A communication method and apparatus

Abstract

A communication method and device, the method comprising: a first network element sending a first request message to a second network element, the first request message being used to request data of a terminal device group, wherein the terminal devices in the terminal device group are terminal devices participating in horizontal federated learning. The first network element receives the data of the terminal device group from the second network element, the data of the terminal device group comprising aggregated information of measurement results of QoS parameters, the first network element adjusts the QoS parameters of a first terminal device in the terminal device group according to the data of the terminal device group, and the first network element sends the adjusted QoS parameters of the first terminal device to a policy control network element. The above method can improve the efficiency of model training of federated learning.

Description

[0001] This application claims priority to the Chinese Patent Application No. 202210108437.1 filed on January 28, 2022, entitled "A Communication Method and Device", the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0002] Embodiments of the present application relate to the field of wireless communication, and in particular to a communication method and device. BACKGROUND

[0003] Federated learning (FL) is a machine learning framework that can effectively help multiple users to use data and build machine learning models while meeting the requirements of user privacy protection, data security and government regulations. As a distributed machine learning paradigm, federated learning can effectively solve the problem of data silos, build joint models without sharing user data, and thus technically break down data silos and achieve AI collaboration.

[0004] In the model training process of federated learning, the existing quality of service (QoS) parameter adjustment method may affect the efficiency of model training of federated learning. SUMMARY

[0005] The present application provides a communication method and device to improve the efficiency of model training of federated learning.

[0006] In a first aspect, the present application provides a communication method, comprising: a first network element sending a first request message to a second network element, the first request message being used to request data of a terminal device group, wherein the terminal devices in the terminal device group are terminal devices participating in horizontal federated learning; the first network element receiving the data of the terminal device group from the second network element, the data of the terminal device group comprising aggregated information of measurement results of QoS parameters; the first network element adjusting the QoS parameters of a first terminal device in the terminal device group according to the data of the terminal device group; and the first network element sending the adjusted QoS parameters of the first terminal device to a policy control network element.

[0007] By using the above method, the first network element subscribes to the data of the terminal device group from the second network element, the second network element sends the data of the terminal device group to the first network element, and the first network element adjusts the QoS parameters of at least one terminal device in the terminal device group according to the data of the terminal device group, thereby improving the efficiency of model training of federated learning.

[0008] In a possible design, the first request message includes indication information, where the indication information is used to indicate measurement results of QoS parameters of the feedback terminal device; and the data of the terminal device group further includes measurement results of QoS parameters of a second terminal device in the terminal device group, where the second terminal device includes the first terminal device.

[0009] With the above design, the second network element can further send, to the first network element, the measurement results of the QoS parameters of the terminal devices in the terminal device group.

[0010] In a possible design, the first request message further includes a feedback condition, where the feedback condition is used to indicate a condition that needs to be met for feeding back the measurement results of the QoS parameters of the terminal device; and the measurement results of the QoS parameters of the second terminal device meet the feedback condition.

[0011] With the above design, the second network element can feed back the measurement results of the QoS parameters of the terminal device that meets the feedback condition.

[0012] In a possible design, the aggregation information includes an aggregated bit rate, where the aggregated bit rate is used to represent a sum of bit rates of QoS flows of the terminal devices in the terminal device group; and when the first network element adjusts the QoS parameters of the first terminal device in the terminal device group according to the data of the terminal device group, if the aggregated bit rate is greater than a maximum subscription bit rate of the terminal device group, the first network element adjusts the bit rates of the QoS flows of the first terminal device according to the data of the terminal device group.

[0013] For example, the first network element can adjust the bit rates of the QoS flows of at least one terminal device according to the data of the terminal device group, where the at least one terminal device includes the first terminal device, and after the bit rates of the QoS flows of the at least one terminal device are adjusted, a sum of the bit rates of the QoS flows of each terminal device in the terminal device group is less than or equal to the maximum subscription bit rate.

[0014] With the above design, the QoS parameters of the terminal device can be adjusted, so that the sum of the bit rates of the QoS flows of each terminal device in the terminal device group is less than or equal to the maximum subscription bit rate.

[0015] In a possible design, the aggregation information includes a statistical value of a transmission delay of the terminal devices in the terminal device group; and when the first network element adjusts the QoS parameters of the first terminal device in the terminal device group according to the data of the terminal device group, if the statistical value of the transmission delay is greater than a first preset threshold, the first network element adjusts the bit rates of the QoS flows of the first terminal device according to the data of the terminal device group.

[0016] The statistical value of the transmission delay of the terminal devices in the terminal device group can be a variance determined according to the transmission delays of the terminal devices in the terminal device group. For example, the statistical value of the transmission delay of the terminal devices in the terminal device group refers to a variance determined by the weighted transmission delay or the differential transmission delay of the terminal devices in the terminal device group.

[0017] By using the above design, the QoS parameter of the terminal device can be adjusted, so that the statistical value of the transmission delay is less than or equal to the first preset threshold.

[0018] In a possible design, the first terminal device is a terminal device whose transmission delay is greater than a second preset threshold or less than a third preset threshold, where the second preset threshold is greater than the third preset threshold.

[0019] By using the above design, the transmission delay of the terminal device with a longer transmission delay can be reduced, and the transmission delay of the terminal device with a shorter transmission delay can be increased to ensure that the sum of the bit rates of the QoS flows of all the terminal devices in the terminal device group is less than or equal to the subscription maximum bit rate, so that the data sent by all the terminal devices in the terminal device group to the server can arrive at the server at almost the same time, thereby improving the efficiency of model training of federated learning.

[0020] In a possible design, the first request message further includes the first preset threshold and / or the second preset threshold.

[0021] In a possible design, the first request message includes at least one of the following: an identifier of a terminal device included in the terminal device group, an identifier of the terminal device group, identifier information of an analysis type, and a type of the aggregation information.

[0022] In a possible design, the first request message further includes at least one of the following: a trigger condition for sending the aggregation information, and an application identifier used to indicate an application to which a measurement result of the QoS parameter of the terminal device corresponds.

[0023] The trigger condition for sending the aggregation information can include periodically feeding back the aggregation information, or feeding back the aggregation information when a preset condition is met. For example, the preset condition can include that the aggregated bit rate is greater than the subscription maximum bit rate of the terminal device group, and / or the statistical value of the transmission delay of the terminal devices in the terminal device group is greater than the first preset threshold.

[0024] In a possible design, the first network element is an application function network element, and the second network element is a data analysis function network element or a policy control network element.

[0025] In a second aspect, the present application provides a communication method, comprising: receiving, by a second network element, a first request message from a first network element, the first request message being used to request data of a terminal device group, wherein a terminal device in the terminal device group is a terminal device participating in horizontal federated learning; and sending, by the second network element, the data of the terminal device group, the data of the terminal device group comprising aggregated information of measurement results of QoS parameters.

[0026] By using the above method, the first network element subscribes to the data of the terminal device group from the second network element, and the second network element sends the data of the terminal device group to the first network element, so that the first network element adjusts the QoS parameters of at least one terminal device in the terminal device group according to the data of the terminal device group, thereby improving the efficiency of model training of federated learning.

[0027] In a possible design, the first request message comprises indication information, the indication information being used to indicate that the measurement results of the QoS parameters of the terminal device are fed back; and the data of the terminal device group further comprises measurement results of QoS parameters of a second terminal device in the terminal device group, the second terminal device comprising the first terminal device.

[0028] By using the above design, the second network element can further send the measurement results of the QoS parameters of the terminal device in the terminal device group to the first network element.

[0029] In a possible design, the first request message further comprises a feedback condition, the feedback condition being used to indicate a condition required to be met for feeding back the measurement results of the QoS parameters of the terminal device; and the measurement results of the QoS parameters of the second terminal device meet the feedback condition.

[0030] By using the above design, the second network element can feed back the measurement results of the QoS parameters of the terminal device meeting the feedback condition.

[0031] In a possible design, the aggregated information comprises an aggregated bit rate, the aggregated bit rate being used to represent a sum of bit rates of QoS flows of the terminal devices in the terminal device group.

[0032] In a possible design, the aggregated information comprises a statistical value of transmission delays of the terminal devices in the terminal device group.

[0033] In a possible design, the first request message comprises at least one of the following: an identifier of a terminal device in the terminal device group, an identifier of the terminal device group, identification information of an analysis type, and a type of the aggregated information.

[0034] In a possible design, the first request message further includes at least one of the following: a trigger condition for sending the aggregated information; and an application identifier, used to indicate that the measurement result of the QoS parameter of the terminal device corresponds to an application.

[0035] In a possible design, the method further includes: sending, by the second network element, a second request message to a user plane network element, where the second request message is used to request data of the terminal device group; and receiving, by the second network element, the data of the terminal device group from the user plane network element.

[0036] With the above design, the second network element can directly obtain the data of the terminal device group from the user plane network element.

[0037] In a possible design, the method further includes: sending, by the second network element, a third request message to a user plane network element, where the third request message is used to request measurement results of QoS parameters of terminal devices in the terminal device group; receiving, by the second network element, the measurement results of the QoS parameters of the terminal devices in the terminal device group from the user plane network element; and determining, by the second network element, the data of the terminal device group according to the measurement results of the QoS parameters of the terminal devices in the terminal device group and the first request message.

[0038] With the above design, the second network element can aggregate the measurement results of the QoS parameters of the terminal devices obtained from the user plane network element, and obtain the data of the terminal device group.

[0039] In a possible design, the second network element is a policy control network element; and the method further includes: receiving, by the second network element, bit rates of QoS flows respectively corresponding to each terminal device included in the terminal device group from the first network element; receiving, by the second network element, a subscription maximum bit rate of the terminal device group from a third network element, where the third network element is a unified data storage network element or a unified data management network element; determining, by the second network element, that a sum of the bit rates of the QoS flows respectively corresponding to each terminal device included in the terminal device group is less than the subscription maximum bit rate of the terminal device group according to the subscription maximum bit rate of the terminal device group and the bit rates of the QoS flows respectively corresponding to each terminal device included in the terminal device group; generating, by the second network element, PCC rules respectively corresponding to each terminal device included in the terminal device group; and sending, by the second network element, the PCC rules to a session management network element.

[0040] In a possible design, the second network element is a policy control network element; and the method further includes: receiving, by the second network element, from the first network element, bit rate reference ranges of QoS flows respectively corresponding to each of the terminal devices included in the terminal device group; receiving, by the second network element, from a third network element, a subscription maximum bit rate of the terminal device group, the third network element being a unified data storage network element or a unified data management network element; determining, by the second network element, according to the subscription maximum bit rate of the terminal device group and the bit rate reference ranges of the QoS flows respectively corresponding to each of the terminal devices included in the terminal device group, that a sum of lower bounds of the bit rate reference ranges of the QoS flows respectively corresponding to each of the terminal devices included in the terminal device group is less than the subscription maximum bit rate of the terminal device group; generating, by the second network element, PCC rules respectively corresponding to each of the terminal devices included in the terminal device group; and sending, by the second network element, the PCC rules to a session management network element.

[0041] With the above design, the policy control network element can determine whether the subscription maximum bit rate of the terminal device group is met according to a value of a bit rate of a QoS flow respectively corresponding to each of the terminal devices included in the terminal device group or a value range of the bit rate of the QoS flow respectively corresponding to each of the terminal devices included in the terminal device group.

[0042] In a possible design, the method further includes: sending, by the second network element, second information to the first network element, the second information indicating that corresponding PCC rules have been generated for each of the terminal devices included in the terminal device group.

[0043] In a possible design, the first network element is an application function network element, and the second network element is a data analytics function network element or a policy control network element.

[0044] In a third aspect, an embodiment of the present application provides a communication apparatus, which is a first network element or an apparatus used for implementing a function of the first network element, and includes a processing module and a transceiver module; the transceiver module is configured to: send a first request message to a second network element, the first request message being used to request data of a terminal device group, wherein a terminal device in the terminal device group is a terminal device participating in horizontal federated learning; and receive the data of the terminal device group from the second network element, the data of the terminal device group including aggregated information of measurement results of QoS parameters; the processing module is configured to: adjust a QoS parameter of a first terminal device in the terminal device group according to the data of the terminal device group; and the transceiver module is configured to: send the adjusted QoS parameter of the first terminal device to a policy control network element.

[0045] In a possible design, the first request message includes indication information, and the indication information indicates that measurement results of QoS parameters of terminal devices are fed back.

[0046] The data of the terminal device group further includes a measurement result of a QoS parameter of a second terminal device in the terminal device group, and the second terminal device includes the first terminal device.

[0047] In a possible design, the first request message further includes a feedback condition, and the feedback condition indicates a condition required to be met for feeding back the measurement result of the QoS parameter of the terminal device; and the measurement result of the QoS parameter of the second terminal device meets the feedback condition.

[0048] In a possible design, the aggregation information includes an aggregated bit rate, and the aggregated bit rate is used to represent a sum of bit rates of QoS flows of terminal devices in the terminal device group; and the processing module is configured to, when adjusting the QoS parameter of the first terminal device in the terminal device group according to the data of the terminal device group, adjust a bit rate of a QoS flow of the first terminal device according to the data of the terminal device group if the aggregated bit rate is greater than a maximum bit rate of a subscription of the terminal device group.

[0049] In a possible design, the aggregation information includes a statistical value of a transmission delay of terminal devices in the terminal device group; and the processing module is configured to, when adjusting the QoS parameter of the first terminal device in the terminal device group according to the data of the terminal device group, adjust a bit rate of a QoS flow of the first terminal device according to the data of the terminal device group if the statistical value of the transmission delay is greater than a first preset threshold.

[0050] In a possible design, the first terminal device is a terminal device with a transmission delay greater than a second preset threshold.

[0051] In a possible design, the first request message further includes the first preset threshold and / or the second preset threshold.

[0052] In a possible design, the first request message includes at least one of the following: an identity of a terminal device in the terminal device group, an identity of the terminal device group, identity information of an analysis type, and a type of the aggregation information.

[0053] In a possible design, the first request message further includes at least one of the following: a trigger condition for sending the aggregation information; and an application identity used to indicate an application corresponding to the measurement result of the QoS parameter of the terminal device.

[0054] In a possible design, the first network element is an application function network element, and the second network element is a data analysis function network element or a policy control network element.

[0055] In a fourth aspect, the present application provides a communication device, which is a second network element or a device for implementing the function of the second network element, comprising a processing module and a transceiver module; the processing module invokes the transceiver module to perform: receiving a first request message from a first network element, the first request message being used to request data of a terminal device group, wherein the terminal devices in the terminal device group are terminal devices participating in horizontal federated learning; and sending the data of the terminal device group, the data of the terminal device group comprising aggregated information of measurement results of QoS parameters.

[0056] In a possible design, the first request message comprises indication information indicating that the measurement results of the QoS parameters of the terminal device are fed back; and the data of the terminal device group further comprises measurement results of QoS parameters of a second terminal device in the terminal device group, the second terminal device comprising the first terminal device.

[0057] In a possible design, the first request message further comprises a feedback condition indicating a condition required to be met for feeding back the measurement results of the QoS parameters of the terminal device; and the measurement results of the QoS parameters of the second terminal device meet the feedback condition.

[0058] In a possible design, the aggregated information comprises an aggregated bit rate used to represent a sum of bit rates of QoS flows of the terminal devices in the terminal device group.

[0059] In a possible design, the aggregated information comprises a statistical value of transmission delays of the terminal devices in the terminal device group.

[0060] In a possible design, the first request message comprises at least one of the following: an identity of a terminal device in the terminal device group, an identity of the terminal device group, identity information of an analysis type, and a type of the aggregated information.

[0061] In a possible design, the first request message further comprises at least one of the following: a trigger condition for sending the aggregated information; and an application identity used to indicate an application corresponding to the measurement results of the QoS parameters of the terminal device.

[0062] In a possible design, the transceiver module is further configured to send a second request message to a user plane network element, the second request message being used to request the data of the terminal device group; and receive the data of the terminal device group from the user plane network element.

[0063] In a possible design, the transceiver module is further configured to: send, to a user plane network element, a third request message for requesting measurement results of QoS parameters of terminal devices in the terminal device group; and receive, from the user plane network element, the measurement results of the QoS parameters of the terminal devices in the terminal device group; and the processing module is further configured to determine data of the terminal device group according to the measurement results of the QoS parameters of the terminal devices in the terminal device group and the first request message.

[0064] In a possible design, the first network element is an application function network element, and the second network element is a data analytics function network element or a policy control network element.

[0065] In a fifth aspect, the present application provides a communication method, which includes: an application function network element obtaining transmission time delays and local calculation time delays of each terminal device in a terminal device group, the terminal devices in the terminal device group being terminal devices participating in horizontal federated learning, the application function network element adjusting a QoS parameter of a first terminal device in the terminal device group according to the transmission time delays and the local calculation time delays of the terminal devices, and the application function network element sending the adjusted QoS parameter of the first terminal device to a policy control network element.

[0066] With the above method, the application function network element can also collect data without the NWDAF or the PCF, but itself statistically or obtains the transmission time delays and the local calculation time delays of each terminal device in the terminal device group, and then determines an adjustment strategy of the QoS parameter of the terminal device in the terminal device group, thereby reducing the number of messages exchanged between network elements, reducing the network element load, and improving the QoS parameter adjustment efficiency of the terminal device, so that the sum of the local calculation time delays and the transmission time delays of different terminal devices is roughly the same, thereby improving the model training efficiency of the horizontal federated learning.

[0067] In a possible design, when the application function network element adjusts the QoS parameter of the first terminal device in the terminal device group according to the transmission time delays and the local calculation time delays of the terminal devices, the application function network element determines a statistical value of total time delays of the terminal devices according to the transmission time delays and the local calculation time delays of the terminal devices, where the total time delay of each terminal device is the sum of the transmission time delay and the local calculation time delay of the terminal device, and the application function network element adjusts the QoS parameter of the first terminal device in a case where the total time delay of the first terminal device is greater than a first preset threshold or smaller than a second preset threshold, where the first preset threshold and the second preset threshold are determined according to the statistical value.

[0068] In one possible design, the adjusted QoS parameter of the first terminal device includes at least one of a resource type, a priority, a packet delay budget, a guaranteed flow bit rate, a maximum flow bit rate, an allocation and pre-emption priority.

[0069] In one possible design, the local computation delay is a length of time needed to determine an updated parameter of a model of the federated learning, and the transmission delay is a length of time needed to transmit the updated parameter to the application function network element.

[0070] In a sixth aspect, the present disclosure provides a communication apparatus, which is an application function network element, or an apparatus for implementing a function of an application function network element, and the apparatus includes a processing module and a transceiver module. The processing module is configured to obtain a transmission delay and a local computation delay of each terminal device in a terminal device group, the terminal devices in the terminal device group being terminal devices participating in a federated learning, and adjust a QoS parameter of a first terminal device in the terminal device group according to the transmission delay and the local computation delay of each terminal device. The transceiver module is configured to send the adjusted QoS parameter of the first terminal device to a policy control network element.

[0071] In one possible design, when adjusting the QoS parameter of the first terminal device according to the transmission delay and the local computation delay of each terminal device, the processing module is configured to determine a statistical value of a total delay of each terminal device according to the transmission delay and the local computation delay of each terminal device, where the total delay of each terminal device is a sum of the transmission delay and the local computation delay of the terminal device, and adjust the QoS parameter of the first terminal device in a case where the total delay of the first terminal device is greater than a first preset threshold or less than a second preset threshold, where the first preset threshold and the second preset threshold are determined according to the statistical value.

[0072] In one possible design, the adjusted QoS parameter of the first terminal device includes at least one of a resource type, a priority, a packet delay budget, a guaranteed flow bit rate, a maximum flow bit rate, an allocation and pre-emption priority.

[0073] In one possible design, the local computation delay is a length of time needed to determine an updated parameter of a model of the federated learning, and the transmission delay is a length of time needed to transmit the updated parameter to the application function network element.

[0074] In a seventh aspect, the present disclosure also provides an apparatus. The apparatus can perform the method designs described above. The apparatus can be a chip or circuit that can perform the functions corresponding to the method described above, or a device including the chip or circuit.

[0075] In a possible implementation, the apparatus includes a memory for storing computer executable program code, and a processor coupled with the memory. The program code stored in the memory includes instructions that, when executed by the processor, cause the apparatus or a device in which the apparatus is installed to perform the method in any of the possible designs described above.

[0076] In a possible implementation, the apparatus can further include a communication interface, which can be a transceiver, or if the apparatus is a chip or circuit, the communication interface can be an input / output interface of the chip, such as an input / output pin or the like.

[0077] In a possible design, the apparatus includes corresponding functional units respectively for implementing the steps in the above method. The functions can be implemented by hardware, or by hardware executing corresponding software. The hardware or software includes one or more units corresponding to the above functions.

[0078] In an eighth aspect, the present application provides a computer readable storage medium, which stores a computer program, when the computer program is run on an apparatus, performs the method in any of the possible designs described above.

[0079] In a ninth aspect, the present application provides a computer program product, which includes a computer program, when the computer program is run on an apparatus, performs the method in any of the possible designs described above.

[0080] In a tenth aspect, the present application provides a communication system, which includes a first network element and a second network element, the first network element is configured to perform the method in any of the possible designs of the first aspect, and the second network element is configured to perform the method in any of the possible designs of the second aspect. BRIEF DESCRIPTION OF DRAWINGS

[0081] Figure 1 Architecture diagram of a mobile communication system to which the present application is applied;

[0082] Figure 2 Diagram of a dataset participating in horizontal federated learning in an embodiment of the present application;

[0083] Figure 3 Diagram of a model training process of horizontal federated learning in an embodiment of the present application;

[0084] Figure 4 Overview flowchart of a communication method in an embodiment of the present application;

[0085] Figure 5A Diagram of transmission delay of terminal devices in a terminal device group before adjustment in an embodiment of the present application;

[0086] Figure 5B a schematic diagram of adjusting the transmission delay of the terminal device in the adjusted terminal device group in an embodiment of the present application;

[0087] Figure 6 a flowchart of the interaction between the first network element and other network elements before the first network element sends the first request message to the second network element in an embodiment of the present application;

[0088] Figure 7 a schematic diagram of another communication method in an embodiment of the present application;

[0089] Figure 8 a schematic diagram of another communication method in an embodiment of the present application;

[0090] Figure 9 a schematic diagram of another communication method in an embodiment of the present application;

[0091] Figure 10 a flowchart of the interaction between the AF and the PCF in an embodiment of the present application;

[0092] Figure 11 a schematic diagram of one of the communication apparatuses in an embodiment of the present application;

[0093] Figure 12 a schematic diagram of another of the communication apparatuses in an embodiment of the present application. DETAILED DESCRIPTION

[0094] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. The terms “first”, “second” and corresponding terms of reference in the specification and claims of the present application and the above drawings are used to distinguish similar objects, not necessarily to describe a specific order or sequence. It should be understood that the terms used in this way can be interchanged under appropriate circumstances, which is only a distinguishing way used in the description of the embodiments of the present application to describe the same attribute objects. In addition, the terms “include” and “have” and any variations thereof are intended to cover non-exclusive inclusion, so that the processes, methods, systems, products or devices containing a series of units do not necessarily limit to those units, but can include other units not clearly listed or inherent to these processes, methods, products or devices.

[0095] In the description of the present application, unless otherwise specified, " / " means or, for example, A / B can mean A or B; "and / or" in the present application is only a description of the relationship between the associated objects, which means that there can be three relationships, for example, A and / or B, which can mean: A alone, A and B together, B alone. In addition, in the description of the present application, "at least one" means one or more, and "more" means two or more. "At least one" or the like means any combination of these items, including single or multiple combinations. For example, at least one of a, b, or c can mean a, b, c, a-b, a-c, b-c, or a-b-c, where a, b, and c can be single or multiple.

[0096] The technical solutions provided by the embodiments of the present application can be applied to various communication systems. For example: it can be applied to LTE system or 5G system, and can also be applied to other future-oriented new systems, etc. The embodiments of the present application do not make specific limitations on this. In addition, the term "system" can be replaced by "network". The following will be described only by taking the communication system architecture of 5G as an example.

[0097] As shown in Figure 1 The 5G communication system formulated by the 3rd generation partnership project (3GPP) standard. The communication system includes terminal equipment (for example, user equipment (UE)), radio access network (RAN), core network (CN). The data network (DN) is logically, the network elements of the core network can be divided into user plane and control plane, the control plane is responsible for the management of the mobile network, and the user plane is responsible for the transmission of service data.

[0098] Among them, the terminal equipment is the entrance for mobile users to interact with the network, which can provide basic computing power, storage capacity, display service window to users, and receive user operation input. The next generation terminal equipment (NextGen UE) can use new air interface technology to establish signal connection with RAN, data connection, so as to transmit control signal and service data to mobile network. The terminal equipment can include various handheld devices with wireless communication function, vehicle-mounted equipment, wearable equipment, computing equipment or other processing equipment connected to wireless modem, and various forms of terminals, mobile stations (MS), terminals, soft terminals, etc., such as water meters, electricity meters, sensors, etc.

[0099] RAN: deployed in a location close to the terminal device, provides network access function for authorized users in a specific area, and can determine different quality transmission tunnels according to the level of users, the demand of services, etc. to transmit user data. RAN can manage its own resources, rationally utilize, provide access services for terminal devices on demand, and is responsible for forwarding control signals and user data between terminal devices and core network.

[0100] Core network: responsible for maintaining the subscription data of the mobile network, managing the network elements of the mobile network, providing session management, mobility management, policy management, security authentication, etc. for terminal devices. When the terminal device is attached, it provides network access authentication for the terminal device; when the terminal device has a service request, it allocates network resources for the terminal device; when the terminal device moves, it updates the network resources for the terminal device; when the terminal device is idle, it provides a fast recovery mechanism for the terminal device; when the terminal device detaches, it releases the network resources for the terminal device; when the terminal device has service data, it provides data routing functions for the terminal device, such as forwarding uplink data to the data network; or receiving downlink data of the terminal device from the data network, and forwarding it to the RAN, so as to send it to the terminal device by the RAN.

[0101] Data network (DN): a data network that provides service to users, generally with the client located in the terminal device and the server located in the data network. The data network can be a private network such as a local area network, or an external network not controlled by the operator, such as the Internet, or a dedicated network jointly deployed by the operator, such as a network providing IMS (IP Multimedia Core Network Subsystem) services.

[0102] The core network user plane includes a user plane function (UPF), and the core network control plane includes an access and mobility management function (AMF), a session management function (SMF), a network exposure function (NEF), a network function repository function (NRF), a unified data management (UDM), a unified data repository (UDR), a policy control function (PCF), and an application function (AF).

[0103] The core network control plane adopts a service-oriented architecture, and the interaction between control plane network elements adopts a service call mode to replace the point-to-point communication mode in the traditional architecture. In the service-oriented architecture, a control plane network element exposes services to other control plane network elements for calling by other control plane network elements. In the point-to-point communication, a communication interface between control plane network elements stores a specific set of messages that can only be used by the control plane network elements at both ends of the interface when communicating.

[0104] The functions of the functional entities in the core network are briefly introduced as follows.

[0105] 1. Session management network element: mainly used for session management, IP address allocation and management of terminal devices, selection of manageable user equipment plane functions, policy control, or termination of charging function interfaces, and downlink data notification. In 5G communication, the session management network element can be an SMF network element. In future communication such as 6G communication, the session management function network element can still be an SMF network element, or have other names, which are not limited in the present application. Nsmf is a service-based interface provided by the SMF, and the SMF can communicate with other network functions through Nsmf.

[0106] 2. Access management network element: mainly used for mobility management and access management, for example, it can be a mobility management entity (MME) function in a 4G communication network or an AMF network element in a 5G network. In future communication such as 6G communication, the access management network element can still be an AMF network element, or have other names, which are not limited in the present application. Namf is a service-based interface provided by the AMF, and the AMF can communicate with other network functions through Namf.

[0107] 3、Network Exposure Network Element: used for securely exposing services and capabilities provided by 3GPP network functions to the outside. In 5G communication, the network exposure network element can be a NEF network element, and in future communication such as 6G communication, the network exposure function network element can still be a NEF network element, or have other names, which are not limited in the present application. Nnef is a service-based interface provided by NEF, and NEF can communicate with other network functions through Nnef.

[0108] 4、Network Repository Network Element: used for providing service registration, discovery and authorization, and maintaining available network function (NF) instance information, and can realize on-demand configuration of network functions and services and interconnection between NFs. In 5G communication, the network repository network element can be a NRF network element, and in future communication such as 6G communication, the network repository function network element can still be a NRF network element, or have other names, which are not limited in the present application. Nnrf is a service-based interface provided by NRF, and NRF can communicate with other network functions through Nnrf.

[0109] 5、Policy Control Network Element: used for a unified policy framework to guide network behavior, and provides policy rule information for control plane function network elements (such as AMF, SMF, etc.). In 5G communication, the policy control network element can be a PCF network element, and in future communication such as 6G communication, the policy control network element can still be a PCF network element, or have other names, which are not limited in the present application. Npcf is a service-based interface provided by PCF, and PCF can communicate with other network functions through Npcf.

[0110] 6、Data Management Network Element: used for processing user identification, subscription, access authentication, registration, or mobility management, etc. In 5G communication, the data management network element can be a UDM network element, and in future communication such as 6G communication, the data management network element can still be a UDM network element, or have other names, which are not limited in the present application. Nudm is a service-based interface provided by UDM, and UDM can communicate with other network functions through Nudm.

[0111] 7、Data Storage Network Element: used for performing access functions of subscription data, policy data, application data and other types of data. In 5G communication, the data storage network element can be a UDR network element, and in future communication such as 6G communication, the data storage network element can still be a UDR network element, or have other names, which are not limited in the present application. Nudr is a service-based interface provided by UDR, and UDR can communicate with other network functions through Nudr.

[0112] 8. Application network element: used for data routing of application influence, access network exposure function, or policy control interaction with policy framework, etc. In 5G communication, the application network element can be an AF network element, and in future communication such as 6G communication, the application network element can still be an AF network element or have other names, which are not limited in the present application. Naf is a service-based interface provided by the AF, and the AF can communicate with other network functions through Naf.

[0113] 9. User plane network element: used for packet routing and forwarding, or quality of service (QoS) processing of user plane data, etc. In 5G communication, the user plane network element can be a user plane function (UPF) network element, and in future communication such as 6G communication, the user plane network element can still be a UPF network element or have other names, which are not limited in the present application.

[0114] It can be understood that the core network can also include other network elements, which are not limited in the present application.

[0115] For the convenience of understanding the embodiments of the present application, several basic concepts involved in the embodiments of the present application are briefly described.

[0116] 1. Horizontal federated learning

[0117] According to the different data source characteristics of the participants of federated learning, federated learning can be divided into three categories: horizontal federated learning, vertical federated learning and federated transfer learning. The principle of horizontal federated learning is mainly introduced below.

[0118] Among them, the data set includes multiple groups of data, each group of data corresponds to a user, and each group of data records at least one user feature of the user corresponding to the group of data.

[0119] Among them, the user features (X1, X2, X3…) coincide to a high degree and the users (U1, U2, U3…) coincide to a low degree between the data sets of horizontal federated learning, for example, data set A includes UE1 data and UE2 data, UE1 data includes user features X1, X2 and X3, UE2 data includes user features X1, X2 and X3, data set B includes UE4 data and UE5 data, UE4 data includes user features X1, X2 and X3, and UE5 data includes user features X1, X2 and X3. For example, Figure 2As shown, the intersection of users is small, while the intersection of user features is large for dataset A and dataset B. Therefore, dataset A and dataset B can select the same user features from respective datasets for model training of federated learning. It should be noted that the participants of federated learning are not limited to two parties. For example, in the internet of things (IoT) scenario, the participants of federated learning can be very large.

[0120] As shown in Figure 3 The federated learning model training process with intermediate result transmission via 5G system (5GS) is shown. In the model training process of federated learning, each participant (such as Figure 3 Client 1, Client 2, …, Client K) needs to download the latest model from the server (Server) via 5GS to perform iteration of the model training process. The specific iteration process generally includes the following steps:

[0121] ① Each participant calculates the gradient of the model using the local dataset, and then uploads the encrypted gradient to the Server. The gradient here includes the direction and size of the change of the model parameters.

[0122] ② The Server performs gradient aggregation according to the gradient uploaded by each participant, and updates the parameters of the model. The gradient aggregation here generally refers to averaging the collected gradients.

[0123] ③ The Server distributes the updated model parameters to each participant.

[0124] ④ Each participant updates the local model according to the updated model parameters.

[0125] 2. Quality of service (QoS)

[0126] QoS describes a set of service requirements that must be met by the network to ensure an appropriate service level for data transmission.

[0127] 5G networks need to support a variety of service transmissions, such as video, mobile payment, web browsing, and factory automation control. Different services require different QoS. For example, video services require a large bandwidth, while automation control services generally require low latency and high reliability. Based on the QoS framework of 5G networks, operators can provide different QoS guarantees for different services.

[0128] In 5G network, a Quality of Service Flow (QoS Flow) is the finest granularity for end-to-end QoS control in a protocol data unit (PDU) session. The QoS control of a QoS Flow is mainly determined by the QoS parameters associated with it. The QoS framework of 5G network supports both guaranteed bit rate (GBR) QoS Flows and non-GBR QoS Flows. For GBR QoS Flows, the network needs to reserve resources to guarantee their bandwidth. For non-GBR QoS Flows, no resource reservation is needed. The corresponding QoS parameters of GBR and non-GBR QoS Flows are shown in Table 1.

[0129] Table 1: QoS parameters of different types of QoS Flow

[0130]

[0131] Among them, 5QI is used to identify a set of 5G QoS characteristics. The 5G QoS characteristics are used to describe the message forwarding processing of the QoS Flow in the end-to-end between the UE and the UPF, including resource type (such as Non-GBR and GBR), priority, packet delay budget (PDB) and packet loss rate, etc.

[0132] It should be noted that the above QoS parameters are all for a single QoS Flow. In addition, some aggregate QoS parameters are also defined in 5G network. The aggregate QoS parameters are used to represent the statistical characteristics of the QoS parameters of multiple QoS Flows, for example, the aggregate maximum bit rate (AMBR) is used to limit the maximum value of the total bit rate of a group of QoS Flows.

[0133] According to the different data granularity, some different levels of AMBR parameters are also defined in the existing 5G standard:

[0134] Session-AMBR: limits the aggregate bit rate expected to be provided in all Non-GBR QoS Flows of a specific PDU session.

[0135] UE-level aggregated maximum bit rate (UE-AMBR): applies to Non-GBR QoS Flows.

[0136] Slice-level maximum bit rate per UE (UE-Slice-MBR): limits the aggregated bit rate expected to be provided in all GBR QoS Flows and Non-GBR QoS Flows of a PDU session of a specific slice for a specific UE.

[0137] Slice-level maximum bit rate (Slice-MBR per S-NSSAI): applies to GBR QoS Flows and Non-GBR QoS Flows.

[0138] It can be understood that the method provided by the application can be applied not only to the scene of horizontal federated learning, but also to other model training processes or business processes that require interaction between a server and terminal devices in a terminal device group. The following describes the scene of horizontal federated learning.

[0139] In the model training process of federated learning, in order to adjust the QoS parameters of different participants and improve the efficiency of model training, the application provides a communication method, as shown in Figure 4 The first network element can be an application function network element, and the second network element can be a data analysis function network element or a policy control network element. Illustratively, the data analysis function network element can be a network data analysis function (NWDAF) network element, and the policy control network element can be a PCF network element.

[0140] The method comprises:

[0141] Step 400: The first network element sends a first request message to the second network element, and the first request message is used to request data of a terminal device group, wherein the data of the terminal device group comprises aggregated information of measurement results of QoS parameters.

[0142] The first request message can also be referred to as a first subscription message, and the application does not limit the name of the first request message. Illustratively, when the second network element is an NWDAF network element, the first request message can be an analysis subscription message (Nnwdaf_AnalyticsSubscription_Subscribe). When the second network element is a PCF network element, the first request message can be an event exposure subscription message (Nnwdaf_EventExposure Subscribe).

[0143] The terminal device group includes at least two terminal devices. The terminal devices in the terminal device group can be terminal devices participating in the horizontal federated learning. Exemplarily, the terminal devices in the terminal device group jointly participate in the model training process of the same horizontal federated learning.

[0144] In addition, it needs to be explained that before the first network element sends the first request message to the second network element, the terminal devices in the terminal device group have all established sessions of services associated with the horizontal federated learning. For example, the first service is a certain horizontal federated learning model training service, the terminal device group includes three terminal devices, and the three terminal devices are participants of the first service, and the three terminal devices have all established sessions about the first service.

[0145] The aggregation information can also be referred to as aggregated QoS information, and the aggregation information is determined by aggregating the measurement results of the QoS parameters of each terminal device in the terminal device group. The aggregation information is used to represent the statistical characteristics of the measurement results of the QoS parameters of each terminal device included in the terminal device group. The QoS parameter of the terminal device here can refer to the related description of the QoS parameter described above.

[0146] Exemplarily, the first request message can include but is not limited to the following contents:

[0147] 1. Identification information of the analysis type, used to indicate that the aggregation information is the aggregation information of the terminal device group, i.e., the aggregation information at the terminal device group level.

[0148] Exemplarily, the identification information can be represented by an analysis identification (Analytics ID), for example, Analytics ID = Aggregated QoS for FL group, indicating the aggregation information of the FL group.

[0149] 2. Type of the aggregation information, used to indicate the parameters included in the aggregation information. Exemplarily, the type of the aggregation information can be represented by the aggregated QoS parameters to be measured (Aggregated QoS parameters to be measured).

[0150] The type of the aggregation information can specifically include an aggregated bit rate, a statistical value of the transmission delay of the terminal devices in the terminal device group, etc. It can be understood that according to the model training service demand of the horizontal federated learning, the parameter information can also include other parameters, which are not limited here.

[0151] The aggregated bit rate is used to represent the sum of the bit rates of the QoS flows of the terminal devices in the terminal device group. In addition, according to the conversion relationship between the bit rate and the bandwidth, the aggregated bit rate can be converted into an aggregated bandwidth, which is used to represent the sum of the bandwidths corresponding to the terminal devices in the terminal device group.

[0152] It can be understood that the QoS flow here is the QoS flow corresponding to the federated learning service, and the QoS flow involved in the following description refers to the QoS flow corresponding to the federated learning service unless otherwise specified. The bit rate of the QoS flow here can refer to the GFBR.

[0153] Exemplarily, in the model training process of the federated learning, each participant needs to complete local calculation and upload the calculation result to the server, and therefore a large amount of bandwidth resources will be consumed. In order to avoid affecting other terminal devices or other services, a maximum available bandwidth is generally set for the model training service of the federated learning. In order to ensure that the sum of the bandwidths of the participants does not exceed the set maximum available bandwidth, the first network element can subscribe to the aggregated bandwidth from the second network element, and adjust the bandwidth of at least one participant when the aggregated bandwidth is greater than the set maximum available bandwidth. The specific adjustment manner can refer to the related description in the following step 420.

[0154] The statistical value of the transmission delay of the terminal devices in the terminal device group can be a variance determined according to the transmission delay of the terminal devices in the terminal device group.

[0155] Exemplarily, the statistical value of the transmission delay of the terminal devices in the terminal device group refers to a variance determined by the weighted transmission delay of the terminal devices in the terminal device group (hereinafter referred to as the variance of the weighted transmission delay) or a variance determined by the differential transmission delay of the terminal devices in the terminal device group (hereinafter referred to as the variance of the differential transmission delay). Taking any one terminal device (denoted as terminal device A) in the terminal device group as an example, the weighted transmission delay of the terminal device A is the ratio of the actual transmission delay of the terminal device A to the expected transmission delay of the terminal device A, and the differential transmission delay of the terminal device A is the difference between the actual transmission delay of the terminal device A and the expected transmission delay of the terminal device A. The actual transmission delay of the terminal device A is the time interval between the sending time of the QoS flow of the terminal device A and the time when the QoS flow of the terminal device A arrives at the server, and the expected transmission delay of the terminal device A is the time interval between the sending time of the QoS flow of the terminal device A and the preset time. The actual transmission delay of the terminal device A can be obtained by measuring the QoS parameter of the terminal device A, and the expected transmission delay of the terminal device A can be sent by the first network element to the second network element through the step 400, or can be configured to the second network element or the user plane network element in advance.

[0156] It can be understood that the server herein can be combined with any network element, or can be a separate server. The actual transmission delay of the first terminal device can be the PDB of the first terminal device.

[0157] As shown in Figure 5A , the terminal device group includes UE1, UE2 and UE3, each UE corresponds to a time axis, the time length from the starting point of the time axis to the triangular mark corresponds to the local calculation completion time of the UE, the time length from the triangular mark to the circular mark corresponds to the actual transmission delay of the UE, and the time length from the triangular mark to the intersection of the first dashed line and the time axis corresponds to the expected transmission delay of the UE. Wherein, the position of the first dashed line can be understood as the time when the server receives the calculation result of the last UE (i.e. UE2). Wherein, the local calculation completion time of each UE can be different and difficult to change, i.e. the positions of the triangular marks in Figure 5A are different, and the transmission delays of each UE can also be different, i.e. the positions of the circular marks in Figure 5A are different, but the transmission delay of the UE can be realized by adjusting the bandwidth of the UE.

[0158] Exemplarily, the weighted transmission delay of UE1 = the actual transmission delay of UE1 / the expected transmission delay of UE1, the weighted transmission delay of UE2 = the actual transmission delay of UE2 / the expected transmission delay of UE2, and the weighted transmission delay of UE3 = the actual transmission delay of UE3 / the expected transmission delay of UE3; or, the weighted transmission delay of UE1 = the expected transmission delay of UE1 / the actual transmission delay of UE1, the weighted transmission delay of UE2 = the expected transmission delay of UE2 / the actual transmission delay of UE2, and the weighted transmission delay of UE3 = the expected transmission delay of UE3 / the actual transmission delay of UE3. The greater the variance determined by the weighted transmission delay of UE1, the weighted transmission delay of UE2 and the weighted transmission delay of UE3, the greater the difference in time when the QoS flows sent by the three respectively reach the server, and the smaller the variance, the smaller the difference in time when the QoS flows sent by the three respectively reach the server.

[0159] Exemplarily, the differential transmission latency of UE1 = the actual transmission latency of UE1 - the expected transmission latency of UE1, the differential transmission latency of UE2 = the actual transmission latency of UE2 - the expected transmission latency of UE2, and the differential transmission latency of UE3 = the actual transmission latency of UE3 - the expected transmission latency of UE3; or, the differential transmission latency of UE1 = the expected transmission latency of UE1 - the actual transmission latency of UE1, the differential transmission latency of UE2 = the expected transmission latency of UE2 - the actual transmission latency of UE2, and the differential transmission latency of UE3 = the expected transmission latency of UE3 - the actual transmission latency of UE3. The greater the variance determined by the differential transmission latency of UE1, the differential transmission latency of UE2, and the differential transmission latency of UE3, the greater the difference in time of the QoS flows sent by the three UEs respectively to reach the server, and the smaller the variance, the smaller the difference in time of the QoS flows sent by the three UEs respectively to reach the server.

[0160] Exemplarily, according to the principle of transverse federated learning, the server needs to update the model after receiving the calculation results of each participant. When the transmission times of the calculation results of each participant to the server are greatly different, that is, the variance determined by the weighted transmission latency or the differential transmission latency of each participant is large, the overall training period will be lengthened, and the model training efficiency will be reduced. When the transmission times of the calculation results of each participant to the server are roughly the same, that is, the variance determined by the weighted transmission latency of each participant is small, the waiting time of the server can be reduced, and the model training efficiency can be improved. Therefore, the first network element can subscribe to the variance from the second network element, and adjust the transmission latency of at least one participant when the variance is greater than a set variance threshold. The specific adjustment manner can be referred to the related description in the following step 420.

[0161] Exemplarily, according to the principle of transverse federated learning, the server needs to update the model after receiving the calculation results of each participant. When the transmission times of the calculation results of each participant to the server are greatly different, that is, the variance determined by the weighted transmission latency or the differential transmission latency of each participant is large, the overall training period will be lengthened, and the model training efficiency will be reduced. When the transmission times of the calculation results of each participant to the server are roughly the same, that is, the variance determined by the weighted transmission latency of each participant is small, the waiting time of the server can be reduced, and the model training efficiency can be improved. Therefore, the first network element can subscribe to the variance from the second network element, and adjust the transmission latency of at least one participant when the variance is greater than a set variance threshold. The specific adjustment manner can be referred to the related description in the following step 420.

[0162] 3. Information for indicating the terminal device group. Exemplarily, the information for indicating the terminal device group can comprise an identity of each terminal device in the terminal device group, or a terminal device list corresponding to the terminal device group, or an identity of the terminal device group, etc. For example, the information for indicating the terminal device group can be identified by a target of analytics reporting.

[0163] For example, the terminal device group comprises UE1, UE2 and UE3, and the information for indicating the terminal device group comprises an identity of UE1, an identity of UE2 and an identity of UE3.

[0164] According to the information for indicating the terminal device group, the second network element can determine which QoS parameter measurement results of which terminal devices to collect, or aggregate which QoS parameter measurement results of which terminal devices, and further determine the aggregated information.

[0165] 4. Trigger condition for sending the aggregated information

[0166] The trigger condition for sending the aggregated information can comprise periodically feeding back the aggregated information, or feeding back the aggregated information when a preset condition is met.

[0167] For example, the preset condition can comprise that an aggregated bit rate is greater than a maximum bit rate of the terminal device group in a subscription, and / or a statistical value of a transmission delay of a terminal device in the terminal device group is greater than a first preset threshold, etc.

[0168] The second network element feeds back the aggregated information according to the trigger condition for sending the aggregated information.

[0169] 5. Indication information, the indication information indicating the QoS parameter measurement results of the terminal devices to be fed back.

[0170] It can be understood that the measurement results to be fed back are indicated by the indication information as terminal device level measurement results.

[0171] 6. Feedback condition, the feedback condition indicating a condition to be met for feeding back the QoS parameter measurement results of the terminal devices.

[0172] For example, the feedback condition can indicate that a terminal device with a transmission delay greater than a second preset threshold feeds back the QoS parameter measurement results, or a terminal device with a transmission delay less than a third preset threshold feeds back the QoS parameter measurement results, or a terminal device with a bit rate of a QoS flow greater than a preset bit rate feeds back the QoS parameter measurement results. The second preset threshold is greater than the third preset threshold.

[0173] In some embodiments, when the first request message only includes the indication information without the feedback condition, the data of the terminal device group further includes measurement results of QoS parameters of each terminal device in the terminal device group, i.e., the second network element feeds back the measurement results of QoS parameters of all terminal devices in the terminal device group to the first network element.

[0174] In some embodiments, when the first request message includes the indication information and the feedback condition, the data of the terminal device group further includes measurement results of QoS parameters of at least one terminal device in the terminal device group, and the measurement results of QoS parameters of the at least one terminal device meet the feedback condition, i.e., the second network element feeds back the measurement results of QoS parameters of part of terminal devices in the terminal device group to the first network element.

[0175] 7. An application identifier, used to indicate a range of an application that collects the measurement results of QoS parameters of the terminal device.

[0176] For example, when the terminal devices in the terminal device group are terminal devices participating in horizontal federated learning, the application identifier is an identifier of an application of horizontal federated learning.

[0177] The second network element instructs the user plane network element to collect the measurement results of QoS parameters of the terminal device according to the application identifier.

[0178] Step 410: The second network element sends the data of the terminal device group to the first network element.

[0179] Exemplarily, when the second network element is an NWDAF network element, the second network element sends an analysis notification message (Nnwdaf_AnalyticsSubscription_Notify) to the first network element, and the message carries the data of the terminal device group. When the second network element is a PCF network element, the first request message can be an event exposure notification message (Nnwdaf_EventExposure_Notify), which carries the data of the terminal device group. It should be noted that each communication between the first network element and the second network element needs to be forwarded through an NEF network element.

[0180] In a possible implementation, the second network element sends a second request message to the user plane network element, and the second request message is used to request the data of the terminal device group. The second network element receives the data of the terminal device group from the user plane network element.

[0181] The second request message can include the contents included in the first request message except the type information associated with the aggregation information. For example, when the first request message includes the above-mentioned 1-7 contents, the second request message includes the above-mentioned 2-7 contents.

[0182] According to the implementation manner, the user plane network element collects the measurement results of the QoS parameters of the terminal devices in the terminal device group, and determines the data of the terminal device group according to the collected measurement results, for example, determines the aggregation information according to the parameter information associated with the aggregation information in the second request message, and determines the measurement results of the QoS parameters of the terminal devices that need to be fed back according to the indication information and / or the feedback condition. The user plane network element sends the determined data of the terminal device group to the second network element, and the second network element sends the data of the terminal device group to the first network element.

[0183] For example, the terminal device group includes UE1, UE2 and UE3, the user plane network element receives the second request message, and collects the measurement results of the QoS parameters of UE1, the measurement results of the QoS parameters of UE2, and the measurement results of the QoS parameters of UE3 according to the second request message. Further, the user plane network element determines that the aggregation information includes the statistical values of the aggregated bit rate and the transmission delay according to the parameter information associated with the aggregation information in the second request message, and then sums the bit rates of the QoS flows corresponding to UE1, the bit rates of the QoS flows corresponding to UE2, and the bit rates of the QoS flows corresponding to UE3 as the aggregated bit rate of the terminal device group. The user plane network element also calculates the statistical value of the transmission delay of the terminal device group according to the actual transmission delay of UE1, the actual transmission delay of UE2, the actual transmission delay of UE3, and the expected transmission delay of UE1, the expected transmission delay of UE2, and the expected transmission delay of UE3 received from the second network element or configured in advance. The user plane network element sends the data of the terminal device group to the second network element, wherein the data of the terminal device group includes the aggregated bit rate of the terminal device group and the statistical value of the transmission delay of the terminal device group. The measurement results of the QoS parameters of UE1 include the bit rate of the QoS flow corresponding to UE1 and the actual transmission delay of UE1, the measurement results of the QoS parameters of UE2 include the bit rate of the QoS flow corresponding to UE2 and the actual transmission delay of UE2, and the measurement results of the QoS parameters of UE3 include the bit rate of the QoS flow corresponding to UE3 and the actual transmission delay of UE3.

[0184] In a possible implementation manner, the second network element sends a third request message to the user plane network element, the third request message being used to request the measurement results of the QoS parameters of the terminal devices in the terminal device group, the second network element receives the measurement results of the QoS parameters of the terminal devices in the terminal device group from the user plane network element, and the second network element determines the data of the terminal device group according to the measurement results of the QoS parameters of the terminal devices in the terminal device group and the first request message.

[0185] With the above implementation manner, the user plane network element collects the measurement results of the QoS parameters of the terminal devices in the terminal device group, and sends the collected measurement results to the second network element. The second network element determines the data of the terminal device group according to the collected results, for example, determines the aggregation information according to the parameter information associated with the aggregation information in the first request message, and determines the measurement results of the QoS parameters of the terminal devices that need to be fed back according to the indication information and / or the feedback condition.

[0186] For example, the terminal device group includes UE1, UE2 and UE3, the user plane network element receives the third request message, and collects the measurement results of the QoS parameters of UE1, the measurement results of the QoS parameters of UE2, and the measurement results of the QoS parameters of UE3 according to the third request message. The user plane network element sends the measurement results of the QoS parameters of UE1, the measurement results of the QoS parameters of UE2, and the measurement results of the QoS parameters of UE3 to the second network element. According to the parameter information associated with the aggregation information in the first request message, the second network element determines that the aggregation information includes the statistical values of the aggregated bit rate and the transmission delay, and then sums the bit rates of the QoS flows corresponding to UE1, the bit rates of the QoS flows corresponding to UE2, and the bit rates of the QoS flows corresponding to UE3 as the aggregated bit rate of the terminal device group. The second network element also calculates the statistical value of the transmission delay of the terminal device group according to the actual transmission delay of UE1, the actual transmission delay of UE2, the actual transmission delay of UE3, and the expected transmission delay of UE1, the expected transmission delay of UE2, and the expected transmission delay of UE3 configured in advance. The second network element sends the data of the terminal device group to the first network element, wherein the data of the terminal device group includes the aggregated bit rate of the terminal device group and the statistical value of the transmission delay of the terminal device group. The measurement results of the QoS parameters of UE1 include the bit rate of the QoS flow corresponding to UE1 and the actual transmission delay of UE1, the measurement results of the QoS parameters of UE2 include the bit rate of the QoS flow corresponding to UE2 and the actual transmission delay of UE2, and the measurement results of the QoS parameters of UE3 include the bit rate of the QoS flow corresponding to UE3 and the actual transmission delay of UE3.

[0187] Step 420: The first network element adjusts the QoS parameters of the first terminal device in the terminal device group according to the data of the terminal device group.

[0188] It can be understood that the first network element can adjust the QoS parameters of at least one terminal device in the terminal device group, for example, the QoS parameters of the first terminal device, or the QoS parameters of multiple terminal devices, or the QoS parameters of all terminal devices.

[0189] In adjusting the QoS parameter of the terminal device, the first network element can adjust the value of the QoS parameter, or the value range of the QoS parameter. For example, the first network element can adjust the value of the bit rate of the QoS flow of the first terminal device, or the value range of the bit rate.

[0190] The following describes a possible implementation manner of adjusting the QoS parameter of at least one terminal device in a terminal device group in combination with data of the terminal device group, wherein the at least one terminal device includes the first terminal device.

[0191] Manner 1: When the aggregation information includes the aggregation bit rate, if the aggregation bit rate is greater than the subscribed maximum bit rate of the terminal device group, the first network element adjusts the bit rate of the QoS flow of the at least one terminal device. After adjusting the bit rate of the QoS flow of the at least one terminal device, the sum of the bit rates of the QoS flows of all the terminal devices in the terminal device group is less than or equal to the subscribed maximum bit rate, or the sum of the lower bounds of the value ranges of the bit rates of the QoS flows of all the terminal devices in the terminal device group is less than or equal to the subscribed maximum bit rate.

[0192] In some possible embodiments, if the first request message includes the trigger condition for sending the aggregation information, and the trigger condition for sending the aggregation information includes that the aggregation bit rate of the terminal device group is greater than the subscribed maximum bit rate of the terminal device group, the second network element or the user plane network element determines whether the aggregation bit rate is greater than the subscribed maximum bit rate of the terminal device group according to the subscribed maximum bit rate of the terminal device group. If yes, the second network element sends the data of the terminal device group to the first network element, and the data of the terminal device group includes the statistical value of the transmission delay.

[0193] In some possible embodiments, if the first request message does not include the trigger condition for sending the aggregation information, the first network element or determines whether the aggregation bit rate is greater than the subscribed maximum bit rate of the terminal device group according to the subscribed maximum bit rate of the terminal device group.

[0194] In some possible embodiments, when the first network element adjusts the bit rate of the QoS flow of the at least one terminal device, the first network element can determine an adjustment scheme according to the bit rate of the QoS flow of each terminal device configured last time for the terminal device group, that is, which terminal device's bit rate of the QoS flow is adjusted, and the direction and size of the adjustment, so that after adjusting the bit rate of the QoS flow of the at least one terminal device, the sum of the bit rates of the QoS flows of all the terminal devices in the terminal device group is less than or equal to the subscribed maximum bit rate, or the sum of the lower bounds of the value ranges of the bit rates of the QoS flows of all the terminal devices in the terminal device group is less than or equal to the subscribed maximum bit rate.

[0195] In some possible embodiments, when the first network element adjusts the bitrates of the QoS flows of the at least one terminal device, if the first request message comprises the indication information, the second network element further feeds back, to the first network element, the measurement results of the QoS parameters of the terminal devices in the terminal device group, and the first network element can determine the adjustment scheme in combination with the measurement results of the QoS parameters of the terminal devices in the terminal device group; or, when the first network element adjusts the bitrates of the QoS flows of the at least one terminal device, if the first request message comprises the indication information and the feedback condition, the second network element further feeds back, to the first network element, the measurement results of the QoS parameters of the terminal devices in the terminal device group that meet the feedback condition, and the first network element can determine the adjustment scheme in combination with the measurement results of the QoS parameters of the terminal devices in the terminal device group that meet the feedback condition. For example, the feedback condition can be that the terminal devices whose bitrates of the QoS flows are greater than a preset bitrate feed back the measurement results of the QoS parameters.

[0196] For example, the terminal device group comprises UE1, UE2, and UE3, and the aggregate bitrate of the terminal device group is the sum of the bitrate of the QoS flow corresponding to UE1, the bitrate of the QoS flow corresponding to UE2, and the bitrate of the QoS flow corresponding to UE3. When the aggregate bitrate of the terminal device group is greater than the subscribed maximum bitrate of the terminal device group, the first network element can adjust the values of at least one of UE1, UE2, and UE3, so that the sum of the bitrate of the QoS flow corresponding to UE1, the bitrate of the QoS flow corresponding to UE2, and the bitrate of the QoS flow corresponding to UE3 is less than or equal to the subscribed maximum bitrate of the terminal device group, or the sum of the lower bound of the value range of the bitrate of the QoS flow corresponding to UE1, the lower bound of the value range of the bitrate of the QoS flow corresponding to UE2, and the lower bound of the value range of the bitrate of the QoS flow corresponding to UE3 is less than or equal to the subscribed maximum bitrate of the terminal device group.

[0197] Method 2: When the aggregate information comprises the statistical value of the transmission delay of the terminal devices in the terminal device group, if the statistical value of the transmission delay is greater than a first preset threshold, the first network element adjusts the bitrates of the QoS flows of the at least one terminal device according to the data of the terminal device group.

[0198] For example, the at least one terminal device is a terminal device whose transmission delay is greater than a second preset threshold or less than a third preset threshold, where the second preset threshold is greater than the third preset threshold.

[0199] In addition, similar to method 1, after the bitrates of the QoS flows of the at least one terminal device are adjusted, the sum of the bitrates of the QoS flows of the terminal devices in the terminal device group is less than or equal to the subscribed maximum bitrate, or the sum of the lower bounds of the value ranges of the bitrates of the QoS flows of the terminal devices in the terminal device group is less than or equal to the subscribed maximum bitrate.

[0200] In some possible embodiments, if the first request message includes a trigger condition for sending aggregation information, and the trigger condition for sending aggregation information includes a transmission delay statistic greater than a first preset threshold, then the second network element or the user plane network element determines whether the transmission delay statistic is greater than the first preset threshold based on the first preset threshold. If so, the second network element sends data of the terminal device group to the first network element, and the data of the terminal device group includes the transmission delay statistic. In some possible embodiments, if the first request message does not include a trigger condition for sending aggregation information, then the first network element determines whether the transmission delay statistic is greater than the first preset threshold based on the first preset threshold.

[0201] In some possible embodiments, when the first network element adjusts the bit rate of the QoS stream of at least one terminal device, if the first request message includes indication information, the second network element also feeds back the measurement results of the QoS parameters of each terminal device in the terminal device group to the first network element. The first network element can determine the adjustment scheme by combining the measurement results of the QoS parameters of each terminal device in the terminal device group, that is, which terminal devices' QoS stream bit rates are adjusted, and the direction and magnitude of the adjustment. For example, for terminal devices with long transmission delays (e.g., terminal devices with transmission delays greater than a second preset threshold), the bit rate is increased to reduce their transmission delay, and / or for terminal devices with short transmission delays (e.g., terminal devices with transmission delays less than a third preset threshold), the bit rate is decreased to increase their transmission delay.

[0202] Alternatively, when the first network element adjusts the bit rate of the QoS stream of at least one terminal device, if the first request message includes indication information and feedback conditions, the second network element also feeds back to the first network element the measurement results of the QoS parameters of the terminal devices in the terminal device group that meet the feedback conditions. The first network element can determine the adjustment scheme by combining the measurement results of the QoS parameters of the terminal devices in the terminal device group that meet the feedback conditions. For example, the feedback conditions are that terminal devices whose transmission delay is greater than a second preset threshold feed back the measurement results of their QoS parameters, and / or terminal devices whose transmission delay is less than a third preset threshold feed back the measurement results of their QoS parameters.

[0203] For example, such as Figure 5A As shown, assuming the variance determined by the weighted transmission delay of UE1, the weighted transmission delay of UE2, and the weighted transmission delay of UE3, or the variance determined by the differential transmission delay of UE1, the differential transmission delay of UE2, and the differential transmission delay of UE3, is greater than a first preset threshold, then the first network element lowers the bit rate of the QoS stream of UE1, increasing the actual transmission delay of UE1; raises the bit rate of the QoS stream of UE2, shortening the actual transmission delay of UE2; and lowers the bit rate of the QoS stream of UE3, increasing the actual transmission delay of UE3. Figure 5BAs shown, for each UE, the time length from the triangular mark to the hollow circle mark corresponds to the actual transmission delay of the UE before adjustment, the time length from the triangular mark to the solid circle mark corresponds to the actual transmission delay of the UE after adjustment, the time length from the triangular mark to the intersection of the first dashed line and the time axis corresponds to the expected transmission delay of the UE after adjustment, and the time length from the triangular mark to the intersection of the second dashed line and the time axis corresponds to the expected transmission delay of the UE before adjustment. It can be seen that after the first network element adjusts the bit rates of UE1, UE2 and UE3, the times when the QoS flows of UE1, UE2 and UE3 respectively arrive at the server are approximately the same and are close to each other. Further, before adjustment, the server receives the calculation results of all UEs in the terminal device group at the time indicated by the second dashed line, and after adjustment, the server receives the calculation results of all UEs in the terminal device group at the time indicated by the first dashed line. The time indicated by the first dashed line is earlier than the time indicated by the second dashed line, so that the waiting time of the server can be reduced, and the model training efficiency can be improved.

[0204] Step 430: The first network element sends the adjusted QoS parameter of the first terminal device to the policy control network element.

[0205] In addition, the policy control network element also performs corresponding policy control according to the aggregation information of the terminal device group, so as to ensure that the limitations of the aggregation information are met, for example, the aggregated bit rate needs to be less than or equal to the maximum bit rate subscribed by the terminal device group, and the statistical value of the transmission delay needs to be less than or equal to the first preset threshold. Taking the aggregated bit rate as an example, a related policy control method is briefly introduced below.

[0206] Exemplarily, the policy control method comprises:

[0207] 1. When establishing a session management (SM) policy association, the policy control network element checks whether the aggregated bit rate of the terminal device group is higher than the maximum bit rate subscribed by the terminal device group. If yes, the policy control network element should refuse to establish the SM policy association, otherwise, the policy control network element normally establishes the SM policy association.

[0208] 2. When the aggregated bit rate of the terminal device group is less than the subscribed maximum bit rate of the terminal device group, and the aggregated bit rate of the terminal device group is greater than a preset threshold, the policy control network element can limit the traffic of a protocol data unit (PDU) session or a PCC rule by a policy control method, and perform corresponding interaction with a session management network element, wherein the preset threshold is less than the subscribed maximum bit rate of the terminal device group, and the preset threshold is close to the subscribed maximum bit rate of the terminal device group, for example, the preset threshold is 95% of the subscribed maximum bit rate of the terminal device group; if the policy control network element finds that the aggregated bit rate of the terminal device group is less than the preset threshold, the policy control network element can relax the traffic limitation of the PDU session or the PCC rule.

[0209] By using the above embodiment, the first network element subscribes the data of the terminal device group to the second network element, the second network element sends the data of the terminal device group to the first network element, and the first network element adjusts the QoS parameter of at least one terminal device in the terminal device group according to the data of the terminal device group, thereby improving the model training efficiency.

[0210] In addition, before the first network element sends the first request message to the second network element, the first network element can also perform the operation process as shown in Figure 6 .

[0211] Step 600: The first network element sends the subscribed maximum bit rate of the terminal device group to the third network element, and the third network element is a unified data storage network element or a unified data management network element.

[0212] For example, the first network element can determine the subscribed maximum bit rate of the terminal device group according to the model training business requirement of horizontal federated learning, and send the subscribed maximum bit rate of the terminal device group to the third network element. The subscribed maximum bit rate of the terminal device group can be carried by a data management creation request (Nudr_DM_Create Request) or a data management update request (Nudr_DM_Update Request).

[0213] When the subscribed maximum bit rate of the terminal device group is carried by the Nudr_DM_Create Request, the third network element creates the subscribed maximum bit rate of the terminal device group in the subscription information, and sends a data management creation response (Nudr_DM_Create Response) to the first network element, which indicates that the subscription information has been created.

[0214] When the subscription maximum bit rate of the terminal device group is carried by the Nudr_DM_Update Request, the third network element updates the subscription maximum bit rate of the terminal device group in the subscription information, and sends a data management update response (Nudr_DM_Update Response) to the first network element, indicating that the subscription information has been updated.

[0215] Step 610: The first network element sends first information to the policy control function, and the first information includes reference information of QoS parameters corresponding to each terminal device included in the terminal device group.

[0216] The reference information of the QoS parameters corresponding to each terminal device included in the terminal device group includes at least one of the following:

[0217] The value or value range of the bit rate of the QoS flow corresponding to each terminal device included in the terminal device group, or the value or value range of the packet loss rate corresponding to each terminal device included in the terminal device group, or the value or value range of the transmission delay corresponding to each terminal device included in the terminal device group.

[0218] For example, for a certain terminal device, the bit rate of the QoS flow = 20 Mbps, and the packet loss rate = 0.01%. Or, 10 Mbps < bit rate of QoS flow < 20 Mbps, and packet loss rate < 0.01%.

[0219] Exemplarily, the first network element can determine the first information according to the model training service requirement of horizontal federated learning, and the first information can be carried by a policy authorization creation request (Npcf_PolicyAuthorization_Create Request) or a policy authorization update request (Npcf_PolicyAuthorization_Update Request).

[0220] Step 620: The policy control network element requests the subscription maximum bit rate of the terminal device group from the third network element.

[0221] Exemplarily, the policy control network element requests the subscription maximum bit rate of the terminal device group from the third network element through a data management query request (Nudr_DM_Query Request) service operation.

[0222] Step 630: The third network element sends the subscription maximum bit rate of the terminal device group to the policy control network element.

[0223] Exemplarily, the third network element sends the subscription maximum bit rate of the terminal device group to the policy control network element through a data management query response (Nudr_DM_Query Response) service operation.

[0224] Step 640: The policy control network element verifies the first information according to the subscribed maximum bit rate of the terminal device group.

[0225] Exemplarily, when the first information includes the values of the bit rates of the QoS flows respectively corresponding to the terminal devices included in the terminal device group, the policy control network element determines whether the sum of the bit rates of the QoS flows respectively corresponding to the terminal devices included in the terminal device group is less than or equal to the subscribed maximum bit rate of the terminal device group according to the subscribed maximum bit rate of the terminal device group and the values of the bit rates of the QoS flows respectively corresponding to the terminal devices included in the terminal device group, and if yes, the verification succeeds, otherwise, the verification fails.

[0226] Or, when the first information includes the value ranges of the bit rates of the QoS flows respectively corresponding to the terminal devices included in the terminal device group, the policy control network element determines whether the sum of the lower bounds of the value ranges of the bit rates of the QoS flows respectively corresponding to the terminal devices included in the terminal device group is less than or equal to the subscribed maximum bit rate of the terminal device group according to the subscribed maximum bit rate of the terminal device group and the value ranges of the bit rates of the QoS flows respectively corresponding to the terminal devices included in the terminal device group, and if yes, the verification succeeds, otherwise, the verification fails.

[0227] Step 650: When the verification succeeds, the policy control network element generates the PCC rules respectively corresponding to the terminal devices included in the terminal device group according to the first information, and the second network element sends the PCC rules respectively corresponding to the terminal devices to the session management network element.

[0228] Further, the session management network element can establish or modify the corresponding session according to the PCC rules respectively corresponding to the terminal devices sent by the policy control network element.

[0229] Step 660: The policy control network element sends second information to the first network element, and the second information indicates that the corresponding PCC rules respectively corresponding to the terminal devices included in the terminal device group have been generated.

[0230] Exemplarily, when the first information is carried by the Npcf_PolicyAuthorization_Create Request, the second information can be carried by the Npcf_PolicyAuthorization_Create Response.

[0231] When the first information is carried by the Npcf_PolicyAuthorization_Update Request, the second information can be carried by a policy authorization update response (Npcf_PolicyAuthorization_Update Response).

[0232] In addition, in order to adjust the QoS parameters of different participants, the efficiency of model training of federated learning is improved. The application further provides a communication method, as shown in Figure 7 The first network element can be an application function network element, as shown in Figure 7 In the embodiment shown in the figure, the first network element subscribes to the data analysis function network element for the data of the terminal device group, and provides address information of the policy control network element. The second network element feeds back the data of the terminal device group to the policy control network element. The policy control network element adjusts the QoS parameters of the first terminal device in the terminal device group according to the data of the terminal device group. Compared with Figure 4 In the embodiment shown in the figure, the first network element is responsible for receiving the data of the terminal device group, and adjusting the QoS parameters of the first terminal device in the terminal device group according to the data of the terminal device group. Figure 7 In the embodiment shown in the figure, the policy control network element is responsible for receiving the data of the terminal device group, and adjusting the QoS parameters of the first terminal device in the terminal device group according to the data of the terminal device group.

[0233] Specifically, the method comprises:

[0234] Step 700: The first network element sends a first request message to the data analysis function network element, the first request message being used to request the data of the terminal device group, and the data of the terminal device group comprising aggregated information of measurement results of QoS parameters.

[0235] Wherein, step 700 can refer to the related description in the above step 400, which will not be repeated here.

[0236] It should be noted that, in addition to the above 1-8 contents, the first request message also needs to include the address information of the policy control network element.

[0237] Step 710: The data analysis function network element sends the data of the terminal device group to the policy control network element according to the address information of the policy control network element.

[0238] Wherein, the specific implementation of the data analysis function network element determining the data of the terminal device group can refer to the related description in the above step 410.

[0239] Step 720: The policy control network element receives the information of the terminal device group, and adjusts the QoS parameters of the first terminal device in the terminal device group according to the data of the terminal device group.

[0240] The specific implementation manner in which the policy control network element adjusts the QoS parameter of the first terminal device in the terminal device group according to the data of the terminal device group can refer to the specific implementation manner in which the first network element adjusts the QoS parameter of the first terminal device in the terminal device group according to the data of the terminal device group in step 420.

[0241] With the above embodiments, the first network element subscribes to the data analysis function network element for data of a terminal device group, and notifies the second network element to send the data of the terminal device group to the policy control network element, and the policy control network element adjusts the QoS parameter of at least one terminal device in the terminal device group according to the data of the terminal device group, thereby improving the efficiency of model training of federated learning.

[0242] It can be understood that before step 700 is performed, the first network element can also perform the operation flow as shown in Figure 6 .

[0243] The present application also provides another communication method, as shown in Figure 8 , wherein the first network element can be an application function network element. In the embodiment as shown in Figure 8 , the first network element subscribes to the policy control network element for data of a terminal device group, and the policy control network element obtains the data of the terminal device group and adjusts the QoS parameter of the first terminal device in the terminal device group according to the data of the terminal device group. Therefore, compared with the embodiment as shown in Figure 4 , the first network element is responsible for receiving the data of the terminal device group and adjusting the QoS parameter of the first terminal device in the terminal device group according to the data of the terminal device group, and in the embodiment as shown in Figure 8 , the policy control network element is responsible for obtaining the data of the terminal device group and adjusting the QoS parameter of the first terminal device in the terminal device group according to the data of the terminal device group.

[0244] The method comprises:

[0245] Step 800: The first network element sends a first request message to the policy control network element, the first request message being used to request data of a terminal device group, and the data of the terminal device group comprising aggregated information of measurement results of QoS parameters.

[0246] The step 800 can refer to the related description in the above step 400, which will not be repeated here.

[0247] Step 810: The policy control network element determines the data of the terminal device group.

[0248] The specific implementation manner in which the policy control network element determines the data of the terminal device group can refer to the related description of the specific implementation manner in which the second network element determines the data of the terminal device group in the above step 410.

[0249] Step 820: The policy control network element adjusts the QoS parameter of the first terminal device in the terminal device group according to the data of the terminal device group.

[0250] The specific implementation of the policy control network element adjusting the QoS parameter of the first terminal device in the terminal device group according to the data of the terminal device group can refer to the specific implementation of the first network element adjusting the QoS parameter of the first terminal device in the terminal device group according to the data of the terminal device group in step 420.

[0251] With the above embodiment, the first network element subscribes to the data of the terminal device group from the policy control network element, the policy control network element obtains the data of the terminal device group, and adjusts the QoS parameter of at least one terminal device in the terminal device group according to the data of the terminal device group, thereby improving the model training efficiency.

[0252] It can be understood that before step 800 is performed, the first network element can also perform an operation process as shown in Figure 6 .

[0253] The present application provides another communication method. Different from the above embodiment, the application function network element can also not collect data through the NWDAF or the PCF, but itself statistically or obtain the transmission delay and the local calculation delay of each terminal device in the terminal device group, and then determine the adjustment strategy of the QoS parameter of the terminal device in the terminal device group, thereby reducing the number of messages exchanged between network elements, reducing the network element load, and improving the QoS parameter adjustment efficiency of the terminal device, so that the sum of the local calculation delay and the transmission delay of different terminal devices is roughly the same, thereby improving the model training efficiency of horizontal federated learning.

[0254] As shown in Figure 9 , the method comprises:

[0255] Step 900: The application function network element obtains the transmission delay and the local calculation delay of each terminal device in the terminal device group, and the terminal device in the terminal device group is a terminal device participating in horizontal federated learning.

[0256] The local calculation delay is the time length required to determine the update parameter of the model of horizontal federated learning. The transmission delay is the time length required to transmit the update parameter to the application function network element. It can be understood that for different terminal devices in the terminal device group, the local calculation delay of different terminal devices may be different, and the transmission delay of different terminal devices may also be different. Illustratively, as shown in Figure 5AAs shown, the terminal device group includes UE1, UE2 and UE3, each UE corresponds to a time axis, the time length from the starting point of the time axis (i.e. the local calculation start time) to the triangular mark is the local calculation completion time of the UE, i.e. the local calculation delay, and the time length from the triangular mark to the circular mark is the transmission delay of the UE. The local calculation start time is the starting time of determining the updated parameters of the model of the horizontal federated learning.

[0257] For example, the application function network element can determine the transmission delay of the terminal device sending the data packet carrying the updated parameters according to the sending time of the data packet and the receiving time of the data packet. For details, refer to the related description in Figure 10 .

[0258] For example, the application function network element can obtain the local calculation delay of the terminal device participating in the horizontal federated learning before the horizontal federated learning starts, or obtain the local calculation delay of the terminal device participating in the horizontal federated learning in the model training stage of the horizontal federated learning. For details, refer to the related description in Figure 10 .

[0259] Step 910: The application function network element adjusts the QoS parameter of the first terminal device in the terminal device group according to the transmission delay and the local calculation delay of each terminal device.

[0260] For example, when the application function network element adjusts the QoS parameter of the first terminal device in the terminal device group according to the transmission delay and the local calculation delay of each terminal device, the application function network element determines the statistical value of the total delay of each terminal device according to the transmission delay and the local calculation delay of each terminal device, wherein the total delay of each terminal device is the sum of the transmission delay and the local calculation delay of the terminal device. In the case that the total delay of the first terminal device is greater than a first preset threshold or less than a second preset threshold, the application function network element adjusts the QoS parameter of the first terminal device, wherein the first preset threshold and the second preset threshold are determined according to the statistical value.

[0261] The statistical value can be an arithmetic mean, or a geometric mean, or a root mean square average, or a harmonic mean, or a weighted average, etc., which is not limited in the present application.

[0262] The first preset threshold and the second preset threshold are determined by the application function network element. For example, the first preset threshold can be the statistical value of the total delay + 20ms, or the statistical value of the total delay * 1.12, etc. For example, the second preset threshold can be the statistical value of the total delay - 20ms, or the statistical value of the total delay * 0.88, etc.

[0263] The adjusted QoS parameter of the first terminal device comprises at least one of a resource type, a priority, a packet delay budget, a guaranteed bit rate, a maximum bit rate, and an allocation and pre-emption priority.

[0264] The resource type can comprise GBR, Non-GBR, and Delay-critical GBR. By adjusting the type of the QoS flow of the first terminal device from Non-GBR to GBR or Delay-critical GBR, the network can be ensured to reserve sufficient resources for the QoS flow of the first terminal device to guarantee its bandwidth, thereby improving the probability that the set transmission delay is guaranteed.

[0265] The priority can refer to the priority of resource scheduling, i.e., the priority level parameter in 5QI. The allocation and pre-emption priority refers to the priority of resource allocation and pre-emption. By adjusting the priority of the QoS flow of the first terminal device or the allocation and pre-emption priority, the network can be ensured to preferentially process the QoS flow with a high priority or allocation and pre-emption priority in a congestion scenario, and thus the probability that the set transmission delay is guaranteed can be improved by increasing the priority or the allocation and pre-emption priority of the QoS flow in a case where the network resources are limited.

[0266] The packet delay budget can be adjusted to reduce the transmission delay of the QoS flow of the first terminal device, or the packet delay budget can be adjusted to increase the transmission delay of the QoS flow of the first terminal device.

[0267] The guaranteed bit rate and the maximum bit rate can be adjusted to ensure the stability of the bit rate of the QoS flow, thereby ensuring the stability of the delay.

[0268] In addition, the application function network element can also adjust the parameters such as the packet error rate, the average window, and the maximum data burst.

[0269] Step 920: The application function network element sends the adjusted QoS parameter of the first terminal device to the policy control network element.

[0270] In addition, in a possible implementation, the application function network element also adjusts the QoS reference of the first terminal device in the terminal device group according to the transmission delay and the local calculation delay of each terminal device. The QoS reference does not specify a specific QoS parameter, which is mapped to a specific QoS parameter by the policy control network element. That is, the policy control network element can adjust the QoS parameter of the first terminal device in the terminal device group according to the QoS reference of the first terminal device. At this time, the application function network element sends the adjusted QoS reference of the first terminal device to the policy control network element.

[0271] For example, the application function network element sends the adjusted expected transmission delay of the first terminal device to the policy control network element, and the policy control network element maps the expected transmission delay of the first terminal device to specific QoS parameters, such as at least one of resource type, priority, packet delay budget, guaranteed flow bit rate, maximum flow bit rate, and allocation and preemption priority.

[0272] The following describes the embodiments with reference to the accompanying drawings: Figure 10 The following describes the embodiments with reference to the accompanying drawings: Figure 9 The following describes the embodiments with reference to the accompanying drawings:

[0273] In this embodiment, before the model training phase of the horizontal federated learning, the application function network element can set different expected transmission delays for each terminal device according to the local computing delays of the terminal devices participating in the horizontal federated learning, so that the sum of the local computing delay and the expected transmission delay of different terminal devices is approximately the same, so as to improve the model training efficiency of the horizontal federated learning. In the model training phase of the horizontal federated learning, the application function network element can obtain the sum of the local computing delay and the actual transmission delay of different terminal devices, i.e., the total delay. Further, the application function network element can calculate the statistical value of the total delay of different terminal devices, i.e., the expected total delay, and filter out terminal devices with a large difference between the total delay and the expected total delay, so that the total delay of these terminal devices can be approximately the same as the expected total delay by adjusting the QoS parameters of these terminal devices.

[0274] S1001.AF obtains the expected transmission delay of each UE according to the local computing delay and the initial total delay of each UE in the terminal device group, and sets the QoS reference value or QoS parameter value of the QoS Flow for the federated learning service of each UE accordingly.

[0275] In this embodiment, the AF can obtain the local computing delay and the initial total delay of the UE participating in the horizontal federated learning in the test phase. The test phase refers to the period of time before the AF performs the model training of the horizontal federated learning with the UE after the AF determines the UE participating in the horizontal federated learning and establishes a session connection with the UE participating in the horizontal federated learning. In addition, the test phase can also be referred to as the preparation phase. For example, in the test phase, the AF and the UE interact through application layer data packets, and the local computing delay and the initial total delay of the UE can be obtained, and then the local computing delay and the initial total delay of each UE participating in the horizontal federated learning can be obtained.

[0276] Exemplarily, in order to achieve that the sum of the local computing latency and the transmission latency of different UEs is approximately the same, the AF can set a deadline to achieve that the intermediate result (i.e. the updated parameter) of each UE reaches the AF before the deadline. Further, the AF can calculate the expected transmission latency of each UE according to the deadline and the local computing latency of each UE, and set the QoS reference value or the QoS parameter value of the QoS Flow of each UE according to the expected transmission latency of each UE, for example, the PDB.

[0277] For example, the terminal device group includes UE1, UE2 and UE3, wherein the local computing latency of UE1 is 20 ms, the total latency of UE1 is 90 ms, the local computing latency of UE2 is 40 ms, the total latency of UE2 is 70 ms, the local computing latency of UE3 is 30 ms, and the total latency of UE3 is 80 ms. The AF sets the deadline according to the above three total latencies, for example, the AF calculates the average of the above three total latencies as 80 ms, and sets the deadline as 80 ms, i.e. the time length required from the local computing start time to the time when the data packets of all UEs are received is 80 ms. Then the expected transmission latency of UE1 is 60 ms, the expected transmission latency of UE2 is 40 ms, and the expected transmission latency of UE3 is 50 ms. The AF sets the PDB of the QoS Flow of UE1 according to the expected transmission latency of UE1. The AF sets the PDB of the QoS Flow of UE2 according to the expected transmission latency of UE2, and the AF sets the PDB of the QoS Flow of UE3 according to the expected transmission latency of UE3. For example, the PDB of the QoS Flow of UE2 < the PDB of the QoS Flow of UE3 < the PDB of the QoS Flow of UE1.

[0278] S1002. The AF sends a request message to the PCF, wherein the request message includes the QoS reference value or the QoS parameter value.

[0279] Exemplarily, the AF sends the QoS reference value or the QoS parameter value of each UE in the terminal device group to the PCF through the PCF service interface policy authorization creation request (Npcf_PolicyAuthorization_Creat Request), or the AF sends the QoS reference value or the QoS parameter value of the QoS parameter of each UE in the terminal device group to the PCF through the PCF service interface policy authorization update request (Npcf_PolicyAuthorization_Update Request).

[0280] The request message includes a UE identifier (UE ID), an application session identifier (App session ID), a QoS reference or QoS parameter, wherein the UE identifier is used to indicate the UE whose QoS parameter is adjusted. The application session identifier (App session ID) is used to indicate the QoS parameter of the session of the application being adjusted. The QoS reference or QoS parameter indicates the QoS parameter being adjusted, for example, the QoS reference is the transmission delay of the terminal device, and the QoS parameter is the PDB.

[0281] S1003. The PCF sets the corresponding QoS parameter according to the request message, and generates a PCC rule.

[0282] S1004. The PCF sends the PCC rule to the SMF, so that the SMF binds the PCC rule to the corresponding QoS Flow.

[0283] S1005. The PCF sends a response message to the AF. The response message indicates that the QoS parameter setting is successful.

[0284] For example, when the AF sends the QoS reference value or the QoS parameter value of each UE in the terminal device group to the PCF through the PCF service interface policy authorization creation request, the PCF notifies the AF of the success of the QoS parameter setting through the PCF service interface policy authorization creation response (Npcf_PolicyAuthorization_Create Response). Alternatively, when the AF sends the QoS reference value or the QoS parameter value of the QoS parameter of each UE in the terminal device group to the PCF through the PCF service interface policy authorization update request, the PCF notifies the AF of the success of the QoS parameter setting through the PCF service interface policy authorization update response (Npcf_PolicyAuthorization_Update Response).

[0285] In addition, in some embodiments, the PCF can also notify the AF of the failure of the QoS parameter setting, for example, the network resource is insufficient, and the QoS parameter specified by the AF cannot be met.

[0286] After S1005, when the QoS parameter setting of the QoS Flow of the transverse federated learning of the terminal device is completed, the AF performs model training of the transverse federated learning with each UE in the terminal device group.

[0287] S1006. The AF obtains the transmission delay and the local calculation delay of each UE in the terminal device group.

[0288] In a possible design, when the UE sends a data packet to the AF, the sending time of the data packet is carried in the data packet as a timestamp, and the AF can calculate the transmission delay of the data packet according to the receiving time of the data packet and the timestamp. The data packet carries the updated parameters of the model of the federated learning.

[0289] In a possible design, the AF can obtain the local computing delay of the UE in the test phase.

[0290] In a possible design, when the UE sends a data packet to the AF, the sending time of the data packet is carried in the data packet as a timestamp, and the AF can determine the local computing delay of the UE according to the timestamp and the local computing start time. The data packet carries the updated parameters of the model of the federated learning. For example, the local computing delay of each UE that the AF can obtain can be the average value of the local computing delays of all data packets in a training round. It can be understood that the local computing start times of each UE participating in the federated learning are approximately the same.

[0291] S1007. The AF adjusts the QoS reference value or the QoS parameter value of the UE1 in the terminal device group according to the transmission delay and the local computing delay of each UE.

[0292] The AF calculates the sum of the transmission delay and the local computing delay of each UE as the total delay of the UE, and further determines a statistical value of the total delays of the UEs, i.e., an expected total delay. The statistical value of the total delays of the UEs can be an arithmetic mean, a geometric mean, a root mean square, a harmonic mean, or a weighted average of the total delays of the UEs.

[0293] Further, in a case where the total delay of the UE1 is greater than a first preset threshold or less than a second preset threshold, the AF adjusts the QoS reference value or the QoS parameter value of the UE1.

[0294] For example, assuming that the statistical value is an arithmetic mean, the AF determines that the arithmetic mean of the total delays of the UEs is 50 ms, sets the first preset threshold as the arithmetic mean+10 ms, i.e., 60 ms, and sets the second preset threshold as the arithmetic mean-10 ms, i.e., 40 ms, and the AF adjusts the QoS reference value or the QoS parameter value of the UE whose total delay is less than 40 ms or greater than 60 ms.

[0295] For example, for a UE whose total latency is less than 40 ms, the AF can increase the PDB corresponding to the UE, and / or decrease the ARP of the QoS Flow of the UE.

[0296] For example, for a UE whose total latency is greater than 60 ms, the AF can decrease the PDB corresponding to the UE, and / or increase the ARP of the QoS Flow of the UE. By increasing the ARP of the QoS Flow of the UE, the probability of reaching the set expected transmission latency can be improved. In addition, the AF can set the QoS parameters such as GBR and / or MBR of the QoS Flow to ensure the stability of the QoS Flow bit rate, and thus ensure the stability of the latency. The GBR can also be GFBR, and the MBR can also be MFBR.

[0297] S1008. The AF sends a request message to the PCF, where the request message is used to request to adjust the QoS parameters of UE1.

[0298] For example, the AF sends the adjusted QoS reference value or QoS parameter value of UE1 to the PCF through the PCF service interface policy authorization update request (Npcf_PolicyAuthorization_Update Request).

[0299] The request message includes the UE1 ID, the App session ID, the adjusted QoS reference value or QoS parameter value.

[0300] S1009. The PCF sets the corresponding QoS parameters according to the request message, and generates a PCC rule.

[0301] In addition, the PCF updates the corresponding QoS parameters according to the request message. In some possible embodiments, the PCF can change the type of the QoS Flow from Non-GBR to GBR or Delay-critical GBR, and generate an updated PCC rule. For example, only the GBR and Delay-critical GBR type of QoS Flow has the GBR and / or MBR parameters, so if the AF specifies the value of GBR and / or MBR, or the AF sets the reference value of the flow bit rate, it means that if the type of the QoS Flow is Non-GBR before, the PCF will change the type of the QoS Flow to GBR or Delay-critical GBR.

[0302] S1010. The PCF sends the updated PCC rule to the SMF, so that the SMF binds the updated PCC rule to the corresponding QoS Flow.

[0303] S1011. The PCF sends a response message to the AF. The response message indicates that the QoS parameter setting is successful.

[0304] Exemplarily, the PCF notifies the AF of the successful QoS parameter setting through a PCF service interface policy authorization update response (Npcf_PolicyAuthorizationUpdate Response) service operation.

[0305] By using the above method, when determining the QoS parameter adjustment policy of the terminal device, the application function network element does not need to rely on the NWDAF or the PCF network element to obtain the data of the terminal device from the UPF, which can reduce the burden of other network elements (such as NWDAF, PCF, and UPF), and improve the QoS parameter adjustment efficiency of the terminal device.

[0306] Figure 11 A possible exemplary block diagram of a communication apparatus involved in the embodiments of the present application is shown, the apparatus 1100 includes: a transceiver module 1120 and a processing module 1110, the transceiver module 1120 can include a receiving unit and a sending unit. The processing module 1110 is used to control and manage the actions of the apparatus 1100. The transceiver module 1120 is used to support the communication between the apparatus 1100 and other network entities. Optionally, the apparatus 1100 can also include a storage unit, which is used to store the program code and data of the apparatus 1100.

[0307] Optionally, each module in the apparatus 1100 can be implemented by software.

[0308] Optionally, the processing module 1110 can be a processor or a controller, for example, can be a general central processing unit (CPU), a general processor, a digital signal processing (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, a transistor logic device, a hardware component or any combination thereof. It can implement or execute various exemplary logical blocks, modules and circuits described in combination with the disclosure of the embodiments of the present application. The processor can also be a combination of computing functions, such as a combination of one or more microprocessors, a combination of DSP and microprocessor, etc. The transceiver module 1120 can be a communication interface, a transceiver or a transceiver circuit, etc., wherein the communication interface is collectively referred to, and in specific implementation, the communication interface can include multiple interfaces, and the storage unit can be a memory.

[0309] When the apparatus 1100 is a first network element or a chip in the first network element, the processing module 1110 in the apparatus 1100 can support the apparatus 1100 to perform the actions of the first network element in each method example above, for example, can support the apparatus 1100 to perform step 400 or step 420 in the method in Figure 4 .

[0310] The transceiver module 1120 can support the apparatus 1100 to communicate with a second network element, for example, the transceiver module 1120 can support the apparatus 1100 to perform step 400, step 410, step 430 in the method in Figure 4 .

[0311] For example, the transceiver module 1120 is configured to send a first request message to a second network element, the first request message is used to request data of a terminal device group, wherein the terminal devices in the terminal device group are terminal devices participating in horizontal federated learning; receive the data of the terminal device group from the second network element, the data of the terminal device group includes aggregated information of measurement results of QoS parameters; the processing module 1110 is configured to adjust the QoS parameters of a first terminal device in the terminal device group according to the data of the terminal device group; the transceiver module 1120 is configured to send the adjusted QoS parameters of the first terminal device to a policy control network element.

[0312] In a possible design, the first request message includes indication information used to indicate a measurement result of a QoS parameter of the feedback terminal device; the data of the terminal device group further includes a measurement result of a QoS parameter of a second terminal device in the terminal device group, and the second terminal device includes the first terminal device.

[0313] In a possible design, the first request message further includes a feedback condition used to indicate a condition required to be met for feeding back the measurement result of the QoS parameter of the terminal device; and the measurement result of the QoS parameter of the second terminal device meets the feedback condition.

[0314] In a possible design, the aggregation information includes an aggregated bit rate used to represent a sum of bit rates of QoS flows of terminal devices in the terminal device group; and the processing module 1110 is configured to, when adjusting the QoS parameter of the first terminal device in the terminal device group according to the data of the terminal device group, adjust a bit rate of a QoS flow of the first terminal device according to the data of the terminal device group if the aggregated bit rate is greater than a maximum bit rate subscribed by the terminal device group.

[0315] In a possible design, the aggregation information includes a statistical value of a transmission delay of terminal devices in the terminal device group; and the processing module 1110 is configured to, when adjusting the QoS parameter of the first terminal device in the terminal device group according to the data of the terminal device group, adjust a bit rate of a QoS flow of the first terminal device according to the data of the terminal device group if the statistical value of the transmission delay is greater than a first preset threshold.

[0316] In a possible design, the first terminal device is a terminal device with a transmission delay greater than a second preset threshold.

[0317] It should be understood that the apparatus 1100 according to the embodiments of the present application can correspond to the first network element in the foregoing method embodiments, and the operations and / or functions of each module in the apparatus 1100 are respectively to implement the corresponding steps of the methods of the first network element in the foregoing method embodiments, and thus can also achieve the beneficial effects in the foregoing method embodiments, and for brevity, no further description is given here.

[0318] When the apparatus 1100 is the second network element or a chip in the second network element, the processing module 1110 in the apparatus 1100 can support the apparatus 1100 to perform the actions of the second network element in the foregoing method examples.

[0319] The transceiver module 1120 can support the apparatus 1100 to communicate with the first network element, for example, the transceiver module 1120 can support the apparatus 1100 to perform the following actions. Figure 4Step 400, Step 410, Step 430 in the method 1000.

[0320] For example, the processing module 1110 invokes the transceiver module 1120 to perform: receiving, from a first network element, a first request message, the first request message being used to request data of a terminal device group, wherein terminal devices in the terminal device group are terminal devices participating in horizontal federated learning; and sending the data of the terminal device group, the data of the terminal device group including aggregated information of measurement results of QoS parameters.

[0321] In a possible design, the first request message includes indication information, the indication information being used to indicate that measurement results of QoS parameters of a terminal device are fed back; and the data of the terminal device group further includes measurement results of QoS parameters of a second terminal device in the terminal device group, the second terminal device including the first terminal device.

[0322] In a possible design, the first request message further includes a feedback condition, the feedback condition being used to indicate a condition required to be met for feeding back the measurement results of QoS parameters of the terminal device; and the measurement results of QoS parameters of the second terminal device meet the feedback condition.

[0323] In a possible design, the aggregated information includes an aggregated bit rate, the aggregated bit rate being used to represent a sum of bit rates of QoS flows of terminal devices in the terminal device group.

[0324] In a possible design, the aggregated information includes a statistical value of transmission delays of terminal devices in the terminal device group.

[0325] In a possible design, the transceiver module 1120 is further configured to send, to a user plane network element, a second request message, the second request message being used to request the data of the terminal device group; and receive, from the user plane network element, the data of the terminal device group.

[0326] In a possible design, the transceiver module 1120 is further configured to send, to a user plane network element, a third request message, the third request message being used to request measurement results of QoS parameters of terminal devices in the terminal device group; and receive, from the user plane network element, the measurement results of QoS parameters of the terminal devices in the terminal device group; and the processing module 1110 is further configured to determine the data of the terminal device group according to the measurement results of QoS parameters of the terminal devices in the terminal device group and the first request message.

[0327] It should be noted that the apparatus 1100 according to the embodiments of the present application can correspond to the second network element in the foregoing method embodiments, and the operations and / or functions of each module in the apparatus 1100 are respectively to implement the corresponding steps of the method of the second network element in the foregoing method embodiments, and thus the beneficial effects in the foregoing method embodiments can also be achieved. For brevity, details are not repeated here.

[0328] When the apparatus 1100 is an application function network element or a chip in an application function network element, the processing module 1110 in the apparatus 1100 can support the apparatus 1100 to perform the actions of the application function network element in each of the method examples above. For example, the processing module 1110 can support the apparatus 1100 to perform steps 900 and 910 in the method in Figure 9 .

[0329] The transceiver module 1120 can support the apparatus 1100 to communicate with a policy control network element, for example, the transceiver module 1120 can support the apparatus 1100 to perform step 920 in the method in Figure 9 .

[0330] For example, the processing module 1110 is configured to obtain a transmission delay and a local calculation delay of each terminal device in a terminal device group, the terminal devices in the terminal device group are terminal devices participating in horizontal federated learning, and adjust a QoS parameter of a first terminal device in the terminal device group according to the transmission delay and the local calculation delay of each terminal device; and the transceiver module 1120 is configured to send the adjusted QoS parameter of the first terminal device to a policy control network element.

[0331] In a possible design, when adjusting the QoS parameter of the first terminal device in the terminal device group according to the transmission delay and the local calculation delay of each terminal device, the processing module 1110 is configured to determine a statistical value of a total delay of each terminal device according to the transmission delay and the local calculation delay of each terminal device, wherein the total delay of each terminal device is a sum of the transmission delay and the local calculation delay of the terminal device; and adjust the QoS parameter of the first terminal device in a case where the total delay of the first terminal device is greater than a first preset threshold or less than a second preset threshold, wherein the first preset threshold and the second preset threshold are determined according to the statistical value.

[0332] In a possible design, the adjusted QoS parameter of the first terminal device includes at least one of a resource type, a priority, a packet delay budget, a guaranteed bit rate, a maximum bit rate, an allocation and preemption priority.

[0333] In a possible design, the local computing delay is a length of time required for determining updated parameters of a model of the horizontal federated learning; and the transmission delay is a length of time required for transmitting the updated parameters to the application function network element.

[0334] It should be understood that the apparatus 1100 according to the embodiments of the present application can correspond to the application function network element in the foregoing method embodiments, and the operations and / or functions of various modules in the apparatus 1100 are respectively to implement corresponding steps of the method of the application function network element in the foregoing method embodiments, and thus can also implement the beneficial effects in the foregoing method embodiments, which will not be repeated here for brevity.

[0335] Figure 12 A schematic structural diagram of a communication apparatus 1200 according to an embodiment of the present application is shown. As shown in the figure, the apparatus 1200 includes a processor 1201. Figure 12

[0336] When the apparatus 1200 is the first network element or a chip in the first network element, in a possible implementation, when the processor 1201 is configured to invoke an interface to perform the following actions: sending, to a second network element, a first request message, the first request message being used to request data of a terminal device group, wherein terminal devices in the terminal device group are terminal devices participating in horizontal federated learning; receiving, from the second network element, the data of the terminal device group, the data of the terminal device group including aggregated information of measurement results of QoS parameters; adjusting a QoS parameter of a first terminal device in the terminal device group according to the data of the terminal device group; and sending, to a policy control network element, the adjusted QoS parameter of the first terminal device.

[0337] It should be understood that the apparatus 1200 can also be used to perform other steps and / or operations of the first network element side in the foregoing embodiments, which will not be repeated here for brevity.

[0338] When the apparatus 1200 is the second network element or a chip in the second network element, in a possible implementation, when the processor 1201 is configured to invoke an interface to perform the following actions:

[0339] receiving, from a first network element, a first request message, the first request message being used to request data of a terminal device group, wherein terminal devices in the terminal device group are terminal devices participating in horizontal federated learning; and sending the data of the terminal device group, the data of the terminal device group including aggregated information of measurement results of QoS parameters.

[0340] It should be understood that the apparatus 1200 can also be used to perform other steps and / or operations of the second network element side in the foregoing embodiments, which will not be repeated here for brevity.

[0341] ​When the apparatus 1200 is an application function network element or a chip in an application function network element, in one possible implementation, when the processor 1201 is configured to invoke an interface to perform the following actions:

[0342] obtain a transmission delay and a local calculation delay of each terminal device in a terminal device group, the terminal devices in the terminal device group being terminal devices participating in horizontal federated learning, adjust a QoS parameter of a first terminal device in the terminal device group according to the transmission delay and the local calculation delay of each terminal device, and send the adjusted QoS parameter of the first terminal device to a policy control network element.

[0343] It should be understood that the apparatus 1200 can also be used to perform other steps and / or operations of the application function network element side in the foregoing embodiments, which are not described herein for brevity.

[0344] It should be understood that the processor 1201 can invoke an interface to perform the foregoing transmission and reception actions, wherein the invoked interface can be a logical interface or a physical interface, and no limitation is made in this regard. Optionally, the physical interface can be implemented through a transceiver. Optionally, the apparatus 1200 further includes a transceiver 1203.

[0345] Optionally, the apparatus 1200 further includes a memory 1202, and the memory 1202 can store program codes in the foregoing method embodiments, so as to facilitate the processor 1201 to invoke.

[0346] Specifically, if the apparatus 1200 includes the processor 1201, the memory 1202 and the transceiver 1203, the processor 1201, the memory 1202 and the transceiver 1203 communicate with each other through internal connection paths to transfer control and / or data signals. In one possible design, the processor 1201, the memory 1202 and the transceiver 1203 can be implemented through chips, and the processor 1201, the memory 1202 and the transceiver 1203 can be implemented in the same chip, or can be implemented in different chips respectively, or any two of them can be combined in one chip. The memory 1202 can store program codes, and the processor 1201 invokes the program codes stored in the memory 1202 to implement the corresponding functions of the apparatus 1200.

[0347] The method disclosed by the embodiments of the present application can be applied to a processor or implemented by the processor. The processor can be an integrated circuit chip having a signal processing capability. In the implementation process, each step of the method embodiments can be completed by an integrated logic circuit or an instruction in the form of software in the processor. The processor can be a general processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, and can also be a system chip (SoC), a central processing unit (CPU), a network processor (NP), a digital signal processing circuit (DSP), a micro controller unit (MCU), a programmable logic device (PLD) or other integrated chip. The disclosed methods, steps and logic block diagrams in the embodiments of the present application can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in conjunction with the embodiments of the present application can be directly embodied as a hardware code processor for execution, or a combination of hardware and software modules in the code processor for execution. The software module can be located in a random memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register or other mature storage medium in the art. The storage medium is located in the memory, and the processor reads the information in the memory, and combines the hardware to complete the steps of the above method.

[0348] It is to be understood that the memory in the embodiments of the present application can be a volatile memory or a nonvolatile memory, or can include both volatile and nonvolatile memory. Among them, the nonvolatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically EPROM (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM) used as an external cache. By way of example, and not limitation, many forms of RAM can be used, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchlink DRAM (SLDRAM), and direct Rambus RAM (DR RAM). It should be noted that the memory of the system and method described herein is intended to include, but not be limited to, these and any other suitable types of memory.

[0349] It should be understood that in the embodiments of the present application, the numbers "first", "second" and the like are only used to distinguish different objects, such as to distinguish different parameter information or messages, and do not limit the scope of the embodiments of the present application, and the embodiments of the present application are not limited thereto.

[0350] It should also be understood that in various embodiments of the present application, the size of the serial number of each process described above does not mean the order of execution, and the execution order of each process should be determined by its function and inherent logic. The various numerical designations or serial numbers involved in the above processes are only for the convenience of description, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0351] It should also be understood that the term "and / or" herein is only a description of the association relationship between the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent the three cases of A alone, A and B together, and B alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after it.

[0352] The meaning of the expression similar to "the item includes one or more of the following: A, B, and C" appearing in this application, if not specifically stated, generally means that the item can be any one of the following: A; B; C; A and B; A and C; B and C; A, B and C; A and A; A, A and A; A, A and B; A, A and C, A, B and B; A, C and C; B and B, B, B and B, B, B and C, C and C; C, C and C, and other combinations of A, B and C. The above is an example of three elements A, B and C to illustrate the optional items of the item. When expressed as "the item includes at least one of the following: A, B, …, and X", that is, when there are more elements in the expression, the applicable items of the item can also be obtained according to the foregoing rules.

[0353] Those skilled in the art can appreciate that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized in electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0354] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0355] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be realized by other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0356] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0357] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit.

[0358] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the parts that essentially contribute to the prior art or the parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the embodiments of the present application. The aforementioned storage medium includes a U disk, a mobile hard disk, a read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk, and various media that can store program codes.

Claims

1. A communication method characterized by comprising: The method comprises: The first network element sends a first request message to the second network element, wherein the first request message is used to request data of a terminal device group, and the terminal devices in the terminal device group are terminal devices participating in horizontal federated learning; The first network element receives the data of the terminal device group from the second network element, wherein the data of the terminal device group comprises aggregated information of measurement results of quality of service (QoS) parameters; The first network element adjusts the QoS parameters of a first terminal device in the terminal device group according to the data of the terminal device group; The first network element sends the adjusted QoS parameters of the first terminal device to a policy control network element.

2. The method of claim 1, wherein, The first request message comprises indication information, which is used to indicate the measurement results of the QoS parameters of the terminal device; The data of the terminal device group further comprises measurement results of the QoS parameters of a second terminal device in the terminal device group, and the second terminal device comprises the first terminal device.

3. The method of claim 2, wherein, The first request message further comprises a feedback condition, which is used to indicate a condition required to be met for feeding back the measurement results of the QoS parameters of the terminal device; The measurement results of the QoS parameters of the second terminal device meet the feedback condition.

4. The method according to any one of claims 1 to 3, characterized in that, The aggregated information comprises an aggregated bit rate, which is used to represent the sum of bit rates of QoS flows of the terminal devices in the terminal device group; The first network element adjusts the QoS parameters of the first terminal device in the terminal device group according to the data of the terminal device group, comprising: When the aggregated bit rate is greater than the maximum bit rate of the terminal device group in a subscription, the first network element adjusts the bit rate of the QoS flow of the first terminal device according to the data of the terminal device group.

5. The method according to any one of claims 1 to 3, wherein The aggregated information comprises a statistical value of transmission delay of the terminal devices in the terminal device group; The first network element adjusts the QoS parameters of the first terminal device in the terminal device group according to the data of the terminal device group, comprising: If the statistical value of the transmission delay is greater than a first preset threshold, the first network element adjusts the bit rate of the QoS flow of the first terminal device according to the data of the terminal device group.

6. The method of claim 5, wherein, The first terminal device is a terminal device with a transmission delay greater than a second preset threshold.

7. The method of claim 6, wherein, The first request message further comprises the first preset threshold and / or the second preset threshold.

8. The method of any one of claims 1-3, wherein, The first request message comprises at least one of the following:

9. The method of claim 8, wherein, An identifier of the terminal device in the terminal device group, an identifier of the terminal device group, identification information of an analysis type, and a type of the aggregated information. The first request message further comprises at least one of the following: A trigger condition for sending the aggregated information; 10. The method of any one of claims 1-3, wherein, An application identifier, which is used to indicate an application corresponding to the measurement results of the QoS parameters of the terminal device.

11. A communication method, comprising: The first network element is an application function network element, and the second network element is a data analysis function network element or a policy control network element. The method comprises: The second network element receives a first request message from the first network element, wherein the first request message is used to request data of a terminal device group, and the terminal devices in the terminal device group are terminal devices participating in horizontal federated learning; The second network element sends data of the terminal device group, and the data of the terminal device group includes aggregated information of measurement results of QoS parameters, so that the first network element adjusts QoS parameters of a first terminal device in the terminal device group according to the data of the terminal device group.

12. The method of claim 11, wherein, The first request message includes indication information, and the indication information is used to indicate measurement results of QoS parameters of a feedback terminal device. The data of the terminal device group further includes measurement results of QoS parameters of a second terminal device in the terminal device group, and the second terminal device includes the first terminal device.

13. The method of claim 12, wherein, The first request message further includes a feedback condition, and the feedback condition is used to indicate a condition required to be met for feeding back the measurement results of the QoS parameters of the terminal device. The measurement results of the QoS parameters of the second terminal device meet the feedback condition.

14. The method according to any one of claims 11 to 13, wherein, The aggregated information includes an aggregated bit rate used to represent a sum of bit rates of QoS flows of terminal devices in the terminal device group.

15. The method according to any one of claims 11 to 13, wherein, The aggregated information includes a statistical value of transmission delays of terminal devices in the terminal device group.

16. The method of any one of claims 11-13, wherein, The first request message includes at least one of the following: identification of a terminal device in the terminal device group, identification of the terminal device group, identification information of an analysis type, and a type of the aggregated information.

17. The method of claim 16, wherein, The first request message further includes at least one of the following: a trigger condition for sending the aggregated information; an application identifier used to indicate an application corresponding to the measurement results of the QoS parameters of the terminal device.

18. The method of any one of claims 11-13, wherein, Further comprising: The second network element sends a second request message to a user plane network element, and the second request message is used to request data of the terminal device group. The second network element receives the data of the terminal device group from the user plane network element.

19. The method of any one of claims 11-13, wherein, Further comprising: The second network element sends a third request message to a user plane network element, and the third request message is used to request measurement results of QoS parameters of terminal devices in the terminal device group. The second network element receives the measurement results of the QoS parameters of the terminal devices in the terminal device group from the user plane network element. The second network element determines data of the terminal device group according to the measurement results of the QoS parameters of the terminal devices in the terminal device group and the first request message.

20. The method of any one of claims 11-13, wherein, The first network element is an application function network element, and the second network element is a data analysis function network element or a policy control network element.

21. A method of communication, comprising: The method comprises: An application function network element obtains transmission delays and local calculation delays of each terminal device in a terminal device group, and the terminal devices in the terminal device group are terminal devices participating in horizontal federated learning. The application function network element adjusts QoS parameters of a first terminal device in the terminal device group according to the transmission delays and the local calculation delays of the terminal devices. The application function network element sends the adjusted QoS parameters of the first terminal device to a policy control network element.

22. The method of claim 21, wherein, The application function network element adjusts QoS parameters of a first terminal device in the terminal device group according to transmission delays and local calculation delays of the terminal devices, comprising: The application function network element determines a statistical value of the total latency of each terminal device according to the transmission latency and the local calculation latency of the terminal device, wherein the total latency of each terminal device is the sum of the transmission latency and the local calculation latency of the terminal device; In a case where the total latency of the first terminal device is greater than a first preset threshold or less than a second preset threshold, the application function network element adjusts the QoS parameter of the first terminal device, wherein the first preset threshold and the second preset threshold are determined according to the statistical value.

23. The method of claim 22, wherein, The adjusted QoS parameter of the first terminal device includes at least one of a resource type, a priority, a packet latency budget, a guaranteed flow bit rate, a maximum flow bit rate, an allocation and preemption priority.

24. The method of any one of claims 21-23, wherein, The local calculation latency is a time length required for determining an update parameter of a model of the horizontal federated learning. The transmission latency is a time length required for transmitting the update parameter to the application function network element.

25. A communications device, characterized by The system comprises a first network element and a second network element; The first network element is configured to perform the method in any one of claims 1 to 10; and The second network element is configured to perform the method in any one of claims 11 to 20.

26. A communications device, characterized by ​ 27. A computer readable storage medium, characterized in that, ​ 28. A communication system, characterized by ​ ​ ​

Images