CN116962225A - Network performance monitoring method, device, electronic equipment and computer program product - Google Patents
Network performance monitoring method, device, electronic equipment and computer program product Download PDFInfo
- Publication number
- CN116962225A CN116962225A CN202210379619.2A CN202210379619A CN116962225A CN 116962225 A CN116962225 A CN 116962225A CN 202210379619 A CN202210379619 A CN 202210379619A CN 116962225 A CN116962225 A CN 116962225A
- Authority
- CN
- China
- Prior art keywords
- network
- sample data
- target
- data set
- performance monitoring
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 89
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000004590 computer program Methods 0.000 title claims abstract description 22
- 238000013528 artificial neural network Methods 0.000 claims abstract description 77
- 238000012549 training Methods 0.000 claims abstract description 34
- 238000000605 extraction Methods 0.000 claims description 29
- 230000006870 function Effects 0.000 claims description 27
- 230000003321 amplification Effects 0.000 claims description 15
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 15
- 238000011176 pooling Methods 0.000 claims description 12
- 238000007781 pre-processing Methods 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 7
- 238000005286 illumination Methods 0.000 claims description 5
- 238000013519 translation Methods 0.000 claims description 5
- 239000003550 marker Substances 0.000 claims description 4
- 238000001514 detection method Methods 0.000 abstract description 13
- 238000012806 monitoring device Methods 0.000 abstract description 6
- 230000000007 visual effect Effects 0.000 abstract description 6
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 15
- 238000004891 communication Methods 0.000 description 13
- 238000007689 inspection Methods 0.000 description 9
- 230000004913 activation Effects 0.000 description 8
- 238000004458 analytical method Methods 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 238000013434 data augmentation Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 230000001131 transforming effect Effects 0.000 description 2
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000003190 augmentative effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013079 data visualisation Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/04—Processing captured monitoring data, e.g. for logfile generation
- H04L43/045—Processing captured monitoring data, e.g. for logfile generation for graphical visualisation of monitoring data
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Biophysics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computer Networks & Wireless Communication (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- Signal Processing (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Environmental & Geological Engineering (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The application relates to the technical field of artificial intelligence, and provides a network performance monitoring method, a network performance monitoring device, electronic equipment and a computer program product. The method comprises the following steps: acquiring a network flow data graph of a network to be monitored; inputting the network flow 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, wherein the target neural network is obtained by training based on the network flow sample data graph corresponding to each monitoring node. According to the network performance monitoring method, device, electronic equipment and computer program product provided by the embodiment of the application, network performance detection can be performed through the visual image and the target neural network, and the efficiency and accuracy of network performance detection are improved.
Description
Technical Field
The present application relates to the field of artificial intelligence technology, and in particular, to a network performance monitoring method, apparatus, electronic device, and computer program product.
Background
In recent years, the scale of communication networks is continuously enlarged, the network structure is gradually complicated, the composition of network flow and the complexity of network flow characteristics are gradually improved, and in order to monitor the operation quality of the communication network, various anomalies are timely found, and a communication monitoring system is deployed by domestic operators and communication management departments.
The interface module of the communication monitoring system is used for collecting data from various types of physical links and completing preprocessing, and the communication infrastructure is established based on the interface module. The bandwidth of the transmission line can reach tens of megabytes and hundreds of megabytes at present, so that the interface module needs to adapt to the bandwidth. And aiming at the network performance monitoring of the interface module, the monitoring of the inside of the local area network is needed first. When a network fails, an analysis mechanism for failure judgment and preliminary analysis is lacking, so that the overall analysis of the network state cannot be performed, and reliable and convenient decision support cannot be provided for network maintenance management personnel.
At present, network performance monitoring in a local area network mainly adopts a manual inspection mode, and the manual inspection can have the problems of low inspection efficiency, low inspection accuracy and the like.
Disclosure of Invention
The embodiment of the application provides a network performance monitoring method, a device, electronic equipment and a computer program product, which are used for solving the technical problems of low manual network performance monitoring efficiency and low inspection accuracy.
In a first aspect, an embodiment of the present application provides a network performance monitoring method, including:
acquiring a network flow data graph of a network to be monitored;
inputting the network flow 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, wherein the target neural network is obtained by training based on the network flow sample data graph corresponding to each monitoring node.
In some embodiments, the target neural network is trained by:
preprocessing the network traffic sample data graph to obtain a target sample data set, wherein the network traffic sample data graph is an image obtained by converting network data packets corresponding to all monitoring nodes;
inputting the target sample data set into an initial neural network for feature extraction to obtain a feature data set;
and training the initial neural network based on the characteristic data set and a preset loss function to obtain a trained target neural network.
In some embodiments, the target sample data set comprises a marker information set; the preprocessing of the network traffic sample data graph to obtain a target sample data set comprises the following steps:
coarse positioning is carried out on a first target area in the network flow sample data graph, and a coarse positioning sample data set is obtained;
and marking a second target area in the coarse positioning sample data set, and generating the marking information set.
In some embodiments, the target sample data set includes an amplified sample data set, and the coarsely locating the first target area in the network traffic sample data map, after obtaining the coarsely located sample data set, further includes:
processing the coarse positioning sample data set by utilizing data amplification operation to obtain an amplified sample data set;
the data amplification operation includes at least one of:
image rotation, image translation, image scaling, adjusting image contrast, adjusting image illumination, and increasing image noise.
In some embodiments, the feature data set includes: the first feature map, the second feature map, the third feature map and the fourth feature map, the target sample data set is input into an initial neural network to perform feature extraction, and a feature data set is obtained, including:
inputting the target sample data set into a first convolution layer for feature extraction to obtain the first feature map;
inputting the target sample data set into a second convolution layer for feature extraction to obtain a second feature map;
inputting the target sample data set into a pooling layer for pooling, and inputting the pooled data into a third convolution layer for feature extraction to obtain the third feature map;
and inputting the target sample data set into a fourth convolution layer for feature extraction to obtain the fourth feature map.
In some embodiments, after the network traffic data map of the network to be monitored is input to a target neural network to obtain a network performance monitoring result output by the target neural network, the method further includes:
determining whether the network performance of the network to be monitored is faulty or not based on the network performance monitoring result;
and under the condition that the network performance fault of the network to be monitored is determined, determining the network fault type of the network to be monitored and fault handling measures corresponding to the network fault type.
In some embodiments, the preset loss function is:
wherein p is i To predict the probability that the sample data belongs to class i,for the probability that the real sample data belongs to class i, t i Predictive offset representing the RPN training phase, +.>Representing the actual offset of the RPN training phase, N cls For the size of the feature map, N reg Lambda is target data, L, is the size of the target area cls To classify the loss function, L reg Is a regression loss function.
In a second aspect, an embodiment of the present application provides a network performance monitoring apparatus, including:
the acquisition module is used for acquiring a network flow data graph of the network to be monitored;
the output module is used for inputting the network flow 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, and the target neural network is obtained by training based on the network flow sample data graph corresponding to each monitoring node.
In a third aspect, an embodiment of the present application provides an electronic device, including a processor and a memory storing a computer program, where the processor implements the network performance monitoring method according to the first aspect when executing the program.
In a fourth aspect, embodiments of the present application provide a computer program product comprising a computer program which, when executed by a processor, implements the network performance monitoring method of the first or second aspect.
According to the network performance monitoring method, the device, the electronic equipment and the computer program product, the network flow data diagram of the network to be monitored is input into the target neural network to obtain the network performance detection result, so that the network performance detection based on the visual image is realized, the efficiency and the accuracy of the network performance detection are improved, the manual inspection cost is reduced, and the intelligent network performance monitoring is realized.
Drawings
In order to more clearly illustrate the application or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the application, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flow chart of a network performance monitoring method according to an embodiment of the present application;
fig. 2 is a schematic diagram of a neural network structure applying the network performance monitoring method provided by the embodiment of the present application;
fig. 3 is a schematic structural diagram of a network performance monitoring device according to an embodiment of the present application;
fig. 4 is a schematic diagram of an entity structure of an electronic device according to an embodiment of the present application;
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present application more apparent, the technical solutions of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
The execution subject of the network performance monitoring method provided by the application can be an electronic device, a component in the electronic device, an integrated circuit, or a chip. The electronic device may be a mobile electronic device or a non-mobile electronic device. By way of example, the mobile electronic device may be a cell phone, tablet computer, notebook computer, palm top computer, vehicle mounted electronic device, wearable device, ultra-mobile personal computer (ultra-mobile personal computer, UMPC), netbook or personal digital assistant (personal digital assistant, PDA), etc., and the non-mobile electronic device may be a server, network attached storage (Network Attached Storage, NAS), personal computer (personal computer, PC), teller machine or self-service machine, etc., and the application is not limited in particular.
The following describes the technical scheme of the present application in detail by taking a computer to execute the network performance monitoring method provided by the embodiment of the present application as an example.
Fig. 1 is a flow chart of a network performance monitoring method according to an embodiment of the present application. Referring to fig. 1, an embodiment of the present application provides a network performance monitoring method, which may include: step 110 and step 120.
Step 110, a network flow data graph of the network to be monitored is obtained.
It should be noted that the network to be detected may correspond to a network monitored by a corresponding node to be monitored in the local area network. Network traffic on a network line corresponding to a node to be monitored in a target period can be acquired, wherein the network traffic comprises input traffic data, output traffic and traffic data components of the network to be monitored.
And generating a network flow data graph according to the network flow data of the network to be monitored, which is acquired in real time. In actual implementation, data visualization may be implemented through a vc++6.0 network graph control or a calit network traffic monitoring graph analysis tool. The tool can be called according to the requirement to obtain the network flow data graph, or the background can be utilized to directly and automatically grasp the network flow data graph, which is not particularly limited herein.
The method for representing the network flow data graph can be based on a statistical method, and can carry out classified statistical display on various characteristic attributes of the network flow, such as a meter panel, a statistical data table, a histogram, a pie chart, a dot diagram, a waterfall diagram and the like.
Step 120, inputting the network flow 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, wherein the target neural network is obtained by training based on the network flow sample data graph corresponding to each monitoring node.
In the step, a target neural network after training is obtained, and then a network flow data graph of the network to be monitored is input into the target neural network, and the target neural network can directly output a network performance monitoring result.
According to the network performance monitoring result, network performance evaluation can be performed on the network to be monitored. Specifically, if the network performance monitoring result indicates that no fault exists in the network to be monitored, judging that the network performance of the network to be monitored is excellent; if the network performance monitoring result indicates that more than 3 times of faults occur in the network to be monitored, the network performance of the network to be monitored is judged to be poor.
It can be appreciated that when the network performance is determined to be poor, a threshold value of the number of times of occurrence of faults in the network to be detected may be preset. For example: the frequency threshold may be set to 5 times, and if more than 5 times of faults occur in the network to be monitored, it is determined that the network performance of the network to be monitored is poor.
According to the network performance monitoring method provided by the embodiment of the application, the network performance detection result is obtained by inputting the network flow data graph of the network to be monitored into the target neural network, so that the network performance detection is realized based on the visual image, the efficiency and the accuracy of the network performance detection are improved, the manual inspection cost can be reduced, and the intelligent network performance monitoring is realized.
In some embodiments, the target neural network is trained by:
preprocessing a network traffic sample data graph to obtain a target sample data set, wherein the network traffic sample data graph is an image obtained by converting network data packets corresponding to all monitoring nodes;
inputting the target sample data set into an initial neural network for feature extraction to obtain a feature data set;
and training the initial neural network based on the characteristic data set and a preset loss function to obtain a target neural network after training.
In actual implementation, network data packets of each monitoring node in a preset history period are acquired first. The historical period may be determined according to actual needs and is not particularly limited herein.
The data packet is a data unit in the communication transmission of the TCP/IP protocol, which operates in the third layer (network layer) and the fourth layer (transport layer) of the open system interconnection (Open System Interconnection, OSI) model, and the frame operates in the second layer (data link layer). The data packets are contained in frames, and the network data packets comprise four-layer protocol data packets of a data link layer, a network layer, a transmission layer and an application layer.
The OSI model is a standard defined by the international organization for standardization, which defines a layered architecture in which each layer defines protocols for different communication levels. The OSI model has 7 layers, 1 to 7 layers being respectively: a physical layer, a data link layer, a network layer, a transport layer, a session layer, a presentation layer, and an application layer.
The characteristic attributes of the network data packet are not manually formulated, but are network traffic from a network line, and may include: the number of network data packets, the size distribution of the network data packets, the distribution of the datagram sending intervals and other characteristic attribute information in each time window.
After obtaining the network data packet of each monitoring node in the preset history period, the network data packet can be converted into a visual image, namely a network traffic sample data graph.
And preprocessing the network traffic sample data graph to obtain a target sample data set. And taking the target sample data set as an input of the initial neural network to train the initial neural network.
The initial neural network firstly performs feature extraction on a target sample data set to obtain a feature data set, then trains the initial neural network based on a preset loss function, performs error calculation by using the loss function in the training process, and continuously updates weight parameters of the model by using a back propagation algorithm until the neural network converges to reach an expected target, and finally completes training.
It is understood that the initial neural network may be an acceptance series network, for example, an acceptance V1, an acceptance V2, an acceptance V3, an acceptance V4, and an acceptance-ResNet-V2, which are not particularly limited herein.
According to the network performance monitoring method provided by the embodiment of the application, the target sample data set is obtained by preprocessing the network flow sample data diagram, and then the initial neural network is trained according to the target sample training data set, and the target neural network is obtained, so that the network flow data diagram of the network to be detected is conveniently monitored, the monitoring result can be quickly obtained, and the efficiency and the accuracy of network performance monitoring can be improved compared with the method of directly adopting manual inspection.
In some embodiments, the feature data set includes: the first feature map, the second feature map, the third feature map and the fourth feature map, the target sample data set is input into an initial neural network for feature extraction, and a feature data set is obtained, comprising:
inputting the target sample data set into a first convolution layer for feature extraction to obtain a first feature map;
inputting the target sample data set into a second convolution layer for feature extraction to obtain a second feature map;
inputting the target sample data set into a pooling layer for pooling, and inputting the pooled data into a third convolution layer for feature extraction to obtain a third feature map;
and inputting the target sample data set into a fourth convolution layer for feature extraction to obtain a fourth feature map.
In actual implementation, the initial neural network may be an admission V2 network, and the structure of the admission V2 network is shown in fig. 2.
The acceptance V2 network includes a Base layer for importing relevant data. Four branches can be led from the Base layer: a first layer, a second layer, a third layer, and a fourth layer.
The first layer may be a first convolution layer comprising a convolution kernel of size 1*1 followed by a ReLU activation function. The target sample dataset may be input to a first convolution layer for feature extraction to obtain a first feature map.
Wherein, 1*1 convolution kernel realizes dimension reduction, and simultaneously uses ReLU activation function to enhance the non-linear capability of the network.
The second layer may be a second convolution layer comprising a convolution kernel of size 1*1 followed by a ReLU activation function, followed by a convolution kernel of size 3*3, followed by a ReLU activation function. The target sample dataset may be input to a second convolution layer for feature extraction to obtain a second feature map.
Compared with the second layer, as one layer of convolution operation is added, the method corresponds to one more ReLU, namely one layer of nonlinear mapping is added, so that the characteristic information is more discriminant.
The third layer may be a pooling layer followed by a convolution kernel of size 1*1 followed by a ReLU activation function and a third convolution layer comprising a pooling layer. And inputting the target sample data set into a pooling layer for pooling, and inputting the pooled data into a third convolution layer for feature extraction to obtain a third feature map.
The fourth layer may be a fourth convolutional layer: the convolution kernel of size 1*1 is followed by a ReLU activation function, then a convolution kernel of size 3*3 is followed by a ReLU activation function, then a convolution kernel of size 3*3 is followed by a ReLU activation function. And inputting the target sample data set into a fourth convolution layer for feature extraction to obtain a fourth feature map.
Because the convolution of two 3*3 is specifically the same receptive field as the convolution of one 5*5, but the number of parameters is less than the convolution of 5*5. And because a layer of convolution operation is added, the method corresponds to one more ReLU, namely a layer of nonlinear mapping is added, so that the characteristic information is more discriminant
And finally, the first characteristic diagram, the second characteristic diagram, the third characteristic diagram and the fourth characteristic diagram which are output from the first layer to the fourth layer can be combined through a filter.
According to the network performance monitoring method provided by the embodiment of the application, the characteristic extraction is carried out through different convolution kernels, so that the recognition error rate can be reduced.
In some embodiments, the target sample data set includes a marker information set; preprocessing a network traffic sample data graph to obtain a target sample data set, including:
coarse positioning is carried out on a first target area in a network flow sample data graph, and a coarse positioning sample data set is obtained;
and marking the second target area in the coarse positioning sample data set, and generating a marking information set.
The rough image positioning can roughly position the target in the case that the target position is not determined. Therefore, the method is mostly used in the conditions of complex background, unfixed target position and clear detection target.
If the characteristic attribute of the network traffic is determined to be required to be identified and located according to the requirement, the first target area in all the network traffic sample data graphs can be coarsely located to obtain a coarsely located sample data set.
For example: the network traffic data map includes input traffic data and output traffic data, and the first target area is an 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 flow data is positioned.
After the coarse positioning sample dataset is acquired, a second target area in the coarse positioning sample dataset is marked, and a marking information set is generated. For example: and (3) marking the data area of interest in the roughly positioned visual image by a rectangular frame, and forming a marking information set, wherein the data area of interest is the second target area. Wherein the rectangular box mark is used as an example only and is not particularly limited herein.
The image area corresponding to the network traffic characteristic attribute can be explicitly marked in the marking information set, and the corresponding network traffic characteristic attribute information is indicated. For example: the second target area may be an image area where traffic data is input for a certain period of time on a certain day, may be marked by a rectangular frame, and indicates detailed input traffic data information.
According to the network performance monitoring method provided by the embodiment of the application, the network flow sample data graph is subjected to coarse positioning, and then the coarse positioning data set is marked, so that the fine positioning of the image is realized, and the image identification accuracy is improved.
In some embodiments, the target sample data set comprises an amplified sample data set, and after coarsely locating the first target region in the network traffic sample data map, the method further comprises:
processing the coarse positioning sample data set by utilizing data amplification operation to obtain an amplified sample data set;
the data amplification operation includes at least one of:
image rotation, image translation, image scaling, adjusting image contrast, adjusting image illumination, and increasing image noise.
It will be appreciated that during training of the neural network, data augmentation is the operation of data augmentation on the reading, and therefore needs to be done at the time of data reading.
The data amplification has certain randomness, and different pictures can be obtained from the same picture through the data amplification.
The data amplification can be expanded from a color space, a scale space to a sample space, and meanwhile, the data amplification is correspondingly different according to different tasks. For example: for image classification, data amplification generally does not change the label; for object detection, data amplification can change the object coordinate position; for image segmentation, the data is augmented with pixel labels.
In actual implementation, the data augmentation operation may be:
image translation, image scaling, adjusting or transforming contrast, saturation, and zero degrees of image color, cropping image center, cropping four corners and center of image to obtain a five-part image, gray scale transforming image, pixel filling using fixed values, random affine transformation, random region cropping, image rotation (e.g., random horizontal flip, random rotation, or random vertical flip), adjusting image illumination, and increasing image noise, etc
According to the network performance monitoring method provided by the embodiment of the application, the samples of the training set can be increased through data amplification, meanwhile, the situation of model overfitting can be effectively relieved, and stronger generalization capability can be brought to the model.
In some embodiments, the network traffic data map of the network to be monitored is input to the target neural network, and after obtaining the network performance monitoring result output by the target neural network, the method further includes:
based on the network performance monitoring result, determining whether the network performance of the network to be monitored is faulty;
and under the condition of determining network performance faults of the network to be monitored, determining the network fault type of the network to be monitored and fault handling measures corresponding to the network fault type.
It can be understood that, according to the network performance monitoring result output by the target neural network, whether the network performance of the network to be monitored is faulty can be directly determined, and if the network performance of the network to be monitored is faulty, the number of faults of the network to be monitored, the network fault type and the fault handling measures corresponding to the network fault type can be determined.
The network failure types are mainly: device connectivity faults caused by abnormal physical interface states; network link failure caused by route configuration errors of network switching equipment; network congestion failure due to bursty traffic data traffic, and the like.
After determining the network failure type, the network failure model may be utilized to determine corresponding failure handling measures to handle the current network failure type.
The network performance monitoring method provided by the embodiment of the application can determine the corresponding fault processing measures according to the network fault type, so that the network fault processing can be automatically and rapidly carried out, and the automatic intelligent processing of the network fault is completed.
In some embodiments, the preset loss function is:
wherein p is i To predict the probability that the sample data belongs to class i,for the probability that the real sample data belongs to class i, t i Predictive offset representing the RPN training phase, +.>Representing the actual offset of the RPN training phase, N cls For the size of the feature map, N reg Lambda is target data, L, is the size of the target area cls To classify the loss function, L reg Is a regression loss function.
Here, λ may be a constant set in advance, and is not particularly limited here. The input to the region generation network (Region Proposal Network, RPN) is a convolved feature map and the output is a candidate block diagram, i.e. an image labeled with the target region.
In actual execution, a network flow data graph of a network to be monitored is input into a target neural network for network performance monitoring, and characteristic attributes of network flow corresponding to a target area in an image can be obtained through the target neural network.
The network performance monitoring method provided by the embodiment of the application can identify the network flow data graph of the network to be monitored according to the preset loss function, can rapidly train the target neural network and improves the training quality of the target neural network.
The network performance monitoring device provided by the embodiment of the application is described below, and the network performance monitoring device described below and the network performance monitoring method described above can be referred to correspondingly.
Fig. 3 is a schematic structural diagram of a network performance monitoring device according to an embodiment of the present application. Referring to fig. 3, an embodiment of the present application provides a network performance monitoring apparatus, which may include: an acquisition module 310 and an output module 320.
An obtaining module 310, configured to obtain a network traffic data map of a network to be monitored;
and the output module 320 is configured to input the network traffic data graph of the network to be monitored to a target neural network, and obtain a network performance monitoring result output by the target neural network, where the target neural network is obtained by training based on the network traffic sample data graph corresponding to each monitoring node.
According to the network performance monitoring device provided by the embodiment of the application, the network performance detection result is obtained by inputting the network flow data graph of the network to be monitored into the target neural network, so that the network performance detection is realized based on the visual image, the efficiency and the accuracy of the network performance detection are improved, the manual inspection cost can be reduced, and the intelligent network performance monitoring is realized.
In some embodiments, the apparatus further comprises:
the training module is used for preprocessing the network flow sample data graph to obtain a target sample data set, and the network flow sample data graph is an image obtained by converting the network data packets corresponding to the monitoring nodes;
inputting the target sample data set into an initial neural network for feature extraction to obtain a feature data set;
and training the initial neural network based on the characteristic data set and a preset loss function to obtain a trained target neural network.
In some embodiments, the target sample data set comprises a marker information set; the training module is also configured to:
coarse positioning is carried out on a first target area in the network flow sample data graph, and a coarse positioning sample data set is obtained;
and marking a second target area in the coarse positioning sample data set, and generating the marking information set.
In some embodiments, the target sample data set comprises an amplified sample data set, the training module further to:
processing the coarse positioning sample data set by utilizing data amplification operation to obtain an amplified sample data set;
the data amplification operation includes at least one of:
image rotation, image translation, image scaling, adjusting image contrast, adjusting image illumination, and increasing image noise.
In some embodiments, the feature data set includes: the training module is further configured to:
inputting the target sample data set into a first convolution layer for feature extraction to obtain the first feature map;
inputting the target sample data set into a second convolution layer for feature extraction to obtain a second feature map;
inputting the target sample data set into a pooling layer for pooling, and inputting the pooled data into a third convolution layer for feature extraction to obtain the third feature map;
and inputting the target sample data set into a fourth convolution layer for feature extraction to obtain the fourth feature map.
In some embodiments, the apparatus further comprises:
a first determining module, configured to determine, based on the network performance monitoring result, whether the network performance of the network to be monitored is faulty;
and the second determining module is used for determining the network fault type of the network to be monitored and fault handling measures corresponding to the network fault type under the condition that the network performance fault of the network to be monitored is determined.
In some embodiments, the preset loss function is:
wherein p is i To predict the probability that the sample data belongs to class i,for the probability that the real sample data belongs to class i, t i Predictive offset representing the RPN training phase, +.>Representing the actual offset of the RPN training phase, N cls For the size of the feature map, N reg Lambda is target data, L, is the size of the target area cls To classify the loss function, L reg Is a regression loss function.
Fig. 4 illustrates a physical schematic diagram of an electronic device, as shown in fig. 4, which may include: processor 410, communication interface (Communication Interface) 420, memory 430 and communication bus 440, wherein processor 410, communication interface 420 and memory 430 communicate with each other via communication bus 440. The processor 410 may invoke a computer program in the memory 430 to perform the steps of the network performance monitoring method, including, for example:
acquiring a network flow data graph of a network to be monitored;
inputting the network flow 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, wherein the target neural network is obtained by training based on the network flow sample data graph corresponding to each monitoring node.
Further, the logic instructions in the memory 430 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on this understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In another aspect, embodiments of the present application further provide a computer program product, where the computer program product includes a computer program, where the computer program may be stored on a non-transitory computer readable storage medium, where the computer program when executed by a processor is capable of executing the steps of the network performance monitoring method provided by the foregoing embodiments, for example, including:
acquiring a network flow data graph of a network to be monitored;
inputting the network flow 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, wherein the target neural network is obtained by training based on the network flow sample data graph corresponding to each monitoring node.
In another aspect, embodiments of the present application further provide a processor-readable storage medium storing a computer program for causing a processor to execute the steps of the method provided in the above embodiments, for example, including:
acquiring a network flow data graph of a network to be monitored;
inputting the network flow 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, wherein the target neural network is obtained by training based on the network flow sample data graph corresponding to each monitoring node.
The processor-readable storage medium may be any available medium or data storage device that can be accessed by a processor, including, but not limited to, magnetic storage (e.g., floppy disks, hard disks, magnetic tape, magneto-optical disks (MOs), etc.), optical storage (e.g., CD, DVD, BD, HVD, etc.), semiconductor storage (e.g., ROM, EPROM, EEPROM, nonvolatile storage (NAND FLASH), solid State Disk (SSD)), and the like.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present application without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application, and are not limiting; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims (10)
1. A method for monitoring network performance, comprising:
acquiring a network flow data graph of a network to be monitored;
inputting the network flow 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, wherein the target neural network is obtained by training based on the network flow sample data graph corresponding to each monitoring node.
2. The network performance monitoring method of claim 1, wherein the target neural network is trained by:
preprocessing the network traffic sample data graph to obtain a target sample data set, wherein the network traffic sample data graph is an image obtained by converting network data packets corresponding to all monitoring nodes;
inputting the target sample data set into an initial neural network for feature extraction to obtain a feature data set;
and training the initial neural network based on the characteristic data set and a preset loss function to obtain a trained target neural network.
3. The network performance monitoring method of claim 2, wherein the target sample data set comprises a marker information set; the preprocessing of the network traffic sample data graph to obtain a target sample data set comprises the following steps:
coarse positioning is carried out on a first target area in the network flow sample data graph, and a coarse positioning sample data set is obtained;
and marking a second target area in the coarse positioning sample data set, and generating the marking information set.
4. The network performance monitoring method of claim 3, wherein the target sample data set comprises an amplified sample data set, and wherein the coarse positioning of the first target region in the network traffic sample data map, after obtaining the coarse positioning sample data set, further comprises:
processing the coarse positioning sample data set by utilizing data amplification operation to obtain an amplified sample data set;
the data amplification operation includes at least one of:
image rotation, image translation, image scaling, adjusting image contrast, adjusting image illumination, and increasing image noise.
5. The network performance monitoring method of claim 2, wherein the feature data set comprises: the first feature map, the second feature map, the third feature map and the fourth feature map, the target sample data set is input into an initial neural network to perform feature extraction, and a feature data set is obtained, including:
inputting the target sample data set into a first convolution layer for feature extraction to obtain the first feature map;
inputting the target sample data set into a second convolution layer for feature extraction to obtain a second feature map;
inputting the target sample data set into a pooling layer for pooling, and inputting the pooled data into a third convolution layer for feature extraction to obtain the third feature map;
and inputting the target sample data set into a 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, wherein the inputting the network traffic data map of the network to be monitored into a target neural network, after obtaining the network performance monitoring result output by the target neural network, further includes:
determining whether the network performance of the network to be monitored is faulty or not based on the network performance monitoring result;
and under the condition that the network performance fault of the network to be monitored is determined, determining the network fault type of the network to be monitored and fault handling measures corresponding to the network fault type.
7. The network performance monitoring method according to any one of claims 2 to 5, wherein the predetermined loss function is:
wherein p is i To predict the probability that the sample data belongs to class i,for the probability that the real sample data belongs to class i, t i Predictive offset representing the RPN training phase, +.>Representing the actual offset of the RPN training phase, N cls For the size of the feature map, N reg Lambda is target data, L, is the size of the target area cls To classify the loss function, L reg Is a regression loss function.
8. A network performance monitoring apparatus, comprising:
the acquisition module is used for acquiring a network flow data graph of the network to be monitored;
the output module is used for inputting the network flow 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, and the target neural network is obtained by training based on the network flow sample data graph corresponding to each monitoring node.
9. An electronic device comprising a processor and a memory storing a computer program, characterized in that the processor implements the network performance monitoring method of any one of claims 1 to 7 when executing the computer program.
10. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements the network performance monitoring method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210379619.2A CN116962225A (en) | 2022-04-12 | 2022-04-12 | Network performance monitoring method, device, electronic equipment and computer program product |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210379619.2A CN116962225A (en) | 2022-04-12 | 2022-04-12 | Network performance monitoring method, device, electronic equipment and computer program product |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116962225A true CN116962225A (en) | 2023-10-27 |
Family
ID=88442990
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210379619.2A Pending CN116962225A (en) | 2022-04-12 | 2022-04-12 | Network performance monitoring method, device, electronic equipment and computer program product |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116962225A (en) |
-
2022
- 2022-04-12 CN CN202210379619.2A patent/CN116962225A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110943857B (en) | Power communication network fault analysis and positioning method based on convolutional neural network | |
| CN110827251B (en) | Power transmission line locking pin defect detection method based on aerial image | |
| Xu et al. | Detecting tiny objects in aerial images: A normalized Wasserstein distance and a new benchmark | |
| CN113095438B (en) | Wafer defect classification method, device and system thereof, electronic equipment and storage medium | |
| CN107330731B (en) | Method and device for identifying click abnormity of advertisement space | |
| CN112581463A (en) | Image defect detection method and device, electronic equipment, storage medium and product | |
| CN111353413A (en) | Low-missing-report-rate defect identification method for power transmission equipment | |
| JP2012032369A (en) | Defect identification method, defect identification apparatus, program, and recording medium | |
| CN112669323B (en) | Image processing method and related equipment | |
| CN113469951B (en) | Hub defect detection method based on cascade region convolutional neural network | |
| CN117156442B (en) | Cloud data security protection method and system based on 5G network | |
| WO2021027157A1 (en) | Vehicle insurance claim settlement identification method and apparatus based on picture identification, and computer device and storage medium | |
| CN112613569A (en) | Image recognition method, and training method and device of image classification model | |
| JP2019067299A (en) | Label estimating apparatus and label estimating program | |
| CN114048546A (en) | Graph convolution network and unsupervised domain self-adaptive prediction method for residual service life of aircraft engine | |
| CN114332457A (en) | Image instance segmentation model training method, image instance segmentation method and device | |
| CN114639102A (en) | Cell segmentation method and device based on key point and size regression | |
| CN114708247A (en) | Cigarette case packaging defect identification method and device based on deep learning | |
| CN112419243B (en) | Power distribution room equipment fault identification method based on infrared image analysis | |
| CN116959099B (en) | Abnormal behavior identification method based on space-time diagram convolutional neural network | |
| CN113012107A (en) | Power grid defect detection method and system | |
| CN116962225A (en) | Network performance monitoring method, device, electronic equipment and computer program product | |
| CN116797551A (en) | Peak-shaving power supply equipment state evaluation method considering image data privacy | |
| CN115620083A (en) | Model training method, face image quality evaluation method, device and medium | |
| CN117011216A (en) | Defect detection method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |