels.

[0377] In one embodiment, initiating longitudinal federated learning training based on the comparison results includes:

[0378] If the comparison results show that the estimated training time is no later than the first required time, then initiate vertical federated learning training.

[0379] In one embodiment, initiating longitudinal federated learning training based on the comparison results includes:

[0380] If the comparison result shows that the estimated training time is later than the first requirement time, a second requirement instruction is sent to the second communication node.

[0381] Receive the new first demand time sent by the second communication node;

[0382] The estimated training duration is compared with the new first demand time, and if the comparison result shows that the estimated training duration is no later than the new first demand time, vertical federated learning training is initiated.

[0383] In one embodiment, initiating longitudinal federated learning training based on the comparison results includes:

[0384] If the comparison result shows that the estimated training time is later than the first requirement time, a second requirement instruction is sent to the second communication node.

[0385] If no new first demand time is received from the second communication node, vertical federated learning training will not be initiated.

[0386] In one embodiment, the second demand indication includes at least one of the following:

[0387] The first requirement cannot be met within the specified timeframe;

[0388] Estimated training duration;

[0389] The first required time cannot meet the instructions and the estimated training duration.

[0390] In one embodiment, the first communication node is a vertical federated training server; the second communication node is a model training logic function.

[0391] In one exemplary implementation Figure 15 This is a schematic diagram of a communication device provided in an embodiment of this application. The communication device is applied to a second communication node, such as... Figure 15 As shown, the device includes:

[0392] The second request receiving module 910 is configured to receive a first model subscription request sent by a third communication node; wherein the first model subscription request includes the model requirement time.

[0393] The second time sending module 920 is configured to send the first required time to the first communication node.

[0394] In one embodiment, the first required time includes at least one of the following:

[0395] Model time requirement;

[0396] Training duration requirements;

[0397] Training completion time required.

[0398] In one embodiment, the method includes at least one of the following:

[0399] The first model subscription request also includes an analytics identifier;

[0400] Send the analysis identifier to the first communication node.

[0401] In one embodiment, the method further includes:

[0402] Receive the second demand indication sent by the first communication node, and send the second demand indication to the third communication node;

[0403] Receive a second model subscription request sent by a third communication node; wherein the second model subscription request includes the adjusted model requirement time;

[0404] The new first demand time is sent to the first communication node according to the second model subscription request.

[0405] In one embodiment, the second demand indication includes at least one of the following:

[0406] The first requirement cannot be met within the specified timeframe;

[0407] Estimated training duration;

[0408] The first requirement time cannot meet the instructions and estimated training duration.

[0409] In one embodiment, estimating the training duration includes at least one of the following:

[0410] Training duration for vertical federated learning;

[0411] The time required to complete the longitudinal federated learning training.

[0412] In one embodiment, the first model subscription request includes at least one of the following:

[0413] The model subscription request sent initially by the third communication node;

[0414] The third communication node retransmits the model subscription request; wherein, prior to the retransmission, the third communication node has received a second demand indication sent by another second communication node.

[0415] In one embodiment, the first communication node is a vertical federated training server; the second communication node is a model training logic function; and the third communication node is an analysis logic function.

[0416] In one exemplary implementation Figure 16This is a schematic diagram of a communication device provided in an embodiment of this application. The communication device is applied to a third communication node, such as... Figure 16 As shown, the device includes:

[0417] The third instruction receiving module 1010 is configured to receive a first instruction; wherein the first instruction includes a first identifier.

[0418] The third request sending module 1020 is configured to send a first analysis request to the first communication node; wherein the first analysis request includes a first identifier.

[0419] In one embodiment, the first instruction includes at least one of the following:

[0420] Training completion instructions;

[0421] The model can be indicated.

[0422] In one embodiment, receiving a first instruction includes at least one of the following:

[0423] Receive the first instruction sent by the first communication node;

[0424] Receive the first instruction sent by the second communication node.

[0425] In one embodiment, before sending the first analysis request to the first communication node, the method further includes:

[0426] Received the second analysis request.

[0427] In one embodiment, sending a first analysis request to a first communication node includes:

[0428] The first analysis request is sent to the first communication node according to the second analysis request.

[0429] In one embodiment, the second analysis request includes an analysis identifier.

[0430] In one embodiment, sending a first analysis request to a first communication node according to a second analysis request includes:

[0431] The first analysis request is sent to the first communication node based on the analysis identifier.

[0432] In one embodiment, the first identifier includes at least one of the following:

[0433] Analytics ID;

[0434] Vertical Federated Learning Identifier (VFL ID);

[0435] Vertical federated learning correlation ID (VFL correlation ID);

[0436] Vertical federated learning model ID (VFL model ID);

[0437] Machine Learning Model ID (ML ID).

[0438] In one embodiment, when the first identifier includes an analysis identifier, the third communication node stores the mapping relationship between the analysis identifier and the vertical federated learning relationship identifier.

[0439] In one embodiment, after sending the first analysis request to the first communication node, the method further includes:

[0440] Receive the reasoning or analysis results fed back by the first communication node.

[0441] In one embodiment, the first communication node is a vertical federated training server; the second communication node is a model training logic function; and the third communication node is an analysis logic function.

[0442] This application also provides a communication node. Figure 17 This is a schematic diagram of the structure of a communication node provided in an embodiment of this application, such as... Figure 17 As shown, the communication node provided in this application embodiment includes a memory 1120, a processor 1110, and a computer program stored in the memory and executable on the processor. When the processor 1110 executes the program, it implements the above-described communication method.

[0443] The communication node may also include a memory 1120; the processor 1110 in the communication node may be one or more. Figure 17 Taking a processor 1110 as an example; memory 1120 is used to store one or more programs; the one or more programs are executed by the one or more processors 1110, so that the one or more processors 1110 implement the communication method as described in the embodiments of this application.

[0444] The communication node also includes: a communication device 1130, an input device 1140, and an output device 1150.

[0445] The processor 1110, memory 1120, communication device 1130, input device 1140, and output device 1150 in the communication node can be connected via a bus or other means. Figure 17 Taking the example of a connection between China and Israel via a bus.

[0446] Input device 1140 can be used to receive input digital or character information, and to generate key signal inputs related to user settings and function control of the communication node. Output device 1150 may include display devices such as a display screen.

[0447] The communication device 1130 may include a receiver and a transmitter. The communication device 1130 is configured to perform information transmission and reception communication under the control of the processor 1110.

[0448] The memory 1120, as a computer-readable storage medium, can be configured to store software programs, computer-executable programs, and modules, such as program instructions / modules corresponding to the communication method described in the embodiments of this application (e.g., a first request receiving module 610, a first duration estimation module 620; or a first request sending module 710, a first duration receiving module 720, a first indication sending module 730; or a second time receiving module 810, a second duration estimation module 820, a training initiation module 830; or a second request receiving module 910, a second time sending module 920; or a third indication receiving module 1010, a third request sending module 1020). The memory 1120 may include a program storage area and a data storage area, wherein the program storage area may store the operating system and at least one application program required for a function; the data storage area may store data created according to the use of the communication node, etc. In addition, the memory 1120 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some instances, memory 1120 may further include memory remotely configured relative to processor 1110, which can be connected to a communication node via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0449] This application also provides a storage medium storing a computer program, which, when executed by a processor, implements any of the communication methods described in this application.

[0450] Optionally, the communication method, applied to the first communication node, includes: receiving a training duration request sent by the second communication node; estimating the training duration based on the training duration request, determining the estimated training duration, and feeding back the estimated training duration to the second communication node.

[0451] Optionally, the communication method, applied to the second communication node, includes: receiving a first model subscription request sent by the third communication node, and sending a training duration request to the first communication node according to the first model subscription request; wherein the first model subscription request includes the model required time; receiving an estimated training duration sent by the first communication node; comparing the estimated training duration with the model required time, and sending a longitudinal federated learning training instruction to the first communication node according to the comparison result.

[0452] Optionally, the communication method, applied to the first communication node, includes: receiving a first required time sent by the second communication node; estimating the training duration to determine the estimated training duration; comparing the estimated training duration with the first required time, and initiating longitudinal federated learning training based on the comparison result.

[0453] Optionally, the communication method, applied to a second communication node, includes: receiving a first model subscription request sent by a third communication node; wherein the first model subscription request includes a model demand time; and sending the first demand time to the first communication node.

[0454] Optionally, the communication method, applied to a third communication node, includes: receiving a first instruction; wherein the first instruction includes a first identifier; and sending a first analysis request to a first communication node; wherein the first analysis request includes the first identifier.

[0455] The computer storage medium in this application embodiment can be any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. For example, a computer-readable storage medium can be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable CD-ROM, optical storage device, magnetic storage device, or any suitable combination thereof. The computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0456] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit programs for use by or in connection with an instruction execution system, apparatus, or device.

[0457] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, radio frequency (RF), etc., or any suitable combination thereof.

[0458] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0459] Optionally, embodiments of the present invention also provide a computer program product, including a computer program that, when executed by a processor, implements the communication method provided in any embodiment of the present invention.

[0460] The above description is merely an exemplary embodiment of this application and is not intended to limit the scope of protection of this application.

[0461] Those skilled in the art will understand that the term user terminal encompasses any suitable type of wireless user equipment, such as mobile phones, portable data processing devices, portable web browsers, or vehicle-mounted mobile stations.

[0462] Generally, the various embodiments of this application can be implemented in hardware or dedicated circuitry, software, logic, or any combination thereof. For example, some aspects can be implemented in hardware, while others can be implemented in firmware or software that can be executed by a controller, microprocessor, or other computing device, although this application is not limited thereto.

[0463] Embodiments of this application can be implemented by executing computer program instructions through the data processor of a mobile device, for example, in a processor entity, or through hardware, or through a combination of software and hardware. The computer program instructions can be assembly instructions, Instruction Set Architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages.

[0464] Any block diagram of logical flow in the accompanying drawings of this application may represent program steps, or may represent interconnected logic circuits, modules, and functions, or may represent a combination of program steps and logic circuits, modules, and functions. The computer program may be stored on memory. Memory may be of any type suitable to the local technical environment and may be implemented using any suitable data storage technology, such as, but not limited to, read-only memory (ROM), random access memory (RAM), optical storage devices and systems (Digital Video Disc (DVD) or Compact Disk (CD), etc.). Computer-readable media may include non-transitory storage media. Data processors may be of any type suitable to the local technical environment, such as, but not limited to, general-purpose computers, special-purpose computers, microprocessors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), and processors based on multi-core processor architectures.

[0465] A detailed description of exemplary embodiments of this application has been provided above through exemplary and non-limiting examples. However, various modifications and adjustments to the above embodiments will be apparent to those skilled in the art when considered in conjunction with the accompanying drawings and claims, without departing from the scope of this application. Therefore, the proper scope of this application will be determined by the claims.

Claims

1. A communication method, characterized in that, Applied to the first communication node, including: Receive training duration request sent by the second communication node; Based on the training duration request, the training duration is estimated, the estimated training duration is determined, and the estimated training duration is fed back to the second communication node.

2. The communication method according to claim 1, characterized in that, The estimated training duration includes at least one of the following: Training duration for vertical federated learning; The time required to complete the longitudinal federated learning training.

3. The communication method according to claim 1, characterized in that, The training duration request includes an analysis identifier.

4. The communication method according to claim 3, characterized in that, The step of estimating the training duration based on the training duration request includes: The training time is estimated based on the analysis identifier.

5. The communication method according to any one of claims 1-4, characterized in that, The first communication node is a vertical federated training server; the second communication node is the model training logic function.

6. A communication method, characterized in that, Applied to the second communication node, including: The system receives a first model subscription request sent by a third communication node and sends a training duration request to the first communication node according to the first model subscription request; wherein the first model subscription request includes the model's required time. Receive the estimated training duration sent by the first communication node; The estimated training time is compared with the model's required time, and a vertical federated learning training instruction is sent to the first communication node based on the comparison result.

7. The communication method according to claim 6, characterized in that, The estimated training duration includes at least one of the following: Training duration for vertical federated learning; The time required to complete the longitudinal federated learning training.

8. The communication method according to claim 6, characterized in that, The step of sending a longitudinal federated learning training instruction to the first communication node based on the comparison result includes: If the comparison result indicates that the estimated training time is not later than the model's required time, a longitudinal federated learning training instruction is sent to the first communication node.

9. The communication method according to claim 6, characterized in that, The step of sending a longitudinal federated learning training instruction to the first communication node based on the comparison result includes: If the comparison result indicates that the estimated training time is later than the model's required time, a first requirement indication is fed back to the third communication node. Receive a second model subscription request sent by the third communication node; wherein the second model subscription request includes the adjusted model demand time; The estimated training time is compared with the adjusted model requirement time, and if the comparison result shows that the estimated training time is not later than the adjusted model requirement time, a vertical federated learning training instruction is sent to the first communication node.

10. The communication method according to claim 6, characterized in that, The step of sending a longitudinal federated learning training instruction to the first communication node based on the comparison result includes: If the comparison result indicates that the estimated training time is later than the model's required time, a first requirement indication is fed back to the third communication node. If a second model subscription request is not received from the third communication node, no longitudinal federated learning training instruction will be sent to the first communication node.

11. The communication method according to claim 9 or 10, characterized in that, The first demand indication includes at least one of the following: The model's required timeframe cannot meet the indication; The estimated training duration; The model's required time cannot meet the indication and the estimated training duration.

12. The communication method according to claim 6, characterized in that, Includes at least one of the following: The first model subscription request also includes an analysis identifier; The training duration request includes an analysis identifier; The longitudinal federated learning training instructions include analysis identifiers.

13. The communication method according to any one of claims 6-10 and 12, characterized in that, The first model subscription request includes at least one of the following: The model subscription request sent initially by the third communication node; The third communication node retransmits the model subscription request; wherein, prior to the retransmission, the third communication node has received a first demand indication sent by another second communication node.

14. The communication method according to any one of claims 6-10 and 12, characterized in that, The first communication node is a vertical federated training server; the second communication node is the model training logic function; and the third communication node is the analysis logic function.

15. A communication method, characterized in that, Applied to the first communication node, including: Receive the first request time sent by the second communication node; Perform training duration estimation to determine the estimated training duration; The estimated training time is compared with the first required time, and vertical federated learning training is initiated based on the comparison result.

16. The communication method according to claim 15, characterized in that, The first required time includes at least one of the following: Model time requirement; Training completion time required; Training duration requirements.

17. The communication method according to claim 15, characterized in that, The estimated training duration includes at least one of the following: Training duration for vertical federated learning; The time required to complete the longitudinal federated learning training.

18. The communication method according to claim 15, characterized in that, Prior to the training duration estimation, the method also includes receiving the analysis identifier.

19. The communication method according to claim 18, characterized in that, The training time estimation includes: The training time is estimated based on the analysis identifier.

20. The communication method according to claim 15, characterized in that, The step of initiating vertical federated learning training based on the comparison results includes: If the comparison result shows that the estimated training time is not later than the first required time, then initiate longitudinal federated learning training.

21. The communication method according to claim 15, characterized in that, The step of initiating vertical federated learning training based on the comparison results includes: If the comparison result indicates that the estimated training time is later than the first required time, a second required instruction is sent to the second communication node. Receive the new first demand time sent by the second communication node; The estimated training duration is compared with the new first requirement time, and if the comparison result shows that the estimated training duration is not later than the new first requirement time, vertical federated learning training is initiated.

22. The communication method according to claim 15, characterized in that, The step of initiating vertical federated learning training based on the comparison results includes: If the comparison result indicates that the estimated training time is later than the first required time, a second required instruction is sent to the second communication node. If no new first demand time is received from the second communication node, longitudinal federated learning training will not be initiated.

23. The communication method according to claim 21 or 22, characterized in that, The second demand indication includes at least one of the following: The first requirement cannot be met within the specified timeframe; The estimated training duration; The first required time cannot meet the instructions and the estimated training duration.

24. The communication method according to any one of claims 15-22, characterized in that, The first communication node is a vertical federated training server; the second communication node is the model training logic function.

25. A communication method, characterized in that, Applied to the second communication node, including: Receive a first model subscription request sent by a third communication node; wherein the first model subscription request includes the model requirement time; Send the first required time to the first communication node.

26. The communication method according to claim 25, characterized in that, The first required time includes at least one of the following: The model requires time; Training duration requirements; Training completion time required.

27. The communication method according to claim 25, characterized in that, Includes at least one of the following: The first model subscription request also includes an analysis identifier; Send the analysis identifier to the first communication node.

28. The communication method according to claim 25, characterized in that, Also includes: Receive the second demand indication sent by the first communication node, and send the second demand indication to the third communication node; Receive a second model subscription request sent by the third communication node; wherein the second model subscription request includes the adjusted model demand time; The new first demand time is sent to the first communication node according to the second model subscription request.

29. The communication method according to claim 28, characterized in that, The second demand indication includes at least one of the following: The first requirement cannot be met within the specified timeframe; Estimated training duration; The first requirement time cannot meet the instructions and estimated training duration.

30. The communication method according to claim 29, characterized in that, The estimated training duration includes at least one of the following: Training duration for vertical federated learning; The time required to complete the longitudinal federated learning training.

31. The communication method according to any one of claims 25-30, characterized in that, The first model subscription request includes at least one of the following: The model subscription request sent initially by the third communication node; The third communication node retransmits the model subscription request; wherein, prior to the retransmission, the third communication node has received a second demand indication sent by another second communication node.

32. The communication method according to any one of claims 25-30, characterized in that, The first communication node is a vertical federated training server; the second communication node is the model training logic function; and the third communication node is the analysis logic function.

33. A communication method, characterized in that, Applied to third communication nodes, including: Receive a first instruction; wherein the first instruction includes a first identifier; Send the first analysis request to the first communication node; The first analysis request includes the first identifier.

34. The communication method according to claim 33, characterized in that, The first instruction includes at least one of the following: Training completion instructions; The model can be indicated.

35. The communication method according to claim 33, characterized in that, The receipt of the first instruction includes at least one of the following: Receive the first instruction sent by the first communication node; Receive the first instruction sent by the second communication node.

36. The communication method according to claim 33, characterized in that, Before sending the first analysis request to the first communication node, the method further includes: Received the second analysis request.

37. The communication method according to claim 36, characterized in that, Sending the first analysis request to the first communication node includes: The first analysis request is sent to the first communication node according to the second analysis request.

38. The communication method according to claim 37, characterized in that, The second analysis request includes an analysis identifier.

39. The communication method according to claim 38, characterized in that, Sending the first analysis request to the first communication node according to the second analysis request includes: A first analysis request is sent to the first communication node based on the analysis identifier.

40. The communication method according to any one of claims 33-39, characterized in that, The first identifier includes at least one of the following: Analysis identifiers; Vertical federated learning identifiers; Vertical federated learning relationship identifier; Vertical federated learning model identifier; Machine learning model identifier.

41. The communication method according to claim 40, characterized in that, If the first identifier includes the analysis identifier, the third communication node stores the mapping relationship between the analysis identifier and the vertical federated learning relationship identifier.

42. The communication method according to any one of claims 33-39, characterized in that, The first instruction further includes: the address information of the first communication node.

43. The communication method according to any one of claims 33-39, characterized in that, After sending the first analysis request to the first communication node, the method further includes: Receive the reasoning or analysis results fed back by the first communication node.

44. The communication method according to any one of claims 33-39, characterized in that, The first communication node is a vertical federated training server; the second communication node is the model training logic function; and the third communication node is the analysis logic function.

45. A communication node, characterized in that, include: The program includes a memory, a processor, a program stored in the memory and executable on the processor, and a data bus for implementing communication between the processor and the memory, wherein the program, when executed by the processor, implements the steps of the communication method as described in any one of claims 1-44.

46. ​​A storage medium for computer-readable storage, characterized in that, The storage medium stores one or more programs, which can be executed by one or more processors to implement the steps of the communication method as described in any one of claims 1-44.

47. A computer program product comprising a computer program that, when executed by a processor, implements the steps of the communication method as described in any one of claims 1-44.

Images