CN116962225A - Network performance monitoring methods, devices, electronic equipment and computer program products - Google Patents

Abstract

This application relates to the field of artificial intelligence technology and provides a network performance monitoring method, device, electronic equipment and computer program products. The method includes: obtaining a network traffic data graph of the network to be monitored; inputting the network traffic data graph of the network to be monitored into a target neural network to obtain a network performance monitoring result output by the target neural network, where the target neural network It is trained based on the network traffic sample data graph corresponding to each monitoring node. The network performance monitoring methods, devices, electronic equipment and computer program products provided by the embodiments of the present application can perform network performance detection through visual images and target neural networks, improving the efficiency and accuracy of network performance detection.

Description

Network performance monitoring methods, devices, electronic equipment and computer program products

Technical field

This application relates to the field of artificial intelligence technology, specifically to a network performance monitoring method, device, electronic equipment and computer program product.

Background technique

In recent years, the scale of communication networks has continued to expand and the network structure has become increasingly complex, resulting in the increasing complexity of network traffic composition and network traffic characteristics. In order to monitor the operating quality of communication networks and discover various abnormalities in a timely manner, domestic operators and communication management departments Communication monitoring systems have been deployed.

The interface module of the communication monitoring system is responsible for collecting data from various types of physical links and completing preprocessing. The communication infrastructure is established based on the interface module. At present, the bandwidth of transmission lines can reach tens of megabits, hundreds of megabits, or even gigabit. The interface module needs to be adapted to the bandwidth. For network performance monitoring of interface modules, it is first necessary to monitor the inside of the LAN. When a network failure occurs, the lack of an analysis mechanism for fault judgment and preliminary analysis makes it impossible to conduct an overall analysis of the network status, and it cannot provide reliable and convenient decision support for network maintenance managers.

At present, network performance monitoring in LAN mainly adopts manual inspection. Manual inspection has problems such as low inspection efficiency and low inspection accuracy.

Contents of the invention

Embodiments of the present application provide a network performance monitoring method, device, electronic equipment and computer program products to solve the technical problems of low manual monitoring efficiency and low inspection accuracy of network performance.

In the first aspect, embodiments of the present application provide a network performance monitoring method, including:

Obtain the network traffic data graph of the network to be monitored;

Input the network traffic data graph of the network to be monitored into the target neural network to obtain the network performance monitoring results output by the target neural network. The target neural network is trained based on the network traffic sample data graph corresponding to each monitoring node. .

In some embodiments, the target neural network is trained through the following steps:

Preprocess the network traffic sample data graph to obtain a target sample data set, where the network traffic sample data graph is an image converted based on the network data packets corresponding to each monitoring node;

Input the target sample data set into the initial neural network for feature extraction to obtain a feature data set;

Based on the feature data set and the preset loss function, the initial neural network is trained to obtain a trained target neural network.

In some embodiments, the target sample data set includes a mark information set; the preprocessing of the network traffic sample data graph to obtain the target sample data set includes:

Perform rough positioning on the first target area in the network traffic sample data map to obtain a rough positioning sample data set;

Mark the second target area in the coarse positioning sample data set to generate the marking information set.

In some embodiments, the target sample data set includes an amplified sample data set, and rough positioning of the first target area in the network traffic sample data map to obtain the rough positioning sample data set further includes:

Use a data amplification operation to process the rough positioning sample data set to obtain an amplified sample data set;

The data amplification operation includes at least one of the following:

Image rotation, image translation, image scaling, adjusting image contrast, adjusting image lighting and adding image noise.

In some embodiments, the feature data set includes: a first feature map, a second feature map, a third feature map and a fourth feature map, and the target sample data set is input into an initial neural network for feature processing. Extract and obtain feature data sets, including:

Input the target sample data set to the first convolution layer for feature extraction to obtain the first feature map;

Input the target sample data set to the second convolution layer for feature extraction to obtain the second feature map;

Input the target sample data set to the pooling layer for pooling, and input the pooled data to the third convolution layer for feature extraction to obtain the third feature map;

The target sample data set is input to the fourth convolution layer for feature extraction to obtain the fourth feature map.

In some embodiments, after inputting the network traffic data graph of the network to be monitored to the target neural network and obtaining the network performance monitoring results output by the target neural network, the method further includes:

Based on the network performance monitoring results, determine whether the network performance of the network to be monitored is faulty;

When a network performance failure of the network to be monitored is determined, a network failure type of the network to be monitored and a fault handling measure corresponding to the network failure type are determined.

In some embodiments, the preset loss function is:

Among them, p _i is the probability that the predicted sample data belongs to category i, is the probability that the real sample data belongs to category i, t _i represents the prediction offset in the RPN training stage,/> Represents the actual offset of the RPN training phase, N _cls is the size of the feature map, N _reg is the size of the target area, λ is the target data, L _cls is the classification loss function, and L _reg is the regression loss function.

In a second aspect, embodiments of the present application provide a network performance monitoring device, including:

The acquisition module is used to obtain the network traffic data graph of the network to be monitored;

The output module is used to input the network traffic data graph of the network to be monitored to the target neural network to obtain the network performance monitoring results output by the target neural network. The target neural network is based on the network traffic samples corresponding to each monitoring node. Data graph training.

In a third aspect, embodiments of the present application provide an electronic device, including a processor and a memory storing a computer program. When the processor executes the program, the network performance monitoring method described in the first aspect is implemented.

In a fourth aspect, embodiments of the present application provide a computer program product, including a computer program that implements the network performance monitoring method described in the first aspect or the second aspect when the computer program is executed by a processor.

The network performance monitoring methods, devices, electronic equipment and computer program products provided by the embodiments of the present application obtain network performance detection results by inputting the network traffic data graph of the network to be monitored into the target neural network, thereby realizing network performance based on visual images. Performance testing simultaneously improves the efficiency and accuracy of network performance testing, reduces manual inspection costs, and enables intelligent network performance monitoring.

Description of the drawings

In order to explain the technical solutions in this application or the prior art more clearly, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are of the present invention. For some embodiments of the application, those of ordinary skill in the art can also obtain other drawings based on these drawings without exerting creative efforts.

Figure 1 is a schematic flowchart of a network performance monitoring method provided by an embodiment of the present application;

Figure 2 is a schematic structural diagram of a neural network using the network performance monitoring method provided by the embodiment of the present application;

Figure 3 is a schematic structural diagram of a network performance monitoring device provided by an embodiment of the present application;

Figure 4 is a schematic diagram of the physical structure of an electronic device provided by an embodiment of the present application;

Detailed ways

In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are part of this application. Examples, not all examples. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of this application.

The execution subject of the network performance monitoring method provided by this application may be an electronic device, a component in an electronic device, an integrated circuit, or a chip. The electronic device may be a mobile electronic device or a non-mobile electronic device. For example, the mobile electronic device may be a mobile phone, a tablet computer, a notebook computer, a handheld computer, a vehicle-mounted electronic device, a wearable device, an ultra-mobile personal computer (UMPC), a netbook or a personal digital assistant (personal digital assistant). assistant, PDA), etc. The non-mobile electronic device can be a server, Network Attached Storage (NAS), personal computer (PC), teller machine or self-service machine, etc. This application does not make specific limitations.

The technical solution of the present application will be described in detail below by taking a computer to execute the network performance monitoring method provided by the embodiment of the present application as an example.

Figure 1 is a schematic flowchart of a network performance monitoring method provided by an embodiment of the present application. Referring to Figure 1, an embodiment of the present application provides a network performance monitoring method, which may include: step 110 and step 120.

Step 110: Obtain the network traffic data graph of the network to be monitored.

It should be noted that the network to be detected may correspond to the network monitored by the corresponding node to be monitored in the local area network. The network traffic on the network line corresponding to the node to be monitored during the target period can be collected, where the network traffic includes input traffic data, output traffic and traffic data components of the network to be monitored.

Generate network traffic data graphs based on the real-time collected network traffic data of the network to be monitored. In actual implementation, data visualization can be achieved through the VC++6.0 network graph control or the Cacti network traffic monitoring graphical analysis tool. You can call the above tools to obtain the network traffic data graph according to your needs, or you can use the background to directly and automatically capture the network traffic data graph, which is not specifically limited here.

The representation method of the network traffic data graph can be based on statistical methods, and statistical display can be carried out according to various characteristics and attributes of the network traffic, such as using dashboards, statistical data tables, histograms, pie charts, dot plots, waterfall charts, etc.

Step 120: Input the network traffic data graph of the network to be monitored into the target neural network to obtain the network performance monitoring results output by the target neural network. The target neural network is trained based on the network traffic sample data graph corresponding to each monitoring node.

In this step, the trained target neural network is first obtained, and then the network traffic data graph of the network to be monitored is input to the target neural network. The target neural network can directly output the network performance monitoring results.

Based on the network performance monitoring results, network performance evaluation of the network to be monitored can be performed. Specifically, if the network performance monitoring results indicate that there are no faults in the network to be monitored, the network performance of the network to be monitored is determined to be excellent; if the network performance monitoring results indicate that there are more than 3 faults in the network to be monitored, the network performance to be monitored is determined to be excellent. Network performance is poor.

It can be understood that when the network performance is judged to be poor, a threshold for the number of faults in the network to be detected can be set in advance. For example, the frequency threshold can be set to 5 times. If more than 5 faults occur in the network to be monitored, the network performance of the network to be monitored is judged to be poor.

The network performance monitoring method provided by the embodiments of this application obtains network performance detection results by inputting the network traffic data graph of the network to be monitored into the target neural network, thereby realizing network performance detection based on visual images and improving network performance detection at the same time. The efficiency and accuracy can also reduce the cost of manual inspection and realize intelligent network performance monitoring.

In some embodiments, the target neural network is trained by the following steps:

Preprocess the network traffic sample data graph to obtain the target sample data set. The network traffic sample data graph is an image converted based on the network data packets corresponding to each monitoring node;

Input the target sample data set into the initial neural network for feature extraction to obtain a feature data set;

Based on the feature data set and the preset loss function, the initial neural network is trained to obtain the trained target neural network.

In actual execution, the network data packets of each monitoring node within the preset historical period are first obtained. The historical period can be determined according to actual needs and is not specifically limited here.

It should be noted that data packets are data units in TCP/IP protocol communication and transmission. TCP/IP protocol works on the third layer (network layer) and fourth layer of the Open System Interconnection (OSI) model. (Transport layer), frames work on the second layer (Data Link layer). Data packets are included in frames, and network data packets include four-layer protocol data packets: data link layer, network layer, transport layer, and application layer.

The OSI model is a standard defined by the International Organization for Standardization. It defines a layered architecture in which each layer defines protocols for different communication levels. The OSI model has 7 layers. Layers 1 to 7 are: physical layer, data link layer, network layer, transport layer, session layer, presentation layer and application layer.

The characteristic attributes of network data packets are not manually prepared, but come from the network traffic on the network line. Characteristic attributes can include: the number of network data packets in each time window, the size distribution of network data packets, the distribution of datagram sending intervals, etc. Feature attribute information.

After obtaining the network data packets of each monitoring node within the preset historical period, the network data packets can be converted into visual images, that is, network traffic sample data graphs.

Preprocess the network traffic sample data graph to obtain the target sample data set. And use the target sample data set as the input of the initial neural network to train the initial neural network.

The initial neural network first performs feature extraction on the target sample data set to obtain the feature data set, and then trains the initial neural network based on the preset loss function. During the training process, the loss function is used to calculate the error, and the back propagation algorithm is used to continuously Update the weight parameters of the model until the neural network converges to reach the expected goal, and finally completes the training.

It can be understood that the initial neural network may be an Inception series network, such as InceptionV1, Inception V2, Inception V3, Inception V4, and Inception-ResNet-V2, which are not specifically limited here.

The network performance monitoring method provided by the embodiment of the present application obtains the target sample data set by preprocessing the network traffic sample data graph, and then trains the initial neural network based on the target sample training data set and obtains the target neural network, thereby facilitating the detection of the network to be detected. Monitoring network traffic data graphs can quickly obtain monitoring results. Compared with direct manual inspection, it can improve the efficiency and accuracy of network performance monitoring.

In some embodiments, the feature data set includes: a first feature map, a second feature map, a third feature map and a fourth feature map. The target sample data set is input into the initial neural network for feature extraction to obtain the feature data set. ,include:

Input the target sample data set to the first convolution layer for feature extraction to obtain the first feature map;

Input the target sample data set to the second convolution layer for feature extraction to obtain the second feature map;

Input the target sample data set to the pooling layer for pooling, and input the pooled data to the third convolution layer for feature extraction to obtain the third feature map;

The target sample data set is input to the fourth convolution layer for feature extraction to obtain the fourth feature map.

In actual execution, the initial neural network can be the Inception V2 network. The structure of the Inception V2 network is shown in Figure 2.

The Inception V2 network includes a Base layer, which is used to import related data. Four branches can be derived from the Base layer: the first layer, the second layer, the third layer and the fourth layer.

The first layer can be the first convolution layer, including a convolution kernel of size 1*1, followed by a ReLU activation function. The target sample data set can be input to the first convolution layer for feature extraction to obtain the first feature map.

Among them, the 1*1 convolution kernel achieves dimensionality reduction, and the use of ReLU activation function can also enhance the nonlinear capability of the network.

The second layer can be the second convolution layer, including a convolution kernel of size 1*1, followed by a ReLU activation function, then a convolution kernel of size 3*3, and then a ReLU activation function. The target sample data set can be input to the second convolution layer for feature extraction to obtain the second feature map.

Compared with the second layer, because a layer of convolution operation is added, there is one more ReLU, that is, a layer of nonlinear mapping is added to make the feature information more discriminative.

The third layer can be a pooling layer and a third convolution layer, including a pooling layer. The pooling layer is followed by a convolution kernel of size 1*1, and then followed by a ReLU activation function. The target sample data set can be input to the pooling layer for pooling, and the pooled data can be input to the third convolution layer for feature extraction to obtain a third feature map.

The fourth layer can be the fourth convolution layer: a convolution kernel of size 1*1, followed by a ReLU activation function, then a convolution kernel of size 3*3, followed by a ReLU activation function, and then a convolution kernel of size 3 *3 convolution kernel, finally connected to the ReLU activation function. The target sample data set can be input to the fourth convolution layer for feature extraction to obtain the fourth feature map.

Because two 3*3 convolutions have the same receptive field as a 5*5 convolution, but the number of parameters is less than that of a 5*5 convolution. And because a layer of convolution operation is added, it corresponds to one more ReLU, that is, a layer of nonlinear mapping is added to make the feature information more discriminative.

Finally, the first feature map, the second feature map, the third feature map and the fourth feature map output from the first layer to the fourth layer can be combined through filters.

The network performance monitoring method provided by the embodiments of this application can reduce the recognition error rate by performing feature extraction through different convolution kernels.

In some embodiments, the target sample data set includes a mark information set; preprocessing the network traffic sample data graph to obtain the target sample data set includes:

Perform rough positioning on the first target area in the network traffic sample data map to obtain a rough positioning sample data set;

Mark the second target area in the coarse positioning sample data set to generate a marking information set.

It should be noted that image coarse positioning can roughly locate the target when the target position is uncertain. Therefore, this method is mostly used when the background is complex, the target position is not fixed, but the detection target is clear.

The characteristic attributes of the network traffic may be determined based on the requirements to determine the rough area that needs to be identified and located. Then the first target area in all network traffic sample data maps may be roughly positioned to obtain a rough positioning sample data set.

For example: the network traffic data graph includes input traffic data and output traffic data, and the first target area is the image area corresponding to the input traffic data. Therefore, the coarse positioning data set is the data set in which the image area corresponding to the input traffic data is located.

After the coarse positioning sample data set is obtained, the second target area in the coarse positioning sample data set is marked to generate a marking information set. For example: mark the data area of interest in the visual image after rough positioning with a rectangular frame and form a mark information set. The data area of interest is the second target area. Among them, the rectangular frame mark is only used as an example and is not specifically limited here.

The mark information set can clearly mark the image area corresponding to the network traffic characteristic attribute, and indicate the corresponding network traffic characteristic attribute information. For example, the second target area can be an image area where traffic data is input during a certain period of a certain day. It can be marked by a rectangular frame and indicate detailed input traffic data information.

The network performance monitoring method provided by the embodiment of the present application first performs rough positioning on the network traffic sample data map, and then marks the rough positioning data set to achieve fine positioning of the image and improve the accuracy of image recognition.

In some embodiments, the target sample data set includes an amplified sample data set, and rough positioning is performed on the first target area in the network traffic sample data graph. After obtaining the rough positioning sample data set, it also includes:

Use the data amplification operation to process the coarse positioning sample data set to obtain the amplified sample data set;

Data amplification operations include at least one of the following:

Image rotation, image translation, image scaling, adjusting image contrast, adjusting image lighting and adding image noise.

It is understandable that during the training process of neural network, data amplification is a data enhancement operation for reading, so it needs to be completed when the data is read.

Data amplification has a certain degree of randomness, and the same picture may obtain different pictures after data amplification.

Data amplification can be expanded from color space, scale space to sample space. At the same time, data amplification has corresponding differences according to different tasks. For example: for image classification, data amplification generally does not change the label; for object detection, data amplification will change the object coordinate position; for image segmentation, data amplification will change the pixel label.

In actual implementation, the data augmentation operation can be:

Image translation, image scaling, adjusting or transforming the contrast, saturation and zero degree of the image color, cropping the image center, cropping the four corners and center of the image to obtain a five-point image, grayscale transformation of the image, using fixed Values can be used to perform pixel filling, random affine transformation, random area cropping, image rotation (for example: random horizontal flip, random rotation or random vertical flip), adjust image lighting and increase image noise, etc.

The network performance monitoring method provided by the embodiments of the present application can increase the number of samples in the training set through data amplification, and can also effectively alleviate the over-fitting of the model, and can also bring stronger generalization capabilities to the model.

In some embodiments, after inputting the network traffic data graph of the network to be monitored to the target neural network and obtaining the network performance monitoring results output by the target neural network, the method further includes:

Based on the network performance monitoring results, determine whether the network performance of the network to be monitored is faulty;

When a network performance fault of the network to be monitored is determined, the network fault type of the network to be monitored and the fault handling measures corresponding to the network fault type are determined.

It can be understood that through the network performance monitoring results output by the target neural network, it can be directly determined whether the network performance of the network to be monitored is faulty. If the network performance of the network to be monitored is faulty, the number of faults of the network to be monitored and the number of network faults can also be determined. Type and troubleshooting measures corresponding to the network fault type.

The main types of network failures include: device connectivity failures caused by abnormal physical interface status; network link failures caused by routing configuration errors of network switching equipment; network congestion failures caused by sudden business data traffic, etc.

After the network fault type is determined, the network fault processing model can be used to determine the corresponding fault processing measures to handle the current network fault type.

The network performance monitoring method provided by the embodiments of this application can determine corresponding fault handling measures according to the type of network fault, thereby automatically and quickly handling network faults and completing automated and intelligent processing of network faults.

In some embodiments, the preset loss function is:

Among them, p _i is the probability that the predicted sample data belongs to category i, is the probability that the real sample data belongs to category i, t _i represents the prediction offset in the RPN training stage,/> Represents the actual offset of the RPN training phase, N _cls is the size of the feature map, N _reg is the size of the target area, λ is the target data, L _cls is the classification loss function, and L _reg is the regression loss function.

It should be noted that λ can be a preset constant and is not specifically limited here. The input of the Region Proposal Network (RPN) is the convolved feature map, and the output is the candidate box map, that is, the image marked with the target area.

In actual execution, the network traffic data graph of the network to be monitored is input into the target neural network for network performance monitoring. Through the target neural network, the characteristic attributes of the network traffic corresponding to the target area in the image can be obtained. Here, the classification loss function is mainly used. Training or learning with the detection frame regression loss function can ultimately retain the rectangular frame and the position of the rectangular frame marking the network traffic corresponding to the target area.

The network performance monitoring method provided by the embodiments of the present application can identify the network traffic data graph of the network to be monitored based on the preset loss function, quickly train the target neural network, and improve the training quality of the target neural network.

The network performance monitoring device provided by the embodiment of the present application is described below. The network performance monitoring device described below and the network performance monitoring method described above can be mutually referenced.

Figure 3 is a schematic structural diagram of a network performance monitoring device provided by an embodiment of the present application. Referring to Figure 3, an embodiment of the present application provides a network performance monitoring device, which may include: an acquisition module 310 and an output module 320.

The acquisition module 310 is used to obtain the network traffic data graph of the network to be monitored;

The output module 320 is used to input the network traffic data graph of the network to be monitored to the target neural network to obtain the network performance monitoring results output by the target neural network. The target neural network is based on the network traffic corresponding to each monitoring node. Obtained by training on sample data graph.

The network performance monitoring device provided by the embodiment of the present application inputs the network traffic data graph of the network to be monitored into the target neural network to obtain network performance detection results, thereby realizing network performance detection based on visual images and improving network performance detection at the same time. The efficiency and accuracy can also reduce the cost of manual inspection and realize intelligent network performance monitoring.

In some embodiments, the device further includes:

A training module, used to preprocess the network traffic sample data graph to obtain a target sample data set. The network traffic sample data graph is an image converted based on the network data packets corresponding to each monitoring node;

Input the target sample data set into the initial neural network for feature extraction to obtain a feature data set;

Based on the feature data set and the preset loss function, the initial neural network is trained to obtain a trained target neural network.

In some embodiments, the target sample data set includes a label information set; the training module is also used to:

Perform rough positioning on the first target area in the network traffic sample data map to obtain a rough positioning sample data set;

Mark the second target area in the coarse positioning sample data set to generate the marking information set.

In some embodiments, the target sample data set includes an amplification sample data set, and the training module is also used to:

Use a data amplification operation to process the rough positioning sample data set to obtain an amplified sample data set;

The data amplification operation includes at least one of the following:

Image rotation, image translation, image scaling, adjusting image contrast, adjusting image lighting and adding image noise.

In some embodiments, the feature data set includes: a first feature map, a second feature map, a third feature map and a fourth feature map, and the training module is also used to:

Input the target sample data set to the first convolution layer for feature extraction to obtain the first feature map;

Input the target sample data set to the second convolution layer for feature extraction to obtain the second feature map;

Input the target sample data set to the pooling layer for pooling, and input the pooled data to the third convolution layer for feature extraction to obtain the third feature map;

The target sample data set is input to the fourth convolution layer for feature extraction to obtain the fourth feature map.

In some embodiments, the device further includes:

A first determination module, configured to determine whether the network performance of the network to be monitored is faulty based on the network performance monitoring results;

The second determination module is configured to determine the network fault type of the network to be monitored and the fault handling measures corresponding to the network fault type when it is determined that the network performance fault of the network to be monitored is faulty.

In some embodiments, the preset loss function is:

Among them, p _i is the probability that the predicted sample data belongs to category i, is the probability that the real sample data belongs to category i, t _i represents the prediction offset in the RPN training stage,/> Represents the actual offset of the RPN training phase, N _cls is the size of the feature map, N _reg is the size of the target area, λ is the target data, L _cls is the classification loss function, and L _reg is the regression loss function.

Figure 4 illustrates a schematic diagram of the physical structure of an electronic device. As shown in Figure 4, the electronic device may include: a processor (processor) 410, a communication interface (Communication Interface) 420, a memory (memory) 430 and a communication bus 440. Among them, the processor 410, the communication interface 420, and the memory 430 complete communication with each other through the communication bus 440. The processor 410 can call the computer program in the memory 430 to perform the steps of the network performance monitoring method, for example, including:

Obtain the network traffic data graph of the network to be monitored;

Input the network traffic data graph of the network to be monitored into the target neural network to obtain the network performance monitoring results output by the target neural network. The target neural network is trained based on the network traffic sample data graph corresponding to each monitoring node. .

In addition, the above-mentioned logical instructions in the memory 430 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the existing technology or the part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code. .

On the other hand, embodiments of the present application also provide a computer program product. The computer program product includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. The computer program is executed by a processor. When , the computer can perform the steps of the network performance monitoring method provided by the above embodiments, including, for example:

Obtain the network traffic data graph of the network to be monitored;

Input the network traffic data graph of the network to be monitored into the target neural network to obtain the network performance monitoring results output by the target neural network. The target neural network is trained based on the network traffic sample data graph corresponding to each monitoring node. .

On the other hand, embodiments of the present application also provide a processor-readable storage medium. The processor-readable storage medium stores a computer program. The computer program is used to cause the processor to execute the methods provided in the above embodiments. Steps, for example include:

Obtain the network traffic data graph of the network to be monitored;

Input the network traffic data graph of the network to be monitored into the target neural network to obtain the network performance monitoring results output by the target neural network. The target neural network is trained based on the network traffic sample data graph corresponding to each monitoring node. .

The processor-readable storage medium may be any available media or data storage device that the processor can access, including but not limited to magnetic storage (such as floppy disks, hard disks, tapes, magneto-optical disks (MO), etc.), optical storage (such as CD, DVD, BD, HVD, etc.), and semiconductor memories (such as ROM, EPROM, EEPROM, non-volatile memory (NANDFLASH), solid state drive (SSD)), etc.

The device embodiments described above are only illustrative. The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in One location, or it can be distributed across multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Persons of ordinary skill in the art can understand and implement the method without any creative effort.

Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the part of the above technical solution that essentially contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., including a number of instructions to cause a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments or certain parts of the embodiments.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present application, but not to limit it; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be Modifications are made to the technical solutions described in the foregoing embodiments, or equivalent substitutions are made to some of the technical features; however, these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions in the embodiments of the present application.

Claims (10)

1. A network performance monitoring method, characterized by comprising: Obtain the network traffic data graph of the network to be monitored; Input the network traffic data graph of the network to be monitored into the target neural network to obtain the network performance monitoring results output by the target neural network. The target neural network is trained based on the network traffic sample data graph corresponding to each monitoring node. . 2. The network performance monitoring method according to claim 1, characterized in that the target neural network is trained through the following steps: Preprocess the network traffic sample data graph to obtain a target sample data set, where the network traffic sample data graph is an image converted based on the network data packets corresponding to each monitoring node; Input the target sample data set into the initial neural network for feature extraction to obtain a feature data set; Based on the feature data set and the preset loss function, the initial neural network is trained to obtain a trained target neural network. 3. The network performance monitoring method according to claim 2, characterized in that the target sample data set includes a mark information set; the preprocessing of the network traffic sample data graph to obtain the target sample data set includes: Perform rough positioning on the first target area in the network traffic sample data map to obtain a rough positioning sample data set; Mark the second target area in the coarse positioning sample data set to generate the marking information set. 4. The network performance monitoring method according to claim 3, wherein the target sample data set includes an amplified sample data set, and the first target area in the network traffic sample data map is roughly positioned. , after obtaining the rough positioning sample data set, it also includes: Use a data amplification operation to process the rough positioning sample data set to obtain an amplified sample data set; The data amplification operation includes at least one of the following: Image rotation, image translation, image scaling, adjusting image contrast, adjusting image lighting and adding image noise. 5. The network performance monitoring method according to claim 2, characterized in that the feature data set includes: a first feature map, a second feature map, a third feature map and a fourth feature map, and the The target sample data set is input into the initial neural network for feature extraction, and a feature data set is obtained, including: Input the target sample data set to the first convolution layer for feature extraction to obtain the first feature map; Input the target sample data set to the second convolution layer for feature extraction to obtain the second feature map; Input the target sample data set to the pooling layer for pooling, and input the pooled data to the third convolution layer for feature extraction to obtain the third feature map; The target sample data set is input to the fourth convolution layer for feature extraction to obtain the fourth feature map. 6. The network performance monitoring method according to any one of claims 1 to 5, characterized in that the network traffic data graph of the network to be monitored is input to the target neural network to obtain the output of the target neural network. After the network performance monitoring results, it also includes: Based on the network performance monitoring results, determine whether the network performance of the network to be monitored is faulty; When a network performance failure of the network to be monitored is determined, a network failure type of the network to be monitored and a fault handling measure corresponding to the network failure type are determined. 7. The network performance monitoring method according to any one of claims 2 to 5, characterized in that the preset loss function is: Among them, p _i is the probability that the predicted sample data belongs to category i, is the probability that the real sample data belongs to category i, t _i represents the prediction offset in the RPN training stage,/> Represents the actual offset in the RPN training phase, N _cls is the size of the feature map, N _reg is the size of the target area, λ is the target data, L _cls is the classification loss function, and L _reg is the regression loss function. 8. A network performance monitoring device, characterized by including: The acquisition module is used to obtain the network traffic data graph of the network to be monitored; The output module is used to input the network traffic data graph of the network to be monitored to the target neural network to obtain the network performance monitoring results output by the target neural network. The target neural network is based on the network traffic samples corresponding to each monitoring node. Data graph training. 9. An electronic device, comprising a processor and a memory storing a computer program, characterized in that when the processor executes the computer program, the network performance monitoring method according to any one of claims 1 to 7 is implemented. 10. A computer program product, comprising a computer program, characterized in that when the computer program is executed by a processor, the network performance monitoring method according to any one of claims 1 to 7 is implemented.