CN114692898A

CN114692898A - MEC federal learning method, device and computer readable storage medium

Info

Publication number: CN114692898A
Application number: CN202210331863.1A
Authority: CN
Inventors: 马丽萌; 王达; 欧阳晔; 杨爱东; 叶晓舟
Original assignee: Asiainfo Technologies China Inc
Current assignee: Asiainfo Technologies China Inc
Priority date: 2022-03-30
Filing date: 2022-03-30
Publication date: 2022-07-01

Abstract

The embodiment of the application provides an MEC (Multi-agent computer) federal learning method, an MEC federal learning device and a computer readable storage medium. The method comprises the following steps: when any multi-access edge computing MEC system in any federal learning group sends a federal learning request, determining an MEC system coordinator from each MEC system based on the resource capacity utilization rate of each MEC system in the federal learning group; receiving model information sent by other MEC systems in a federal learning group through an MEC system coordinator, and acquiring the resource capacity utilization rate of the MEC system coordinator in real time in the receiving process; and determining whether coordination party switching is needed or not based on the resource capacity utilization rate of the MEC system coordination party, and after the MEC system coordination party aggregates the model information, respectively sending the aggregated model information to each MEC system in the federal learning group. The scheme realizes the MEC federal learning, ensures the performance of the coordination party of the MEC system, and realizes the resource capacity usage balance of the MEC system in the MEC system federal learning group.

Description

MEC federal learning method, device and computer readable storage medium

Technical Field

The application relates to the technical field of 5G edge computing, in particular to an MEC federal learning method, an MEC federal learning device and a computer readable storage medium.

Background

The MEC (Multi-access Edge Computing) is a product of the evolution of the mobile base station and the natural development of the fusion of the IT and the telecommunication network, and by deploying various services and cache contents at the network Edge, the congestion of the mobile core network can be further relieved, and the mobile core network can be efficiently served locally. The MEC provides a new ecosystem and value chain, and operators can open their Radio Access Network (RAN) edge to authenticated and certified third parties, enabling them to be deployed flexibly and quickly to mobile users, enterprises, and vertical industries, providing innovative applications and services, including, for example, video analytics, location services, enhanced displays, local content distribution, and data caching.

Based on the problems of data fragmentation, data orphism, user privacy disclosure, data shortage and the like faced by machine learning, federal learning comes from this, and the federal learning is a distributed machine learning framework, which can allow a plurality of participants to protect own data in local private areas of the participants under the requirements of user privacy protection, data safety and government regulations, and can collaborate and safely establish a federal learning model, and the technology can effectively solve the problem of data island and realize intelligent cooperation among the participants.

Based on the characteristics of MEC such as real-time, agility, intelligence and safety, MEC queues are added into mobile network operators, enterprises and vertical industries, MEC systems are established, and communication among different MEC systems is an essential requirement in the current and future edge computing industries and ecosystems. The federation of MEC systems may enable shared use of MEC services and applications. New functionality is provided by cooperating with other services rather than developing all the services, for example, the speech recognition functionality may serve as a key function for other services (such as navigation applications), in which case the speech recognition service provider need not be the same as the navigation service provider, and each service may be deployed on a different MEC system in the MEC environment.

However, at present, there is no mature MEC federal learning scheme, so it is urgently needed to provide a new MEC federal learning scheme.

Disclosure of Invention

The purpose of this application is to solve at least one of the above technical defects, and the technical solution provided by this application embodiment is as follows:

in a first aspect, an embodiment of the present application provides an MEC federation learning method, including:

when any multi-access edge computing MEC system in any federal learning group sends a federal learning request, determining an MEC system coordinator from each MEC system based on the resource capacity utilization rate of each MEC system in the federal learning group;

receiving model information sent by other MEC systems in a federal learning group through an MEC system coordinator, and acquiring the resource capacity utilization rate of the MEC system coordinator in real time in the receiving process;

if the resource capacity utilization rate of the MEC system coordinator does not exceed a first preset threshold value in the receiving process, after the MEC system coordinator aggregates the model information, respectively sending the aggregated model information to each MEC system in the federal learning group;

if the resource capacity utilization rate of the MEC system coordinator exceeds a first preset threshold value in the receiving process, determining a new MEC system coordinator from each MEC system and switching to the new MEC system coordinator to receive each model information, repeating the determination and switching of the new MEC system coordinator until the resource capacity utilization rate of the new MEC system coordinator does not exceed the first preset threshold value in the receiving process, aggregating each model information through the new MEC system coordinator, and then respectively sending the aggregated model information to each MEC system in the federal learning group.

In an optional embodiment of the present application, determining an MEC system coordinator from each MEC system based on a resource capacity utilization rate of each MEC system in a federal learning group includes:

acquiring a preset available MEC system coordinator set, wherein each MEC system in the preset available MEC system coordinator set carries a corresponding priority;

and determining an MEC system with the resource capacity utilization rate not exceeding a second preset threshold and the highest priority in the MEC systems belonging to the available MEC system coordinator set in the federal learning group as the MEC system coordinator, wherein the second preset threshold is not more than the first preset threshold.

In an optional embodiment of the present application, the method further comprises:

and if the resource capacity utilization rate of all MEC systems belonging to the available MEC system coordinator set in the federal learning group exceeds a second preset threshold, determining the MEC system with the minimum resource capacity utilization rate in the MEC systems not belonging to the available MEC system coordinator set in the federal learning group as the MEC system coordinator.

In an optional embodiment of the present application, the obtaining a preset available MEC system coordinator set includes:

acquiring a plurality of federal learning groups from a preset number of MEC systems;

and determining the MEC systems appearing in any one federal learning group as elements of the available MEC coordinator set, constructing the available MEC coordinator set, wherein the priority of each MEC system in the MEC coordinator set is in direct proportion to the repeated occurrence times of the MEC system in each federal learning group.

In an alternative embodiment of the present application, a plurality of federal learning groups are obtained from a preset number of MEC systems, including:

determining the MEC systems with the same data source in each MEC system as a federal learning group; and/or the presence of a gas in the gas,

and determining the MEC system with the similarity of the model features meeting the preset conditions in each MEC system as a federal learning group.

In an optional embodiment of the present application, the receiving, by the MEC system coordinator, model information sent by other MEC systems in the federal learning group includes:

determining whether an existing MEC system coordinator is included in a federal learning group;

if not, directly receiving model information sent by other MEC systems in the federal learning group through an MEC system coordinator;

and if so, obtaining the model information received by the existing MEC system coordinator through the MEC system coordinator, and continuously receiving the model information of other MEC systems through the MEC system.

In an optional embodiment of the present application, determining a new MEC system coordinator from the MEC systems and switching to the new MEC system coordinator to receive the model information includes:

determining a new MEC system coordinator from each MEC system based on the resource capacity utilization rate of each MEC system in the federal learning group;

and obtaining the model information received by the MEC system coordinator through the new MEC system coordinator, and continuously receiving the model information of other MEC systems through the new MEC system.

In a second aspect, an embodiment of the present application provides an MEC federal learning device, including:

the MEC system coordinator determining module is used for determining an MEC system coordinator from each MEC system based on the resource capacity utilization rate of each MEC system in any federated learning group when any multi-access edge computing MEC system in any federated learning group sends a federated learning request;

the model information receiving module is used for receiving model information sent by other MEC systems in the Federal learning group through the MEC system coordinator and acquiring the resource capacity utilization rate of the MEC system coordinator in real time in the receiving process;

the first model information aggregation module is used for aggregating each model information through the MEC system coordinator and then respectively sending the aggregated model information to each MEC system in the federal learning group if the resource capacity utilization rate of the MEC system coordinator does not exceed a first preset threshold value in the receiving process;

in a third aspect, an embodiment of the present application provides an electronic device, including a memory and a processor;

the memory has a computer program stored therein;

a processor configured to execute a computer program to implement the method provided in the embodiment of the first aspect or any optional embodiment of the first aspect.

In a fourth aspect, this application provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the method provided in the embodiments of the first aspect or any optional embodiment of the first aspect.

In a fifth aspect, embodiments of the present application provide a computer program product or a computer program comprising computer instructions stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device when executing implements the method provided in the embodiment of the first aspect or any optional embodiment of the first aspect.

The beneficial effect that technical scheme that this application provided brought is:

the method comprises the steps of determining an MEC system coordinator based on the resource capacity utilization rate of each MEC system in a initiated Federal learning request group, acquiring the resource capacity utilization rate of the MEC system coordinator in the receiving process in real time in the subsequent process of receiving model information of other MEC systems through the MEC system coordinator, and determining whether to switch the MEC system coordinator based on the resource capacity utilization rate again. According to the scheme, the MEC federal learning is realized, the resource capacity utilization rate is considered in the process of confirming the MEC system coordinator, the performance of the MEC system coordinator is guaranteed, and the balance of resource capacity use of the MEC system in the MEC system federal learning group is realized.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings used in the description of the embodiments of the present application will be briefly described below.

Fig. 1 is a schematic flowchart of an MEC federal learning method according to an embodiment of the present application;

FIG. 2 is a flow diagram of detecting that a requested service is not within an MEC system in an example of an embodiment of the present application;

fig. 3 is a flowchart illustrating federal learning when no handover occurs in the MEC system coordinator in an example of an embodiment of the present application;

FIG. 4 is a flow chart of Federal learning in which a coordinating party for a handover is located in a set of available MEC system coordinators in an example of an embodiment of the present application;

FIG. 5 is a flow chart of Federal learning in which a coordinator of a handover is not located in a set of available MEC system coordinators in an example of an embodiment of the present application;

fig. 6 is a diagram of a MEC system coordinator switching interaction in an example of the embodiment of the present application;

fig. 7 is a federal learning interaction diagram when no handover occurs for the coordination party of the MEC system in an example of the embodiment of the present application;

fig. 8 is a federal learning interaction diagram when a coordination party of the MEC system switches over in an example of the embodiment of the present application

Fig. 9 is a schematic structural diagram of an MEC federal learning device provided in an embodiment of the present application;

fig. 10 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

Reference will now be made in detail to the embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the drawings are exemplary only for the purpose of explaining the present application and are not to be construed as limiting the present application.

As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Further, "connected" or "coupled" as used herein may include wirelessly connected or wirelessly coupled. As used herein, the term "and/or" includes all or any element and all combinations of one or more of the associated listed items.

To make the objects, technical solutions and advantages of the present application more clear, the following detailed description of the embodiments of the present application will be made with reference to the accompanying drawings.

Fig. 1 is a schematic flowchart of an MEC federal learning method provided in an embodiment of the present application, and as shown in fig. 1, the method may include:

step S101, when any multi-access edge computing MEC system in any federal learning group sends a federal learning request, an MEC system coordinator is determined from each MEC system based on the frequency of appearance of the MEC system in each federal learning group and the resource capacity utilization rate.

Specifically, when any MEC system in any federal learning group sends out a federal learning request, the MEC federal learning method provided by the embodiment of the present application is triggered. And then, acquiring the resource capacity utilization rate of each MEC system in the federal learning group, and determining the MEC system serving as the MEC system coordinating party according to the resource capacity utilization rate of each MEC system. The MEC system coordinator will act as the executive body in the federal learning to aggregate all the federal learning participants, i.e., all the MEC systems in the federal learning group.

It can be understood that, in the solution of the present application, when an MEC system coordinator is selected from the federal learning group, in order to ensure the performance of the MEC system coordinator, the resource capacity utilization rate of each MEC system in the federal learning group needs to be considered.

It should be noted that the MEC system is composed of an MEC host and MEC management, and the MEC host is an entity including an MEC platform and a virtualization infrastructure, and provides computing, storage and network resources for running the MEC application. The MEC platform is a collection of basic functions required to run MEC applications on a specific virtualization infrastructure and to enable it to provide and use MEC services, which may also provide services. The MEC application is instantiated on the virtualized infrastructure of the MEC host according to the configuration or request verified by the MEC management layer. The MEC management includes MEC system level management and MEC host level management. The MEC system level management includes a multi-access edge orchestrator as its core component, which outlines the entire MEC system. The MEC host level management includes an MEC platform manager and a virtualization infrastructure manager, responsible for managing MEC specific functions of a specific MEC host and applications running thereon. The interface Mp1 is an interface between the MEC platform and an MEC application program, the interface Mm5 is an interface between the MEC platform manager and the MEC platform, and the interface Mm3 is an interface between the MEC orchestrator and the MEC platform manager.

When an instantiated MEC application in an MEC system in the federal learning group initiates a service request, the service does not exist in the detected MEC system, as shown in fig. 2, the specific detection process is as follows:

1. the service consumer (i.e., an MEC application instantiated in the MEC system) sends a request to the MEC platform using its ID over the Mp1 reference point requesting the desired service.

Service of a corresponding MEC platform discovery request in the MEC system is not available locally.

3. The service request is forwarded to the MEC platform manager of the MEC system over the Mm5 interface.

The MEC platform manager in turn forwards the service request to the MEC orchestrator through the Mm3 interface.

The MEC orchestrator detects topology and available services of the MEC system and discovers that the requested service is not available in the MEC system.

And S102, receiving model information sent by other MEC systems in the federal learning group through the MEC system coordinator, and acquiring the resource capacity utilization rate of the MEC system coordinator in real time in the receiving process.

Specifically, the MEC system coordinator determined in the previous step receives model information sent by other MEC systems in the federal learning group, and after the model information of each MEC system is received, the model information can be aggregated to obtain aggregated model information, so that the federal learning is completed. However, in order to further ensure the performance of the MEC system coordinator, it is necessary to monitor the resource capacity utilization rate of the MEC system coordinator in real time during the receiving process of the data model information, and when the resource capacity utilization rate exceeds a first preset threshold, deactivate the current MEC system coordinator, determine a new MEC system coordinator, and switch from the current MEC system coordinator to the new MEC system coordinator. The subsequent steps of the handover of the MEC system coordinator will be described in detail.

And step S103, if the resource capacity utilization rate of the MEC system coordinator does not exceed a first preset threshold value in the receiving process, aggregating the model information through the MEC system coordinator, and then respectively sending the aggregated model information to each MEC system in the federal learning group.

In one case, the resource capacity utilization rate of the current MEC system coordinator does not exceed the first preset threshold, that is, after the current MEC system coordinator receives all the model information, the resource capacity utilization rate is not greater than the first preset threshold, and the performance of the MEC system coordinator can be ensured. The method can further aggregate the model information through an MEC system coordinator, and then respectively send the aggregated model information to each MEC system in the federal learning group, and each MEC system updates the local model information after receiving the aggregated model information.

It should be noted that the first preset threshold may be set according to experience and actual requirements, for example, set to 60%.

And step S104, if the resource capacity utilization rate of the MEC system coordinator exceeds a first preset threshold value in the receiving process, determining a new MEC system coordinator from each MEC system and switching to receive each model information through the new MEC system coordinator, repeating the determination and the switching of the new MEC system coordinator until the resource capacity utilization rate of the new MEC system coordinator does not exceed the first preset threshold value in the receiving process, aggregating each model information through the new MEC system coordinator, and then respectively sending the aggregated model information to each MEC system in the federated learning group.

In another case, the resource capacity utilization rate of the current MEC system coordinator exceeds a first preset threshold, that is, after the current MEC system coordinator receives all the model information, the resource capacity utilization rate is greater than the first preset threshold, and the performance of the MEC system coordinator may not be guaranteed. Therefore, a new MEC coordinator needs to be determined again, and after the new MEC system coordinator aggregates the model information, the aggregated model information is sent to each MEC system in the federal learning group respectively.

Specifically, the manner of re-determining the new MEC system coordinator may be the same as that in step S101, also taking into account the resource capacity utilization of each MEC system in the federal learning group. The process is repeated, namely after a new MEC system coordinator is determined each time, the MEC system coordinator is switched to the new MEC system coordinator from the previous MEC system coordinator, model information is received through the new MEC system coordinator, the resource capacity utilization rate of the new MEC system coordinator in the process of butt joint is monitored in real time, if the resource capacity utilization rate exceeds a first preset threshold value, the new MEC system coordinator is determined again, the resource capacity utilization rate in the switching and receiving processes is monitored, and the step of determining the new MEC system coordinator, switching and receiving process monitoring can be repeated for multiple times until the resource capacity utilization rate does not exceed the first preset threshold value after the finally determined new MEC system coordinator receives all the model information.

According to the scheme provided by the application, the MEC system coordinator is determined based on the resource capacity utilization rate of each MEC system in the initiated Federal learning request group, the resource capacity utilization rate of the MEC system coordinator in the receiving process is obtained in real time in the subsequent process of receiving model information of other MEC systems through the MEC system coordinator, and whether the MEC system coordinator needs to be switched or not is determined based on the resource capacity utilization rate again. According to the scheme, the MEC federal learning is realized, the resource capacity utilization rate is considered in the process of confirming the MEC system coordination party, the performance of the MEC system coordination party is guaranteed, and the resource capacity use balance of the MEC system in the MEC system federal learning group is realized.

Wherein, obtain and predetermine available MEC system coordinator set, include:

Further, a plurality of federal learning groups are obtained from a preset number of MEC systems, including:

Specifically, firstly, assuming that there are m MEC systems, p MEC system federal learning groups are formed based on the method of the same data source of the business model or based on the similarity of the business model features.

Wherein, the same source of the sample refers to: different MEC systems may have different model samples, and data of different AI model samples are sourced from the same physical device. For example, one MEC system has an image recognition model and another MEC system has a speech recognition model, and although the models are different in different MEC systems, the data of the two models originate from the same physical terminal equipment. The similarity of model features refers to: each MEC system has different models but the same characteristic attributes in all models, for example, different MEC systems have image recognition models from different visual terminals, and although the sources of collected image/video sample data are different, the model samples have the same characteristic attributes.

Then, counting the number N of each MEC system in the p MEC system federal learning groups, wherein each MEC system repeats the number N in the p MEC system federal learning groups, and calculating the weight value of each MEC system in the p MEC system groups as R ═ N/p, wherein the weight value represents the importance of the MEC system in the p MEC system groups, and the larger the weight value is, the more model information the MEC system has is, the more the MEC system has is trained by using a federal learning model, and the federal learning efficiency is improved.

And finally, based on the weight values R of the MEC systems calculated in the steps, sorting the weight values R from large to small, selecting the corresponding MEC system with the weight value R larger than 0, and after the weight values R are sorted, collecting the corresponding MEC systems as { MEC₁，MEC₂……，MEC_n(ii) a n is less than m, defining the MEC system set as an available MEC system coordinator set, and simultaneously determining the priority of the MEC systems contained in the MEC system set, wherein the higher the weight value is, the higher the priority of the MEC systems is.

As can be seen from the foregoing description, the MEC federal learning scheme of the present application can be divided into two cases: one is that the MEC system coordinator does not switch after the MEC system coordinator is determined. Another is that after determining that the MEC system coordinator is switched, the MEC system coordinator further includes: and if the resource capacity utilization rate of all MEC systems belonging to the available MEC system coordinator set in the federal learning group exceeds a second preset threshold, determining the MEC system with the minimum resource capacity utilization rate in the MEC systems not belonging to the available MEC system coordinator set in the federal learning group as the MEC system coordinator. Next, the two cases will be described in detail separately.

As shown in fig. 3, the scheme that no handover occurs (no handover occurs) for the MEC system coordinator may include the following steps:

an MEC system in an MEC federated learning group initiates a federated learning request.

2. And setting a first preset threshold and a second preset threshold of the MEC system resource capacity utilization rate as C1 and C2 respectively, and selecting the MEC system from the federal learning group according to the priority of the available MEC system coordinator set.

3. And judging whether the capacity utilization rate of the currently selected MEC system resources exceeds a second preset threshold value C2.

If the resource capacity utilization rate of the currently selected MEC system exceeds a second preset threshold value C2, returning to step 2, and selecting other MEC systems from the federal learning group again according to the priority of the available MEC system coordinator set.

If the currently selected MEC system does not exceed the second preset threshold C2, determining that the currently selected MEC system is the MEC system coordinator.

4. Each MEC system participant encrypts and sends local calculation model information to the MEC system coordinator, wherein the model information comprises model characteristics, model parameters and other information.

5. In the process of sending the model information, the resource capacity utilization rate of the MEC system coordinator is monitored in real time, the resource capacity utilization rate of the MEC system coordinator is always smaller than a first preset threshold value C1, and at the moment, the MEC system coordinator is not switched.

And 6, the MEC system coordinator aggregates the received model information and then sends the aggregated model information to each MEC system participant.

7. And each MEC participant receives the aggregated model information and updates the local model information.

As shown in fig. 4, the scheme that the MEC system coordinator performs handover and the new MEC system coordinator to be handed over is located in the set of available MEC system coordinators may include the following steps:

And if the resource capacity utilization rate of the currently selected MEC system exceeds a second preset threshold value C2, returning to the step 2, and selecting the MEC system from the federal learning group again according to the priority of the available MEC system coordinator set.

And if the capacity utilization rate of the currently selected MEC system resources does not exceed a second preset threshold value C2, judging whether an MEC system federal learning group contains an existing MEC system coordinator.

Further, if the existing MEC system coordinator is not included in the MEC system federal learning group, the selected MEC system is determined to be the MEC system coordinator.

If the existing MEC system coordination party is included in the MEC system federal learning group, the selected MEC system is determined to be the MEC system switching coordination party, and the existing MEC system coordination party is switched to the determined MEC system coordination party

5. And in the transmission process of the model information, monitoring the resource capacity utilization rate of the MEC system coordinator in real time.

6. And judging whether the resource capacity utilization rate of the MEC system coordinator exceeds a first preset threshold C1.

And if the resource capacity utilization rate of the MEC system coordinator exceeds a first preset threshold value C1, returning to the step 2 for execution.

And if the resource capacity utilization rate of the MEC system coordinator does not exceed a first preset threshold value C1, the MEC system coordinator aggregates the received model information and sends the aggregated model information to each MEC system participant.

As shown in fig. 5, the scheme that the MEC system coordinator is switched to and the new MEC system coordinator to be switched to is not located in the set of available MEC system coordinators may include the following steps:

2. And setting a first preset threshold and a second preset threshold of the MEC system resource capacity utilization rate as C1 and C2 respectively, and selecting the MEC system from the federal learning group according to the priority combined by available MEC system coordinators.

3. And judging whether the capacity utilization rate of the currently selected MEC system resources exceeds a second preset rate threshold value C2.

And if the resource capacity utilization rate of the currently selected MEC system exceeds a second preset threshold value C2, judging whether the resource capacity utilization rates of the MEC systems belonging to preset available MEC coordinators in the federal learning group are all larger than C2.

If all the resource capacity utilization rates of the MEC systems in the federal learning group are greater than C2, selecting the MEC system with the minimum resource capacity utilization rate from the available MEC system coordination set, and executing step 4 to judge whether the federal learning group contains the existing MEC system coordination party.

And if the resource capacity utilization rate of the MEC systems in the federal learning group is not all larger than C2, returning to the step 2 execution, and selecting the MEC systems from the federal learning group again according to the priority of the available MEC system coordinator set.

4. And if the resource capacity utilization rate of the currently selected MEC system does not exceed a second preset threshold value C2, judging whether the Federal learning group contains the existing MEC system coordinator.

And if the existing MEC system coordinator is not contained in the federal learning group, determining that the MEC system selected currently is the MEC system coordinator.

And if the existing MEC system coordinator is included in the federal learning group, determining that the selected MEC system is the MEC system switching coordinator, and switching to the currently selected MEC system coordinator by the existing MEC system coordinator.

5. Each MEC system participant encrypts and sends local calculation model information to the MEC system coordinator, wherein the model information comprises model characteristics, model parameters and other information.

6. And in the transmission process of the model information, monitoring the resource capacity utilization rate of the MEC system coordinator in real time.

7. And judging whether the resource capacity utilization rate of the MEC system coordinator exceeds a first preset threshold C1.

And if the resource capacity utilization rate of the MEC system coordinator exceeds a first preset threshold value C1, returning to the step 2 for execution, and selecting the MEC system from the federal learning group again according to the priority of the available MEC system coordinator set.

8. And each MEC participant receives the aggregated model information and updates the local model information.

In an optional embodiment of the present application, determining a new MEC system coordinator from the MEC systems and switching to receiving the model information through the new MEC system coordinator includes:

Wherein, when an MEC system in the federal learning group initiates a federal learning request, the MEC system coordinator 1 is determined, and other participating parties in the federal learning group transmit models to the MEC system coordinator 1. At this time, it is determined that the handover of the MEC system coordinator needs to be performed, as shown in fig. 6, the process may include the following steps:

1. in the process of transmitting the model information, the resource capacity utilization rate of the MEC system coordinator 1 reaches a first preset threshold value C1.

The MEC system coordinator 1 will inform the various participants to stop sending model information.

And 3, the MEC system coordinator 1 determines the MEC system coordinator 2 (namely a new MEC system coordinator) according to the method for determining the MEC system coordinator.

And 4, the MEC system coordinator 1 sends a switching request to the MEC system coordinator 2.

And 5, the MEC system coordinator 2 sends a response to the MEC system coordinator 1 to agree with the switching request.

And 6, the MEC system coordinator 1 sends the received model information and the model information to the MEC system coordinator 2.

The MEC system coordinator 1 will notify the various participants that the coordinator has changed at this time.

8. And each coordinator 2 participating in the MEC system switching continues to send the model.

The scheme of the present application is further described below by two examples, as shown in fig. 7, no switching occurs between the coordination parties of the MEC system, and the method for the MEC system federal learning may include:

and 1, initiating a service request by an MEC application program, detecting that the service does not exist in an MEC system where the MEC is positioned, and triggering federal learning.

2. In a constructed federal learning group, an MEC system coordinator 1 is determined in an MEC system according to a method for determining the MEC system coordinator.

And 3, the MEC system coordinator 1 sends a federal learning request to each participant in the federal learning group.

4. And the participation direction of each MEC system sends model information to the MEC system coordinator 1, wherein the model information comprises model characteristics, model parameters and other information.

And 5, the MEC system coordinator 1 aggregates the model information sent by each participant.

And 6, the MEC system coordinator 1 sends the aggregated model information to each participant.

7. Each MEC system participant updates local model information.

As shown in fig. 7, the MEC system coordinator performs a handover, and the MEC system federal learning method may include:

the MEC application program initiates a service request, detects that the service does not exist in an MEC system where the MEC is located, and triggers federal learning.

2. In a constructed federal learning group, an MEC system coordinator 1 is determined in an MEC system according to the determined MEC system coordinator.

The MEC system coordinator 1 sends a federated learning request to each participant in the federated learning group.

4. Each MEC system participant sends model information to the MEC system coordinator 1, wherein the model information comprises model types and model parameter information.

5. And in the process of sending the model information, switching of the MEC system coordinator occurs, and the MEC system coordinator 2 is determined according to the second method for determining the MEC system coordinator.

6. According to the MEC system coordinator switching method, the MEC system coordinator is switched from the MEC system coordinator 1 to the MEC system coordinator 2.

7. Each MEC system participant sends model information to the MEC system coordinator 2, wherein the model information comprises model types and model parameter information.

The MEC system coordinator 2 aggregates the received model information.

And 9, the MEC system coordinator 2 sends the aggregated model information to each MEC system participant.

10. Each MEC participant updates local model information.

Fig. 9 is a schematic structural diagram of an MEC federal learning apparatus provided in an embodiment of the present application, and as shown in fig. 9, the apparatus 900 may include: an MEC system coordinator determining module 901, a model information receiving module 902, a first model information aggregating module 903, and a second model information aggregating module 904, where:

the MEC system coordinator determining module 901 is configured to determine an MEC system coordinator from each MEC system based on a resource capacity utilization rate of each MEC system in any federal learning group when any multi-access edge computing MEC system in any federal learning group sends a federal learning request;

the model information receiving module 902 is configured to receive, by an MEC system coordinator, model information sent by other MEC systems in the federal learning group, and obtain, in real time, a resource capacity utilization rate of the MEC system coordinator in a receiving process;

the first model information aggregation module 083 is configured to, if the resource capacity utilization rate of the MEC system coordinator does not exceed a preset threshold value in the receiving process, aggregate each model information by the MEC system coordinator, and then send the aggregated model information to each MEC system in the federal learning group;

the second model information aggregation module 904 is configured to determine a new MEC system coordinator from each MEC system and switch to receive each model information through the new MEC system coordinator if the resource capacity utilization rate of the MEC system coordinator exceeds a preset threshold in the receiving process, repeat the determination and the switching of the new MEC system coordinator until the resource capacity utilization rate of the new MEC system coordinator does not exceed the preset threshold in the receiving process, aggregate each model information through the new MEC system coordinator, and then send the aggregated model information to each MEC system in the federal learning group.

In an optional embodiment of the present application, the MEC system coordinator determination module is specifically configured to:

and determining an MEC system with the resource capacity utilization rate not exceeding a preset threshold and the highest priority in the MEC systems belonging to the available MEC system coordinator set in the federal learning group as the MEC system coordinator.

In an optional embodiment of the present application, the MEC system coordinator determination module is further configured to:

and if the resource capacity utilization rate of all MEC systems belonging to the available MEC system coordinator set in the federal learning group exceeds a preset threshold, determining the MEC system with the minimum resource capacity utilization rate in the MEC systems not belonging to the available MEC system coordinator set in the federal learning group as the MEC system coordinator.

In an optional embodiment of the present application, the model information receiving module is specifically configured to:

In an optional embodiment of the present application, the second model information aggregation module is specifically configured to:

Referring now to fig. 10, a schematic diagram of an electronic device (e.g., a terminal device or a server executing the method shown in fig. 1) 1000 suitable for implementing embodiments of the present application is shown. The electronic device in the embodiments of the present application may include, but is not limited to, a mobile terminal such as a mobile phone, a notebook computer, a digital broadcast receiver, a PDA (personal digital assistant), a PAD (tablet), a PMP (portable multimedia player), a vehicle terminal (e.g., a car navigation terminal), a wearable device, and the like, and a stationary terminal such as a digital TV, a desktop computer, and the like. The electronic device shown in fig. 10 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present application.

The electronic device includes: a memory for storing a program for executing the method of the above-mentioned method embodiments and a processor; the processor is configured to execute programs stored in the memory. The processor may be referred to as a processing device 1001 described below, and the memory may include at least one of a Read Only Memory (ROM)1002, a Random Access Memory (RAM)1003, and a storage device 1008, which are described below:

as shown in fig. 10, the electronic device 1000 may include a processing means (e.g., a central processing unit, a graphics processor, etc.) 1001 that may perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM)1002 or a program loaded from a storage means 1008 into a Random Access Memory (RAM) 1003. In the RAM1003, various programs and data necessary for the operation of the electronic apparatus 1000 are also stored. The processing device 1001, the ROM 1002, and the RAM1003 are connected to each other by a bus 1004. An input/output (I/O) interface 1005 is also connected to bus 1004.

Generally, the following devices may be connected to the I/O interface 1005: input devices 1006 including, for example, a touch screen, touch pad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; an output device 1007 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, and the like; storage devices 1008 including, for example, magnetic tape, hard disk, and the like; and a communication device 1009. The communications apparatus 1009 may allow the electronic device 1000 to communicate wirelessly or by wire with other devices to exchange data. While fig. 10 illustrates an electronic device having various means, it is to be understood that not all illustrated means are required to be implemented or provided. More or fewer devices may alternatively be implemented or provided.

In particular, according to embodiments of the application, the processes described above with reference to the flow diagrams may be implemented as computer software programs. For example, embodiments of the present application include a computer program product comprising a computer program carried on a non-transitory computer readable medium, the computer program containing program code for performing the method illustrated by the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication means 1009, or installed from the storage means 1008, or installed from the ROM 1002. When executed by the processing device 1001, the computer program performs the above-described functions defined in the method of the embodiment of the present application.

It should be noted that the computer readable storage medium mentioned above in the present application may be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In this application, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, optical cables, RF (radio frequency), etc., or any suitable combination of the foregoing.

In some embodiments, the clients, servers may communicate using any currently known or future developed network Protocol, such as HTTP (HyperText Transfer Protocol), and may interconnect with any form or medium of digital data communication (e.g., a communications network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), the Internet (e.g., the Internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future developed network.

The computer readable medium may be embodied in the electronic device; or may exist separately without being assembled into the electronic device.

The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to:

when any multi-access edge computing MEC system in any federal learning group sends a federal learning request, determining an MEC system coordinator from each MEC system based on the resource capacity utilization rate of each MEC system in the federal learning group; receiving model information sent by other MEC systems in a federal learning group through an MEC system coordinator, and acquiring the resource capacity utilization rate of the MEC system coordinator in real time in the receiving process; if the resource capacity utilization rate of the MEC system coordinator does not exceed a preset threshold value in the receiving process, after the MEC system coordinator aggregates the model information, respectively sending the aggregated model information to each MEC system in the federal learning group; if the resource capacity utilization rate of the MEC system coordination party exceeds a preset threshold value in the receiving process, determining a new MEC system coordination party from each MEC system, switching to receive each model information through the new MEC system coordination party, repeating the determination and switching of the new MEC system coordination party until the resource capacity utilization rate of the new MEC system coordination party does not exceed the preset threshold value in the receiving process, aggregating each model information through the new MEC system coordination party, and then respectively sending the aggregated model information to each MEC system in the federal learning group.

Computer program code for carrying out operations for aspects of the present application may be written in any combination of one or more programming languages, including but not limited to an object oriented programming language such as Java, Smalltalk, C + +, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

The modules or units described in the embodiments of the present application may be implemented by software or hardware. Where the name of a module or unit does not in some cases constitute a limitation of the unit itself, for example, the first program switching module may also be described as a "module for switching the first program".

The functions described herein above may be performed, at least in part, by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), systems on a chip (SOCs), Complex Programmable Logic Devices (CPLDs), and the like.

In the context of this application, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific method implemented by the computer-readable medium described above when executed by the electronic device may refer to the corresponding process in the foregoing method embodiments, and will not be described herein again.

Embodiments of the present application provide a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device realizes the following when executed:

The foregoing is only a partial embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.

Claims

1. An MEC federal learning method, comprising:

when any multi-access edge computing MEC system in any federated learning group sends a federated learning request, determining an MEC system coordinator from each MEC system based on the resource capacity utilization rate of each MEC system in the federated learning group;

receiving model information sent by other MEC systems in the federal learning group through the MEC system coordinator, and acquiring the resource capacity utilization rate of the MEC system coordinator in real time in the receiving process;

if the resource capacity utilization rate of the MEC system coordinator does not exceed a first preset threshold value in the receiving process, the MEC system coordinator aggregates the model information and sends the aggregated model information to each MEC system in the Federal learning group respectively;

if the resource capacity utilization rate of the MEC system coordinator exceeds the first preset threshold value in the receiving process, determining a new MEC system coordinator from each MEC system and switching to the new MEC system coordinator to receive each model information, repeating the determination and switching of the new MEC system coordinator until the resource capacity utilization rate of the new MEC system coordinator does not exceed the first preset threshold value in the receiving process, and after the new MEC system coordinator aggregates each model information, respectively sending the aggregated model information to each MEC system in the federal learning group.

2. The method of claim 1, wherein determining an MEC system coordinator from each MEC system based on resource capacity utilization of each MEC system in the federal learning group comprises:

and determining an MEC system with a resource capacity utilization rate not exceeding a second preset threshold and the highest priority in the MEC systems belonging to the available MEC system coordinator set in the federal learning group as the MEC system coordinator, wherein the second preset threshold is not more than the first preset threshold.

3. The method of claim 2, further comprising:

and if the resource capacity utilization rates of all MEC systems belonging to the available MEC system coordinator set in the federal learning group exceed the second preset threshold, determining an MEC system with the minimum resource capacity utilization rate in the MEC systems not belonging to the available MEC system coordinator set in the federal learning group as the MEC system coordinator.

4. The method of claim 2, wherein the obtaining a preset set of available MEC system coordinators comprises:

and determining the MEC systems appearing in any one federated learning group as elements of the available MEC coordinator set, constructing the available MEC coordinator set, wherein the priority of each MEC system in the MEC coordinator set is in direct proportion to the repeated occurrence times of each MEC system in each federated learning group.

5. The method according to claim 4, wherein said obtaining a plurality of federal learning groups from a preset number of MEC systems comprises:

6. The method of claim 1, wherein the receiving, by the MEC system coordinator, model information sent by other MEC systems in the federated learning group comprises:

determining whether an existing MEC system coordinator is included in the federal learning group;

if not, directly receiving model information sent by other MEC systems in the federal learning group through the MEC system coordinator;

7. The method of claim 1, wherein the determining a new MEC system coordinator from the MEC systems and switching to the new MEC system coordinator to receive the model information comprises:

determining the new MEC system coordinator from each MEC system based on the resource capacity utilization rate of each MEC system in the federal learning group;

8. An MEC federal learning device, comprising:

the MEC system coordinator determination module is used for determining an MEC system coordinator from each MEC system based on the resource capacity utilization rate of each MEC system in any federated learning group when any multi-access edge computing MEC system in any federated learning group sends a federated learning request;

the first model information aggregation module is used for aggregating each model information through the MEC system coordinator and then respectively sending the aggregated model information to each MEC system in the federated learning group if the resource capacity utilization rate of the MEC system coordinator does not exceed a first preset threshold value in the receiving process;

and the second model information aggregation module is used for determining a new MEC system coordinator from each MEC system and switching to the new MEC system coordinator to receive each model information if the resource capacity utilization rate of the MEC system coordinator exceeds the first preset threshold in the receiving process, repeating the determination and the switching of the new MEC system coordinator until the resource capacity utilization rate of the new MEC system coordinator does not exceed the first preset threshold in the receiving process, and after the new MEC system coordinator aggregates each model information, respectively sending the aggregated model information to each MEC system in the federated learning group.

9. An electronic device comprising a memory and a processor;

the memory has stored therein a computer program;

the processor for executing the computer program to implement the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, which computer program, when being executed by a processor, carries out the method of any one of claims 1 to 7.