A federated learning group processing method, device and functional entity

Abstract

This application provides a method, apparatus, and functional entity for processing federated learning groups. The method includes: obtaining characteristic information of a federated learning group (FL); determining a second functional entity based on the characteristic information of the FL group; and adding the second functional entity to the FL group. This solution enables dynamic processing of FL groups, thereby better meeting the requirements of intelligent networks and services and improving the intelligent performance of communications and applications.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a federated learning group processing method, device and functional entity. Background Art

[0002] To enable intelligent data analysis and control of mobile communication networks, the fifth-generation mobile communication system (5G System) introduced the Network Data Analytics Function (NWDAF). Based on artificial intelligence (AI) algorithms, NWDAF interacts with other network functions (NFs) to provide network data analysis (hereinafter referred to as "analysis" or "network analysis") services for other network functions.

[0003] NWDAF can conduct federated learning with other NWDAFs to obtain an optimized machine learning ML (Machine Learning) model for AI reasoning. An existing process for federated learning between NWDAFs is as follows: one or more ML model users NWDAF (ML model consumer) request a global model from the ML model provider NWDAF (ML model provider), and then use the locally collected data to train the model, generate a local model, and send the locally updated model parameters to the ML model provider NWDAF. The ML model provider NWDAF updates the global model based on the locally updated model parameters provided by the ML model user NWDAF, and then provides the updated global model to the ML model user NWDAF. This cycle is repeated multiple times, and eventually a better model for the network analysis is obtained.

[0004] In addition to NWDAF, user equipment (UE) and other network functions can also support AI capabilities, such as:

[0005] (1) UE can have AI functional modules to realize the intelligence of different applications and communications;

[0006] (2) Other network functions (such as access and mobility management function AMF and session management function SMF) can be co-located with NWDAF to achieve intelligent control functions based on network data analysis.

[0007] However, in 5G mobile communication systems, how to establish federated learning between network functions (such as NWDAF and other NFs) or between network functions and UEs (such as NWDAF and UE) becomes a problem to be solved. Summary of the Invention

[0008] The purpose of this application is to provide a federated learning group processing method, device and functional entity to solve the above problems.

[0009] To achieve the above objectives, an embodiment of the present application provides a federated learning group processing method, which is applied to a first functional entity. The method includes:

[0010] Get the feature information of the federated learning FL group;

[0011] determining a second functional entity according to the characteristic information of the FL group;

[0012] Add the second functional entity to the FL group.

[0013] Optionally, obtaining characteristic information of a federated learning FL group includes:

[0014] receiving a first request sent by an application function entity or a network open function entity, and acquiring characteristic information of the FL group according to application information or service information related to the FL group carried in the first request;

[0015] or,

[0016] A second request sent by the policy control function entity is received, and characteristic information of the FL group is acquired according to the control policy related to the FL group carried in the second request.

[0017] Optionally, the characteristic information of the FL group includes the first information and / or identification information of a functional entity allowed to join the FL group;

[0018] The first information includes any one or more of the following:

[0019] Information about the application or service corresponding to the FL group;

[0020] Area information corresponding to the FL group;

[0021] Network slice information corresponding to the FL group;

[0022] The data network information corresponding to the FL group.

[0023] Optionally, determining the second functional entity according to the characteristic information of the FL group includes:

[0024] determining, according to the characteristic information of the FL group, a second functional entity that meets a first condition;

[0025] The first condition includes one or a combination of the following conditions:

[0026] Support the applications or services corresponding to the FL group;

[0027] Serving the data network, network slice or area corresponding to the FL group;

[0028] Access the data network or network slice corresponding to the FL group;

[0029] The position is within the region corresponding to the FL group;

[0030] The identification information is included in the identification information of the functional entity allowed to join the FL group.

[0031] Optionally, adding the second functional entity to the FL group includes:

[0032] Sending a first FL join request to the second functional entity, where the first FL join request is used to request the second functional entity to join the FL group, and the first FL join request carries information about the first functional entity and first information about the FL group;

[0033] Receive a first FL joining response sent by the second functional entity.

[0034] Optionally, the determining the second functional entity according to the characteristic information of the FL group specifically includes:

[0035] receiving a second FL join request sent by the second functional entity, where the second FL join request is used to request to join the FL group, and the second FL join request carries the first information of the FL group and identification information of the second functional entity;

[0036] determining, according to the first information, an FL group to be joined by the second functional entity;

[0037] determining whether the identification information of the second functional entity is included in the identification information of the functional entities allowed to join the FL group;

[0038] If so, it is determined that the second functional entity is the second functional entity corresponding to the characteristic information of the FL group.

[0039] Optionally, the method further includes:

[0040] Storing second information of the FL group, wherein the second information includes at least one of the following:

[0041] information of the first functional entity;

[0042] information of the second functional entity;

[0043] identification information of the FL group;

[0044] Information about the application or service corresponding to the FL group;

[0045] Network slice information corresponding to the FL group;

[0046] Data network information corresponding to the FL group;

[0047] Information about the machine learning (ML) model of the FL group.

[0048] Optionally, the information of the first functional entity includes: at least one of identification information, address information and location information of the first functional entity; and / or,

[0049] The information of the second functional entity includes: at least one of identification information, address information and location information of the second functional entity.

[0050] Optionally, the method further includes (specifically, after adding the second functional entity to the FL group, the method further includes):

[0051] determining whether the second functional entity meets a condition for leaving the FL group;

[0052] If the second functional entity meets a condition for leaving the FL group, the second functional entity is removed from the FL group.

[0053] Optionally, the second functional entity satisfies a condition for leaving the FL group, including at least one of the following:

[0054] The second functional entity no longer supports the application or service corresponding to the FL group;

[0055] The second functional entity no longer serves the data network, network slice or area corresponding to the FL group;

[0056] The second functional entity no longer accesses the data network or network slice corresponding to the FL group;

[0057] The second functional entity leaves the area corresponding to the FL group;

[0058] The identification information of the second functional entity is no longer included in the identification information of functional entities allowed to join the FL group;

[0059] receiving a first FL leave request sent by the second functional entity, where the first FL leave request is used to request leaving the FL group;

[0060] The second functional entity has abnormal behavior, and the abnormal behavior includes: the frequency of establishing a connection between the second functional entity and the first functional entity is higher than a first threshold, or the amount of data transmitted between the second functional entity and the first functional entity exceeds a second threshold.

[0061] Optionally, also include:

[0062] receiving updated feature information of the FL group sent by the third functional entity;

[0063] The determining whether the second functional entity meets the condition of leaving the FL group includes:

[0064] Determining whether the second functional entity meets a condition for leaving the FL group based on the updated characteristic information of the FL group. Specifically, determining whether the second functional entity meets the condition for leaving the FL group may include: receiving updated characteristic information of the FL group sent by a third functional entity; and determining whether the second functional entity meets the condition for leaving the FL group based on the updated characteristic information of the FL group.

[0065] Optionally, removing the second functional entity from the FL group includes:

[0066] sending a second FL leave request to the second functional entity, where the second FL leave request is used to request the second functional entity to leave the FL group;

[0067] Receive a second FL leave response sent by the second functional entity.

[0068] Optionally, removing the second functional entity from the FL group includes:

[0069] The information of the second functional entity is deleted from the second information of the FL group.

[0070] This embodiment of the present application further provides a federated learning group processing method, which is applied to a second functional entity. The method includes:

[0071] Get the feature information of the federated learning FL group;

[0072] determining, based on the characteristic information of the FL group, whether a condition for joining the FL group is met;

[0073] If the conditions for joining the FL group are met, the user joins the FL group.

[0074] Optionally, obtaining characteristic information of a federated learning FL group includes:

[0075] receiving a third request sent by the application function entity or the network open function entity, and acquiring characteristic information of the FL group according to application information or service information related to the FL group carried in the third request;

[0076] or,

[0077] The fourth request sent by the policy control function entity is received, and characteristic information of the FL group is acquired according to the control policy related to the FL group carried in the fourth request.

[0078] Optionally, the characteristic information of the FL group includes the first information and / or identification information of a functional entity allowed to join the FL group;

[0079] The first information includes any one or more of the following:

[0080] Information about the application or service corresponding to the FL group;

[0081] Area information corresponding to the FL group;

[0082] Network slice information corresponding to the FL group;

[0083] The data network information corresponding to the FL group.

[0084] Optionally, the determining, according to the characteristic information of the FL group, whether a condition for joining the FL group is met includes:

[0085] determining, according to the characteristic information of the FL group, whether the second functional entity meets a second condition;

[0086] The second condition includes one or a combination of the following conditions:

[0087] Support the applications or services corresponding to the FL group;

[0088] Serving the data network, network slice or area corresponding to the FL group;

[0089] Access the data network or network slice corresponding to the FL group;

[0090] The position is within the region corresponding to the FL group;

[0091] The identification information is included in the identification information of the functional entity allowed to join the FL group.

[0092] Optionally, before joining the FL group, the following steps are further included:

[0093] Sending a first query request to a fourth functional entity, where the first query request carries the third information of the FL group;

[0094] receiving a first query response sent by the fourth functional entity, where the first query response carries information of at least one first functional entity that satisfies the third information;

[0095] determining, from the at least one first functional entity, a first functional entity to establish the FL group with the second functional entity;

[0096] The third information of the FL group includes at least one of the following:

[0097] Information about the application or service corresponding to the FL group;

[0098] Network slice information corresponding to the FL group;

[0099] Data network information corresponding to the FL group;

[0100] Information about the machine learning (ML) model of the FL group.

[0101] Optionally, the information of the first functional entity includes: at least one of address information, location information, and service area information of the first functional entity;

[0102] The determining, from the at least one first functional entity, a first functional entity to establish the FL group with the second functional entity according to the information of the at least one first functional entity includes:

[0103] selecting, according to the information of the at least one first functional entity, a first functional entity that meets a third condition from the at least one first functional entity as the first functional entity for establishing the FL group with the second functional entity;

[0104] The third condition includes at least one of the following:

[0105] Belonging to the same local network as the second functional entity;

[0106] Serving the area where the second functional entity is located;

[0107] Among the at least one first functional entity, the distance between the second functional entity and the first functional entity is the shortest.

[0108] Optionally, joining the FL group includes:

[0109] Sending a second FL join request to the first functional entity, where the second FL join request is used to request to join the FL group, and the second FL join request carries the first information and information about the second functional entity;

[0110] Receive a second FL joining response sent by the first functional entity.

[0111] Optionally, the method further includes:

[0112] Storing second information of the FL group, wherein the second information includes at least one of the following:

[0113] Information of the first functional entity;

[0114] information of the second functional entity;

[0115] identification information of the FL group;

[0116] Information about the application or service corresponding to the FL group;

[0117] Network slice information corresponding to the FL group;

[0118] Data network information corresponding to the FL group;

[0119] Information about the machine learning (ML) model of the FL group.

[0120] Optionally, the information of the first functional entity includes: at least one of identification information, address information and location information of the first functional entity; and / or,

[0121] The information of the second functional entity includes: at least one of identification information, address information and location information of the second functional entity.

[0122] Optionally, after joining the FL group, the method further includes:

[0123] determining whether the second functional entity meets a condition for leaving the FL group;

[0124] If the second functional entity meets the conditions for leaving the FL group, the second functional entity sends a first FL leave request to the first functional entity, where the first FL leave request is used to request leaving the FL group. Specifically, if the second functional entity meets the conditions for leaving the FL group, the second functional entity leaves the FL group; wherein leaving the FL group includes sending the first FL leave request to the first functional entity, where the first FL leave request is used to request leaving the FL group.

[0125] Optionally, the second functional entity satisfies a condition for leaving the FL group, including at least one of the following:

[0126] The second functional entity no longer supports the application or service corresponding to the FL group;

[0127] The second functional entity no longer serves the data network, network slice or area corresponding to the FL group;

[0128] The second functional entity no longer accesses the data network or network slice corresponding to the FL group;

[0129] The second functional entity leaves the area corresponding to the FL group.

[0130] Optionally, the method further includes: receiving updated feature information of the FL group sent by a third functional entity;

[0131] The determining whether the second functional entity meets the condition of leaving the FL group includes:

[0132] Determining whether the second functional entity meets a condition for leaving the FL group based on the updated characteristic information of the FL group. Specifically, determining whether the second functional entity meets the condition for leaving the FL group may include: receiving updated characteristic information of the FL group sent by a third functional entity; and determining whether the second functional entity meets the condition for leaving the FL group based on the updated characteristic information of the FL group.

[0133] Optionally, when the second functional entity meets the condition of leaving the FL group, the method further includes:

[0134] Deleting the second information of the FL group stored in the second functional entity. Specifically, the leaving the FL group further includes: deleting the second information of the FL group stored in the second functional entity.

[0135] This embodiment of the present application further provides a federated learning group processing method, which is applied to a third functional entity. The method includes:

[0136] Sending a first request or a second request to the first functional entity;

[0137] The first request carries application information or service information related to the federated learning FL group, and the second request carries a control policy related to the federated learning FL group; so that the first functional entity creates or updates the FL group according to the application information, service information or control policy related to the FL group.

[0138] Optionally, the application information, service information or control policy related to the FL group includes: characteristic information of the FL group;

[0139] The characteristic information of the FL group includes the first information and / or identification information of a functional entity allowed to join the FL group;

[0140] The first information includes any one or more of the following:

[0141] Information about the application or service corresponding to the FL group;

[0142] Area information corresponding to the FL group;

[0143] Network slice information corresponding to the FL group;

[0144] The data network information corresponding to the FL group.

[0145] The embodiment of the present application further provides a functional entity, which is a first functional entity and includes a memory, a transceiver, and a processor:

[0146] A memory for storing a computer program; a transceiver for transmitting and receiving data under the control of the processor; and a processor for reading the computer program in the memory and performing the following operations:

[0147] Get the feature information of the federated learning FL group;

[0148] determining a second functional entity according to the characteristic information of the FL group;

[0149] Add the second functional entity to the FL group.

[0150] Optionally, obtaining characteristic information of a federated learning FL group includes:

[0151] receiving, by the transceiver, a first request sent by an application function entity or a network open function entity, and acquiring characteristic information of the FL group according to application information or service information related to the FL group carried in the first request;

[0152] or,

[0153] A second request sent by the policy control function entity is received through the transceiver, and characteristic information of the FL group is acquired according to the control policy related to the FL group carried in the second request.

[0154] Optionally, the characteristic information of the FL group includes the first information and / or identification information of a functional entity allowed to join the FL group;

[0155] The first information includes any one or more of the following:

[0156] Information about the application or service corresponding to the FL group;

[0157] Area information corresponding to the FL group;

[0158] Network slice information corresponding to the FL group;

[0159] The data network information corresponding to the FL group.

[0160] Optionally, determining the second functional entity according to the characteristic information of the FL group includes:

[0161] determining, according to the characteristic information of the FL group, a second functional entity that meets a first condition;

[0162] The first condition includes one or a combination of the following conditions:

[0163] Support the applications or services corresponding to the FL group;

[0164] Serving the data network, network slice or area corresponding to the FL group;

[0165] Access the data network or network slice corresponding to the FL group;

[0166] The position is within the region corresponding to the FL group;

[0167] The identification information is included in the identification information of the functional entity allowed to join the FL group.

[0168] Optionally, adding the second functional entity to the FL group includes:

[0169] sending, by the transceiver, a first FL joining request to the second functional entity, where the first FL joining request is used to request the second functional entity to join the FL group, and the first FL joining request carries information about the first functional entity and first information about the FL group;

[0170] A first FL joining response sent by the second functional entity is received through the transceiver.

[0171] Optionally, the determining the second functional entity according to the characteristic information of the FL group specifically includes:

[0172] receiving, by the transceiver, a second FL joining request sent by the second functional entity, where the second FL joining request is used to request to join a FL group, and the second FL joining request carries the first information of the FL group and identification information of the second functional entity;

[0173] determining, according to the first information, an FL group to be joined by the second functional entity;

[0174] determining whether the identification information of the second functional entity is included in the identification information of the functional entities allowed to join the FL group;

[0175] If so, it is determined that the second functional entity is the second functional entity corresponding to the characteristic information of the FL group.

[0176] Optionally, the processor is further configured to:

[0177] The memory is used to store second information of the FL group, wherein the second information includes at least one of the following:

[0178] information of the first functional entity;

[0179] information of the second functional entity;

[0180] identification information of the FL group;

[0181] Information about the application or service corresponding to the FL group;

[0182] Network slice information corresponding to the FL group;

[0183] Data network information corresponding to the FL group;

[0184] Information about the machine learning (ML) model of the FL group.

[0185] Optionally, the information of the first functional entity includes: at least one of identification information, address information and location information of the first functional entity; and / or,

[0186] The information of the second functional entity includes: at least one of identification information, address information and location information of the second functional entity.

[0187] Optionally, the processor is further configured to:

[0188] After adding the second functional entity to the FL group, determining whether the second functional entity meets a condition for leaving the FL group;

[0189] If the second functional entity meets a condition for leaving the FL group, the second functional entity is removed from the FL group.

[0190] Optionally, the second functional entity satisfies a condition for leaving the FL group, including at least one of the following:

[0191] The second functional entity no longer supports the application or service corresponding to the FL group;

[0192] The second functional entity no longer serves the data network, network slice or area corresponding to the FL group;

[0193] The second functional entity no longer accesses the data network or network slice corresponding to the FL group;

[0194] The second functional entity leaves the area corresponding to the FL group;

[0195] The identification information of the second functional entity is no longer included in the identification information of functional entities allowed to join the FL group;

[0196] receiving a first FL leave request sent by the second functional entity, where the first FL leave request is used to request leaving the FL group;

[0197] The second functional entity has abnormal behavior, and the abnormal behavior includes: the frequency of establishing a connection between the second functional entity and the first functional entity is higher than a first threshold, or the amount of data transmitted between the second functional entity and the first functional entity exceeds a second threshold.

[0198] Optionally, the determining whether the second functional entity meets a condition for leaving the FL group includes:

[0199] receiving, through the transceiver, updated feature information of the FL group sent by a third functional entity;

[0200] Determine, according to the updated characteristic information of the FL group, whether the second functional entity meets a condition for leaving the FL group.

[0201] Optionally, removing the second functional entity from the FL group includes:

[0202] sending, by the transceiver, a second FL leave request to the second functional entity, where the second FL leave request is used to request the second functional entity to leave the FL group;

[0203] A second FL leave response sent by the second functional entity is received through the transceiver.

[0204] Optionally, removing the second functional entity from the FL group includes:

[0205] The information of the second functional entity is deleted from the second information of the FL group.

[0206] The embodiment of the present application further provides a functional entity, which is a second functional entity and includes a memory, a transceiver, and a processor:

[0207] A memory for storing a computer program; a transceiver for transmitting and receiving data under the control of the processor; and a processor for reading the computer program in the memory and performing the following operations:

[0208] Get the feature information of the federated learning FL group;

[0209] determining, based on the characteristic information of the FL group, whether a condition for joining the FL group is met;

[0210] If the conditions for joining the FL group are met, the user joins the FL group.

[0211] Optionally, obtaining characteristic information of a federated learning FL group includes:

[0212] receiving, by the transceiver, a third request sent by the application function entity or the network open function entity, and acquiring characteristic information of the FL group according to application information or service information related to the FL group carried in the third request;

[0213] or,

[0214] The fourth request sent by the policy control function entity is received through the transceiver, and characteristic information of the FL group is acquired according to the control policy related to the FL group carried in the fourth request.

[0215] Optionally, the characteristic information of the FL group includes the first information and / or identification information of a functional entity allowed to join the FL group;

[0216] The first information includes any one or more of the following:

[0217] Information about the application or service corresponding to the FL group;

[0218] Area information corresponding to the FL group;

[0219] Network slice information corresponding to the FL group;

[0220] The data network information corresponding to the FL group.

[0221] Optionally, the determining, according to the characteristic information of the FL group, whether a condition for joining the FL group is met includes:

[0222] determining, according to the characteristic information of the FL group, whether the second functional entity meets a second condition;

[0223] The second condition includes one or a combination of the following conditions:

[0224] Support the applications or services corresponding to the FL group;

[0225] Serving the data network, network slice or area corresponding to the FL group;

[0226] Access the data network or network slice corresponding to the FL group;

[0227] The position is within the region corresponding to the FL group;

[0228] The identification information is included in the identification information of the functional entity allowed to join the FL group.

[0229] Optionally, the processor is further configured to:

[0230] Before joining the FL group, sending a first query request to a fourth functional entity through the transceiver, where the first query request carries third information of the FL group;

[0231] receiving, by the transceiver, a first query response sent by the fourth functional entity, where the first query response carries information of at least one first functional entity that satisfies the third information;

[0232] determining, from the at least one first functional entity, a first functional entity to establish the FL group with the second functional entity;

[0233] The third information of the FL group includes at least one of the following:

[0234] Information about the application or service corresponding to the FL group;

[0235] Network slice information corresponding to the FL group;

[0236] Data network information corresponding to the FL group;

[0237] Information about the machine learning (ML) model of the FL group.

[0238] Optionally, the information of the first functional entity includes: at least one of address information, location information, and service area information of the first functional entity;

[0239] The determining, from the at least one first functional entity, a first functional entity to establish the FL group with the second functional entity according to the information of the at least one first functional entity includes:

[0240] selecting, according to the information of the at least one first functional entity, a first functional entity that meets a third condition from the at least one first functional entity as the first functional entity for establishing the FL group with the second functional entity;

[0241] The third condition includes at least one of the following:

[0242] Belonging to the same local network as the second functional entity;

[0243] Serving the area where the second functional entity is located;

[0244] Among the at least one first functional entity, the distance between the second functional entity and the first functional entity is the shortest.

[0245] Optionally, joining the FL group includes:

[0246] sending, by the transceiver, a second FL join request to the first functional entity, where the second FL join request is used to request to join the FL group, and the second FL join request carries the first information and information about the second functional entity;

[0247] A second FL join response sent by the first functional entity is received through the transceiver.

[0248] Optionally, the processor is further configured to:

[0249] The memory is used to store second information of the FL group, wherein the second information includes at least one of the following:

[0250] information of the first functional entity;

[0251] information of the second functional entity;

[0252] identification information of the FL group;

[0253] Information about the application or service corresponding to the FL group;

[0254] Network slice information corresponding to the FL group;

[0255] Data network information corresponding to the FL group;

[0256] Information about the machine learning (ML) model of the FL group.

[0257] Optionally, the information of the first functional entity includes: at least one of identification information, address information and location information of the first functional entity; and / or,

[0258] The information of the second functional entity includes: at least one of identification information, address information and location information of the second functional entity.

[0259] Optionally, the processor is further configured to:

[0260] After joining the FL group, determining whether the second functional entity meets a condition for leaving the FL group;

[0261] If the second functional entity meets the condition for leaving the FL group, leaving the FL group;

[0262] The leaving the FL group includes: sending a first FL leave request to a first functional entity, where the first FL leave request is used to request to leave the FL group.

[0263] Optionally, the second functional entity satisfies a condition for leaving the FL group, including at least one of the following:

[0264] The second functional entity no longer supports the application or service corresponding to the FL group;

[0265] The second functional entity no longer serves the data network, network slice or area corresponding to the FL group;

[0266] The second functional entity no longer accesses the data network or network slice corresponding to the FL group;

[0267] The second functional entity leaves the area corresponding to the FL group.

[0268] Optionally, the determining whether the second functional entity meets a condition for leaving the FL group includes:

[0269] receiving, through the transceiver, updated feature information of the FL group sent by a third functional entity;

[0270] Determine, according to the updated characteristic information of the FL group, whether the second functional entity meets a condition for leaving the FL group.

[0271] Optionally, the leaving the FL group further includes:

[0272] The second information of the FL group stored in the second functional entity is deleted.

[0273] The embodiment of the present application further provides a functional entity, which is a third functional entity and includes a memory, a transceiver, and a processor:

[0274] A memory for storing a computer program; a transceiver for transmitting and receiving data under the control of the processor; and a processor for reading the computer program in the memory and performing the following operations:

[0275] Sending a first request or a second request to the first functional entity through the transceiver;

[0276] The first request carries application information or service information related to the federated learning FL group, and the second request carries a control policy related to the federated learning FL group; so that the first functional entity creates or updates the FL group according to the application information, service information or control policy related to the FL group.

[0277] Optionally, the application information, service information or control policy related to the FL group includes: characteristic information of the FL group;

[0278] The characteristic information of the FL group includes the first information and / or identification information of a functional entity allowed to join the FL group;

[0279] The first information includes any one or more of the following:

[0280] Information about the application or service corresponding to the FL group;

[0281] Area information corresponding to the FL group;

[0282] Network slice information corresponding to the FL group;

[0283] The data network information corresponding to the FL group.

[0284] The embodiment of the present application further provides a federated learning group processing device, which is applied to a first functional entity, and the device includes:

[0285] A first acquisition unit is used to acquire feature information of a federated learning FL group;

[0286] a first determining unit, configured to determine a second functional entity according to the characteristic information of the FL group;

[0287] The first processing unit is configured to add the second functional entity to the FL group.

[0288] Optionally, obtaining characteristic information of a federated learning FL group includes:

[0289] receiving a first request sent by an application function entity or a network open function entity, and acquiring characteristic information of the FL group according to application information or service information related to the FL group carried in the first request;

[0290] or,

[0291] A second request sent by the policy control function entity is received, and characteristic information of the FL group is acquired according to the control policy related to the FL group carried in the second request.

[0292] Optionally, the characteristic information of the FL group includes the first information and / or identification information of a functional entity allowed to join the FL group;

[0293] The first information includes any one or more of the following:

[0294] Information about the application or service corresponding to the FL group;

[0295] Area information corresponding to the FL group;

[0296] Network slice information corresponding to the FL group;

[0297] The data network information corresponding to the FL group.

[0298] Optionally, determining the second functional entity according to the characteristic information of the FL group includes:

[0299] determining, according to the characteristic information of the FL group, a second functional entity that meets a first condition;

[0300] The first condition includes one or a combination of the following conditions:

[0301] Support the applications or services corresponding to the FL group;

[0302] Serving the data network, network slice or area corresponding to the FL group;

[0303] Access the data network or network slice corresponding to the FL group;

[0304] The position is within the region corresponding to the FL group;

[0305] The identification information is included in the identification information of the functional entity allowed to join the FL group.

[0306] Optionally, adding the second functional entity to the FL group includes:

[0307] Sending a first FL join request to the second functional entity, where the first FL join request is used to request the second functional entity to join the FL group, and the first FL join request carries information about the first functional entity and first information about the FL group;

[0308] Receive a first FL joining response sent by the second functional entity.

[0309] Optionally, the determining the second functional entity according to the characteristic information of the FL group specifically includes:

[0310] receiving a second FL join request sent by the second functional entity, where the second FL join request is used to request to join the FL group, and the second FL join request carries the first information of the FL group and identification information of the second functional entity;

[0311] determining, according to the first information, an FL group to be joined by the second functional entity;

[0312] determining whether the identification information of the second functional entity is included in the identification information of the functional entities allowed to join the FL group;

[0313] If so, it is determined that the second functional entity is the second functional entity corresponding to the characteristic information of the FL group.

[0314] Optionally, the device further includes:

[0315] The first storage unit is configured to store second information of the FL group, wherein the second information includes at least one of the following:

[0316] information of the first functional entity;

[0317] information of the second functional entity;

[0318] identification information of the FL group;

[0319] Information about the application or service corresponding to the FL group;

[0320] Network slice information corresponding to the FL group;

[0321] Data network information corresponding to the FL group;

[0322] Information about the machine learning (ML) model of the FL group.

[0323] Optionally, the information of the first functional entity includes: at least one of identification information, address information and location information of the first functional entity; and / or,

[0324] The information of the second functional entity includes: at least one of identification information, address information and location information of the second functional entity.

[0325] Optionally, also include:

[0326] a second determining unit, configured to determine whether the second functional entity meets a condition for leaving the FL group after adding the second functional entity to the FL group;

[0327] The second processing unit is configured to remove the second functional entity from the FL group if the second functional entity meets a condition for leaving the FL group.

[0328] Optionally, the second functional entity satisfies a condition for leaving the FL group, including at least one of the following:

[0329] The second functional entity no longer supports the application or service corresponding to the FL group;

[0330] The second functional entity no longer serves the data network, network slice or area corresponding to the FL group;

[0331] The second functional entity no longer accesses the data network or network slice corresponding to the FL group;

[0332] The second functional entity leaves the area corresponding to the FL group;

[0333] The identification information of the second functional entity is no longer included in the identification information of functional entities allowed to join the FL group;

[0334] receiving a first FL leave request sent by the second functional entity, where the first FL leave request is used to request leaving the FL group;

[0335] The second functional entity has abnormal behavior, and the abnormal behavior includes: the frequency of establishing a connection between the second functional entity and the first functional entity is higher than a first threshold, or the amount of data transmitted between the second functional entity and the first functional entity exceeds a second threshold.

[0336] Optionally, the determining whether the second functional entity meets a condition for leaving the FL group includes:

[0337] receiving updated characteristic information of the FL group sent by the third functional entity;

[0338] Determine, according to the updated characteristic information of the FL group, whether the second functional entity meets a condition for leaving the FL group.

[0339] Optionally, removing the second functional entity from the FL group includes:

[0340] sending a second FL leave request to the second functional entity, where the second FL leave request is used to request the second functional entity to leave the FL group;

[0341] Receive a second FL leave response sent by the second functional entity.

[0342] Optionally, removing the second functional entity from the FL group includes:

[0343] The information of the second functional entity is deleted from the second information of the FL group.

[0344] This embodiment of the present application further provides a federated learning group processing device, which is applied to a second functional entity. The device includes:

[0345] The second acquisition unit is used to obtain feature information of the federated learning FL group;

[0346] a third determining unit, configured to determine whether a condition for joining the FL group is met based on the characteristic information of the FL group;

[0347] The third processing unit is configured to join the FL group if the conditions for joining the FL group are met.

[0348] Optionally, obtaining characteristic information of a federated learning FL group includes:

[0349] receiving a third request sent by the application function entity or the network open function entity, and acquiring characteristic information of the FL group according to application information or service information related to the FL group carried in the third request;

[0350] or,

[0351] The fourth request sent by the policy control function entity is received, and characteristic information of the FL group is acquired according to the control policy related to the FL group carried in the fourth request.

[0352] Optionally, the characteristic information of the FL group includes the first information and / or identification information of a functional entity allowed to join the FL group;

[0353] The first information includes any one or more of the following:

[0354] Information about the application or service corresponding to the FL group;

[0355] Area information corresponding to the FL group;

[0356] Network slice information corresponding to the FL group;

[0357] The data network information corresponding to the FL group.

[0358] Optionally, the determining, according to the characteristic information of the FL group, whether a condition for joining the FL group is met includes:

[0359] determining, according to the characteristic information of the FL group, whether the second functional entity meets a second condition;

[0360] The second condition includes one or a combination of the following conditions:

[0361] Support the applications or services corresponding to the FL group;

[0362] Serving the data network, network slice or area corresponding to the FL group;

[0363] Access the data network or network slice corresponding to the FL group;

[0364] The position is within the region corresponding to the FL group;

[0365] The identification information is included in the identification information of the functional entity allowed to join the FL group.

[0366] Optionally, also include:

[0367] a first sending unit, configured to send a first query request to a fourth functional entity before joining the FL group, where the first query request carries third information of the FL group;

[0368] a first receiving unit, configured to receive a first query response sent by the fourth functional entity, where the first query response carries information of at least one first functional entity that satisfies the third information;

[0369] a fourth determining unit, configured to determine, from the at least one first functional entity, a first functional entity that is to establish the FL group with the second functional entity;

[0370] The third information of the FL group includes at least one of the following:

[0371] Information about the application or service corresponding to the FL group;

[0372] Network slice information corresponding to the FL group;

[0373] Data network information corresponding to the FL group;

[0374] Information about the machine learning (ML) model of the FL group.

[0375] The embodiment of the present application further provides that the information of the first functional entity includes: at least one of address information, location information, and service area information of the first functional entity;

[0376] The determining, from the at least one first functional entity, a first functional entity to establish the FL group with the second functional entity according to the information of the at least one first functional entity includes:

[0377] selecting, according to the information of the at least one first functional entity, a first functional entity that meets a third condition from the at least one first functional entity as the first functional entity for establishing the FL group with the second functional entity;

[0378] The third condition includes at least one of the following:

[0379] Belonging to the same local network as the second functional entity;

[0380] Serving the area where the second functional entity is located;

[0381] Among the at least one first functional entity, the distance between the second functional entity and the first functional entity is the shortest.

[0382] Optionally, joining the FL group includes:

[0383] Sending a second FL join request to the first functional entity, where the second FL join request is used to request to join the FL group, and the second FL join request carries the first information and information about the second functional entity;

[0384] Receive a second FL joining response sent by the first functional entity.

[0385] Optionally, the device further includes:

[0386] The second storage unit is configured to store second information of the FL group, wherein the second information includes at least one of the following:

[0387] information of the first functional entity;

[0388] information of the second functional entity;

[0389] identification information of the FL group;

[0390] Information about the application or service corresponding to the FL group;

[0391] Network slice information corresponding to the FL group;

[0392] Data network information corresponding to the FL group;

[0393] Information about the machine learning (ML) model of the FL group.

[0394] Optionally, the information of the first functional entity includes: at least one of identification information, address information and location information of the first functional entity; and / or,

[0395] The information of the second functional entity includes: at least one of identification information, address information and location information of the second functional entity.

[0396] Optionally, also include:

[0397] a fifth determining unit, configured to determine whether the second functional entity meets a condition for leaving the FL group after joining the FL group;

[0398] a fourth processing unit, configured to leave the FL group if the second functional entity meets a condition for leaving the FL group;

[0399] The leaving the FL group includes: sending a first FL leave request to a first functional entity, where the first FL leave request is used to request to leave the FL group.

[0400] Optionally, the second functional entity satisfies a condition for leaving the FL group, including at least one of the following:

[0401] The second functional entity no longer supports the application or service corresponding to the FL group;

[0402] The second functional entity no longer serves the data network, network slice or area corresponding to the FL group;

[0403] The second functional entity no longer accesses the data network or network slice corresponding to the FL group;

[0404] The second functional entity leaves the area corresponding to the FL group.

[0405] Optionally, the determining whether the second functional entity meets a condition for leaving the FL group includes:

[0406] receiving updated characteristic information of the FL group sent by the third functional entity;

[0407] Determine, according to the updated characteristic information of the FL group, whether the second functional entity meets a condition for leaving the FL group.

[0408] Optionally, the leaving the FL group further includes:

[0409] The second information of the FL group stored in the second functional entity is deleted.

[0410] This embodiment of the present application further provides a federated learning group processing device, which is applied to a third functional entity. The device includes:

[0411] A second sending unit, configured to send the first request or the second request to the first functional entity;

[0412] The first request carries application information or service information related to the federated learning FL group, and the second request carries a control policy related to the federated learning FL group; so that the first functional entity creates or updates the FL group according to the application information, service information or control policy related to the FL group.

[0413] Optionally, the application information, service information or control policy related to the FL group includes: characteristic information of the FL group;

[0414] The characteristic information of the FL group includes the first information and / or identification information of a functional entity allowed to join the FL group;

[0415] The first information includes any one or more of the following:

[0416] Information about the application or service corresponding to the FL group;

[0417] Area information corresponding to the FL group;

[0418] Network slice information corresponding to the FL group;

[0419] The data network information corresponding to the FL group.

[0420] The embodiment of the present application further provides a processor-readable storage medium, wherein the processor-readable storage medium stores a computer program, wherein the computer program is configured to cause the processor to execute the above-mentioned federated learning group processing method on the first functional entity side; or

[0421] The computer program is used to enable the processor to execute the above-mentioned federated learning group processing method on the second functional entity side; or,

[0422] The computer program is used to enable the processor to execute the above-mentioned federated learning group processing method on the third functional entity side.

[0423] The beneficial effects of the above technical solution of this application are as follows:

[0424] In the above scheme, the federated learning group processing method obtains characteristic information of the federated learning FL group; determines a second functional entity based on the characteristic information of the FL group; and adds the second functional entity to the FL group. This can achieve dynamic processing of the FL group, thereby better meeting the needs of network and business intelligence and improving the intelligent performance of communications and applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0425] Figure 1 A schematic diagram of the wireless communication system architecture according to an embodiment of the present application;

[0426] Figure 2 This is a schematic diagram of the process of the federated learning group processing method of the embodiment of the present application Figure 1 ;

[0427] Figure 3 This is a schematic diagram of the process of the federated learning group processing method of the embodiment of the present application Figure 2 ;

[0428] Figure 4 This is a schematic diagram of the process of the federated learning group processing method of the embodiment of the present application Figure 3 ;

[0429] Figure 5 This is a schematic diagram of the specific implementation process of the federated learning group processing method of the embodiment of the present application Figure 1 ;

[0430] Figure 6 This is a schematic diagram of the specific implementation process of the federated learning group processing method of the embodiment of the present application Figure 2 ;

[0431] Figure 7 This is a schematic diagram of the specific implementation process of the federated learning group processing method of the embodiment of the present application Figure 3 ;

[0432] Figure 8 This is a schematic diagram of the specific implementation process of the federated learning group processing method of the embodiment of the present application Figure 4 ;

[0433] Figure 9 This is a schematic diagram of the specific implementation process of the federated learning group processing method of the embodiment of the present application Figure 5 ;

[0434] Figure 10 This is a schematic diagram of the specific implementation process of the federated learning group processing method of the embodiment of the present application Figure 6 ;

[0435] Figure 11 This is a schematic diagram of the specific implementation process of the federated learning group processing method of the embodiment of the present application Figure 7 ;

[0436] Figure 12 This is a schematic diagram of the specific implementation process of the federated learning group processing method of the embodiment of the present application Figure 8 ;

[0437] Figure 13 This is a schematic diagram of the specific implementation process of the federated learning group processing method of the embodiment of the present application Figure 9 ;

[0438] Figure 14 This is a schematic diagram of the specific implementation process of the federated learning group processing method of the embodiment of the present application Figure 10 ;

[0439] Figure 15 Schematic diagram of the functional entity structure of the embodiment of this application Figure 1 ;

[0440] Figure 16 Schematic diagram of the functional entity structure of the embodiment of this application Figure 2 ;

[0441] Figure 17 Schematic diagram of the functional entity structure of the embodiment of this application Figure 3 ;

[0442] Figure 18 Schematic diagram of the functional entity structure of the embodiment of this application Figure 4 ;

[0443] Figure 19 This is a schematic diagram of the structure of the federated learning group processing device in the embodiment of the present application. Figure 1 ;

[0444] Figure 20 This is a schematic diagram of the structure of the federated learning group processing device in the embodiment of the present application. Figure 2 ;

[0445] Figure 21 This is a schematic diagram of the structure of the federated learning group processing device in the embodiment of the present application. Figure 3 . DETAILED DESCRIPTION

[0446] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0447] In the embodiments of this application, the term "and / or" describes the association relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally indicates that the associated objects are in an "or" relationship.

[0448] In the embodiments of the present application, the term "plurality" refers to two or more than two, and other quantifiers are similar.

[0449] It is explained here that the technical solutions provided in the embodiments of the present application can be applicable to a variety of systems, especially 5G systems. For example, applicable systems may be global system of mobile communication (GSM) systems, code division multiple access (CDMA) systems, wideband code division multiple access (WCDMA) general packet radio service (GPRS) systems, long term evolution (LTE) systems, LTE frequency division duplex (FDD) systems, LTE time division duplex (TDD) systems, long term evolution advanced (LTE-A) systems, universal mobile telecommunication systems (UMTS), worldwide interoperability for microwave access (WiMAX) systems, 5G new air interface (NR) systems, etc. These various systems include terminal devices and network devices. The system may also include a core network part, such as an evolved packet system (EPS), a 5G system (5GS), etc.

[0450] Figure 1 A block diagram of a wireless communication system applicable to embodiments of the present application is shown. The wireless communication system includes a first functional entity and a second functional entity, and the first functional entity and the second functional entity can be implemented as a terminal device and / or a network device respectively.

[0451] The terminal device involved in the embodiments of the present application may be a device that provides voice and / or data connectivity to a user, a handheld device with wireless connection function, or other processing device connected to a wireless modem. In different systems, the name of the terminal device may also be different. For example, in a 5G system, the terminal device may be called User Equipment (UE). A wireless terminal device can communicate with one or more core networks (CN) via a radio access network (RAN). The wireless terminal device may be a mobile terminal device, such as a mobile phone (or "cellular" phone) and a computer with a mobile terminal device. For example, it may be a portable, pocket-sized, handheld, computer-built-in or vehicle-mounted mobile device that exchanges voice and / or data with a radio access network. For example, personal communication service (PCS) phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), and other devices. The wireless terminal device may also be referred to as a system, a subscriber unit, a subscriber station, a mobile station, a mobile station, a remote station, an access point, a remote terminal device, an access terminal device, a user terminal device, a user agent, or a user device, but is not limited in the embodiments of the present application.

[0452] The network device involved in the embodiments of the present application may be a base station, which may include multiple cells providing services to terminals. Depending on the specific application scenario, the base station may also be called an access point, or may be a device in an access network that communicates with a wireless terminal device through one or more sectors on an air interface, or may be named otherwise. The network device may be used to interchange received air frames with Internet Protocol (IP) packets, acting as a router between the wireless terminal device and the rest of the access network, wherein the rest of the access network may include an Internet Protocol (IP) communication network. The network device may also coordinate attribute management of the air interface. For example, the network device involved in the embodiments of the present application may be a network device (Base Transceiver Station, BTS) in the Global System for Mobile communications (GSM) or Code Division Multiple Access (CDMA), or a network device (NodeB) in Wide-band Code Division Multiple Access (WCDMA), or an evolutionary network device (eNB or e-NodeB) in the Long Term Evolution (LTE) system, a 5G base station (gNB) in the 5G network architecture (next generation system), or a home evolved Node B (HeNB), a relay node, a home base station (femto), a pico base station (pico), etc., and is not limited in the embodiments of the present application. In some network structures, the network device may include a centralized unit (CU) node and a distributed unit (DU) node, and the centralized unit and the distributed unit may also be geographically separated.

[0453] Network devices and terminal devices can each use one or more antennas for Multiple Input Multiple Output (MIMO) transmission. MIMO transmission can be either Single User MIMO (SU-MIMO) or Multi User MIMO (MU-MIMO). Depending on the configuration and number of antenna combinations, MIMO transmission can be 2D-MIMO, 3D-MIMO, FD-MIMO, or Massive-MIMO. It can also use diversity transmission, precoding, or beamforming.

[0454] Based on the above, embodiments of the present application provide a method, apparatus, and functional entity for processing a federated learning group, for implementing dynamic processing of a federated learning group. The method, apparatus, and functional entity are based on the same patent application concept. Since the principles for solving the problems of the method, apparatus, and functional entity are similar, their implementation can refer to each other, and any repetitions will not be repeated.

[0455] The federated learning group processing method provided in the embodiment of the present application is applied to the first functional entity, such as Figure 2 As shown, the method includes:

[0456] Step 21: Obtain the feature information of the federated learning FL group;

[0457] Step 22: Determine a second functional entity based on the characteristic information of the FL group;

[0458] Step 23: Add the second functional entity to the FL group.

[0459] The federated learning group processing method provided in the embodiment of the present application obtains characteristic information of a federated learning FL group; determines a second functional entity based on the characteristic information of the FL group; and adds the second functional entity to the FL group. This method can achieve dynamic processing of the FL group, thereby better meeting the needs of network and business intelligence and improving the intelligent performance of communications and applications.

[0460] The obtaining of characteristic information of the federated learning FL group includes: receiving a first request sent by an application function entity or a network open function entity, and obtaining the characteristic information of the FL group according to application information or service information related to the FL group carried in the first request; or receiving a second request sent by a policy control function entity, and obtaining the characteristic information of the FL group according to a control policy related to the FL group carried in the second request.

[0461] In this way, the first functional entity can dynamically process the FL group according to network policies or application requirements.

[0462] Regarding "obtaining characteristic information of a FL group," this may involve creating a FL group and obtaining the characteristic information of the FL group, or obtaining characteristic information for an already created FL group, without limitation. Furthermore, when the second functional entity is a terminal, the first request or the second request is received via the application layer of the terminal.

[0463] In an embodiment of the present application, the characteristic information of the FL group includes first information, and / or identification information of functional entities allowed to join the FL group; the first information includes any one or more of the following: information on the application or service corresponding to the FL group; area information corresponding to the FL group; network slice information corresponding to the FL group; and data network information corresponding to the FL group.

[0464] This will allow for more accurate treatment of the FL group.

[0465] The determining of the second functional entity based on the characteristic information of the FL group includes: determining a second functional entity that meets a first condition based on the characteristic information of the FL group; wherein the first condition includes one or a combination of the following conditions: supporting the application or service corresponding to the FL group; serving the data network, network slice or area corresponding to the FL group; accessing the data network or network slice corresponding to the FL group; being located within the area corresponding to the FL group; and the identification information being included in the identification information of the functional entity allowed to join the FL group.

[0466] This allows for more accurate determination of the second functional entity that can join the FL group. "Serving the data network, network slice, or area corresponding to the FL group" can refer to the second functional entity being an NF; "Accessing the data network or network slice corresponding to the FL group" can refer to the second functional entity being a UE, specifically establishing a protocol data unit (PDU) session under the data network or network slice corresponding to the FL group.

[0467] In this embodiment of the present application, adding the second functional entity to the FL group includes: sending a first FL join request to the second functional entity, where the first FL join request is used to request the second functional entity to join the FL group, and the first FL join request carries information about the first functional entity and first information about the FL group; and receiving a first FL join response sent by the second functional entity.

[0468] In this way, the second functional entity can passively join the FL group.

[0469] The determining the second functional entity based on the characteristic information of the FL group specifically includes: receiving a second FL joining request sent by the second functional entity, where the second FL joining request is used to request joining the FL group, and the second FL joining request carries the first information of the FL group and identification information of the second functional entity; determining, based on the first information, a FL group to be joined by the second functional entity; determining whether the identification information of the second functional entity is included in the identification information of the functional entities allowed to join the FL group; and if so, determining that the second functional entity is the second functional entity corresponding to the characteristic information of the FL group.

[0470] In this way, the second functional entity can actively join the FL group. The identification information is included in the characteristic information of the FL group to be joined by the second functional entity; the characteristic information is detailed above.

[0471] Furthermore, the method for processing a federated learning group further includes: storing second information of the FL group, wherein the second information includes at least one of the following: information of the first functional entity; information of the second functional entity; identification information of the FL group; information of the application or service corresponding to the FL group; network slice information corresponding to the FL group; data network information corresponding to the FL group; and information of the machine learning ML model of the FL group.

[0472] This can facilitate operations on the FL group. "Storage" can be storage when creating the FL group, or when a new functional entity joins the FL group; or storage when updating the FL group (such as updating the second information of the FL group), which is not limited here.

[0473] In the embodiment of the present application, the information of the first functional entity includes: at least one of identification information, address information, and location information of the first functional entity;

[0474] The information of the second functional entity includes: at least one of identification information, address information and location information of the second functional entity.

[0475] This makes it easier to accurately operate on the first functional entity and the second functional entity.

[0476] Furthermore, after adding the second functional entity to the FL group, it also includes: determining whether the second functional entity meets the conditions for leaving the FL group; if the second functional entity meets the conditions for leaving the FL group, removing the second functional entity from the FL group.

[0477] This enables the removal of members from the FL group.

[0478] Among them, the second functional entity meets the conditions for leaving the FL group, including at least one of the following situations: the second functional entity no longer supports the application or service corresponding to the FL group; the second functional entity no longer serves the data network, network slice or area corresponding to the FL group; the second functional entity no longer accesses the data network or network slice corresponding to the FL group; the second functional entity leaves the area corresponding to the FL group; the identification information of the second functional entity is no longer included in the identification information of functional entities allowed to join the FL group; a first FL leave request sent by the second functional entity is received, where the first FL leave request is used to request to leave the FL group; the second functional entity has abnormal behavior, and the abnormal behavior includes: the frequency of establishing connections between the second functional entity and the first functional entity is higher than a first threshold, or the amount of data transmitted between the second functional entity and the first functional entity exceeds a second threshold.

[0479] This makes it easier to accurately confirm whether the second functional entity is removed from the FL group. The "second functional entity no longer serves the data network, network slice, or area corresponding to the FL group" can be when the second functional entity is an NF; the "second functional entity no longer accesses the data network or network slice corresponding to the FL group" can be when the second functional entity is a UE, specifically releasing the PDU session established under the data network or network slice corresponding to the FL group.

[0480] In this embodiment of the present application, the method further includes: receiving updated characteristic information of the FL group sent by a third functional entity, so as to update the characteristic information of the FL group. Optionally, determining whether the second functional entity meets the condition for leaving the FL group includes: determining whether the second functional entity meets the condition for leaving the FL group based on the updated characteristic information of the FL group.

[0481] In this way, it is possible to determine whether to remove the second functional entity from the FL group through a third-party trigger. The third functional entity may be an application functional entity, a network open functional entity, or a policy control functional entity, which is not limited here.

[0482] The removing the second functional entity from the FL group includes: sending a second FL leave request to the second functional entity, where the second FL leave request is used to request the second functional entity to leave the FL group; and receiving a second FL leave response sent by the second functional entity.

[0483] In this way, the second functional entity can be removed from the FL group. It should be noted that the solution of "sending a second FL leave request to the second functional entity" cannot be used in combination with the solution of "receiving a first FL leave request sent by the second functional entity".

[0484] In the embodiment of the present application, removing the second functional entity from the FL group includes: deleting information of the second functional entity from second information of the FL group.

[0485] In this way, the relevant information about the FL group is real-time and accurate.

[0486] The present application also provides a method for processing a federated learning group. Figure 3 As shown, applied to the second functional entity, the method includes:

[0487] Step 31: Obtain feature information of the federated learning FL group;

[0488] Step 32: Determine whether the conditions for joining the FL group are met based on the characteristic information of the FL group;

[0489] Step 33: If the conditions for joining the FL group are met, join the FL group.

[0490] The federated learning group processing method provided in the embodiment of the present application obtains characteristic information of a federated learning FL group; determines whether conditions for joining the FL group are met based on the characteristic information of the FL group; and joins the FL group if the conditions for joining the FL group are met. This method can achieve dynamic processing of the FL group, thereby better meeting the needs of network and business intelligence and improving the intelligent performance of communications and applications.

[0491] The obtaining of characteristic information of the federated learning FL group includes: receiving a third request sent by an application function entity or a network open function entity, and obtaining the characteristic information of the FL group according to application information or service information related to the FL group carried in the third request; or receiving a fourth request sent by a policy control function entity, and obtaining the characteristic information of the FL group according to a control policy related to the FL group carried in the fourth request.

[0492] In this way, the second functional entity can flexibly determine the FL group according to network policies or application requirements.

[0493] Regarding "obtaining characteristic information of a FL group," this may involve creating a FL group and obtaining the characteristic information of the FL group, or obtaining characteristic information for an already created FL group, without limitation. Furthermore, when the second functional entity is a terminal, the first request or the second request is received via the application layer of the terminal.

[0494] In an embodiment of the present application, the characteristic information of the FL group includes first information, and / or identification information of functional entities allowed to join the FL group; the first information includes any one or more of the following: information on the application or service corresponding to the FL group; area information corresponding to the FL group; network slice information corresponding to the FL group; and data network information corresponding to the FL group.

[0495] The determining, based on the characteristic information of the FL group, whether the conditions for joining the FL group are met includes: determining, based on the characteristic information of the FL group, whether the second functional entity meets a second condition; wherein the second condition includes one or a combination of the following conditions: supporting the application or service corresponding to the FL group; serving the data network, network slice or area corresponding to the FL group; accessing the data network or network slice corresponding to the FL group; being located within the area corresponding to the FL group; and the identification information being included in the identification information of the functional entity allowed to join the FL group.

[0496] This allows for accurate determination of whether the second functional entity can join the FL group. "Serving the data network, network slice, or area corresponding to the FL group" can be for the second functional entity being an NF; "Accessing the data network or network slice corresponding to the FL group" can be for the second functional entity being a UE, specifically establishing a PDU session under the data network or network slice corresponding to the FL group.

[0497] Furthermore, before joining the FL group, it also includes: sending a first query request to the fourth functional entity, the first query request carrying the third information of the FL group; receiving a first query response sent by the fourth functional entity, the first query response carrying the information of at least one first functional entity that meets the third information; based on the information of the at least one first functional entity, determining a first functional entity from the at least one first functional entity to establish the FL group with the second functional entity; wherein the third information of the FL group includes at least one of the following: information on the application or service corresponding to the FL group; information on the network slice corresponding to the FL group; information on the data network corresponding to the FL group; information on the machine learning ML model of the FL group.

[0498] The information of the first functional entity includes: at least one of the address information, location information and service area information of the first functional entity; determining a first functional entity to establish the FL group with the second functional entity from the at least one first functional entity based on the information of the at least one first functional entity includes: selecting a first functional entity that meets a third condition from the at least one first functional entity based on the information of the at least one first functional entity as the first functional entity to establish the FL group with the second functional entity; wherein the third condition includes at least one of the following: belonging to the same local network as the second functional entity; serving the area where the second functional entity is located; and being the shortest distance to the second functional entity among the at least one first functional entity.

[0499] In this way, a first functional entity with better performance when interacting with the second functional entity can be obtained.

[0500] In the embodiment of the present application, joining the FL group includes: sending a second FL join request to the first functional entity, where the second FL join request is used to request joining the FL group, and the second FL join request carries the first information and information about the second functional entity; and receiving a second FL join response sent by the first functional entity.

[0501] In this way, the second functional entity can actively join the FL group.

[0502] Furthermore, the method also includes: storing second information of the FL group, wherein the second information includes at least one of the following: information of the first functional entity; information of the second functional entity; identification information of the FL group; information of the application or service corresponding to the FL group; network slice information corresponding to the FL group; data network information corresponding to the FL group; information of the machine learning ML model of the FL group.

[0503] Regarding "storage", it can be storage when creating the FL group, or when a new functional entity joins the FL group; or storage when the FL group is updated (such as updating the second information of the FL group), which is not limited here.

[0504] The information of the first functional entity includes: at least one of identification information, address information and location information of the first functional entity;

[0505] The information of the second functional entity includes: at least one of identification information, address information and location information of the second functional entity.

[0506] This makes it easier to accurately operate on the first functional entity and the second functional entity.

[0507] Furthermore, after joining the FL group, the method further includes: determining whether the second functional entity meets a condition for leaving the FL group; and if the second functional entity meets the condition for leaving the FL group, sending a first FL leave request to the first functional entity, where the first FL leave request is used to request leaving the FL group. Specifically, if the second functional entity meets the condition for leaving the FL group, the second functional entity leaves the FL group. Leaving the FL group includes sending the first FL leave request to the first functional entity, where the first FL leave request is used to request leaving the FL group.

[0508] This enables the removal of the second functional entity from the FL group.

[0509] Among them, the second functional entity meets the conditions for leaving the FL group, including at least one of the following situations: the second functional entity no longer supports the application or service corresponding to the FL group; the second functional entity no longer serves the data network, network slice or area corresponding to the FL group; the second functional entity no longer accesses the data network or network slice corresponding to the FL group; the second functional entity leaves the area corresponding to the FL group.

[0510] Regarding "the second functional entity no longer serves the data network, network slice or area corresponding to the FL group", it can be that the second functional entity is NF; regarding "the second functional entity no longer accesses the data network or network slice corresponding to the FL group", it can be that the second functional entity is UE, specifically releasing the PDU session established under the data network or network slice corresponding to the FL group.

[0511] In an embodiment of the present application, the method further includes: receiving updated characteristic information of the FL group sent by a third functional entity; and determining whether the second functional entity meets the condition for leaving the FL group includes: determining whether the second functional entity meets the condition for leaving the FL group based on the updated characteristic information of the FL group.

[0512] In this way, it is possible to determine whether to remove the second functional entity from the FL group through a third-party trigger. The third functional entity may be an application functional entity, a network open functional entity, or a policy control functional entity, which is not limited here.

[0513] Furthermore, when the second functional entity meets the condition for leaving the FL group, the method further includes: deleting second information about the FL group stored in the second functional entity. Specifically, the method may include: deleting the second information about the FL group stored in the second functional entity.

[0514] In this way, the relevant information about the FL group is real-time and accurate.

[0515] The present application embodiment also provides a federated learning group processing method, which is applied to a third functional entity, such as Figure 4 As shown, the method includes:

[0516] Step 41: Send a first request or a second request to the first functional entity; wherein the first request carries application information or service information related to the federated learning FL group, and the second request carries a control policy related to the federated learning FL group; so that the first functional entity creates or updates the FL group according to the application information, service information or control policy related to the FL group.

[0517] The third functional entity may be an application functional entity, a network open functional entity or a policy control functional entity, which is not limited here.

[0518] The federated learning group processing method provided in an embodiment of the present application sends a first request or a second request to a first functional entity; wherein the first request carries application information or service information related to the federated learning FL group, and the second request carries a control policy related to the federated learning FL group; so that the first functional entity creates or updates the FL group based on the application information, service information, or control policy related to the FL group. Dynamic processing of the FL group can be achieved, thereby better meeting the requirements of network and service intelligence and improving the intelligent performance of communications and applications.

[0519] The application information, service information or control policy related to the FL group includes: characteristic information of the FL group; the characteristic information of the FL group includes first information, and / or identification information of functional entities allowed to join the FL group; the first information includes any one or more of the following: information on the application or service corresponding to the FL group; area information corresponding to the FL group; network slice information corresponding to the FL group; data network information corresponding to the FL group.

[0520] This will allow for more accurate treatment of the FL group.

[0521] It should be noted that the names of the requests, responses, or information mentioned above can be adjusted according to their role or function. For example, the first FL join request can be adjusted to the FL group join request, the join FL group request, the join FL group request, etc., which is not limited in the embodiments of this application. In addition, they can be adapted to the names of similar processes or content involved in the following examples, rather than being different simply because of the literal name difference; for example, the first FL leave request mentioned above can be adapted to the federated learning leave request described below. More specifically, they can be consistent, but are not limited to this.

[0522] The following is an example of the federated learning group processing method provided in the embodiment of the present application. The first functional entity takes the network data analysis function NWDAF (entity) as an example, the second functional entity takes the network function NF (entity) or the terminal UE as an example, the third functional entity takes the application function AF (entity), the network (capability) exposure function NEF (entity) or the policy control function PCF (entity) as an example, and the fourth functional entity takes the network database function NRF (entity) as an example.

[0523] To achieve the above objectives, the present application provides a federated learning group processing method, which is described below with specific examples. In the following examples:

[0524] 1. The first functional entity has a federated learning management function (also known as a federated learning group management function (FLMF). The following example uses a NWDAF that provides a global model. However, the federated learning management function is not limited to the NWDAF that provides a global model. It can also be other network functions, such as other network functions that provide a global model, or network functions dedicated to federated learning management.

[0525] 2. The first functional entity may be a network function NF or a user equipment UE. The network function NF includes a core network function and / or an access network function.

[0526] 3. The AF can be an untrusted AF or a trusted AF. If the AF is a trusted AF, then in Examples 3 and 4, the AF can communicate directly with the UDR (Unified Data Repository), NWDAF, etc., without going through the NEF.

[0527] Example 1: The network function NF initiates joining the federated learning FL group;

[0528] Specifically, you can Figure 5 As shown (the second functional entity takes NF as an example), this solution includes:

[0529] Step 51. NWDAF registration operation;

[0530] (1) NWDAF sends a registration request to the network database function NRF to register its own information (NWDAF profile). If NWDAF supports Federated Learning (FL), the following FL-related information may be included:

[0531] 1) FL capabilities of NWDAF, including specific information about the FL supported by NWDAF, such as the applicable applications, services or scenarios (e.g., image recognition, remote control, V2X communication, etc.), regions, time periods, model types (e.g., specific deep neural network types), and / or accuracy or confidence levels for each FL global model that NWDAF can provide;

[0532] 2) FL group information managed by NWDAF, such as FL group ID.

[0533] In addition, NWDAF can also register the following information for other NFs to discover the local NWDAF:

[0534] 1) NWDAF location information, including operator-specific location information, such as geographic location information or data center information.

[0535] 2) IP address information of the NWDAF, such as the NWDAF IPv4 address and / or IPv6 network prefix (IPv6 network prefix).

[0536] (2) NRF feeds back a registration response to NWDAF.

[0537] Step 52: Network function NF query operation;

[0538] (1) The network function NF sends a query request to the NRF. The query request indicates that the target NF type is NWDAF and carries the following information:

[0539] 1) FL feature information determined according to the federated learning strategy, such as the application (or business or scenario), region, time period, model type, and / or accuracy or confidence level applicable to FL; and / or,

[0540] 2) FL Group ID, if the NF has learned this information through previous processes or pre-configuration.

[0541] In addition, the query request may also include a query instruction for NWDAF location information or IP address information.

[0542] (2) The NRF returns a query response based on the query request. This response may include information about NWDAFs that meet the requirements (i.e., support the requested FL). The NWDAF information may also include information about the corresponding FL Group, such as the FL Group ID. Based on the query request, NWDAF information, and / or NRF local policy, the NRF may also provide the NWDAF's location information and / or IP address information in the query response.

[0543] Subsequently, the NF can select an NWDAF that supports the requested FL based on the query response. If the query response contains the location information and / or IP address information of the NWDAF, the NF can select a local NWDAF that supports the requested FL to speed up the FL process.

[0544] Step 53. The NF sends a federated learning join request (FL join request) to the selected NWDAF, which carries the NF ID, the FL group ID requested to join (if the NF has learned this information through previous processes or pre-configuration) or FL feature information, requesting the NWDAF to add it to the corresponding FL group.

[0545] Step 54: The NWDAF adds the NF to the FL group, specifically by adding the information of the NF (eg, NF ID) to the FL group context stored in the NWDAF.

[0546] The FL group context in the NWDAF may include at least one of the following information:

[0547] (1) FL group ID;

[0548] (2) Group member ID list, i.e., the list of NFs that have joined the FL group;

[0549] (3) Global model information provided by NWDAF to NFs in the FL group, such as the global model file address or global model parameter list;

[0550] (4) Information about the local model provided by the NF in the FL group to the NWDAF, such as the file address of the local model or the parameter list of the local model. This information can be used in the NWDAF to store the local model information of multiple NFs and use these models to retrain the global model.

[0551] (5) FL feature information.

[0552] Step 55. NWDAF sends a federated learning join accept (FL join accept) message (i.e., FL join response) to NF, which may carry FL group identifier, global model information, etc.

[0553] Step 56. Manage update operations;

[0554] (1) The NF sends an NF information update request to the NRF, adding the information of the FL group to which the NF joins, such as the FL group ID and the ID of the FL group management function (here, the NWDAF ID), to the NF profile. Subsequently, network functions such as the AF, PCF, or NWDAF can use this information to discover all or some NFs that have joined a specific FL group (for example, NFs that have joined the FL group within a specific area), and thus request or perform analysis of the ML model transmission status or model performance of the FL group.

[0555] (2) NRF feeds back the updated response to NF.

[0556] Example 2: NF initiates leaving the FL group;

[0557] Specifically, you can Figure 6 As shown (the second functional entity takes NF as an example), this solution includes:

[0558] Step 61: The NF sends a federated learning leave request (FL leave request) to the NWDAF, which carries the NF ID and the ID of the FL group requesting to leave.

[0559] Step 62: Based on the received federated learning leave request of the NF, the NWDAF removes the NF from the FL group; specifically, the NWDAF may remove the NF information (eg, NF ID) from the FL group context corresponding to the FL group identifier.

[0560] Step 63: NWDAF sends a federated learning leave accept (FL leave accept) message to NF, i.e., FL leave response.

[0561] Step 64. Manage update operations;

[0562] (1) The NF sends an NF information update request to the NRF to delete the information of the FL group that the NF has left (such as the FL group ID and the ID of the FL group management function) from the NF profile.

[0563] (2) NRF feeds back the updated response to NF.

[0564] Example 3: Other NFs (such as AF or NWDAF) initiate the process of adding NF to the FL group (if the AF is a trusted AF, the AF can communicate directly with the UDR (Unified Database), NWDAF, etc. without going through the NEF);

[0565] The network can initiate the establishment of a specific FL group (that is, adding the first member to the FL group) or request the NF to join a specific FL group based on application requirements and / or network policies. Figure 7 、 Figure 8 、 Figure 9 and Figure 14 Four possible flows are provided.

[0566] The first one, specifically, can be Figure 7 As shown, this program includes:

[0567] Step 71. AF initiates a service parameter generation or update request (i.e., Nnef_ServiceParameter_Create or Nnef_ServiceParameter_Update) to NEF to generate or update service parameters, including service or application-related federated learning parameters, such as: application ID for which federated learning is required, terminal ID UEID list or terminal group ID UE Group ID, area, single network slice selection auxiliary information S-NSSAI, data network name DNN, application features (Features) and other information.

[0568] Step 72: The NEF stores the service parameters provided by the AF in the UDR, including the federated learning parameters related to the service or application. That is, the UDR stores the FL related parameters.

[0569] Step 73. The UDR sends a data management notification Nudr_DM_Notify to the PCF, which includes the federated learning parameters related to the above business or application.

[0570] Step 74. PCF provides or updates the relevant federated learning policy to UE or other NF (such as AMF, SMF) based on the above-mentioned service or application-related federated learning parameters, and requires UE or other NF to join the corresponding federated learning group. For example:

[0571] a) Send an updated UE policy (UE policy) to the corresponding UE based on the UE ID list or UE Group ID, which carries the federated learning policy determined based on the federated learning parameters related to the above-mentioned service or application, and may include: Application ID, region, S-NSSAI, DNN, application characteristics and other information requiring federated learning, and may also include identification information of UEs allowed to join the federated learning group (such as UE Group ID).

[0572] b) Send an updated access and mobility management policy (AM policy) to the AMF, which carries the federated learning policy related to the access and mobility management function determined according to the federated learning parameters related to the above-mentioned business or application, and may include: a list of UE IDs or UE Group IDs, areas, S-NSSAI, DNNs and other information managed by federated learning related to the access and mobility management function, and may also include identification information of AMFs allowed to join the federated learning group (e.g., AMF set identifier AMF Set ID).

[0573] c) Send an updated session management policy (SM policy) to the SMF, which carries the federated learning policy related to the session management function determined according to the federated learning parameters related to the above-mentioned business or application, and may include: Application ID, UE ID list or UE Group ID, area, S-NSSAI, DNN and other information corresponding to the PDU session managed by federated learning related to the session management function, and may also include identification information of SMFs allowed to join the federated learning group (for example, SMF set identifier SMF Set ID).

[0574] Here, the UE or other NFs (such as AMF, SMF) have the federated learning function. Other NFs are relative to NWDAF.

[0575] Before sending the federated learning policy, the PCF can query the NRF for information about the NFs (such as AMF and SMF) that support the federated learning. That is, the PCF sends an NF discovery request Nnrf_NFDiscovery_Request to the NRF, which carries one or more of the above-mentioned federated learning parameters related to the service or application, such as the Application ID, UE ID list or UEGroup ID, region, S-NSSAI, DNN, application features, etc. The NRF responds to the PCF with information about the NFs that support the federated learning (such as NF type, NF instance ID, etc.).

[0576] After the UE or other NF (such as AMF, SMF) receives the federated learning strategy provided by the PCF, it performs the following steps according to the federated learning strategy provided by the PCF:

[0577] Step 75. UE or other NF query operation;

[0578] (1) The UE or other NF sends a query request to the NRF. The query request indicates that the target NF type is NWDAF and carries the following information:

[0579] 1) FL feature information determined according to the federated learning strategy, such as the application (or business or scenario), region, time period, model type, and / or accuracy or confidence level applicable to FL; and / or,

[0580] 2) FL Group ID, if the UE or other NF has learned this information through previous procedures or pre-configuration.

[0581] In addition, the query request may also include a query instruction for NWDAF location information or IP address information.

[0582] (2) The NRF returns a query response based on the query request. This response may include information about NWDAFs that meet the requirements (i.e., support the requested FL). The NWDAF information may also include information about the corresponding FL Group, such as the FL Group ID. Based on the query request, NWDAF information, and / or NRF local policy, the NRF may also provide the NWDAF's location information and / or IP address information in the query response.

[0583] Subsequently, the UE or other NF may select an NWDAF that supports the requested FL based on the query response. If the query response contains the location information and / or IP address information of the NWDAF, the UE or other NF may select a local NWDAF that supports the requested FL to speed up the FL process.

[0584] Step 76. The UE or other NF sends a federated learning join request (FL join request) to the selected NWDAF, which carries the NF ID, the FL group ID requested to join (if the UE or other NF has learned this information through previous procedures or pre-configuration) or FL feature information, requesting the NWDAF to add it to the corresponding FL group.

[0585] Before executing step 75 or 76, if the federated learning strategy provided by the PCF contains identification information of the UE or NF allowed to join the federated learning group, the UE or other NF can also confirm whether it is allowed to join the federated learning group based on the information. If so, execute steps 75, 76 and subsequent steps.

[0586] Step 77. The NWDAF adds the UE or other NF to the FL group, specifically by adding the information of the UE or other NF (eg, NF ID) to the FL group context stored in the NWDAF.

[0587] The FL group context in the NWDAF may include at least one of the following information:

[0588] (1) FL group ID;

[0589] (2) Group member ID list, i.e., the list of NFs that have joined the FL group;

[0590] (3) Global model information provided by NWDAF to NFs in the FL group, such as the global model file address or global model parameter list;

[0591] (4) Information about the local model provided by the NF in the FL group to the NWDAF, such as the file address of the local model or the parameter list of the local model. This information can be used in the NWDAF to store the local model information of multiple NFs and use these models to retrain the global model.

[0592] (5) FL feature information.

[0593] Step 78. The NWDAF sends a federated learning join accept (FL join accept) message (i.e., FL join response) to the UE or other NF, which may carry the FL group identifier, global model information, etc.

[0594] Step 79. Manage update operations;

[0595] (1) The NF sends an NF information update request to the NRF, adding the information of the FL group to which the NF has joined, such as the FL group ID and the ID of the FL group management function (here, the NWDAF ID), to the NF profile. Subsequent network functions such as the AF, PCF, or NWDAF can use this information to discover all or some NFs that have joined a specific FL group (e.g., NFs that have joined the FL group within a specific area), and thus request or perform analysis of the ML model transmission status or model performance of the FL group. The NF here is the same as the other NFs mentioned above.

[0596] (2) NRF feeds back the updated response to NF.

[0597] That is, after step 74, the UE or other NF (such as AMF, SMF) initiates the federated learning group joining process to the NWDAF according to the federated learning strategy provided by the PCF. The steps are similar to steps 52-56 in Example 1. The FL feature information includes one or more of the above-mentioned service or application-related federated learning parameters, such as the Application ID of federated learning, UE ID list or UE Group ID, area, S-NSSAI, DNN, application features (Features), etc.

[0598] The second type, specifically, can be Figure 8 As shown, this program includes:

[0599] Step 81. AF initiates a service parameter generation or update request (i.e., Nnef_ServiceParameter_Create or Nnef_ServiceParameter_Update) to NEF to generate or update service parameters, including service or application-related federated learning parameters, such as: application ID for which federated learning is required, terminal ID UEID list or terminal group ID UE Group ID, area, single network slice selection auxiliary information S-NSSAI, data network name DNN, application features (Features) and other information.

[0600] Step 82: The NEF stores the service parameters provided by the AF in the UDR, including the federated learning parameters related to the service or application. That is, the UDR stores the FL related parameters.

[0601] Step 83. The UDR sends a data management notification Nudr_DM_Notify to the PCF, which includes the federated learning parameters related to the above business or application.

[0602] Step 84. PCF provides or updates the federated learning strategy to NWDAF, which carries the federated learning parameters related to the above services or applications. Before sending the federated learning strategy, PCF can query NRF for information about NWDAF that supports the federated learning, similar to Figure 5 In step 52 (only when the initiator is PCF), the FL feature information includes one or more federated learning parameters related to the above services or applications, such as the Application ID, UE ID list or UE Group ID, area, S-NSSAI, DNN, application features, etc. of federated learning. Based on this, the NRF replies to the PCF with the information of the NWDAF that supports the federated learning (such as NWDAF Instance ID, etc.).

[0603] Step 85. NWDAF initiates a federated learning request to the UE or other NFs (such as AMF, SMF) based on the received federated learning policy. For example:

[0604] a) The NWDAF sends a federated learning request to the corresponding UE based on the UE ID list or UE Group ID. The request carries FL feature information, including the Application ID, region, S-NSSAI, DNN, and application features of the federated learning, instructing the UE to join the federated learning group of the relevant service or application.

[0605] b) NWDAF sends a federated learning request to AMF, which carries FL feature information, including the UE ID list or UE Group ID, area, S-NSSAI, DNN and other information of federated learning, instructing AMF to join the related access management and / or mobility management federated learning group;

[0606] c) NWDAF sends a federated learning request to SMF, which carries FL feature information, including the Application ID, UE ID list or UE Group ID, area, S-NSSAI, DNN and other information corresponding to the PDU session of federated learning, instructing SMF to join the federated learning group managed by the relevant session.

[0607] Here, the UE or other NFs (such as AMF, SMF) have a federated learning function.

[0608] Before sending a federated learning request, NWDAF can query NRF for information about NFs (such as AMF and SMF) that support federated learning. For example, NWDAF sends an NF discovery request Nnrf_NFDiscovery_Request to NRF, which carries one or more of the above-mentioned federated learning parameters related to the service or application, such as the Application ID, UE ID list or UE Group ID, region, S-NSSAI, DNN, application features, etc. of federated learning. NRF then responds to NWDAF with information about the NFs that support federated learning (such as NF type, NF instance ID, etc.).

[0609] Step 86. The UE or other NF sends a federated learning join request (FL join request) to the NWDAF, which carries the NF ID, the FL group ID requested to join (if the UE or other NF has learned this information through previous procedures or pre-configuration) or FL feature information, requesting the NWDAF to add it to the corresponding FL group.

[0610] Step 87. The NWDAF adds the UE or other NF to the FL group, specifically by adding the information of the UE or other NF (eg, NF ID) to the FL group context stored in the NWDAF.

[0611] The FL group context in the NWDAF may include at least one of the following information:

[0612] (1) FL group ID;

[0613] (2) Group member ID list, i.e., the list of NFs that have joined the FL group;

[0614] (3) Global model information provided by NWDAF to NFs in the FL group, such as the global model file address or global model parameter list;

[0615] (4) Information about the local model provided by the NF in the FL group to the NWDAF, such as the file address of the local model or the parameter list of the local model. This information can be used in the NWDAF to store the local model information of multiple NFs and use these models to retrain the global model.

[0616] (5) FL feature information.

[0617] Step 88. The NWDAF sends a federated learning join accept (FL join accept) message (i.e., FL join response) to the UE or other NF, which may carry the FL group identifier, global model information, etc.

[0618] Step 89. Manage update operations;

[0619] (1) The NF sends an NF information update request to the NRF, adding the information of the FL group to which the NF has joined, such as the FL group ID and the ID of the FL group management function (here, the NWDAF ID), to the NF profile. Subsequent network functions such as the AF, PCF, or NWDAF can use this information to discover all or some NFs that have joined a specific FL group (e.g., NFs that have joined the FL group within a specific area), and thus request or perform analysis of the ML model transmission status or model performance of the FL group. The NF here is the same as the other NFs mentioned above.

[0620] (2) NRF feeds back the updated response to NF.

[0621] That is, after step 85, the UE or other NF (such as AMF, SMF) initiates the joining process of the corresponding federated learning group to the NWDAF according to the federated learning request of the NWDAF. The steps are similar to steps 53-56 in Example 1, wherein the FL feature information includes one or more federated learning parameters related to the above-mentioned services or applications, such as the Application ID, UEID list or UE Group ID, area, S-NSSAI, DNN, application features (Features), etc. of the federated learning.

[0622] The third type, specifically, can be as follows Figure 9 As shown, this program includes:

[0623] Step 91. The AF initiates a federated learning request to the NEF, which includes service or application-related federated learning parameters, such as the Application ID, UE ID list or UE Group ID, region, S-NSSAI, DNN, application features, and other information required for federated learning.

[0624] Step 92. NEF performs an authorization check on the federated learning request of AF based on the service agreement, operator policy, etc. If the request is authorized, NEF queries the relevant NWDAF and sends a federated learning request to NWDAF, which carries the federated learning parameters related to the above service or application. The process of NEF querying NWDAF is similar to Figure 5 Step 52 (only when the initiator is NEF), wherein the FL feature information includes the above-mentioned service or application-related federated learning parameters.

[0625] Step 93. NWDAF initiates a federated learning request to the UE or other NFs (such as AMF, SMF) based on the received federated learning request. For example:

[0626] a) The NWDAF sends a federated learning request to the corresponding UE based on the UE ID list or UE Group ID. The request carries FL feature information, including the Application ID, region, S-NSSAI, DNN, and application features of the federated learning, instructing the UE to join the federated learning group of the relevant service or application.

[0627] b) NWDAF sends a federated learning request to AMF, which carries FL feature information, including the UE ID list or UE Group ID, area, S-NSSAI, DNN and other information of federated learning, instructing AMF to join the related access management and / or mobility management federated learning group;

[0628] c) NWDAF sends a federated learning request to SMF, which carries FL feature information, including the Application ID, UE ID list or UE Group ID, area, S-NSSAI, DNN and other information corresponding to the PDU session of federated learning, instructing SMF to join the federated learning group managed by the relevant session.

[0629] Here, the UE or other NFs (such as AMF, SMF) have a federated learning function.

[0630] Before sending a federated learning request, NWDAF can query NRF for information about NFs (such as AMF and SMF) that support federated learning. For example, NWDAF sends an NF discovery request Nnrf_NFDiscovery_Request to NRF, which carries one or more of the above-mentioned federated learning parameters related to the service or application, such as the Application ID, UE ID list or UE Group ID, region, S-NSSAI, DNN, application features, etc. of federated learning. NRF then responds to NWDAF with information about the NFs that support federated learning (such as NF type, NF instance ID, etc.).

[0631] Step 94. The UE or other NF sends a federated learning join request (FL join request) to the NWDAF, which carries the NF ID, the FL group ID requested to join (if the UE or other NF has learned this information through previous procedures or pre-configuration) or FL feature information, requesting the NWDAF to add it to the corresponding FL group.

[0632] Step 95: The NWDAF adds the UE or other NF to the FL group, specifically by adding the information of the UE or other NF (eg, NF ID) to the FL group context stored in the NWDAF.

[0633] The FL group context in the NWDAF may include at least one of the following information:

[0634] (1) FL group ID;

[0635] (2) Group member ID list, i.e., the list of NFs that have joined the FL group;

[0636] (3) Global model information provided by NWDAF to NFs in the FL group, such as the global model file address or global model parameter list;

[0637] (4) Information about the local model provided by the NF in the FL group to the NWDAF, such as the file address of the local model or the parameter list of the local model. This information can be used in the NWDAF to store the local model information of multiple NFs and use these models to retrain the global model.

[0638] (5) FL feature information.

[0639] Step 96: NWDAF sends a federated learning join accept (FL join accept) message (i.e., FL join response) to the UE or other NF, which may carry the FL group identifier, global model information, etc.

[0640] Step 97. Manage update operations;

[0641] (1) The NF sends an NF information update request to the NRF, adding the information of the FL group to which the NF has joined, such as the FL group ID and the ID of the FL group management function (here, the NWDAF ID), to the NF profile. Subsequent network functions such as the AF, PCF, or NWDAF can use this information to discover all or some NFs that have joined a specific FL group (e.g., NFs that have joined the FL group within a specific area), and thus request or perform analysis of the ML model transmission status or model performance of the FL group. The NF here is the same as the other NFs mentioned above.

[0642] (2) NRF feeds back the updated response to NF.

[0643] That is, after step 93, the UE or other NF (such as AMF, SMF) initiates the joining process of the corresponding federated learning group to the NWDAF according to the federated learning request of the NWDAF. The steps are similar to steps 53-56 in Example 1, wherein the FL feature information includes one or more federated learning parameters related to the above-mentioned services or applications, such as the Application ID, UEID list or UE Group ID, area, S-NSSAI, DNN, application features (Features), etc. of the federated learning.

[0644] The fourth type, specifically, can be Figure 14 As shown, this program includes:

[0645] Step 141. The NWDAF determines to initiate a request to the UE or other NF (e.g., AMF, SMF) to join federated learning or establish a federated learning group based on external trigger conditions or local policies. For example, the NWDAF detects the following conditions:

[0646] 1) The UE enters the FL area, or a specific NF state or function, such as the NF serving the FL area;

[0647] That is, it is detected that the UE joins the FL area or the NF serves the FL area, and it is determined that the UE or other NF needs to join the federated learning.

[0648] Step 142. The NWDAF sends a request to join federated learning or establish a federated learning group to the relevant UE or other NF (e.g., AMF, SMF); including sending a request to join federated learning or establish a federated learning group to the following UE or other NF:

[0649] 1) UE entering the FL area, or NF serving the FL area.

[0650] Step 143. The UE or NF that receives the request to join federated learning or establish a federated learning group sends a federated learning join request (FL join request) to the NWDAF, which carries the UE identifier (UE ID) or NF identifier (NF ID) and the FL group identifier requested to join (if the UE or other NF has learned this information through previous procedures or pre-configuration).

[0651] Step 144. Based on the received federated learning joining request of the UE or NF, the NWDAF adds the UE or NF to the FL group; specifically, it can be: adding the information of the UE or NF (such as the UE ID or NF ID) to the FL group context corresponding to the FL group identifier.

[0652] Step 145: The NWDAF sends a federated learning join accept (FL join accept) message, i.e., a FL join response, to the UE or NF.

[0653] Step 146. Manage update operations;

[0654] (1) The NF sends an NF information update request to the NRF, adding the information of the FL group to which the NF joins (such as the FL group ID and the ID of the FL group management function (here, the NWDAF ID)) to the NF profile.

[0655] (2) NRF feeds back the updated response to NF.

[0656] Specifically, regarding steps 143 to 146: the UE or other NF that receives the request to join the federated learning or establish a federated learning group initiates the process of joining the FL group (i.e., the federated learning group joining process); Figure 5 Steps 53-56 are similar.

[0657] Example 4: Other NFs (such as AF or NWDAF) initiate the removal of NF from the FL group (if the AF is a trusted AF, the AF can communicate directly with the UDR (unified database), NWDAF, etc. without going through the NEF);

[0658] The network can initiate the termination of a specific FL group (that is, the last member of the FL group leaves) or request the NF to leave a specific FL group based on application requirements and / or network policies. Figure 10 、 Figure 11 、 Figure 12 and Figure 13 Four possible flows are provided.

[0659] The first one, specifically, can be Figure 10 As shown, this program includes:

[0660] Step 101. The AF initiates a service parameter generation or update request (i.e., Nnef_ServiceParameter_Create or Nnef_ServiceParameter_Update) to the NEF to generate or update service parameters, including service or application-related federated learning parameters, such as: Application ID, UE ID list or UEGroup ID, region, S-NSSAI, DNN, application features, and other information required to terminate federated learning.

[0661] Step 102: The NEF stores the service parameters provided by the AF in the UDR, including the federated learning parameters related to the service or application. That is, the UDR stores the FL related parameters.

[0662] Step 103: The UDR sends a data management notification Nudr_DM_Notify to the PCF, which includes the federated learning parameters related to the above-mentioned business or application.

[0663] Step 104. The PCF provides or updates the relevant federated learning policy to the UE or other NFs (such as AMF, SMF) based on the above-mentioned service or application-related federated learning parameters, requiring the UE or other NFs to terminate the relevant federated learning or leave the corresponding federated learning group (other NFs are relative to NWDAF). For example:

[0664] a) Send an updated UE policy (UE policy) to the corresponding UE based on the UE ID list or UE Group ID, which carries the federated learning policy determined based on the federated learning parameters related to the above-mentioned service or application, and may include: Application ID, region, S-NSSAI, DNN, application characteristics, etc. that require termination of federated learning.

[0665] b) Send an updated access and mobility management policy (AM policy) to the AMF, which carries the federated learning policy related to the access and mobility management function determined according to the federated learning parameters related to the above-mentioned business or application, and may include: a list of UE IDs or UEGroup IDs, areas, S-NSSAI, DNNs, etc. for which management of federated learning related to access management and / or mobility management functions is required to be terminated.

[0666] c) Send an updated session management policy (SM policy) to the SMF, which carries the federated learning policy related to the session management function determined according to the federated learning parameters related to the above-mentioned business or application, and may include: the Application ID, UE ID list or UEGroup ID, area, S-NSSAI, DNN and other information corresponding to the PDU session managed by the federated learning related to the session management function that requires termination.

[0667] Before sending the federated learning strategy, the PCF may query the NRF for information about the NFs (such as AMF and SMF) that have joined the federated learning group.

[0668] Step 105. The UE or NF that receives the request to terminate federated learning or leave the federated learning group sends a federated learning leave request (FL leave request) to the NWDAF, which carries the UE identifier (UE ID) or NF identifier (NF ID) and the FL group identifier requested to leave.

[0669] Step 106. Based on the received federated learning leave request of the UE or NF, the NWDAF removes the UE or NF from the FL group; specifically, it may be: deleting the UE information (e.g., UE ID) or NF information (e.g., NF ID) from the FL group context corresponding to the FL group identifier.

[0670] Step 107: The NWDAF sends a federated learning leave accept (FL leave accept) message, i.e., a FL leave response, to the UE or NF.

[0671] Step 108: Manage update operations.

[0672] (1) The NF sends an NF information update request to the NRF to delete the information of the FL group that the UE or NF has left (such as the FL group identifier and the identifier of the FL group management function (here, the NWDAF identifier)) from the NF profile.

[0673] (2) NRF feeds back the updated response to NF.

[0674] Specifically, 1. Steps 101 to 104 are Figure 7 Steps 71-74 are similar, except that:

[0675] 1) The service or application-related federated learning parameters provided by the AF include: Application ID, UE ID list or UE Group ID, region, S-NSSAI, DNN, application features, and other information requiring termination of federated learning.

[0676] 2) Based on the above-mentioned service or application-related federated learning parameters, PCF provides or updates the relevant federated learning strategy to UE or other NF (such as AMF, SMF), requiring UE or other NF to terminate the relevant federated learning or leave the corresponding federated learning group. For example:

[0677] a) Send an updated UE policy (UE policy) to the corresponding UE based on the UE ID list or UE Group ID, which carries the federated learning policy determined based on the federated learning parameters related to the above-mentioned service or application, and may include: Application ID, region, S-NSSAI, DNN, application characteristics, etc. that require termination of federated learning.

[0678] b) Send an updated access and mobility management policy (AM policy) to the AMF, which carries the federated learning policy related to the access and mobility management function determined according to the federated learning parameters related to the above-mentioned business or application, and may include: a list of UE IDs or UEGroup IDs, areas, S-NSSAI, DNNs, etc. for which management of federated learning related to access management and / or mobility management functions is required to be terminated.

[0679] c) Send an updated session management policy (SM policy) to the SMF, which carries the federated learning policy related to the session management function determined according to the federated learning parameters related to the above-mentioned business or application, and may include: the Application ID, UE ID list or UEGroup ID, area, S-NSSAI, DNN and other information corresponding to the PDU session managed by the federated learning related to the session management function that requires termination.

[0680] Before sending the federated learning policy, the PCF may query the NRF for information about the UE or other NFs (such as AMF, SMF) that have joined the federated learning group.

[0681] 2. Regarding steps 105 to 108: the UE or other NF that receives the request to terminate federated learning or leave the federated learning group initiates the process of leaving the FL group (i.e., the federated learning group leaving process); Figure 6 Steps 61-64 are similar.

[0682] The second type, specifically, can be Figure 11 As shown, this program includes:

[0683] Step 111. The AF initiates a service parameter generation or update request (i.e., Nnef_ServiceParameter_Create or Nnef_ServiceParameter_Update) to the NEF to generate or update service parameters, including service or application-related federated learning parameters, such as: Application ID, UE ID list or UEGroup ID, region, S-NSSAI, DNN, application features, and other information required to terminate federated learning.

[0684] Step 112: The NEF stores the service parameters provided by the AF in the UDR, including the federated learning parameters related to the service or application. That is, the UDR stores the FL related parameters.

[0685] Step 113. The UDR sends a data management notification Nudr_DM_Notify to the PCF, which includes the federated learning parameters related to the above business or application.

[0686] Step 114: The PCF provides or updates the federated learning strategy to the NWDAF, which carries the federated learning parameters related to the above services or applications.

[0687] Step 115. NWDAF initiates a request to terminate federated learning or leave the federated learning group to the UE or other NF (such as AMF, SMF) based on the received federated learning policy (other NF refers to NWDAF). For example:

[0688] a) The NWDAF sends a request to terminate federated learning or leave the federated learning group to the corresponding UE based on the UE ID list or UE Group ID. The request carries information such as the Application ID, region, S-NSSAI, DNN, and application features of the federated learning, instructing the UE to leave the federated learning group of the relevant service or application.

[0689] b) NWDAF sends a request to AMF to terminate federated learning or leave the federated learning group, which carries the UEID list or UE Group ID, area, S-NSSAI, DNN and other information of federated learning, instructing AMF to leave the related access management and / or mobility management federated learning group;

[0690] c) NWDAF sends a request to SMF to terminate federated learning or leave the federated learning group, which carries the Application ID, UE ID list or UE Group ID, area, S-NSSAI, DNN and other information corresponding to the PDU session of federated learning, instructing SMF to leave the federated learning group managed by the relevant session.

[0691] Step 116. The UE or NF that receives the request to terminate federated learning or leave the federated learning group sends a federated learning leave request (FL leave request) to the NWDAF, which carries the UE identifier (UE ID) or NF identifier (NF ID) and the FL group identifier requested to leave.

[0692] Step 117. Based on the received federated learning leave request of the UE or NF, the NWDAF removes the UE or NF from the FL group; specifically, it may be: deleting the UE information (e.g., UE ID) or NF information (e.g., NF ID) from the FL group context corresponding to the FL group identifier.

[0693] Step 118: The NWDAF sends a federated learning leave accept (FL leave accept) message, i.e., a FL leave response, to the UE or NF.

[0694] Step 119. Manage update operations;

[0695] (1) The NF sends an NF information update request to the NRF, deleting the information of the FL group that the NF has left (such as the FL group ID and the ID of the FL group management function (here, the NWDAF ID)) from the NF profile.

[0696] (2) NRF feeds back the updated response to NF.

[0697] Specifically, 1. Steps 111 to 115 are Figure 8 Steps 81-85 are similar, except that:

[0698] 1) The service or application-related federated learning parameters provided by the AF include: Application ID, UE ID list or UE Group ID, region, S-NSSAI, DNN, application features, and other information requiring termination of federated learning.

[0699] 2) Based on the received federated learning policy, NWDAF initiates a request to terminate federated learning or leave the federated learning group to the UE or other NFs (such as AMF, SMF). For example:

[0700] a) The NWDAF sends a request to terminate federated learning or leave the federated learning group to the corresponding UE based on the UE ID list or UE Group ID. The request carries information such as the Application ID, region, S-NSSAI, DNN, and application features of the federated learning, instructing the UE to leave the federated learning group of the relevant service or application.

[0701] b) NWDAF sends a request to AMF to terminate federated learning or leave the federated learning group, which carries the UEID list or UE Group ID, area, S-NSSAI, DNN and other information of federated learning, instructing AMF to leave the related access management and / or mobility management federated learning group;

[0702] c) NWDAF sends a request to SMF to terminate federated learning or leave the federated learning group, which carries the Application ID, UE ID list or UE Group ID, area, S-NSSAI, DNN and other information corresponding to the PDU session of federated learning, instructing SMF to leave the federated learning group managed by the relevant session.

[0703] 2. Regarding steps 116 to 119: the UE or other NF that receives the request to terminate federated learning or leave the federated learning group initiates the process of leaving the FL group (i.e., the federated learning group leaving process); Figure 6 Steps 61-64 are similar.

[0704] The third type, specifically, can be as follows Figure 12 As shown, this program includes:

[0705] Step 121. The AF initiates a request to the NEF to terminate federated learning or leave the federated learning group, which includes service or application-related federated learning parameters, such as: Application ID for which federated learning termination is required, UE ID list or UEGroup ID, region, S-NSSAI, DNN, application features, and other information.

[0706] Step 122. NEF performs an authorization check on the request from AF based on the service agreement, operator policy, etc. If the request is authorized, NEF queries the relevant NWDAF and sends a request to terminate federated learning or leave the federated learning group to the NWDAF, which carries the federated learning parameters related to the above service or application. The process of NEF querying NWDAF is similar to Figure 5 Step 52 (only when the initiator is NEF), wherein the FL feature information includes the above-mentioned service or application-related federated learning parameters.

[0707] Step 123: NWDAF initiates a request to terminate federated learning or leave the federated learning group to the UE or other NF (such as AMF, SMF) based on the received request to terminate federated learning or leave the federated learning group (other NF refers to NWDAF). For example:

[0708] a) The NWDAF sends a request to terminate federated learning or leave the federated learning group to the corresponding UE based on the UE ID list or UE Group ID. The request carries information such as the Application ID, region, S-NSSAI, DNN, and application features of the federated learning, instructing the UE to leave the federated learning group of the relevant service or application.

[0709] b) NWDAF sends a request to AMF to terminate federated learning or leave the federated learning group, which carries the UEID list or UE Group ID, area, S-NSSAI, DNN and other information of federated learning, instructing AMF to leave the related access management and / or mobility management federated learning group;

[0710] c) NWDAF sends a request to SMF to terminate federated learning or leave the federated learning group, which carries the Application ID, UE ID list or UE Group ID, area, S-NSSAI, DNN and other information corresponding to the PDU session of federated learning, instructing SMF to leave the federated learning group managed by the relevant session.

[0711] Step 124. The UE or NF that receives the request to terminate federated learning or leave the federated learning group sends a federated learning leave request (FL leave request) to the NWDAF, which carries the UE identifier (UE ID) or NF identifier (NF ID) and the FL group identifier requested to leave.

[0712] Step 125. Based on the received federated learning leave request of the UE or NF, the NWDAF removes the UE or NF from the FL group; specifically, it may be: deleting the UE information (e.g., UE ID) or NF information (e.g., NF ID) from the FL group context corresponding to the FL group identifier.

[0713] Step 126: The NWDAF sends a federated learning leave accept (FL leave accept) message, i.e., a FL leave response, to the UE or NF.

[0714] Step 127. Manage update operations;

[0715] (1) The NF sends an NF information update request to the NRF to delete the information of the FL group that the NF has left (such as the FL group ID and the ID of the FL group management function) from the NF profile.

[0716] (2) NRF feeds back the updated response to NF.

[0717] Specifically, 1. Steps 11 to 112 and Figure 9 Steps 91-92 are similar, except that the service or application-related federated learning parameters provided by the AF include: Application ID, UE ID list or UEGroup ID, region, S-NSSAI, DNN, application features, and other information requiring termination of federated learning.

[0718] 2. Based on the received request to terminate federated learning or leave the federated learning group, NWDAF initiates a request to terminate federated learning or leave the federated learning group to the UE or other NF (such as AMF, SMF). For example:

[0719] a) The NWDAF sends a request to terminate federated learning or leave the federated learning group to the corresponding UE based on the UE ID list or UE Group ID. The request carries information such as the Application ID, region, S-NSSAI, DNN, and application features of the federated learning, instructing the UE to leave the federated learning group of the relevant service or application.

[0720] b) NWDAF sends a request to AMF to terminate federated learning or leave the federated learning group, which carries the UEID list or UE Group ID, area, S-NSSAI, DNN and other information of federated learning, instructing AMF to leave the related access management and / or mobility management federated learning group;

[0721] c) NWDAF sends a request to SMF to terminate federated learning or leave the federated learning group, which carries the Application ID, UE ID list or UE Group ID, area, S-NSSAI, DNN and other information corresponding to the PDU session of federated learning, instructing SMF to leave the federated learning group managed by the relevant session.

[0722] 3. Regarding steps 124 to 127: the UE or other NF that receives the request to terminate federated learning or leave the federated learning group initiates the process of leaving the FL group (i.e., the federated learning group leaving process); Figure 6 Steps 61-64 are similar.

[0723] The fourth type, specifically, can be as follows Figure 13 As shown, this program includes:

[0724] Step 131. The NWDAF determines to initiate a request to terminate federated learning or leave the federated learning group to the UE or other NF (e.g., AMF, SMF) based on external trigger conditions or local policies. For example, the NWDAF detects one of the following conditions:

[0725] 1) The UE leaves the FL area, or the NF status changes, for example, the NF no longer serves the FL area;

[0726] 2) UEs or other NFs with abnormal behaviors (e.g., frequent connection establishment (corresponding to the frequency of connection establishment being higher than the first threshold), or the amount of transmitted data exceeding a threshold (corresponding to the amount of transmitted data exceeding the second threshold));

[0727] 3) The load of NWDAF for a specific FL is large.

[0728] That is, it is detected that the UE leaves the FL area or the NF no longer serves the FL area, or the UE or other NF has abnormal behavior, or due to load balancing and other reasons, it is determined that the UE or other NF needs to terminate federated learning.

[0729] Step 132. The NWDAF sends a request to terminate federated learning or leave the federated learning group to the relevant UE or other NF (e.g., AMF, SMF); including sending a request to terminate federated learning or leave the federated learning group to the following UE or other NF:

[0730] 1) The UE leaves the FL area, or the NF no longer serves the FL area.

[0731] 2) UE or other NF with abnormal behavior.

[0732] 3) UEs or other NFs with lower priorities in the FL group with heavy load.

[0733] Step 133. The UE or NF that receives the request to terminate federated learning or leave the federated learning group sends a federated learning leave request (FL leave request) to the NWDAF, which carries the UE identifier (UE ID) or NF identifier (NF ID) and the FL group identifier requested to leave.

[0734] Step 134. Based on the received federated learning leave request of the UE or NF, the NWDAF removes the UE or NF from the FL group; specifically, it may be: deleting the UE information (e.g., UE ID) or NF information (e.g., NF ID) from the FL group context corresponding to the FL group identifier.

[0735] Step 135: The NWDAF sends a federated learning leave accept (FL leave accept) message, i.e., a FL leave response, to the UE or NF.

[0736] Step 136. Manage update operations;

[0737] (1) The NF sends an NF information update request to the NRF, deleting the information of the FL group that the NF has left (such as the FL group ID and the ID of the FL group management function (here, the NWDAF ID)) from the NF profile.

[0738] (2) NRF feeds back the updated response to NF.

[0739] Specifically, regarding steps 133 to 136: the UE or other NF that receives the request to terminate federated learning or leave the federated learning group initiates the process of leaving the FL group (i.e., the federated learning group leaving process); Figure 6 Steps 61-64 are similar.

[0740] It is noted here that in some of the drawings involved in the above examples, only the NF is simplified for the UE or other NF, and the NF in the drawings actually corresponds to the above UE or other NF.

[0741] Based on the above, the solutions provided in the embodiments of this application mainly involve:

[0742] 1) The network function (NF) sends a federated learning join or leave request to the federated learning management function (NWDAF, for example). The request carries the NF ID, the FL group ID (or FL feature information) to join or leave the corresponding FL group. The federated learning management function establishes and maintains the federated learning group context.

[0743] 2) The AF provides service or application-related federated learning parameters, and the PCF generates the corresponding federated learning policy based on these parameters. The network function NF joins or leaves the corresponding federated learning group based on the federated learning policy provided by the PCF.

[0744] 3) The federated learning management function (taking NWDAF as an example) requests a specific NF to join or leave the corresponding federated learning group based on the service or application-related federated learning parameters provided by AF, or the federated learning policy provided by PCF, or other trigger conditions or local policies.

[0745] In summary, the solution provided by the embodiment of the present application can realize the dynamic management of the federated learning group in the mobile communication network (including the establishment or termination of the federated learning group (specifically, when the member joining the FL group is the first member, it is the establishment of the FL group; when the last member removes the FL group, it is the termination of the FL group), the joining or leaving of members, etc.), which can better meet the needs of network and business intelligence and improve the intelligent performance of communications and applications.

[0746] The embodiment of the present application further provides a functional entity, which is a first functional entity, such as Figure 15As shown, the functional entity includes a memory 151, a transceiver 152, and a processor 153:

[0747] The memory 151 is used to store computer programs; the transceiver 152 is used to send and receive data under the control of the processor 153; the processor 153 is used to read the computer program in the memory 151 and perform the following operations:

[0748] Get the feature information of the federated learning FL group;

[0749] determining a second functional entity according to the characteristic information of the FL group;

[0750] Add the second functional entity to the FL group.

[0751] The functional entity provided in the embodiment of the present application obtains characteristic information of a federated learning FL group; determines a second functional entity based on the characteristic information of the FL group; and adds the second functional entity to the FL group; thereby enabling dynamic processing of the FL group, thereby better meeting the requirements of network and business intelligence and improving the intelligent performance of communications and applications;.

[0752] Specifically, the transceiver 152 is configured to receive and send data under the control of the processor 153 .

[0753] Among them, Figure 15 In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically linking together various circuits of one or more processors represented by processor 153 and memory represented by memory 151. The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are all well known in the art and, therefore, will not be described further herein. The bus interface provides an interface. The transceiver 152 may be a plurality of components, i.e., a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium, such as a wireless channel, a wired channel, an optical cable, and the like. The processor 153 is responsible for managing the bus architecture and general processing, and the memory 151 may store data used by the processor 153 when performing operations.

[0754] The processor 153 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or a complex programmable logic device (CPLD). The processor may also adopt a multi-core architecture.

[0755] The obtaining of characteristic information of the federated learning FL group includes: receiving, through the transceiver, a first request sent by an application function entity or a network open function entity, and obtaining the characteristic information of the FL group according to application information or service information related to the FL group carried in the first request; or receiving, through the transceiver, a second request sent by a policy control function entity, and obtaining the characteristic information of the FL group according to a control policy related to the FL group carried in the second request.

[0756] In an embodiment of the present application, the characteristic information of the FL group includes first information, and / or identification information of functional entities allowed to join the FL group; the first information includes any one or more of the following: information on the application or service corresponding to the FL group; area information corresponding to the FL group; network slice information corresponding to the FL group; and data network information corresponding to the FL group.

[0757] The determining of the second functional entity based on the characteristic information of the FL group includes: determining a second functional entity that meets a first condition based on the characteristic information of the FL group; wherein the first condition includes one or a combination of the following conditions: supporting the application or service corresponding to the FL group; serving the data network, network slice or area corresponding to the FL group; accessing the data network or network slice corresponding to the FL group; being located within the area corresponding to the FL group; and the identification information being included in the identification information of the functional entity allowed to join the FL group.

[0758] In this embodiment of the present application, adding the second functional entity to the FL group includes: sending, through the transceiver, a first FL join request to the second functional entity, where the first FL join request is used to request the second functional entity to join the FL group, and the first FL join request carries information about the first functional entity and first information about the FL group; and receiving, through the transceiver, a first FL join response sent by the second functional entity.

[0759] The determining the second functional entity based on the characteristic information of the FL group specifically includes: receiving, through the transceiver, a second FL join request sent by the second functional entity, where the second FL join request is used to request joining the FL group, and the second FL join request carries the first information of the FL group and identification information of the second functional entity; determining, based on the first information, a FL group to be joined by the second functional entity; determining whether the identification information of the second functional entity is included in the identification information of the functional entities allowed to join the FL group; and if so, determining that the second functional entity is the second functional entity corresponding to the characteristic information of the FL group.

[0760] Furthermore, the processor is also used to: use the memory to store second information of the FL group, wherein the second information includes at least one of the following: information of the first functional entity; information of the second functional entity; identification information of the FL group; information of the application or service corresponding to the FL group; network slice information corresponding to the FL group; data network information corresponding to the FL group; information of the machine learning ML model of the FL group.

[0761] The information of the first functional entity includes: at least one of the identification information, address information and location information of the first functional entity; and / or the information of the second functional entity includes: at least one of the identification information, address information and location information of the second functional entity.

[0762] Furthermore, the processor is also used to: after adding the second functional entity to the FL group, determine whether the second functional entity meets the conditions for leaving the FL group; if the second functional entity meets the conditions for leaving the FL group, remove the second functional entity from the FL group.

[0763] Among them, the second functional entity meets the conditions for leaving the FL group, including at least one of the following situations: the second functional entity no longer supports the application or service corresponding to the FL group; the second functional entity no longer serves the data network, network slice or area corresponding to the FL group; the second functional entity no longer accesses the data network or network slice corresponding to the FL group; the second functional entity leaves the area corresponding to the FL group; the identification information of the second functional entity is no longer included in the identification information of functional entities allowed to join the FL group; a first FL leave request sent by the second functional entity is received, where the first FL leave request is used to request to leave the FL group; the second functional entity has abnormal behavior, and the abnormal behavior includes: the frequency of establishing connections between the second functional entity and the first functional entity is higher than a first threshold, or the amount of data transmitted between the second functional entity and the first functional entity exceeds a second threshold.

[0764] In an embodiment of the present application, determining whether the second functional entity meets the conditions for leaving the FL group includes: receiving, through the transceiver, updated characteristic information of the FL group sent by the third functional entity; and determining, based on the updated characteristic information of the FL group, whether the second functional entity meets the conditions for leaving the FL group.

[0765] The removing the second functional entity from the FL group includes: sending a second FL leave request to the second functional entity through the transceiver, where the second FL leave request is used to request the second functional entity to leave the FL group; and receiving a second FL leave response sent by the second functional entity through the transceiver.

[0766] In the embodiment of the present application, removing the second functional entity from the FL group includes: deleting information of the second functional entity from second information of the FL group.

[0767] It should be noted here that the above-mentioned functional entity provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned first functional entity side method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be described in detail here.

[0768] The embodiment of the present application further provides a functional entity, which is a second functional entity, such as Figure 16 and Figure 17 As shown, the functional entity includes a memory 161, a transceiver 162, and a processor 163:

[0769] The memory 161 is used to store computer programs; the transceiver 162 is used to send and receive data under the control of the processor 163; the processor 163 is used to read the computer program in the memory 161 and perform the following operations:

[0770] Get the feature information of the federated learning FL group;

[0771] determining, based on the characteristic information of the FL group, whether a condition for joining the FL group is met;

[0772] If the conditions for joining the FL group are met, the user joins the FL group.

[0773] The functional entity provided in the embodiment of the present application obtains characteristic information of a federated learning FL group; determines whether conditions for joining the FL group are met based on the characteristic information of the FL group; and joins the FL group if the conditions for joining the FL group are met. This enables dynamic processing of the FL group, thereby better meeting the requirements of network and service intelligence and improving the intelligent performance of communications and applications.

[0774] The second functional entity in the embodiment of the present application can be implemented as a terminal (device), specifically as follows Figure 16 As shown, the transceiver 162 is used to receive and send data under the control of the processor 163.

[0775] Among them, Figure 16In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically one or more processors represented by processor 163 and various circuits of memory represented by memory 161 are linked together. The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. The transceiver 162 may be a plurality of components, i.e., a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium, such as a wireless channel, a wired channel, an optical cable, and the like. For different user devices, the user interface 164 may also be an interface capable of connecting external or internal devices as required, and the connected devices include but are not limited to a keypad, a display, a speaker, a microphone, a joystick, and the like.

[0776] The processor 163 is responsible for managing the bus architecture and general processing, and the memory 161 can store data used by the processor 163 when performing operations.

[0777] Optionally, the processor 163 may be a CPU (central processing unit), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array) or a CPLD (Complex Programmable Logic Device), and the processor may also adopt a multi-core architecture.

[0778] The processor calls the computer program stored in the memory to execute any of the methods provided in the embodiments of the present application according to the obtained executable instructions. The processor and the memory can also be arranged physically separately.

[0779] In addition, the second functional entity in the embodiment of the present application can also be implemented as a network (side) device, specifically as follows Figure 17 As shown, the transceiver 162 is used to receive and send data under the control of the processor 163.

[0780] Among them, Figure 17In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically linking together various circuits of one or more processors represented by processor 163 and memory represented by memory 161. The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are all well known in the art and, therefore, will not be described further herein. The bus interface provides an interface. The transceiver 162 may be a plurality of components, i.e., a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium, such as a wireless channel, a wired channel, an optical cable, and the like. The processor 163 is responsible for managing the bus architecture and general processing, and the memory 161 may store data used by the processor 163 when performing operations.

[0781] The processor 163 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or a complex programmable logic device (CPLD). The processor may also adopt a multi-core architecture.

[0782] In an embodiment of the present application, obtaining characteristic information of the federated learning FL group includes: receiving, by the transceiver, a third request sent by an application function entity or a network open function entity, and obtaining the characteristic information of the FL group based on application information or service information related to the FL group carried in the third request; or receiving, by the transceiver, a fourth request sent by a policy control function entity, and obtaining the characteristic information of the FL group based on a control policy related to the FL group carried in the fourth request.

[0783] The characteristic information of the FL group includes first information and / or identification information of functional entities allowed to join the FL group; the first information includes any one or more of the following: information on the application or service corresponding to the FL group; area information corresponding to the FL group; network slice information corresponding to the FL group; and data network information corresponding to the FL group.

[0784] In an embodiment of the present application, determining whether the conditions for joining the FL group are met according to the characteristic information of the FL group includes: determining whether the second functional entity meets the second condition according to the characteristic information of the FL group; wherein the second condition includes one or a combination of the following conditions: supporting the application or service corresponding to the FL group; serving the data network, network slice or area corresponding to the FL group; accessing the data network or network slice corresponding to the FL group; being located within the area corresponding to the FL group; and the identification information is included in the identification information of the functional entity allowed to join the FL group.

[0785] Furthermore, the processor is also used to: before joining the FL group, send a first query request to the fourth functional entity through the transceiver, the first query request carrying the third information of the FL group; receive a first query response sent by the fourth functional entity through the transceiver, the first query response carrying information of at least one first functional entity that meets the third information; based on the information of the at least one first functional entity, determine a first functional entity from the at least one first functional entity to establish the FL group with the second functional entity; wherein the third information of the FL group includes at least one of the following: information on the application or service corresponding to the FL group; information on the network slice corresponding to the FL group; information on the data network corresponding to the FL group; information on the machine learning ML model of the FL group.

[0786] The information of the first functional entity includes: at least one of the address information, location information and service area information of the first functional entity; determining a first functional entity to establish the FL group with the second functional entity from the at least one first functional entity based on the information of the at least one first functional entity includes: selecting a first functional entity that meets a third condition from the at least one first functional entity based on the information of the at least one first functional entity as the first functional entity to establish the FL group with the second functional entity; wherein the third condition includes at least one of the following: belonging to the same local network as the second functional entity; serving the area where the second functional entity is located; and being the shortest distance to the second functional entity among the at least one first functional entity.

[0787] In an embodiment of the present application, joining the FL group includes: sending, through the transceiver, a second FL join request to the first functional entity, where the second FL join request is used to request joining the FL group, and the second FL join request carries the first information and information of the second functional entity; and receiving, through the transceiver, a second FL join response sent by the first functional entity.

[0788] Furthermore, the processor is also used to: use the memory to store second information of the FL group, wherein the second information includes at least one of the following: information of the first functional entity; information of the second functional entity; identification information of the FL group; information of the application or service corresponding to the FL group; network slice information corresponding to the FL group; data network information corresponding to the FL group; information of the machine learning ML model of the FL group.

[0789] The information of the first functional entity includes: at least one of the identification information, address information and location information of the first functional entity; and / or the information of the second functional entity includes: at least one of the identification information, address information and location information of the second functional entity.

[0790] Furthermore, the processor is further configured to: after joining the FL group, determine whether the second functional entity meets a condition for leaving the FL group; and if the second functional entity meets the condition for leaving the FL group, leave the FL group; wherein leaving the FL group includes: sending a first FL leave request to the first functional entity, where the first FL leave request is used to request to leave the FL group.

[0791] Among them, the second functional entity meets the conditions for leaving the FL group, including at least one of the following situations: the second functional entity no longer supports the application or service corresponding to the FL group; the second functional entity no longer serves the data network, network slice or area corresponding to the FL group; the second functional entity no longer accesses the data network or network slice corresponding to the FL group; the second functional entity leaves the area corresponding to the FL group.

[0792] In an embodiment of the present application, determining whether the second functional entity meets the conditions for leaving the FL group includes: receiving, through the transceiver, updated characteristic information of the FL group sent by the third functional entity; and determining, based on the updated characteristic information of the FL group, whether the second functional entity meets the conditions for leaving the FL group.

[0793] Furthermore, the leaving the FL group further includes: deleting second information of the FL group stored in the second functional entity.

[0794] It should be noted here that the above-mentioned functional entity provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned second functional entity side method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be described in detail here.

[0795] The embodiment of the present application further provides a functional entity, which is a third functional entity, such as Figure 18 As shown, the functional entity includes a memory 181, a transceiver 182, and a processor 183:

[0796] The memory 181 is used to store computer programs; the transceiver 182 is used to send and receive data under the control of the processor 183; the processor 183 is used to read the computer program in the memory 181 and perform the following operations:

[0797] Sending a first request or a second request to the first functional entity through the transceiver 182;

[0798] The first request carries application information or service information related to the federated learning FL group, and the second request carries a control policy related to the federated learning FL group; so that the first functional entity creates or updates the FL group according to the application information, service information or control policy related to the FL group.

[0799] The functional entity provided in the embodiment of the present application sends a first request or a second request to a first functional entity; wherein the first request carries application information or service information related to the federated learning FL group, and the second request carries a control policy related to the federated learning FL group; so that the first functional entity creates or updates the FL group based on the application information, service information, or control policy related to the FL group. Dynamic processing of the FL group can be implemented, thereby better meeting the requirements of network and service intelligence and improving the intelligent performance of communications and applications.

[0800] Specifically, the transceiver 182 is used to receive and send data under the control of the processor 183.

[0801] Among them, Figure 18 In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically linking together various circuits of one or more processors represented by processor 183 and memory represented by memory 181. The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are all well known in the art and, therefore, will not be described further herein. The bus interface provides an interface. The transceiver 182 may be a plurality of components, i.e., a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium, such as a wireless channel, a wired channel, an optical cable, and the like. The processor 183 is responsible for managing the bus architecture and general processing, and the memory 181 may store data used by the processor 183 when performing operations.

[0802] The processor 183 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or a complex programmable logic device (CPLD). The processor may also adopt a multi-core architecture.

[0803] The application information, service information or control policy related to the FL group includes: characteristic information of the FL group; the characteristic information of the FL group includes first information, and / or identification information of functional entities allowed to join the FL group; the first information includes any one or more of the following: information on the application or service corresponding to the FL group; area information corresponding to the FL group; network slice information corresponding to the FL group; data network information corresponding to the FL group.

[0804] It should be noted here that the above-mentioned functional entity provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned third functional entity side method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be described in detail here.

[0805] The embodiment of the present application also provides a federated learning group processing device, which is applied to a first functional entity, such as Figure 19 As shown, the device includes:

[0806] A first acquiring unit 191 is configured to acquire feature information of a federated learning FL group;

[0807] A first determining unit 192 is configured to determine a second functional entity according to the characteristic information of the FL group;

[0808] The first processing unit 193 is configured to add the second functional entity to the FL group.

[0809] The federated learning group processing device provided in the embodiment of the present application obtains characteristic information of a federated learning FL group; determines a second functional entity based on the characteristic information of the FL group; and adds the second functional entity to the FL group; thereby enabling dynamic processing of the FL group, thereby better meeting the requirements of network and business intelligence and improving the intelligent performance of communications and applications.

[0810] The obtaining of characteristic information of the federated learning FL group includes: receiving a first request sent by an application function entity or a network open function entity, and obtaining the characteristic information of the FL group according to application information or service information related to the FL group carried in the first request; or receiving a second request sent by a policy control function entity, and obtaining the characteristic information of the FL group according to a control policy related to the FL group carried in the second request.

[0811] In an embodiment of the present application, the characteristic information of the FL group includes first information, and / or identification information of functional entities allowed to join the FL group; the first information includes any one or more of the following: information on the application or service corresponding to the FL group; area information corresponding to the FL group; network slice information corresponding to the FL group; and data network information corresponding to the FL group.

[0812] The determining of the second functional entity based on the characteristic information of the FL group includes: determining a second functional entity that meets a first condition based on the characteristic information of the FL group; wherein the first condition includes one or a combination of the following conditions: supporting the application or service corresponding to the FL group; serving the data network, network slice or area corresponding to the FL group; accessing the data network or network slice corresponding to the FL group; being located within the area corresponding to the FL group; and the identification information being included in the identification information of the functional entity allowed to join the FL group.

[0813] In this embodiment of the present application, adding the second functional entity to the FL group includes: sending a first FL join request to the second functional entity, where the first FL join request is used to request the second functional entity to join the FL group, and the first FL join request carries information about the first functional entity and first information about the FL group; and receiving a first FL join response sent by the second functional entity.

[0814] The determining the second functional entity based on the characteristic information of the FL group specifically includes: receiving a second FL joining request sent by the second functional entity, where the second FL joining request is used to request joining the FL group, and the second FL joining request carries the first information of the FL group and identification information of the second functional entity; determining, based on the first information, a FL group to be joined by the second functional entity; determining whether the identification information of the second functional entity is included in the identification information of the functional entities allowed to join the FL group; and if so, determining that the second functional entity is the second functional entity corresponding to the characteristic information of the FL group.

[0815] Furthermore, the device also includes: a first storage unit, used to store second information of the FL group, wherein the second information includes at least one of the following: information of the first functional entity; information of the second functional entity; identification information of the FL group; information of the application or service corresponding to the FL group; network slice information corresponding to the FL group; data network information corresponding to the FL group; information of the machine learning ML model of the FL group.

[0816] The information of the first functional entity includes: at least one of the identification information, address information and location information of the first functional entity; and / or the information of the second functional entity includes: at least one of the identification information, address information and location information of the second functional entity.

[0817] Furthermore, the federated learning group processing device also includes: a second determination unit, used to determine whether the second functional entity meets the conditions for leaving the FL group after adding the second functional entity to the FL group; and a second processing unit, used to remove the second functional entity from the FL group if the second functional entity meets the conditions for leaving the FL group.

[0818] Among them, the second functional entity meets the conditions for leaving the FL group, including at least one of the following situations: the second functional entity no longer supports the application or service corresponding to the FL group; the second functional entity no longer serves the data network, network slice or area corresponding to the FL group; the second functional entity no longer accesses the data network or network slice corresponding to the FL group; the second functional entity leaves the area corresponding to the FL group; the identification information of the second functional entity is no longer included in the identification information of functional entities allowed to join the FL group; a first FL leave request sent by the second functional entity is received, where the first FL leave request is used to request to leave the FL group; the second functional entity has abnormal behavior, and the abnormal behavior includes: the frequency of establishing connections between the second functional entity and the first functional entity is higher than a first threshold, or the amount of data transmitted between the second functional entity and the first functional entity exceeds a second threshold.

[0819] In this embodiment of the present application, determining whether the second functional entity meets the conditions for leaving the FL group includes: receiving updated characteristic information of the FL group sent by a third functional entity; and determining, based on the updated characteristic information of the FL group, whether the second functional entity meets the conditions for leaving the FL group.

[0820] The removing the second functional entity from the FL group includes: sending a second FL leave request to the second functional entity, where the second FL leave request is used to request the second functional entity to leave the FL group; and receiving a second FL leave response sent by the second functional entity.

[0821] In the embodiment of the present application, removing the second functional entity from the FL group includes: deleting information of the second functional entity from second information of the FL group.

[0822] It should be noted here that the above-mentioned device provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned first functional entity side method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be described in detail here.

[0823] The embodiment of the present application also provides a federated learning group processing device, which is applied to a second functional entity, such as Figure 20 As shown, the device includes:

[0824] The second acquisition unit 201 is used to acquire feature information of the federated learning FL group;

[0825] a third determining unit 202, configured to determine whether a condition for joining the FL group is met based on the characteristic information of the FL group;

[0826] The third processing unit 203 is configured to join the FL group if the conditions for joining the FL group are met.

[0827] The federated learning group processing device provided in the embodiment of the present application obtains characteristic information of a federated learning FL group; determines whether conditions for joining the FL group are met based on the characteristic information of the FL group; and joins the FL group if the conditions for joining the FL group are met. This can achieve dynamic processing of the FL group, thereby better meeting the needs of network and business intelligence and improving the intelligent performance of communications and applications.

[0828] The obtaining of characteristic information of the federated learning FL group includes: receiving a third request sent by an application function entity or a network open function entity, and obtaining the characteristic information of the FL group according to application information or service information related to the FL group carried in the third request; or receiving a fourth request sent by a policy control function entity, and obtaining the characteristic information of the FL group according to a control policy related to the FL group carried in the fourth request.

[0829] In an embodiment of the present application, the characteristic information of the FL group includes first information, and / or identification information of functional entities allowed to join the FL group; the first information includes any one or more of the following: information on the application or service corresponding to the FL group; area information corresponding to the FL group; network slice information corresponding to the FL group; and data network information corresponding to the FL group.

[0830] The determining, based on the characteristic information of the FL group, whether the conditions for joining the FL group are met includes: determining, based on the characteristic information of the FL group, whether the second functional entity meets a second condition; wherein the second condition includes one or a combination of the following conditions: supporting the application or service corresponding to the FL group; serving the data network, network slice or area corresponding to the FL group; accessing the data network or network slice corresponding to the FL group; being located within the area corresponding to the FL group; and the identification information being included in the identification information of the functional entity allowed to join the FL group.

[0831] Furthermore, the federated learning group processing device also includes: a first sending unit, used to send a first query request to the fourth functional entity before joining the FL group, the first query request carrying the third information of the FL group; a first receiving unit, used to receive a first query response sent by the fourth functional entity, the first query response carrying information of at least one first functional entity that meets the third information; a fourth determination unit, used to determine a first functional entity from the at least one first functional entity to establish the FL group with the second functional entity based on the information of the at least one first functional entity; wherein the third information of the FL group includes at least one of the following: information on the application or service corresponding to the FL group; information on the network slice corresponding to the FL group; information on the data network corresponding to the FL group; information on the machine learning ML model of the FL group.

[0832] The information of the first functional entity includes: at least one of the address information, location information and service area information of the first functional entity; determining a first functional entity to establish the FL group with the second functional entity from the at least one first functional entity based on the information of the at least one first functional entity includes: selecting a first functional entity that meets a third condition from the at least one first functional entity based on the information of the at least one first functional entity as the first functional entity to establish the FL group with the second functional entity; wherein the third condition includes at least one of the following: belonging to the same local network as the second functional entity; serving the area where the second functional entity is located; and being the shortest distance to the second functional entity among the at least one first functional entity.

[0833] In the embodiment of the present application, joining the FL group includes: sending a second FL join request to the first functional entity, where the second FL join request is used to request joining the FL group, and the second FL join request carries the first information and information about the second functional entity; and receiving a second FL join response sent by the first functional entity.

[0834] Furthermore, the device also includes: a second storage unit, used to store second information of the FL group, wherein the second information includes at least one of the following: information of the first functional entity; information of the second functional entity; identification information of the FL group; information of the application or service corresponding to the FL group; network slice information corresponding to the FL group; data network information corresponding to the FL group; information of the machine learning ML model of the FL group.

[0835] The information of the first functional entity includes: at least one of the identification information, address information and location information of the first functional entity; and / or the information of the second functional entity includes: at least one of the identification information, address information and location information of the second functional entity.

[0836] Furthermore, the federated learning group processing device also includes: a fifth determination unit, configured to determine whether the second functional entity meets the conditions for leaving the FL group after joining the FL group; and a fourth processing unit, configured to leave the FL group if the second functional entity meets the conditions for leaving the FL group; wherein leaving the FL group includes: sending a first FL leave request to the first functional entity, wherein the first FL leave request is used to request to leave the FL group.

[0837] Among them, the second functional entity meets the conditions for leaving the FL group, including at least one of the following situations: the second functional entity no longer supports the application or service corresponding to the FL group; the second functional entity no longer serves the data network, network slice or area corresponding to the FL group; the second functional entity no longer accesses the data network or network slice corresponding to the FL group; the second functional entity leaves the area corresponding to the FL group.

[0838] In this embodiment of the present application, determining whether the second functional entity meets the conditions for leaving the FL group includes: receiving updated characteristic information of the FL group sent by a third functional entity; and determining, based on the updated characteristic information of the FL group, whether the second functional entity meets the conditions for leaving the FL group.

[0839] Furthermore, the leaving the FL group further includes: deleting second information of the FL group stored in the second functional entity.

[0840] It should be noted here that the above-mentioned device provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned second functional entity side method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be described in detail here.

[0841] The embodiment of the present application also provides a federated learning group processing device, which is applied to a third functional entity, such as Figure 21 As shown, the device includes:

[0842] The second sending unit 211 is configured to send the first request or the second request to the first functional entity;

[0843] The first request carries application information or service information related to the federated learning FL group, and the second request carries a control policy related to the federated learning FL group; so that the first functional entity creates or updates the FL group according to the application information, service information or control policy related to the FL group.

[0844] The federated learning group processing device provided in an embodiment of the present application sends a first request or a second request to a first functional entity; wherein the first request carries application information or service information related to the federated learning FL group, and the second request carries a control policy related to the federated learning FL group; so that the first functional entity creates or updates the FL group based on the application information, service information, or control policy related to the FL group. Dynamic processing of the FL group can be achieved, thereby better meeting the needs of network and service intelligence and improving the intelligent performance of communications and applications.

[0845] The application information, service information or control policy related to the FL group includes: characteristic information of the FL group; the characteristic information of the FL group includes first information, and / or identification information of functional entities allowed to join the FL group; the first information includes any one or more of the following: information on the application or service corresponding to the FL group; area information corresponding to the FL group; network slice information corresponding to the FL group; data network information corresponding to the FL group.

[0846] It should be noted here that the above-mentioned device provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned third functional entity side method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be described in detail here.

[0847] It should be noted that the division of units in the embodiments of the present application is schematic and is merely a logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0848] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a processor-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0849] An embodiment of the present application also provides a processor-readable storage medium, which stores a computer program, and the computer program is used to enable the processor to execute the above-mentioned federated learning group processing method on the first functional entity side; or, the computer program is used to enable the processor to execute the above-mentioned federated learning group processing method on the second functional entity side; or, the computer program is used to enable the processor to execute the above-mentioned federated learning group processing method on the third functional entity side.

[0850] The processor-readable storage medium can be any available medium or data storage device that can be accessed by the processor, including but not limited to magnetic storage (such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO)), optical storage (such as CDs, DVDs, BDs, HVDs, etc.), and semiconductor storage (such as ROMs, EPROMs, EEPROMs, non-volatile memories (NANDFLASH), solid-state drives (SSDs)), etc.

[0851] Among them, the implementation embodiments of the above-mentioned federated learning group processing method on the first functional entity side, the second functional entity side or the third functional entity side are all applicable to the embodiments of the processor-readable storage medium, and can also achieve the same technical effect.

[0852] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage and optical storage, etc.) that contain computer-usable program code.

[0853] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer-executable instructions. These computer-executable instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0854] These processor-executable instructions may also be stored in a processor-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the processor-readable memory produce an article of manufacture comprising an instruction device that implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0855] These processor-executable instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are performed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable device to implement the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0856] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A federated learning group processing method, characterized in that: Applied to the first functional entity, the method includes: Get the feature information of the federated learning FL group; determining a second functional entity according to the characteristic information of the FL group; adding the second functional entity to the FL group; The characteristic information of the FL group includes the first information; or the characteristic information of the FL group includes the first information and identification information of a functional entity allowed to join the FL group; The first information includes any one or more of the following: Information about the application or service corresponding to the FL group; Area information corresponding to the FL group; Network slice information corresponding to the FL group; The data network information corresponding to the FL group.

2. The method for processing a federated learning group according to claim 1, wherein: The obtaining of characteristic information of the federated learning FL group includes: receiving a first request sent by an application function entity or a network open function entity, and acquiring characteristic information of the FL group according to application information or service information related to the FL group carried in the first request; or, A second request sent by the policy control function entity is received, and characteristic information of the FL group is acquired according to the control policy related to the FL group carried in the second request.

3. The method for processing a federated learning group according to claim 1, wherein: The determining the second functional entity according to the characteristic information of the FL group includes: determining, according to the characteristic information of the FL group, a second functional entity that meets a first condition; The first condition includes one or a combination of the following conditions: Support the applications or services corresponding to the FL group; Serving the data network, network slice or area corresponding to the FL group; Access the data network or network slice corresponding to the FL group; The position is within the region corresponding to the FL group; The identification information is included in the identification information of the functional entity allowed to join the FL group.

4. The method for processing a federated learning group according to claim 1, wherein: The adding the second functional entity to the FL group includes: Sending a first FL join request to the second functional entity, where the first FL join request is used to request the second functional entity to join the FL group, and the first FL join request carries information about the first functional entity and first information about the FL group; Receive a first FL joining response sent by the second functional entity.

5. The method for processing a federated learning group according to claim 1, wherein: The determining the second functional entity according to the characteristic information of the FL group specifically includes: receiving a second FL join request sent by the second functional entity, where the second FL join request is used to request to join the FL group, and the second FL join request carries the first information of the FL group and identification information of the second functional entity; determining, according to the first information, an FL group to be joined by the second functional entity; determining whether the identification information of the second functional entity is included in the identification information of the functional entities allowed to join the FL group; If so, it is determined that the second functional entity is the second functional entity corresponding to the characteristic information of the FL group.

6. The method for processing a federated learning group according to claim 1, wherein: The method further comprises: Storing second information of the FL group, wherein the second information includes at least one of the following: information of the first functional entity; information of the second functional entity; identification information of the FL group; Information about the application or service corresponding to the FL group; Network slice information corresponding to the FL group; Data network information corresponding to the FL group; Information about the machine learning (ML) model of the FL group.

7. The federated learning group processing method according to claim 4 or 6, characterized in that: The information of the first functional entity includes: at least one of identification information, address information and location information of the first functional entity; The information of the second functional entity includes: at least one of identification information, address information and location information of the second functional entity.

8. The method for processing a federated learning group according to claim 1 or 6, wherein: The method further comprises: determining whether the second functional entity meets a condition for leaving the FL group; If the second functional entity meets a condition for leaving the FL group, the second functional entity is removed from the FL group.

9. The method for processing a federated learning group according to claim 8, wherein: The second functional entity satisfies a condition for leaving the FL group, including at least one of the following situations: The second functional entity no longer supports the application or service corresponding to the FL group; The second functional entity no longer serves the data network, network slice or area corresponding to the FL group; The second functional entity no longer accesses the data network or network slice corresponding to the FL group; The second functional entity leaves the area corresponding to the FL group; The identification information of the second functional entity is no longer included in the identification information of functional entities allowed to join the FL group; receiving a first FL leave request sent by the second functional entity, where the first FL leave request is used to request leaving the FL group; The second functional entity has abnormal behavior, and the abnormal behavior includes: the frequency of establishing a connection between the second functional entity and the first functional entity is higher than a first threshold, or the amount of data transmitted between the second functional entity and the first functional entity exceeds a second threshold.

10. The method for processing a federated learning group according to claim 8, wherein: The method further comprises: receiving updated feature information of the FL group sent by the third functional entity; The determining whether the second functional entity meets the condition of leaving the FL group includes: Determine, according to the updated characteristic information of the FL group, whether the second functional entity meets a condition for leaving the FL group.

11. The method for processing a federated learning group according to claim 8, wherein: The removing the second functional entity from the FL group includes: sending a second FL leave request to the second functional entity, where the second FL leave request is used to request the second functional entity to leave the FL group; Receive a second FL leave response sent by the second functional entity.

12. The method for processing a federated learning group according to claim 8, wherein: The removing the second functional entity from the FL group includes: The information of the second functional entity is deleted from the second information of the FL group.

13. A method for processing a federated learning group, characterized in that: Applied to the second functional entity, the method includes: Get the feature information of the federated learning FL group; determining, based on the characteristic information of the FL group, whether a condition for joining the FL group is met; If the conditions for joining the FL group are met, join the FL group; The characteristic information of the FL group includes the first information; or the characteristic information of the FL group includes the first information and identification information of a functional entity allowed to join the FL group; The first information includes any one or more of the following: Information about the application or service corresponding to the FL group; Area information corresponding to the FL group; Network slice information corresponding to the FL group; The data network information corresponding to the FL group.

14. The method for processing a federated learning group according to claim 13, wherein: The obtaining of characteristic information of the federated learning FL group includes: receiving a third request sent by the application function entity or the network open function entity, and acquiring characteristic information of the FL group according to application information or service information related to the FL group carried in the third request; or, The fourth request sent by the policy control function entity is received, and characteristic information of the FL group is acquired according to the control policy related to the FL group carried in the fourth request.

15. The method for processing a federated learning group according to claim 13, wherein: Before joining the FL group, also include: Sending a first query request to a fourth functional entity, where the first query request carries the third information of the FL group; receiving a first query response sent by the fourth functional entity, where the first query response carries information of at least one first functional entity that satisfies the third information; determining, from the at least one first functional entity, a first functional entity to establish the FL group with the second functional entity; The third information of the FL group includes at least one of the following: Information about the application or service corresponding to the FL group; Network slice information corresponding to the FL group; Data network information corresponding to the FL group; Information about the machine learning (ML) model of the FL group.

16. The method for processing a federated learning group according to claim 15, wherein: The information of the first functional entity includes: at least one of address information, location information, and service area information of the first functional entity; The determining, from the at least one first functional entity, a first functional entity to establish the FL group with the second functional entity according to the information of the at least one first functional entity includes: selecting, according to the information of the at least one first functional entity, a first functional entity that meets a third condition from the at least one first functional entity as the first functional entity for establishing the FL group with the second functional entity; The third condition includes at least one of the following: Belonging to the same local network as the second functional entity; Serving the area where the second functional entity is located; Among the at least one first functional entity, the distance between the second functional entity and the first functional entity is the shortest.

17. The method for processing a federated learning group according to claim 13, wherein: The method further comprises: determining whether the second functional entity meets a condition for leaving the FL group; In a case where the second functional entity meets the condition for leaving the FL group, the second functional entity sends a first FL leave request to the first functional entity, where the first FL leave request is used to request to leave the FL group.

18. The federated learning group processing method according to claim 17, characterized in that: The second functional entity satisfies a condition for leaving the FL group, including at least one of the following situations: The second functional entity no longer supports the application or service corresponding to the FL group; The second functional entity no longer serves the data network, network slice or area corresponding to the FL group; The second functional entity no longer accesses the data network or network slice corresponding to the FL group; The second functional entity leaves the area corresponding to the FL group.

19. A method for processing a federated learning group, characterized in that: Applied to the third functional entity, the method includes: Sending a first request or a second request to the first functional entity; The first request carries application information or service information related to the federated learning FL group, and the second request carries a control policy related to the federated learning FL group; so that the first functional entity creates or updates the FL group according to the application information, service information or control policy related to the FL group.

20. A functional entity, characterized in that The functional entity is a first functional entity, and the functional entity includes a memory, a transceiver, and a processor: A memory for storing a computer program; a transceiver for transmitting and receiving data under the control of the processor; and a processor for reading the computer program in the memory and performing the following operations: Get the feature information of the federated learning FL group; determining a second functional entity according to the characteristic information of the FL group; adding the second functional entity to the FL group; The characteristic information of the FL group includes the first information; or the characteristic information of the FL group includes the first information and identification information of a functional entity allowed to join the FL group; The first information includes any one or more of the following: Information about the application or service corresponding to the FL group; Area information corresponding to the FL group; Network slice information corresponding to the FL group; The data network information corresponding to the FL group.

21. A functional entity, characterized in that The functional entity is a second functional entity, and the functional entity includes a memory, a transceiver, and a processor: A memory for storing a computer program; a transceiver for transmitting and receiving data under the control of the processor; and a processor for reading the computer program in the memory and performing the following operations: Get the feature information of the federated learning FL group; determining, based on the characteristic information of the FL group, whether a condition for joining the FL group is met; If the conditions for joining the FL group are met, join the FL group; The characteristic information of the FL group includes the first information; or the characteristic information of the FL group includes the first information and identification information of a functional entity allowed to join the FL group; The first information includes any one or more of the following: Information about the application or service corresponding to the FL group; Area information corresponding to the FL group; Network slice information corresponding to the FL group; The data network information corresponding to the FL group.

22. A functional entity, characterized in that The functional entity is a third functional entity, and the functional entity includes a memory, a transceiver, and a processor: a memory for storing computer programs; a transceiver for transmitting and receiving data under the control of the processor; A processor is configured to read the computer program in the memory and perform the following operations: Sending a first request or a second request to the first functional entity through the transceiver; The first request carries application information or service information related to the federated learning FL group, and the second request carries a control policy related to the federated learning FL group; so that the first functional entity creates or updates the FL group according to the application information, service information or control policy related to the FL group.

23. A federated learning group processing device, characterized in that Applied to a first functional entity, the apparatus includes: A first acquisition unit is used to acquire feature information of a federated learning FL group; a first determining unit, configured to determine a second functional entity according to the characteristic information of the FL group; a first processing unit, configured to add the second functional entity to the FL group; The characteristic information of the FL group includes the first information; or the characteristic information of the FL group includes the first information and identification information of a functional entity allowed to join the FL group; The first information includes any one or more of the following: Information about the application or service corresponding to the FL group; Area information corresponding to the FL group; Network slice information corresponding to the FL group; The data network information corresponding to the FL group.

24. A federated learning group processing device, characterized in that Applied to the second functional entity, the device includes: The second acquisition unit is used to obtain feature information of the federated learning FL group; a third determining unit, configured to determine whether a condition for joining the FL group is met based on the characteristic information of the FL group; a third processing unit, configured to join the FL group if the conditions for joining the FL group are met; The characteristic information of the FL group includes the first information; or the characteristic information of the FL group includes the first information and identification information of a functional entity allowed to join the FL group; The first information includes any one or more of the following: Information about the application or service corresponding to the FL group; Area information corresponding to the FL group; Network slice information corresponding to the FL group; The data network information corresponding to the FL group.

25. A federated learning group processing device, characterized in that: Applied to a third functional entity, the device includes: A second sending unit, configured to send the first request or the second request to the first functional entity; The first request carries application information or service information related to the federated learning FL group, and the second request carries a control policy related to the federated learning FL group; so that the first functional entity creates or updates the FL group according to the application information, service information or control policy related to the FL group.

26. A processor-readable storage medium, characterized in that: The processor-readable storage medium stores a computer program, and the computer program is used to enable the processor to execute the federated learning group processing method according to any one of claims 1 to 12; or The computer program is used to enable the processor to execute the federated learning group processing method according to any one of claims 13 to 18; or The computer program is configured to cause the processor to execute the federated learning group processing method according to claim 19.

Images