Business application type identification method and device, equipment, storage medium and program product

Abstract

The embodiment of the invention provides a business application type identification method and device, equipment, a storage medium and a program product. The method comprises the following steps: according to a set transmission rule, target model parameters trained by a plurality of network element management systems are uploaded to a network management system, and model selection adopts combination of a generative adversarial network and a classifier model; aggregating target model parameters pre-trained by the plurality of network element management systems through the network management system to obtain global model parameters; and identifying the service application type through a plurality of optical network units based on the global model parameters. According to the embodiment of the invention, target model parameters trained by a plurality of network element management systems are aggregated through a network management system to obtain global model parameters; by means of the mode that the multiple optical network units conduct service application type recognition based on the global model parameters, the recognition accuracy of the service application type can be improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method, device, equipment, storage medium and program product for identifying a business application type. Background Art

[0002] In the face of diverse demands such as improving home broadband quality, differentiated bearer services for small and medium-sized enterprises, and digital transformation of enterprise parks, more accurate and real-time business perception is needed on the access side. With the advent of 5G commercial use and the 6G era, the types of business applications that the network needs to support have expanded to e-sports, augmented reality / virtual reality, live broadcasting, etc., which have higher requirements for network latency and packet loss rate. There are many types and versions of applications, and user traffic is complex and changeable. It is difficult to identify and classify business application types through fixed packet features.

[0003] With the development of artificial intelligence technology, machine learning models are deployed to identify the types of business applications accessing the network. However, the business application data samples of different regions and users accessing the network are different, and new business application data is constantly generated. The machine learning model is degraded and has weak generalization ability, resulting in low accuracy in identifying business application types. Summary of the invention

[0004] The present invention provides a method, device, equipment, storage medium and program product for identifying a business application type, which can improve the accuracy of identifying the business application type.

[0005] In a first aspect, an embodiment of the present disclosure provides a method for identifying a service application type, comprising: uploading target model parameters pre-trained by multiple network element management systems to a network management system according to set transmission rules; aggregating the target model parameters pre-trained by the multiple network element management systems through the network management system to obtain global model parameters; and identifying the service application type based on the global model parameters through multiple optical network units.

[0006] In the second aspect, the embodiments of the present disclosure also provide a device for identifying business application types, including: a parameter uploading module, used to upload target model parameters pre-trained by multiple network element management systems to a network management system according to set transmission rules; an aggregation module, used to aggregate the target model parameters pre-trained by the multiple network element management systems through the network management system to obtain global model parameters; an identification module, used to identify business application types based on the global model parameters through multiple optical network units.

[0007] In a third aspect, an embodiment of the present disclosure further provides an electronic device, comprising: one or more processors; a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement a method for identifying a business application type as described in an embodiment of the present disclosure.

[0008] In a fourth aspect, an embodiment of the present disclosure further provides a storage medium comprising computer executable instructions, which, when executed by a computer processor, are used to execute the method for identifying the business application type as described in the embodiment of the present disclosure.

[0009] In a fifth aspect, an embodiment of the present disclosure further provides a computer program product, including a computer program, which, when executed by a processor, implements the method for identifying the business application type as described in the embodiment of the present disclosure.

[0010] The technical solution of the disclosed embodiment is to upload target model parameters pre-trained by multiple network element management systems to the network management system according to the set transmission rules; the target model parameters pre-trained by the multiple network element management systems are aggregated by the network management system to obtain global model parameters; and the service application types are identified based on the global model parameters by multiple optical network units. The disclosed embodiment is to improve the recognition accuracy of service application types by aggregating target model parameters pre-trained by multiple network element management systems through the network management system to obtain global model parameters; and the service application types are identified based on the global model parameters by multiple optical network units. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0012] Figure 1 A schematic diagram of a flow chart of a method for identifying a service application type provided by an embodiment of the present invention;

[0013] Figure 2 A schematic diagram of a structure of a device for identifying a business application type provided by an embodiment of the present disclosure;

[0014] Figure 3 The present invention is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0015] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0016] It should be understood that the various steps recorded in the method implementation of the present disclosure can be performed in different orders and / or in parallel. In addition, the method implementation may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect. The term "including" and its variations used herein are open inclusions, that is, "including but not limited to". It should be noted that the concepts of "first", "second", etc. mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units. It should be noted that the modifications of "one" and "multiple" mentioned in the present disclosure are illustrative and not restrictive, and those skilled in the art should understand that unless otherwise clearly stated in the context, it should be understood as "one or more". It is understandable that the data involved in the present technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of the corresponding laws, regulations and relevant provisions.

[0017] Figure 1 The present invention provides a flow chart of a method for identifying a service application type. The present invention is applicable to the case of identifying the service application type of optical access network traffic. The method can be executed by a service application type identification device, which can be implemented in the form of software and / or hardware. Optionally, it can be implemented by an electronic device, which can be a mobile terminal, a PC, or a server. Figure 1 As shown, the method includes:

[0018] S110 . Upload target model parameters pre-trained by multiple network element management systems to the network management system according to the set transmission rules.

[0019] In this embodiment, there is no specific restriction on setting transmission rules. For example, setting transmission rules may include a method for multiple network element management systems to upload the target model parameters and a method for encrypting the target model parameters. In this embodiment, there is no restriction on the method for each network element management system to upload the target model parameters. For example, each network element management system may synchronously upload its own target model parameters. There is no restriction on the time point of synchronization. For example, it may be agreed that each target model parameter is uploaded together after the same round of iterations. For another example, each network element management system may asynchronously upload its own target model parameters. For example, each network element management system may upload its own target model parameters immediately after each round of iterations. In this embodiment, there is no restriction on the method for encrypting the target model parameters, such as symmetric encryption, asymmetric encryption, hash algorithm, digital signature, etc.

[0020] In this embodiment, there can be multiple network element management systems. The network element management system is used to receive the initial business application identification model issued by the network management system, and to train the initial business application identification model based on the business application training sample set. After the business application identification model converges, the target model parameters can be obtained.

[0021] The business application identification model includes a generative adversarial network and a classifier; the generative adversarial network includes a generator and a discriminator; the target model parameters include a first target model parameter, a second target model parameter and a third target model parameter; the first target model parameter is a model parameter of a pre-trained intermediate layer of the discriminator; the second target model parameter is a model parameter of a pre-trained generator; the third target model parameter is a model parameter of a pre-trained intermediate layer of the classifier. The first target model parameter and the third target model parameter are the same, that is, the discriminator and the classifier share one model parameter.

[0022] Among them, the number of network management systems is 1, which is used to send the pre-trained initial business application identification model to multiple network element management systems pre-registered in the network management system, and to aggregate the target model parameters uploaded by all network element management systems, and then send the aggregated target model parameters to each network element management system.

[0023] In this embodiment, there is no restriction on the training method of the initial business application identification model, for example, it can be supervised learning, unsupervised learning, reinforcement learning, transfer learning, federated learning, etc. After the initial business application identification model is trained, the target model parameters can be obtained.

[0024] Optionally, the target model parameters are obtained through the following training method: the target model parameters are obtained through the following training method, including: for any network element management system, obtaining a corresponding business application training sample set collected by at least one optical network unit; obtaining a business application identification model pre-issued by the network management system; for the current round of iterative training, training the business application identification model based on the business application training sample set and the aggregated model training parameters obtained in the previous round to obtain corresponding model training parameters; updating the model training parameters by setting an iterative optimization method; after the current round of iterative training is completed, uploading the updated model training parameters to the network management system in accordance with the set transmission rules to obtain the aggregated model training parameters of the current round; wherein the aggregated model training parameters of the current round are obtained by aggregating the updated model training parameters uploaded by multiple network element management systems of the current round by the network management system; when the business application identification model converges, the corresponding model parameters are used as the target model parameters.

[0025] In this embodiment, the training method of each network element management system is similar. Taking any network element management system as an example, the network element management system can receive a service application training sample set sent by an optical line terminal. The optical line terminal can correspond to at least one optical network unit, and the optical line terminal can collect a local service application training sample set through at least one optical network unit. The service application training sample set sent by each optical network unit to the optical line terminal is a pre-processed service application training sample set. In this embodiment, there is no restriction on the pre-processing method, for example, it can be a noise addition process. The service application training sample set consists of multiple service application samples, and each service application sample is the actual traffic distribution of the local optical network unit.

[0026] In this embodiment, the initial business application identification model pre-issued by the network management system is obtained, and the initial business application identification model is trained using the business application training sample set combined with federated learning (the iterative training process of each round is similar): for the current round of iterative training, the business application identification model is trained based on the business application training sample set and the aggregate model training parameters obtained in the previous round to obtain the corresponding model training parameters, that is, to obtain the first model parameters and the second model parameters. When training the business application identification model, multiple generators are trained together with the discriminator, so that the generator can generate synthetic data samples that match the real process distribution. Among them, each generator is responsible for sample generation of one type of business application, and each generator is responsible for different types of business applications. The first model parameters and the second model parameters are updated by setting an iterative optimization method. In this embodiment, there is no restriction on setting the iterative optimization method, for example, it can be a stochastic gradient descent method, a learning rate optimization algorithm, etc. For the training of the classifier, the second model parameter can be directly used as the model parameter of the classifier, that is, the third model parameter. After the current round of iterative training is completed, the updated first model parameters, second model parameters and third model parameters are uploaded to the network management system in accordance with the set transmission rules to obtain the aggregated model training parameters of the current round; wherein the aggregated model training parameters of the current round are obtained by aggregating the updated model training parameters uploaded by multiple network element management systems of the current round through the network management system; wherein the aggregated model training parameters may include the first aggregated model training parameters, the second aggregated model training parameters and the third aggregated model training parameters. Specifically, the updated first model parameters uploaded by multiple network element management systems of the current round may be aggregated by the network management system to obtain the first aggregated model training parameters; the updated second model parameters uploaded by multiple network element management systems of the current round may be aggregated by the network management system to obtain the second aggregated model training parameters, and the first aggregated model training parameters may be directly used as the third aggregated model training parameters.

[0027] In this embodiment, when the business application identification model converges, the corresponding model parameters are used as target model parameters, that is, the first model parameters obtained in the last round of iterative training are used as the first target model parameters, the second model parameters are used as the second target model parameters, and the third model parameters are used as the third target model parameters.

[0028] It should be noted that each discriminator is used to determine whether the synthetic data sample belongs to real traffic or a synthetic data sample generated by the generator. After aggregation, each discriminator has the ability to distinguish between global real data and all data generated by the generator. Similarly, the generator will be able to capture richer data flow distribution features (it can identify features that are not included in the business application training sample set corresponding to its own network element management system, and this feature is captured from the business application training sample set corresponding to other network element management systems). Since the classifier actually shares the same model parameters as the discriminator, no additional computing power is required for training.

[0029] S120: Aggregate the target model parameters pre-trained by the multiple network element management systems through the network management system to obtain global model parameters.

[0030] The global model parameters include a first global model parameter, a second global model parameter and a third global model parameter.

[0031] In this embodiment, after obtaining the target model parameters pre-trained by multiple network element management systems, there is no restriction on the way of aggregating the target model parameters pre-trained by the multiple network element management systems, for example, the target model parameters pre-trained by the multiple network element management systems can be obtained by weighted averaging. Exemplarily, the first target model parameters pre-trained by multiple network element management systems can be aggregated to obtain the first global model parameters; the second target model parameters pre-trained by multiple network element management systems can be aggregated to obtain the second global model parameters; the first global model parameters can be directly used as the third global model parameters.

[0032] Optionally, the target model parameters pre-trained by the multiple network element management systems are aggregated through the network management system to obtain global model parameters, including: obtaining the model parameter weights corresponding to each of the multiple network element management systems; performing weighted summation based on the first target model parameters corresponding to each of the multiple network element management systems and the model parameter weights corresponding to each of the multiple network element management systems to obtain the first global model parameters and use them as the third global model parameters; performing weighted summation based on the second target model parameters corresponding to each of the multiple network element management systems and the model parameter weights corresponding to each of the multiple network element management systems to obtain the second global model parameters.

[0033] Exemplarily, the first target model parameter corresponding to the i-th network element management system is denoted as ω i The second target model parameter corresponding to the i-th network element management system is denoted as θ i The first global model parameter is denoted as ω0, and the second global model parameter is denoted as θ0. The model parameter weight corresponding to the i-th network element management system is denoted as p i, that is, the model parameter weights corresponding to the first target model parameter and the second target model parameter are the same, both p i .

[0034]

[0035] ω0=∑ i p i ω i ;

[0036] θ0=∑ i p i θ i ;

[0037] Wherein, N represents the total number of network element management systems.

[0038] S130. Identify the service application type based on the global model parameters respectively through multiple optical network units.

[0039] In this embodiment, the "pseudo application label" in the classifier can also be changed to the corresponding "real application label" according to the real annotated data. The "real application label" can be understood as a category label that can be understood by the user. Before this, that is, during the training process, the first target model parameters of the discriminator can be directly used to train the classifier, and the classifier outputs a "pseudo application label", which can be understood as a category label that can be recognized by the machine.

[0040] In this embodiment, after the global model parameters are obtained through the network management system, the global model parameters are sent to multiple network element management systems, and the multiple network element management systems send the global model parameters to at least one optical line terminal corresponding to each network element management system. Each optical line terminal forwards the global model parameters to at least one corresponding optical network unit, and each optical network unit identifies the service application type based on the global model parameters.

[0041] Optionally, multiple optical network units are used to identify service application types based on the global model parameters, including: for each optical network unit, collecting service application test samples; identifying new service application types for the service application test samples based on the first global model parameters and the second global model parameters to obtain a first identification result; if the first identification result is not a new service application type, identifying the service application type for the service application test samples based on the third global model parameters; if the first identification result is a new service application type, performing incremental learning based on the service application test samples through the corresponding network element management system.

[0042] In this embodiment, the process of identifying the service application type based on the global model parameters of each optical network unit is similar. Exemplarily, for each optical network unit, collect local real service application test samples; identify the new service application type of the service application test samples based on the first global model parameter and the second global model parameter to obtain a first identification result. Among them, the identification of the new service application type can be understood as the detection of feature drift. If the first identification result is a non-new service application type, the service application test sample is identified based on the third global model parameter; if the first identification result is a new service application type, the corresponding service application test sample is forwarded to the corresponding optical line terminal, and forwarded to the corresponding network element management system through the optical line terminal, and the corresponding network element management system performs incremental learning based on the service application test sample, and the target model parameters obtained after incremental learning are uploaded to the network management system again, and aggregated again to obtain new global model parameters; and then multiple optical network units are used to identify the service application type based on the new global model parameters.

[0043] The technical solution of the disclosed embodiment is to upload target model parameters pre-trained by multiple network element management systems to the network management system according to the set transmission rules; the target model parameters pre-trained by the multiple network element management systems are aggregated by the network management system to obtain global model parameters; and the service application types are identified based on the global model parameters by multiple optical network units. The disclosed embodiment is to improve the recognition accuracy of service application types by aggregating target model parameters pre-trained by multiple network element management systems through the network management system to obtain global model parameters; and the service application types are identified based on the global model parameters by multiple optical network units.

[0044] Figure 2 A schematic diagram of a device structure for identifying a type of business application provided by an embodiment of the present disclosure, such as Figure 2 As shown, the device includes: a parameter uploading module 210, an aggregation module 220 and an identification module 230;

[0045] The parameter uploading module 210 is used to upload the target model parameters pre-trained by multiple network element management systems to the network management system according to the set transmission rules;

[0046] Aggregation module 220, configured to aggregate target model parameters pre-trained by the plurality of network element management systems through the network management system to obtain global model parameters;

[0047] The identification module 230 is used to identify the service application type based on the global model parameters through multiple optical network units.

[0048] The technical solution of the disclosed embodiment is to upload the target model parameters pre-trained by multiple network element management systems to the network management system through the parameter upload module according to the set transmission rules; the aggregation module aggregates the target model parameters pre-trained by the multiple network element management systems through the network management system to obtain global model parameters; the identification module identifies the service application type based on the global model parameters through multiple optical network units. The disclosed embodiment is to aggregate the target model parameters pre-trained by multiple network element management systems through the network management system to obtain global model parameters; the identification of service application types based on the global model parameters by multiple optical network units can improve the recognition accuracy of service application types.

[0049] Optionally, the above-mentioned device also includes a training module, which is used to: for any network element management system, obtain a business application training sample set collected by the corresponding at least one optical network unit; obtain a business application identification model pre-issued by the network management system; for the current round of iterative training, train the business application identification model based on the business application training sample set and the aggregated model training parameters obtained in the previous round to obtain corresponding model training parameters; update the model training parameters by setting an iterative optimization method; after the current round of iterative training is completed, upload the updated model training parameters to the network management system in accordance with the set transmission rules to obtain the aggregated model training parameters of the current round; wherein the aggregated model training parameters of the current round are obtained by aggregating the updated model training parameters uploaded by multiple network element management systems of the current round by the network management system; when the business application identification model converges, the corresponding model parameters are used as target model parameters.

[0050] Optionally, the business application identification model includes a generative adversarial network and a classifier; the generative adversarial network includes a generator and a discriminator; the target model parameters include a first target model parameter, a second target model parameter and a third target model parameter; the first target model parameter is a model parameter of a pre-trained intermediate layer of the discriminator; the second target model parameter is a model parameter of a pre-trained intermediate layer of the generator; the third target model parameter is a model parameter of a pre-trained intermediate layer of the classifier; the first target model parameter and the third target model parameter are the same.

[0051] Optionally, the aggregation module is specifically used to: obtain the model parameter weights corresponding to each of the multiple network element management systems; perform weighted summation based on the first target model parameters corresponding to each of the multiple network element management systems and the model parameter weights corresponding to each of the multiple network element management systems to obtain the first global model parameter, and use it as the third global model parameter; perform weighted summation based on the second target model parameters corresponding to each of the multiple network element management systems and the model parameter weights corresponding to each of the multiple network element management systems to obtain the second global model parameter.

[0052] Optionally, the identification module is specifically used to: collect business application test samples for each optical network unit; identify new business application types for the business application test samples based on the first global model parameters and the second global model parameters to obtain a first identification result; if the first identification result is not a new business application type, identify the business application type for the business application test samples based on the third global model parameters; if the first identification result is a new business application type, perform incremental learning based on the business application test samples through the corresponding network element management system.

[0053] The setting transmission rule includes a method for uploading the target model parameters by multiple network element management systems and a method for encrypting the target model parameters.

[0054] The device for identifying the business application type provided in the embodiments of the present disclosure can execute the method for identifying the business application type provided in any embodiment of the present disclosure, and has the functional modules and beneficial effects corresponding to the execution method.

[0055] Figure 3 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0056] like Figure 3As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0057] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0058] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 11 performs the various methods and processes described above, such as the identification of the method business application type.

[0059] In some embodiments, the identification of the method business application type can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the identification of the method business application type described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the identification of the method business application type in any other appropriate manner (for example, by means of firmware).

[0060] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0061] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0062] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, 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 disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0063] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0064] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0065] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0066] An embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the method for identifying the business application type provided in any embodiment of the present application.

[0067] In the process of implementation, the computer program product can be written in one or more programming languages ​​or a combination thereof to perform the computer program code of the present invention, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can 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 can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0068] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for identifying a business application type, characterized in that: include: According to the set transmission rules, the target model parameters pre-trained by multiple network element management systems are uploaded to the network management system; Aggregating the target model parameters pre-trained by the multiple network element management systems through the network management system to obtain global model parameters; The service application type is identified based on the global model parameters respectively by multiple optical network units.

2. The method according to claim 1, characterized in that The target model parameters are obtained by the following training method, including: For any network element management system, obtaining a service application training sample set collected by at least one corresponding optical network unit; Obtaining a business application identification model pre-issued by the network management system; For the current round of iterative training, the business application recognition model is trained based on the business application training sample set and the aggregate model training parameters obtained in the previous round to obtain corresponding model training parameters; Updating the model training parameters by setting an iterative optimization method; After the current round of iterative training is completed, the updated model training parameters are uploaded to the network management system according to the set transmission rules to obtain the aggregated model training parameters of the current round; wherein the aggregated model training parameters of the current round are obtained by aggregating the updated model training parameters uploaded by multiple network element management systems of the current round by the network management system; When the business application recognition model converges, the corresponding model parameters are used as target model parameters.

3. The method according to claim 2, characterized in that The business application identification model includes a generative adversarial network and a classifier; the generative adversarial network includes a generator and a discriminator; the target model parameters include a first target model parameter, a second target model parameter and a third target model parameter; the first target model parameter is a model parameter of a pre-trained intermediate layer of the discriminator; the second target model parameter is a model parameter of a pre-trained generator; the third target model parameter is a model parameter of a pre-trained intermediate layer of the classifier; the first target model parameter and the third target model parameter are the same.

4. The method according to claim 3, characterized in that in, The global model parameters include a first global model parameter, a second global model parameter and a third global model parameter; The target model parameters pre-trained by the plurality of network element management systems are aggregated by the network management system to obtain global model parameters, including: Obtaining model parameter weights corresponding to each of the plurality of network element management systems; Based on the weighted sum of the first target model parameters corresponding to each of the multiple network element management systems and the model parameter weights corresponding to each of the multiple network element management systems, the first global model parameter is obtained and used as the third global model parameter; The second global model parameter is obtained by performing weighted summation based on the second target model parameters corresponding to each of the multiple network element management systems and the model parameter weights corresponding to each of the multiple network element management systems.

5. The method according to claim 4, characterized in that The method further comprises: identifying the service application type based on the global model parameters by using multiple optical network units respectively; For each optical network unit, collect service application test samples; Identify a new business application type for the business application test sample based on the first global model parameter and the second global model parameter to obtain a first identification result; If the first identification result is not a new business application type, identifying the business application type of the business application test sample based on the third global model parameter; If the first identification result is a new service application type, incremental learning is performed based on the service application test sample through the corresponding network element management system.

6. The method according to claim 1, characterized in that The setting transmission rule includes a method for uploading the target model parameters by multiple network element management systems and a method for encrypting the target model parameters.

7. A device for identifying a business application type, characterized in that: include: A parameter uploading module is used to upload target model parameters pre-trained by multiple network element management systems to the network management system according to the set transmission rules; An aggregation module, used for aggregating target model parameters pre-trained by the plurality of network element management systems through the network management system to obtain global model parameters; The identification module is used to identify the service application type based on the global model parameters through multiple optical network units.

8. An electronic device, characterized in that: The electronic device comprises: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method for identifying the business application type as described in any one of claims 1 to 6.

9. A storage medium comprising computer executable instructions, wherein the computer executable instructions are used to perform the method for identifying a business application type according to any one of claims 1 to 6 when executed by a computer processor.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the computer program implements the method for identifying the service application type according to any one of claims 1 to 6.

Images