Network security situation assessment method and device, medium and product
By combining full-dimensional dynamic convolution and multi-scale efficient channel attention mechanism with recurrent neural networks for network security situation assessment, the problem of inaccurate assessment in complex network environments of existing models is solved, and higher assessment accuracy is achieved.
Patent Information
- Application Number
- CN202411035397.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-30
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2044-07-30
AI Technical Summary
Existing cybersecurity posture assessment models struggle to accurately assess security conditions in complex network environments, especially when time-dependent data exists and they fail to capture complex patterns.
We employ full-dimensional dynamic convolution and multi-scale efficient channel attention mechanisms to extract features from network traffic data, combine them with recurrent neural networks for time series modeling, determine the categories of network traffic data through feature extraction networks and fully connected layers, and calculate network security posture values.
It improves the accuracy of cybersecurity situation assessment, enables a better understanding of temporal relationships and patterns in time series data, and enhances the model's ability to capture data features.
Smart Images

Figure CN119051899A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of network security technology, and in particular to a network security situation assessment method, device, medium and product. Background Art
[0002] With the continuous development of Internet infrastructure and the emergence of new applications, the scale of the network continues to expand, the topology structure is becoming increasingly complex, and network security issues are becoming increasingly prominent. Traditional network security means can no longer meet the current complex network security needs. Therefore, there is an urgent need for advanced technologies that can comprehensively evaluate network security situations. Network security situation awareness technology has the advantage of integrating information resources. Through comprehensive analysis of data information such as network traffic, behavior logs, and security events, it can promptly discover abnormal activities and potential threats in the network, so that timely countermeasures can be taken to improve the security and protection of the network system.
[0003] Network security situation assessment technology is a technology that assesses the security status of network systems. It assesses the security status of networks by collecting and analyzing information such as the operating status, security events, and attack behaviors of each component in the network system. Existing network security situation assessment models are generally built based on convolutional neural networks. However, when convolutional neural networks have a strong time dependency on data, they may not be able to capture complex patterns in time series. As network security becomes increasingly complex, existing network security situation assessment models are difficult to accurately assess network security. Summary of the invention
[0004] The purpose of this application is to provide a network security situation assessment method, device, medium and product, which can improve the accuracy of network security situation assessment.
[0005] To achieve the above objectives, this application provides the following solutions:
[0006] In a first aspect, the present application provides a network security situation assessment method, comprising:
[0007] Obtain network traffic data within a set period of time to obtain a network traffic data set;
[0008] Based on the network security situation assessment model, each network flow data in the network flow data set is classified respectively to determine the category of each network flow data; the network security situation assessment model is pre-trained using a training sample set, and the training sample set includes multiple groups of network flow sample data and corresponding categories; the network security situation assessment model includes a feature extraction network, a recurrent neural network and a fully connected layer connected in sequence; the feature extraction network uses full-dimensional dynamic convolution and multi-scale efficient channel attention mechanism to extract features from network flow data, the recurrent neural network performs time series modeling on the features output by the feature extraction network, and the fully connected layer determines the category of network flow data according to the time series data output by the recurrent neural network;
[0009] According to the categories of each network traffic data, a network security situation value is calculated to determine the network security situation within the set time period.
[0010] In a second aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned network security situation assessment method.
[0011] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the above-mentioned network security situation assessment method is implemented.
[0012] In a fourth aspect, the present application provides a computer program product, including a computer program, which implements the above-mentioned network security situation assessment method when executed by a processor.
[0013] According to the specific embodiments provided in this application, this application discloses the following technical effects:
[0014] The present application provides a network security situation assessment method, device, medium and product. In the network security situation assessment model, Omni-Dimensional Dynamic Convolution (ODConv) and Multi-Scale Efficient Channel Attention Mechanism (MSECA) are used to extract features from network traffic data. ODConv can gradually reduce the size of the feature map and increase the number of channels. Combined with MSECA, a focus strategy for different channels can be implemented, which can capture more extensive and comprehensive semantic information while retaining detail information, thereby improving the model's ability to capture data features. A recurrent neural network (Bidirectional Recurrent Neural Network, BiGRU) is used to perform time series modeling on the features output by the feature extraction network, and the spatiotemporal correlation of the data is comprehensively considered, so that the network security situation assessment model can better understand the temporal relationship and pattern in the time series data, thereby improving the accuracy of network security situation assessment. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0016] Figure 1 This is an application environment diagram of a network security situation assessment method in one embodiment of the present application;
[0017] Figure 2 A flowchart of a network security situation assessment method provided in one embodiment of the present application;
[0018] Figure 3 A schematic diagram of a calculation process of a network security situation value provided in an embodiment of the present application;
[0019] Figure 4 This is a schematic diagram of the training process of the network security situation assessment model;
[0020] Figure 5 A schematic diagram of the structure of a network security situation assessment model provided in one embodiment of the present application;
[0021] Figure 6 Schematic diagram of a full-dimensional dynamic convolutional layer;
[0022] Figure 7 Schematic diagram of the efficient channel attention mechanism layer;
[0023] Figure 8 Schematic diagram of the multi-scale efficient channel attention mechanism layer;
[0024] Fig. 9 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION
[0025] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0026] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.
[0027] The network security situation assessment method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, or it can be integrated on the server 104, or it can be placed on the cloud or other servers. The terminal 102 can send the network traffic data to be processed to the server 104. After the server 104 receives the network traffic data to be processed, it classifies it based on the network security situation assessment model and calculates the network security situation value. The server 104 can feedback the obtained network security situation value to the terminal 102. In addition, in some embodiments, the network security situation assessment method can also be implemented by the server 104 or the terminal 102 alone.
[0028] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.
[0029] In an exemplary embodiment, Figure 2 and Figure 3As shown, a network security situation assessment method is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate, including the following steps 201 to 203. Among them:
[0030] Step 201, obtaining network traffic data within a set period of time to obtain a network traffic data set.
[0031] Specifically, the network traffic data within a set period is first acquired, and then the network traffic data within the set period is sequentially digitized, standardized and normalized to obtain a network traffic data set.
[0032] The network traffic data includes multiple values. After preprocessing, the network traffic data is 196-feature data and is converted into a 14*14 feature map.
[0033] Step 202: Based on the network security situation assessment model, each network flow data in the network flow data set is classified to determine the category of each network flow data.
[0034] The network security situation assessment model is obtained by pre-training with a training sample set, and the training sample set includes multiple groups of network traffic sample data and corresponding categories.
[0035] Furthermore, if Figure 4 As shown, in the training process of the network security situation assessment model, the network traffic sample data is first preprocessed. This application processes the UNSW-NB15 data set, and the processing operations include: digitization, standardization and normalization operations.
[0036] ① Numericalization: Using UNSW_NB15_testing-set.csv and UNSW_NB15_training-set.csv, the dataset includes 43 features in total. The non-numeric features in the dataset are converted into numerical features using one-hot encoding to ensure that all data are digital, which facilitates the network security situation assessment model to learn data features.
[0037] ② Standardization: In order to reduce the impact of outliers on the classification results, the network traffic sample data is converted into a data set with a mean of 0 and a variance of 1 according to the mean and standard deviation of the network traffic sample data.
[0038] ③ Normalization: Since the values of different eigenvalues may differ greatly, normalization is used to scale the data proportionally and scale them to the same small range. Specifically, the maximum value normalization method is used to normalize the eigenvalues to between 0 and 1.
[0039] After that, the preprocessed data is divided into training set, test set and validation set. A batch of samples are selected from the training set to train the network security situation assessment model. Each round of training determines whether the generalization ability of the model is improved. If yes, the model is saved. If not, the next round of training is continued until the number of training times reaches the upper limit. The optimizer is used to optimize the parameters of the network security situation assessment model, and then the validation set is sent to the network security situation assessment model for training. Then, the result is compared with the previous result. If the result is better than the previous result, the network security situation assessment model of this training is saved and then the training continues. Otherwise, the above process is continued until the number of training cycles is reached. Then the test set is sent to the trained network security situation assessment model for evaluation. Specifically, the test set is first sent to the network security situation assessment model for classification, and then the network security situation value is calculated.
[0040] In an exemplary embodiment, the network status in the UNSW-NB15 data set includes normal, analysis, reconnaissance, fuzzers, denial of service, generic, shellcode, worms, backdoor and exploits. The UNSW-NB15 data set does not contain the true situation value, so this application uses the following calculation method to quantify the situation value representing the degree of network security in a certain period.
[0041] First, the types in the UNSW-NB15 dataset are digitized to correspond to ten types from 0 to 9. The final output of the network security situation assessment model is an integer from 0 to 9 to identify different categories.
[0042] According to the order of each sample collected in the UNSW-NB15 dataset, every 3000 samples are regarded as a period. The impact factors corresponding to the attacks in the dataset are shown in Table 1.
[0043] Table 1 Attack impact factors of UNSW-NB15 dataset
[0044] Attack Type Normal Analysis Reconnaissance Fuzzers DoS Impact Factor 1 2 3 4 5 Attack Type Generic Shellcode Worms Exploits Backdoor Impact Factor 6 7 8 9 10
[0045] This application quantifies the above 10 types of network traffic data and obtains the impact factor of each type of network traffic data. The network security situation value is intended to quantify the network security status and can intuitively reflect the overall security status of the network. The network security situation value proposed in this application is calculated by first calculating the network security situation value based on the impact factor of each type of network traffic data, and then normalizing the network security situation value to obtain the network security situation value.
[0046] like Figure 5 As shown, the network security situation assessment model includes a feature extraction network, a recurrent neural network and a fully connected layer connected in sequence. The feature extraction network uses full-dimensional dynamic convolution and multi-scale efficient channel attention mechanism to extract features from network traffic data, the recurrent neural network performs time series modeling on the features output by the feature extraction network, and the fully connected layer determines the category of network traffic data based on the time series data output by the recurrent neural network.
[0047] In this application, step 202 includes:
[0048] (1) For any network traffic data in the network traffic data set, the feature extraction network is used to extract features of the network traffic data using full-dimensional dynamic convolution and multi-scale efficient channel attention mechanism to obtain network traffic features.
[0049] The feature extraction network includes a first feature extraction unit, a second feature extraction unit and a third feature extraction unit which are connected in sequence.
[0050] The first feature extraction unit includes a convolutional layer, a multi-scale efficient channel attention mechanism layer, a batch normalization layer and an activation function connected in sequence.
[0051] The second feature extraction unit includes a full-dimensional dynamic convolution layer, a multi-scale efficient channel attention mechanism layer, a maximum pooling layer, a batch normalization layer and an activation function connected in sequence.
[0052] The third feature extraction unit includes a full-dimensional dynamic convolution layer, an efficient channel attention mechanism (ECA) layer, a maximum pooling layer, a batch normalization layer and an activation function connected in sequence.
[0053] The scales of the multi-scale efficient channel attention mechanism layers of the first feature extraction unit and the second feature extraction unit are different. The third feature extraction unit only uses the ECA layer to pay attention to the channel, which is more suitable for feature extraction of network security data, and can reduce the amount of calculation and improve efficiency.
[0054] The network traffic data first enters the first feature extraction unit, expands the number of channels through the convolution layer, and then uses the multi-scale efficient channel attention mechanism layer to perform channel attention and feature extraction. Finally, the batch normalization layer and activation function are used to process the features.
[0055] The features extracted by the first feature extraction unit enter the second feature extraction unit, and the full-dimensional dynamic convolution layer dynamically adjusts the convolution kernel for feature extraction, reduces the size of the feature map and expands the number of channels. Then, the multi-scale efficient channel attention mechanism layer focuses on the channel feature extraction again, and the maximum pooling layer retains the main features and reduces the data complexity. Finally, the batch normalization layer and activation function are applied to process the features.
[0056] The features extracted by the second feature extraction unit enter the third feature extraction unit, and the convolution kernel is adjusted through the full-dimensional dynamic convolution layer to extract features, reduce the size of the feature map and expand the number of channels, and then the ECA layer is used to focus on the channels, and the maximum pooling layer is used to retain the main features and reduce the complexity of the data. Finally, the batch normalization layer and activation function are applied to process the features.
[0057] like Figure 6 As shown in the figure, the full-dimensional dynamic convolution layer focuses on the following four dimensions of the kernel space: the spatial kernel size is k×k, the number of input channels is C in , Output channel number C out And the number of convolution kernels n. First, the input data x is compressed to C by channel-by-channel global average pooling. in The feature vector is transformed into a feature vector of length , and passes through a fully connected layer and a rectified linear unit. Then, it passes through four head branches, each of which has an output size of k×k, C in ×1, C out ×1 and n×1 fully connected layers, and an activation function, where the activation functions of the 1st to 3rd head branches are Sigmoid functions, and the activation function of the 4th head branch is Sofmax function, to generate normalized attention weights on each dimension. Next, the attention weights on the four dimensions are applied to the original convolution kernel set respectively, and the adjusted convolution kernel set performs convolution operation on the input data x to obtain the output feature y. Where a s1 ~a sm is the spatial attention weight, a c1 ~a cm is the input channel attention weight, a f1 ~a fm is the output channel attention weight, a w1 ~a wm is the convolution kernel number attention weight, W1~W m is the original convolution kernel set.
[0058] like Figure 7 As shown in the figure, in the efficient channel attention mechanism layer, for each channel of the input feature map Y with a height of H, a width of W, and a number of channels of C, the efficient channel attention mechanism layer first obtains the average response of the channel in the spatial dimension through a global pooling operation (usually average pooling). Then, these pooled features are processed through a one-dimensional convolution with an adaptively selected kernel size (such as selecting a kernel of 5) to calculate the attention weight of each channel. Finally, the calculated channel attention weight is applied to the original feature map, and the final feature representation Y' is generated by weighted summation by channel.
[0059] like Figure 8 As shown in the figure, the multi-scale efficient channel attention mechanism layer first divides the channels into 4 groups by average distribution to perform convolution operations of different scales, and then performs feature fusion at the channel level. Next, according to the previous channel grouping, ECA is applied to each group to enhance the channel attention ability. Specifically, four groups of weight information of size 1×1×C / 4 are first extracted, and feature fusion is performed at the channel level to generate overall weight information of size 1×1×C. Then, these weights are multiplied with the extracted features at the element level, and the obtained channel attention weights are applied to the feature map.
[0060] (2) Time series modeling of the network traffic characteristics is performed through the recurrent neural network to obtain time series data.
[0061] The recurrent neural network includes a first recurrent neural network layer, a first random dropout layer, a second recurrent neural network layer and a second random dropout layer which are connected in sequence.
[0062] The network traffic characteristics are input into the recurrent neural network. Both the first and second recurrent neural network layers need to be randomly inactivated to reduce the dependency between neurons and prevent the network security situation assessment model from overfitting on the training sample set.
[0063] (3) Determine the category of the network traffic data according to the time series data through the fully connected layer.
[0064] The network security situation assessment model provided by this application fully considers the application needs of deep learning in the field of network security. First, ODConv is used to gradually reduce the size of the feature map and increase the number of channels, and MSECA is combined to realize the attention strategy for different channels. Compared with the ordinary channel attention mechanism, by adjusting the channel attention at different scales, it can capture more extensive and comprehensive semantic information while retaining the detailed information, which helps the model to better understand the input data, and realizes efficient allocation of channel attention. Without adding a large number of parameters, the model can adapt to various complex data features, improve the model's ability to capture data features, and thus improve the performance of the model. The result of feature extraction is input into BiGRU for time series modeling. Taking into account the temporal and spatial correlation of the data, the characteristics of the data can be effectively extracted in time and space, so that the network security situation assessment model can better understand the temporal relationship and pattern in the time series data, and improve the modeling ability for time-dependent data, thereby improving the accuracy of network security situation assessment.
[0065] Step 203, calculating the network security situation value according to the category of each network traffic data to determine the network security situation within the set time period.
[0066] The categories of network traffic data include: normal, analysis attack, reconnaissance attack, fuzz testing, denial of service, Trojan attack, vulnerability code, worm, backdoor and penetration attack.
[0067] In an exemplary embodiment, in step 203, firstly, according to the category of each network traffic data, the formula Calculate the network security value; where SV is the network security value, N is the number of network traffic data, T is the number of categories of network traffic data, Q i is the number of the i-th type of network traffic data in the network traffic dataset, E i is the impact factor of the ith type of network traffic data on the network status. Then the network security situation value is normalized to obtain the network security situation value.
[0068] In practical applications, after classifying the actual network traffic data through the network security situation assessment model, it is analyzed according to the actual situation, the number of traffic samples in a period is reset and the impact factor corresponding to the attack is re-quantified, the network security situation value is calculated using the same network security situation value calculation formula, and the network security situation level is divided according to the level classification standard.
[0069] This application further divides the network security level into four levels. The higher the network security situation value, the greater the security risk faced by the network. The level division can enable network security personnel to understand the network security status more intuitively. The corresponding relationship between the network security situation value and the network security situation level is shown in Table 2.
[0070] Table 2 Correspondence between network security situation value and network security situation level
[0071] Network security status value Network security posture level 0.00~0.22 Safety 0.22~0.42 Low risk 0.42~0.72 Medium risk 0.72~1.00 High risk
[0072] This application uses a feature extraction network combined with BiGRU to perform network security situation assessment. After feature extraction in the feature extraction network, since network traffic data has rich time dependencies, these features are input into BiGRU for sequence modeling to better perform network security situation assessment and solve the problem that when convolutional neural networks have strong time dependencies on data, they may not be able to capture complex patterns in time series. In the feature extraction network, ODConv is used to improve the ability of feature extraction, and MSECA is used to enhance the ability to represent important channels, thereby solving the problem of weak feature extraction ability and insufficient attention to channels in ordinary convolution during feature extraction, thereby improving the accuracy of network security situation assessment.
[0073] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Fig. 9 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store network traffic data and a network security situation assessment model. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a network security situation assessment method is implemented.
[0074] Those skilled in the art will understand that Fig. 9The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components. In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.
[0075] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.
[0076] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.
[0077] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.
[0078] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).
[0079] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.
[0080] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0081] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.
Claims
1. A network security situation assessment method, It is characterized in that The network security situation assessment method comprises: Obtain network traffic data within a set period of time to obtain a network traffic data set; Based on the network security situation assessment model, each network flow data in the network flow data set is classified respectively to determine the category of each network flow data; the network security situation assessment model is pre-trained using a training sample set, and the training sample set includes multiple groups of network flow sample data and corresponding categories; the network security situation assessment model includes a feature extraction network, a recurrent neural network and a fully connected layer connected in sequence; the feature extraction network uses full-dimensional dynamic convolution and multi-scale efficient channel attention mechanism to extract features from network flow data, the recurrent neural network performs time series modeling on the features output by the feature extraction network, and the fully connected layer determines the category of network flow data according to the time series data output by the recurrent neural network; According to the categories of each network traffic data, a network security situation value is calculated to determine the network security situation within the set time period.
2. The network security situation assessment method according to claim 1, It is characterized in that Get the network traffic data within the set time period and obtain the network traffic data set, including: Get network traffic data within a set period of time; The network traffic data within the set time period is digitized, standardized and normalized in turn to obtain a network traffic data set.
3. The network security situation assessment method according to claim 1, It is characterized in that Based on the network security situation assessment model, each network flow data in the network flow data set is classified to determine the category of each network flow data, specifically including: For any network traffic data in the network traffic data set, the feature extraction network is used to extract features of the network traffic data by adopting full-dimensional dynamic convolution and multi-scale efficient channel attention mechanism to obtain network traffic features; Performing time series modeling on the network traffic characteristics through the recurrent neural network to obtain time series data; The category of the network traffic data is determined according to the time series data through the fully connected layer.
4. The network security situation assessment method according to claim 3, It is characterized in that The feature extraction network includes a first feature extraction unit, a second feature extraction unit and a third feature extraction unit connected in sequence; The first feature extraction unit includes a convolutional layer, a multi-scale efficient channel attention mechanism layer, a batch normalization layer and an activation function connected in sequence; The second feature extraction unit includes a full-dimensional dynamic convolution layer, a multi-scale efficient channel attention mechanism layer, a maximum pooling layer, a batch normalization layer and an activation function connected in sequence; The third feature extraction unit includes a full-dimensional dynamic convolution layer, an efficient channel attention mechanism layer, a maximum pooling layer, a batch normalization layer and an activation function connected in sequence.
5. The network security situation assessment method according to claim 3, It is characterized in that The recurrent neural network includes a first recurrent neural network layer, a first random dropout layer, a second recurrent neural network layer and a second random dropout layer which are connected in sequence.
6. The network security situation assessment method according to claim 1, It is characterized in that The categories of network traffic data include: normal, analysis attack, reconnaissance attack, fuzz testing, denial of service, Trojan attack, vulnerability code, worm, backdoor and penetration attack.
7. The network security situation assessment method according to claim 1, It is characterized in that According to the categories of each network traffic data, the network security situation value is calculated, including: According to the category of each network traffic data, the formula is used Calculate the network security value; where SV is the network security value, N is the number of network traffic data, T is the number of categories of network traffic data, Q i is the number of the i-th type of network traffic data in the network traffic dataset, E i is the impact factor of the i-th type of network traffic data on the network status; The network security situation value is normalized to obtain a network security situation value.
8. A computer device, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the network security situation assessment method according to any one of claims 1 to 7.
9. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the network security situation assessment method described in any one of claims 1 to 7 is implemented.
10. A computer program product comprising a computer program, It is characterized in that When the computer program is executed by a processor, the network security situation assessment method described in any one of claims 1 to 7 is implemented.
Citation Information
Patent Citations
Method and device for predicting and evaluating operation state of electric power information acquisition system
CN112884008A
Gait recognition method and device based on improved convolutional neural network, and intelligent terminal
CN114582018A
Network security situation prediction method based on CNN and ResGRU
CN116866013A
Abnormal traffic detection method based on cyclic generative adversarial network and multi-head self-attention mechanism
CN117218494A
Network security situation prediction method, system and equipment
CN117439800A
Cited By
Data processing method and device for cyberspace situation analysis
CN120050187A
