Abnormal traffic detection method and related equipment
By acquiring and deeply analyzing system traffic data in real time in the power industrial control network, using the pre-trained abnormal traffic recognition model, the problem of inaccurate identification results in the existing technology is solved, and more efficient malicious attack identification and false alarm leakage is achieved Reduced rate of return.
Patent Information
- Application Number
- CN202410936659.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-12
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-07-12
AI Technical Summary
The prior art identification results are not accurate enough when identifying abnormal traffic in the power industrial control network, resulting in difficulty in identifying malicious attacks and high false alarm and missed alarm rates.
Use real-time acquisition of system traffic data of industrial control systems, identify and analyze protocol data through deep packet detection technology, and enter a pre-trained abnormal traffic recognition model after conversion processing to determine whether the system traffic data is abnormal, expand the perception range, and adapt to different scales. Deployment requirements and deeply analyze conventional network protocols.
It improves the accuracy of abnormal traffic identification, reduces false alarms and missed reports, can identify malicious signaling, adapt to complex environments, and provide efficient and accurate network security detection.
Smart Images

Figure CN119011190A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of power system information security technology, and in particular to an abnormal flow detection method and related equipment. Background Art
[0002] Identifying abnormal behaviors in the network is of great significance for defending against malicious attacks and maintaining network security. When network attacks occur, they often leave behavioral traces in network traffic, so abnormal behavior detection based on traffic is a common network security detection method.
[0003] In the related art, the identification of abnormal traffic in the power industrial control network has the problem of inaccurate identification results. Summary of the invention
[0004] In view of this, the purpose of this application is to propose an abnormal flow detection method and related equipment.
[0005] Based on the above purpose, the present application provides an abnormal flow detection method, which is applied to an industrial control system, and the method includes:
[0006] Acquire system flow data in an industrial control system in real time; the system flow data includes flow data of a kernel layer and flow data of a user layer;
[0007] Identify and parse the system traffic data based on deep data packet inspection technology to obtain protocol data in the system traffic data;
[0008] Converting the protocol data and obtaining a sequence to be tested based on the converted protocol data;
[0009] Inputting the sequence to be tested into a pre-trained abnormal traffic recognition model;
[0010] According to the output result of the abnormal flow identification model, it is determined whether the system flow data is abnormal.
[0011] In some embodiments, the abnormal traffic identification model is an abnormal traffic detection model; and determining whether the system traffic data is abnormal according to the output result of the abnormal traffic identification model includes:
[0012] Get the predicted value of the sequence to be tested;
[0013] According to the detection value output by the abnormal flow detection model and the predicted value, it is determined that the system flow is abnormal flow.
[0014] In some embodiments, the detection value includes the number of abnormal flow data, and determining whether the system flow data is abnormal based on the detection value and the predicted value includes:
[0015] In response to determining that the number of abnormal flow data in the detection value is greater than the number of abnormal flow data in the predicted value, it is determined that the system flow data is abnormal.
[0016] In some embodiments, the real-time acquisition of system flow data in the industrial control system includes:
[0017] Obtain kernel layer traffic data in industrial control systems in real time through the data plane development kit;
[0018] The user layer traffic data in the industrial control system is obtained in real time through XDP based on XML data encapsulation.
[0019] In some embodiments, the converting the protocol data and obtaining the sequence to be tested based on the converted protocol data includes:
[0020] Encode all sub-data fields in the protocol data to obtain status data;
[0021] The state data within the preset time length is sorted according to time to obtain the sequence to be tested.
[0022] In some embodiments, encoding all sub-data fields in the protocol data includes:
[0023] Determining the number of codes of the sub-data field according to the number of changed values of the sub-data field in the protocol data;
[0024] All sub-data fields in the protocol data are encoded according to the number of the codes of the sub-data fields.
[0025] In some embodiments, the protocol data includes application information; the identification and parsing of the system traffic data based on the deep packet inspection technology to obtain the protocol data in the system traffic data includes:
[0026] Obtaining a corresponding protocol data packet in the system flow data by identifying a port number and a data packet structure in the system flow data;
[0027] Extract application information from the corresponding protocol data packet.
[0028] In some embodiments, the protocol data includes Modbus protocol data; the identifying and parsing the system flow data based on deep data packet inspection technology to obtain the protocol data in the system flow data includes:
[0029] According to the port number, data packet structure and data packet property of the Modbus protocol, a Modbus protocol data packet in the system flow data is obtained;
[0030] The device address, function code, data address, data quantity and additional data or parameters in the Modbus protocol data packet are obtained.
[0031] In some embodiments, the abnormal traffic identification model is a traffic threshold prediction model; determining whether the system traffic data is abnormal based on the output result of the abnormal traffic identification model includes: in response to determining that the traffic threshold output by the traffic threshold prediction model exceeds the threshold of a normal baseline, determining that the system traffic data is abnormal.
[0032] The embodiment of the present application also provides an abnormal flow detection device, which is applied to an industrial control system, and the device includes:
[0033] An acquisition module is used to acquire system flow data in the industrial control system in real time; the system flow data includes flow data of the kernel layer and flow data of the user layer;
[0034] A parsing module, used to identify and parse the system flow data based on deep data packet inspection technology to obtain protocol data in the system flow data;
[0035] A processing module, used for converting the protocol data and obtaining a sequence to be tested based on the converted protocol data;
[0036] An input module, used for inputting the sequence to be tested into a pre-trained abnormal traffic recognition model;
[0037] The judgment module is used to determine whether the system flow data is abnormal according to the output result of the abnormal flow recognition model.
[0038] An embodiment of the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any of the methods described above when executing the program.
[0039] An embodiment of the present application also provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable a computer to execute any of the methods described above.
[0040] An embodiment of the present application further provides a computer program product, comprising computer program instructions, which, when executed on a computer, enable the computer to execute any of the methods described above.
[0041] From the above, it can be seen that the abnormal traffic detection method provided by this application can automatically collect system traffic data of the control system in real time and expand the perception range, which can adapt to deployment requirements of different scales; in addition, the application layer data outside the ordinary detection range is analyzed by using deep data packet detection technology, which can deeply analyze conventional network protocols, so it can provide more diverse behavior patterns and adapt to more complex environments. Through the pre-trained abnormal traffic recognition model, more accurate recognition results can be obtained, thereby identifying malicious signaling and reducing false positives and false negatives. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] In order to more clearly illustrate the technical solutions in the present application or related technologies, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0043] Figure 1 A schematic diagram of a flow chart of an abnormal flow detection method according to an embodiment of the present application;
[0044] Figure 2 A schematic diagram of encoding flow data according to an embodiment of the present application;
[0045] Figure 3 This is an architecture diagram of a flow threshold prediction model according to an embodiment of the present application;
[0046] Figure 4 A schematic diagram of the modeling process of the abnormal traffic detection model of an embodiment of the present application;
[0047] Figure 5 A schematic diagram of training an abnormal traffic detection model according to an embodiment of the present application;
[0048] Figure 6 This is another flow chart of the abnormal traffic detection method according to an embodiment of the present application;
[0049] Figure 7 A schematic diagram of the modeling process of the flow threshold prediction model of an embodiment of the present application;
[0050] Figure 8 This is a schematic diagram of an abnormal flow detection device according to an embodiment of the present application;
[0051] Fig. 9 A schematic diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION
[0052] In order to make the objectives, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.
[0053] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present application should be the usual meanings understood by people with ordinary skills in the field to which the present application belongs. The "first", "second" and similar words used in the embodiments of the present application do not represent any order, quantity or importance, but are only used to distinguish different components. "Including" or "comprising" and similar words mean that the elements or objects appearing in front of the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.
[0054] The traffic detection method based on normal behavior baseline is a behavior-based detection method that identifies abnormal behavior by establishing a normal behavior baseline, thereby discovering possible malicious attacks. Before implementing this threat detection method, it is necessary to collect information from multiple sources such as the network, endpoints, telemetry on the network, and identity services and cloud services. After preprocessing and standardization, this information can be used to train and optimize algorithms such as machine learning or statistical analysis to establish a baseline model that reflects normal behavior patterns. Once the normal behavior baseline is established, the collected real-time data can be compared with it to discover any abnormal behavior. If any threat is detected, the system will immediately trigger an alarm response so that the administrator can take appropriate defense and response measures in time. Therefore, the threat detection method based on normal behavior baseline is an efficient and accurate malicious attack detection method that can help organizations detect and respond to potential security threats in a timely manner.
[0055] Traffic-based abnormal behavior is divided into two stages: network traffic data collection and anomaly detection.
[0056] Among them, network traffic data is obtained by actively sending detection messages or passively collecting them through mirrored ports, and traffic features are extracted according to needs, such as basic features of the connection, content features of the connection, traffic statistical features, etc.
[0057] Among them, the core problem of anomaly detection algorithms is the classification of normal traffic and abnormal traffic. Therefore, many security researchers actively explore detection methods that establish a normal behavior baseline to identify abnormal behavior, that is, by correctly establishing a normal traffic behavior baseline and screening out abnormal traffic that deviates from the behavior baseline. Currently, there are mainly detection methods based on statistical models, detection methods based on supervised classification models, and detection methods based on unsupervised models.
[0058] The detection method based on statistical models is mainly used in scenarios where the data at the point of abnormal occurrence changes dramatically. It uses statistical features such as mean, variance, distribution, hidden Markov model, maximum entropy and Gaussian distribution model to describe normal behavior and detect abnormal traffic. The statistical model requires a learning period, during which the extracted traffic data will be recorded and then used to create the statistical model.
[0059] The supervised classification model is trained based on the normal and abnormal traffic currently obtained to obtain a binary classifier, and then the binary classifier is used to classify unknown data. It can use traditional machine learning models such as logistic regression, support vector machine, random forest, gradient boosting tree, etc., or deep learning models such as convolutional neural network, long short-term memory network, deep residual network, etc. Traditional machine learning models have fewer parameters, low requirements for sample data volume, low resource consumption, and fast training speed, but the learning ability is not as strong as the deep learning model, and it needs to rely on manual experience for feature extraction. The deep learning model needs to be trained on a large amount of data, and the resource consumption is large, but the model has strong learning ability and can directly use the original data to achieve end-to-end learning.
[0060] Compared with supervised classification models, unsupervised classification does not require training data and mainly uses online data for learning and detection. Commonly used unsupervised models include clustering, isolation forest, autoencoder, generative adversarial network models, etc. However, this method also requires a learning period. During the learning period, the system will use data to train the model. During this stage, the model does not have detection capabilities. After the learning period is over, the model can be used to detect abnormal traffic.
[0061] However, although the above methods can learn features from traffic data, they do not have a complete preliminary traffic collection solution and fine-grained data identification and analysis strategy. Therefore, the related technology has the problem that the collection of traffic data is not comprehensive enough, and the analysis of traffic data is not accurate enough, resulting in the problem that the obtained model is not accurate enough for traffic detection.
[0062] Based on this, the embodiments of the present application provide an abnormal traffic detection method and related equipment, which can adapt to deployment requirements of different scales by automatically collecting system traffic data of the control system in real time and expanding the perception range; in addition, the application layer data outside the ordinary detection range is analyzed by using deep message detection technology, which can deeply analyze conventional network protocols, thus providing more diverse behavior patterns and adapting to more complex environments. It can solve the problem of inaccurate traffic detection in existing models to a certain extent.
[0063] Figure 1 A flow chart of an exemplary abnormal traffic detection method according to an embodiment of the present application is shown.
[0064] like Figure 1 As shown, an embodiment of the present application provides an abnormal flow detection method, which is applied to an industrial control system. Generally, the industrial control system may be an industrial control system in an electric power industrial control network. The method may include:
[0065] S100, acquiring system flow data in the industrial control system in real time; the system flow data includes flow data of the kernel layer and flow data of the user layer;
[0066] S200, identifying and parsing the system traffic data based on deep data packet inspection technology to obtain protocol data in the system traffic data;
[0067] S300, converting the protocol data, and obtaining a sequence to be tested based on the converted protocol data;
[0068] S400, inputting the sequence to be tested into a pre-trained abnormal traffic recognition model;
[0069] S500: Determine whether the system flow data is abnormal according to the output result of the abnormal flow identification model.
[0070] The abnormal traffic detection method provided in the embodiment of the present application can automatically collect the system traffic data of the control system in real time and expand the perception range, which can adapt to the deployment requirements of different scales; in addition, the application layer data outside the ordinary detection range is analyzed by using deep data packet detection technology, which can deeply analyze the conventional network protocol, so it can provide more diverse behavior patterns and adapt to more complex environments. Through the pre-trained abnormal traffic recognition model, more accurate recognition results can be obtained, thereby identifying malicious signaling and reducing false positives and false negatives.
[0071] In step S100, in some embodiments, the real-time acquisition of system flow data in the industrial control system may include:
[0072] The data plane development kit (dpdk) is used to obtain the kernel layer traffic data in the industrial control system in real time.
[0073] The user layer flow data in the industrial control system is obtained in real time through the XDP (XML Data Package) technology based on XML data encapsulation.
[0074] In some of the embodiments, traffic collection includes a functional interface for collecting mirror data, and uses technologies such as dpdk and xdp to collect traffic from the kernel or physical network card. In terms of hardware, a general hardware network card is used to collect photoelectric signals. After completing the capture and caching of network traffic, multi-queue distribution can be completed through Really Simple Syndication (RSS).
[0075] In this way, the network behavior of all network hosts and online power equipment can be perceived, making it easier to automatically detect and provide alarms for "suspicious" network traffic and behavior through the built-in analysis engine.
[0076] In step S200, in some embodiments, the protocol may include: conventional Internet protocols such as HTTP, FTP, SMTP, POP3, IMAP, SMB, and industrial control field protocol Modbus protocol. The protocol data may include application information; the system flow data is identified and parsed by deep packet inspection technology to obtain the protocol data in the system flow data, which may be obtained by identifying the port number and packet structure in the system flow data, and extracting the user information in the corresponding protocol packet. In this way, the 2-7 layer traditional network flow and industrial control network flow in the collected flow data can be identified and parsed by deep packet inspection technology (Deep Packet Inspection, DPI). Compared with ordinary message detection that only analyzes the content below the 4th layer of IP packets, it has the ability to simultaneously analyze the source address, destination address, source port, destination port, protocol type and application layer in the data packet, etc., and can better identify various application protocols and deeply parse the content of each application protocol.
[0077] In some embodiments, the protocol data may include Modbus protocol data; the identification and parsing of the system flow data based on the deep data packet inspection technology to obtain the protocol data in the system flow data may include:
[0078] According to the port number, data packet structure and data packet property of the Modbus protocol, a Modbus protocol data packet in the system flow data is obtained;
[0079] Obtain one of the additional data and additional parameters, the device address, the function code, the data address and the data quantity in the Modbus protocol data packet. That is, the protocol data packet includes the Modbus protocol data packet; the Modbus protocol data packet includes one of the additional data and additional parameters, the device address, the function code, the data address and the data quantity.
[0080] Usually, the port number of the Modbus (Modbus TCP) protocol can be port 502. The format of the Modbus protocol data packet can include the device address, function code, data, and check (such as CRC check). Among them, the function code can indicate the port number and the format of the data packet, and can determine whether the data packet is a Modbus request. The function code can indicate the type of operation requested (such as reading coil status, writing a single register, etc.), so the nature of the data packet can be obtained by analyzing the function code of the data packet.
[0081] In some embodiments, the protocol data may include HTTP protocol data; the identification and parsing of the system traffic data based on the deep packet inspection technology to obtain the protocol data in the system traffic data may include:
[0082] According to the port number and data packet structure of the HTTP protocol, the port number and data packet structure in the system flow data are identified to obtain the HTTP protocol data packet in the system flow data;
[0083] Obtain the request method, request URL, response status code, response content, etc. in the HTTP protocol data packet.
[0084] Usually, the port of the HTTP protocol may be port 80. The structure of the data packet may be header information of the data packet, etc. By analyzing the header information of the data packet and the commonly used port of the protocol, the HTTP protocol can be identified.
[0085] In some embodiments, the protocol data may include FTP protocol data; the identification and parsing of the system traffic data based on the deep packet inspection technology to obtain the protocol data in the system traffic data may include:
[0086] According to the port number and data packet structure of the FTP protocol, the port number and data packet structure in the system flow data are identified to obtain the FTP protocol data packet in the system flow data;
[0087] Get the command or response in the FTP protocol data packet.
[0088] Usually, the port of FTP protocol can be port 21. The structure of the data packet can be the header information of the data packet, etc. By analyzing the header information of the data packet and the commonly used port of the protocol, the FTP protocol can be identified. The command in the FTP protocol data packet can include RETR or STOR, etc., and the response can include status codes such as 220 or 250. By parsing these commands and responses, the application information in the FTP session can be understood.
[0089] In some embodiments, the protocol data may include SMTP protocol data; the identification and parsing of the system traffic data based on deep packet inspection technology to obtain the protocol data in the system traffic data may include:
[0090] According to the port number and data packet structure of the SMTP protocol, the port number and data packet structure in the system flow data are identified to obtain the SMTP protocol data packet in the system flow data;
[0091] Get the command and response in the SMTP protocol data packet.
[0092] Generally, the SMTP protocol is used for the transmission of e-mails. The port of the SMTP protocol may be port 25. The structure of the data packet may be the header information of the data packet, etc. By analyzing the header information of the data packet and the commonly used ports of the protocol, the SMTP protocol can be identified. The commands in the SMTP protocol data packet may include HELO, MAILFROM, RCPTTO, DATA, etc. By parsing these commands and responses, the sending process of the email can be tracked.
[0093] In some embodiments, the protocol data may include POP3 protocol data; the identification and parsing of the system traffic data based on the deep packet inspection technology to obtain the protocol data in the system traffic data may include:
[0094] According to the port number, data packet structure and data packet property of the POP3 protocol, the port number and data packet structure in the system flow data are identified to obtain the POP3 protocol data packet in the system flow data;
[0095] The command in the POP3 protocol data packet and the server's response are obtained.
[0096] Typically, POP3 is used to retrieve mail from a mail server. The port of the POP3 protocol may be port 110. The structure of the data packet may be the header information of the data packet, etc. By analyzing the header information of the data packet and the commonly used ports of the protocol, the POP3 protocol can be identified. The commands in the POP3 protocol data packet may include commands such as USER, PASS, LIST, or RETR. By parsing these commands and responses, it is possible to understand how the user retrieves and downloads mail.
[0097] In some embodiments, the protocol data may include IMAP protocol data; identifying and parsing the system traffic data based on deep packet inspection technology to obtain the protocol data in the system traffic data may include:
[0098] According to the port number and data packet structure of the IMAP protocol, the port number and data packet structure in the system flow data are identified to obtain the IMAP protocol data packet in the system flow data;
[0099] Get the command in the IMAP protocol data packet.
[0100] Generally, the IMAP protocol allows users to manage emails online. The port of the IMAP protocol may be port 143. The structure of the data packet may be a data packet feature corresponding to the protocol. By analyzing the features corresponding to the protocol in the data packet and the ports commonly used by the protocol, the IMAP protocol can be identified. The commands in the IMAP protocol data packet may include commands such as LOGIN, SELECT, or FETCH. By parsing these commands, it is possible to understand how users perform operations such as reading, moving, and deleting emails online.
[0101] In some embodiments, the protocol data may include SMB protocol data; the identifying and parsing the system traffic data based on the deep packet inspection technology to obtain the protocol data in the system traffic data may include:
[0102] According to the port number and data packet structure of the SMB protocol, the port number and data packet structure in the system flow data are identified to obtain the SMB protocol data packet in the system flow data;
[0103] Get the request and response in the SMB protocol packet.
[0104] Generally, the SMB protocol is mainly used for file and printer sharing. The port of the SMB protocol can be port 445. The structure of the data packet can be the header information of the data packet. By analyzing the header information in the data packet and the commonly used port of the protocol, the SMB protocol can be identified. The request and response in the IMAP protocol data packet can include operations such as file opening, reading, writing, closing, and commands related to printer management. By parsing these commands, operations on files and printers can be understood.
[0105] In some embodiments, when the DPI engine identifies a data packet, it can identify the data packet in the system traffic data by matching the data in the data packet with the information of the corresponding protocol in the protocol library (e.g., port number, protocol characteristics, data packet structure, data packet header information, etc.). Generally, the protocol library can be equipped in the DPI system.
[0106] In some of the embodiments, the parsing of system traffic data can be completed by tools such as Suricata or Wireshark. For example, the collected raw data packets (rawpacket) can be subjected to deep protocol parsing and abnormal behavior analysis by the Suricata network security monitoring engine to obtain high-level events, namely, power industrial control protocol logs and alarm logs. Vector (a high-performance observable data container / pipeline) can be used as a data extraction, transformation and loading tool (ETL, extracttransformload) to convert the high-level events obtained by Suricata (an open source network threat detection engine) parsing network traffic data, and then send them to the Elasticsearch (an open source distributed full-text search engine) system for storage. The Elasticsearch system is used for storage and retrieval of high-level events, and MySQL (an open source cross-platform database management system) is used to store configuration parameters.
[0107] In step S300, usually, a series of data preprocessing processes such as data cleaning, data division, and type conversion are required for the protocol data to obtain state data. The state data within a preset time period is arranged in time series to obtain a sequence to be tested.
[0108] In some embodiments, converting the protocol data and obtaining the sequence to be tested based on the converted protocol data may include:
[0109] Encode all sub-data fields in the protocol data to obtain status data;
[0110] The state data within the preset time length is sorted according to time to obtain the sequence to be tested.
[0111] In some embodiments, for different protocols, corresponding sub-data fields can be extracted respectively, that is, various sub-data fields included in the protocol itself can be extracted. The sub-data fields can be, for example, source address, destination address, protocol identifier, control information, data and checksum or other error detection mechanisms, etc. It can be understood that different protocols may correspond to different field formats or different fields, etc., but for any certain type of protocol, the sub-data fields are usually fixed.
[0112] Taking the industrial control protocol modbus protocol as an example, the sub-data field in the protocol is described. It can include 8 fields, which can be specifically shown in the following Table 1.
[0113] Table 1 Sub-data fields in the industrial control protocol modbus protocol
[0114] Fields meaning source Source IP address destination Target IP address length Data length unitidentifier Disembarkation address functioncode Function code referenceaddress Register start address wordcount Number of operands date Traffic date
[0115] In some of the embodiments, encoding all sub-data fields in the protocol data may include: determining the number of codes of the sub-data fields according to the number of changing values of the sub-data fields in the protocol data; and encoding all sub-data fields in the protocol data according to the number of codes of the sub-data fields.
[0116] Taking specific protocol data as an example, the value of the number of codes for determining the sub-data field is explained. For example, for specific protocol data (i.e. signaling information) "10.10.111.1, 10.10.111.2, 6, 0, 4, 1536, 4, 2024.6.15", it contains 8 segments of sub-data. Encode each segment of sub-data, and use a digital code to represent each sub-data. The number of codes for each sub-data is the number of changes in the value of the sub-data. The specific encoding can be as follows Figure 2 As shown in the figure, each flow can be constructed into a state.
[0117] In some embodiments, the values in the same interval may be encoded as the same value, which can avoid the inability to accurately identify and accurately identify the numerical sequence of approximate periods. Furthermore, the values near the boundary can be determined by voting to solve the problem that the values near the boundary may cross the boundary due to encoding, thereby eliminating the impact of the value floating up and down.
[0118] In some embodiments, the preset unit time length may be five seconds. According to the original traffic packets, five seconds is taken as a unit time, the total number of data packets in the unit time is counted, and the data packets are sorted according to the time series, which can be sorted according to the characteristics of the baseline in the industrial control network to obtain a state sequence.
[0119] In some of the embodiments, the state sequence may be a five-dimensional state sequence, for example, may include the sum of the lengths of data packets within a unit time (five seconds), minimum value, maximum value, average value, and quantity statistics.
[0120] In step S400, the pre-trained abnormal traffic identification model may be an abnormal traffic detection model or a traffic threshold prediction model. In some embodiments, the pre-trained abnormal traffic identification model may be an abnormal traffic detection model. The abnormal traffic detection model may include an LSTM detection model. Among them, the LSTM-based model is used for a recurrent network, each layer uses a layer-by-layer decreasing neuron, the middle layer uses ReLU as an activation function, and the output layer also uses a ReLU activation function. The loss function uses MSE, and the optimizer uses Adam.
[0121] In some embodiments, such as Figure 4As shown, the modeling process of the abnormal traffic detection model may include:
[0122] Add the first LSTM layer, set the first number of neurons, the first activation function and the first input structure; illustratively, the first number of neurons units can be 80; the first activation function activation can be relu, and the input structure input_shape can be (None.5).
[0123] Add a Dropout layer and set the probability value; for example, the probability value can be 0.25.
[0124] Add a second LSTM layer, set the second number of neurons and the second activation function; illustratively, the second number of neurons units can be 80; the second activation function activation can be relu.
[0125] Add a fully connected layer and set the third number of neurons; illustratively, the third number of neurons of the fully connected layer Dense layer can be 1.
[0126] Add a third activation function; the activation function activation can be relu;
[0127] Add the first loss function and set the first optimizer. The first loss function loss can be mse. The first optimizer optimizer can be adam, which completes the construction of the LSTM-based model.
[0128] In some of the embodiments, a Bayesian optimization algorithm may be introduced into the prediction model, wherein the performance and accuracy of the model may be improved by optimizing the data classification prediction research of the long short-term memory network (Bayes-LSTM) based on the Bayesian optimization algorithm.
[0129] Typically, the acquisition of the training data set may be specifically as shown in the aforementioned steps S100-S300, and the obtained training data set may be stored in an elastisearch database.
[0130] In some embodiments, such as Figure 5 and Figure 6 As shown, the training method of the abnormal traffic detection model may include:
[0131] Acquire historical system flow data in the industrial control system; the historical system flow data includes historical flow data of the kernel layer and historical flow data of the user layer.
[0132] The historical system traffic data is identified and parsed based on deep data packet inspection technology to obtain protocol data in the system traffic data.
[0133] The protocol data is converted and processed, and a training data set is obtained based on the converted data. Usually, the training data set can be stored in the elastisearch database. Usually, the protocol data needs to undergo a series of data preprocessing processes such as data cleaning, data partitioning, and data conversion to obtain state data. And according to the sliding window idea, every 61 data of all the state data sorted in time series (the first 60 data are used as input data, and the last data is the label value, indicating the output value) are constructed into a set of training data. Among them, for normal training data, the state data are all positive samples of data, and the generated state is stored as a normal state whitelist, which can facilitate the subsequent state detection process. For the training data obtained from the high-level event conversion, they are all negative sample data.
[0134] According to the training data set, a pre-built model is trained to obtain the anomaly detection model.
[0135] In some of the embodiments, the detection value may include: the number of abnormal traffic data items present in the sequence to be tested, and the number of normal traffic data items present in the sequence to be tested.
[0136] Correspondingly, in some embodiments, in step S500, the abnormal traffic identification model is an abnormal traffic detection model; and determining whether the system traffic data is abnormal according to the output result of the abnormal traffic identification model may include:
[0137] Obtain the predicted value of the sequence to be tested; usually, the predicted value can be obtained from an existing industrial system or from a trained flow threshold prediction model.
[0138] Whether the system flow data is abnormal is determined based on the detection value output by the abnormal flow detection model and the predicted value. Specifically, the system flow data may be determined to be abnormal if the number of abnormal flow data in the detection value is greater than the number of abnormal flow data in the predicted value.
[0139] The embodiment of the present application also provides a method for constructing a traffic threshold prediction model. The traffic threshold prediction model may be a TCN-based model, and its network architecture may be, for example, Figure 3 The TCN-based model can be effectively used in the prediction task of time series data. It mainly combines the extended convolution with the causal convolution, and finally combines two of them into residual modules, and forms a structure by superimposing the residual modules.
[0140] In some possible embodiments, in step S400, Figure 7 As shown, the modeling process of the traffic threshold prediction model may include:
[0141] Create input and set batch_shape of input structure;
[0142] Add TCN layer, set parameters such as nb_fiters and the fourth activation function activation;
[0143] Add a fully connected layer;
[0144] That is, the construction of the TCN-based model is completed;
[0145] Add a second loss function and set a second optimizer, wherein the second loss function loss can be mse. The second optimizer optimizer can be adam, that is, a traffic threshold prediction model anomaly detection model is constructed.
[0146] In some possible embodiments, an improved Bayesian optimization algorithm (IBOA) and a temporal convolutional neural network (TCN) traffic threshold prediction model can be used. By introducing a temporal pattern attention mechanism (TPA) to extract the weights of each input feature, the temporal nature of the traffic threshold historical data can be improved. In addition, a reversible residual network can be introduced to reduce the memory occupied during the TCN model training process. Finally, the Bayesian optimization algorithm is used to optimize the hyperparameters (kernel size K, expansion coefficient d) in the temporal convolutional network, and the traffic threshold prediction can be performed after the optimal model is obtained.
[0147] In some possible embodiments, the training method of the flow threshold prediction model may include:
[0148] Acquire historical system flow data in the industrial control system; the historical system flow data includes historical flow data of the kernel layer and historical flow data of the user layer.
[0149] The historical system traffic data is identified and parsed based on deep data packet inspection technology to obtain protocol data in the system traffic data.
[0150] The protocol data is converted and processed, and a training data set is obtained based on the converted data. Usually, the training data set can be stored in the elastisearch database. Usually, the protocol data needs to undergo a series of data preprocessing processes such as data cleaning, data partitioning, and data conversion to obtain state data. And according to the sliding window idea, every 61 data of all the state data sorted in time series (the first 60 data are used as input data, and the last data is a label value, indicating the output value) are constructed into a set of training data. Exemplarily, when there are 100 statistical data in the original historical traffic, there are 41 training data, and the 41 here is obtained by 100-60+1. When the actual data set size is not enough, the idea of sliding windows can be used to construct a training sample set. The above-mentioned statistics (sum, average, max, min and count) are obtained through various statistical algorithms, and after obtaining these statistics, they are used as the processed observation data set. Then, a window with a width of 60 units is used to slide backward on the observation data set to form the training data set X = {x1, x2, ..., xn}, and the observation value immediately after the window forms the training data set Y = {y1, y2, ...yn}. Usually, the dimension of each data can be 5 bits, that is, sum, average, max, min and count. In actual applications, feature selection can also be performed on it, and data of different dimensions can be used.
[0151] According to the training data set, a pre-built model is trained to obtain the flow threshold prediction model anomaly detection model.
[0152] In this way, by adopting a TCN-based model, it can be effectively used in the prediction task of time series data. It is mainly achieved by combining extended convolution with causal convolution, and finally combining them into residual modules in pairs, and then forming a structure by superimposing residual modules.
[0153] Based on the same inventive concept, the embodiment of the present application also provides an abnormal flow detection method, which is applied to an industrial control system. The method may include:
[0154] Acquire system flow data in an industrial control system in real time; the system flow data includes flow data of a kernel layer and flow data of a user layer;
[0155] Identify and parse the system traffic data based on deep data packet inspection technology to obtain protocol data in the system traffic data;
[0156] Converting the protocol data and obtaining a sequence to be tested based on the converted protocol data;
[0157] Inputting the sequence to be tested into a pre-trained abnormal traffic recognition model;
[0158] Determine whether the system flow data is abnormal based on the output result of the abnormal flow identification model. The abnormal flow identification model may be a flow threshold prediction model. Generally, when the predicted value exceeds the threshold of the normal baseline, it is judged to be abnormal, and when the predicted value does not exceed the threshold of the normal baseline, it is judged to be normal. That is, determining whether the system flow data is abnormal based on the output result of the abnormal flow identification model may include: determining that the system flow is abnormal in response to determining that the flow threshold output by the flow threshold prediction model exceeds the threshold of the normal baseline.
[0159] The baseline can usually show the state of network flow in a normal production environment. Under normal circumstances, the state of the 5G power network system is stable and there will be no drastic fluctuations. When a distributed denial of service attack (DDos) occurs, each attack proxy host will send a large number of service request packets to the target host. The services requested by these packets often consume a large amount of system resources, causing the target host to be unable to provide normal services to users, resulting in a crash. When a large number of data packets gather in a short period of time, resulting in a traffic peak, the large difference from the normal baseline can be used as a sign of abnormal signaling. Since the baseline is a curve obtained by statistics in time order, it can be regarded as a time series data. Packet analysis of industrial control networks can also be called packet capture analysis. The purpose of analyzing data is to have an in-depth understanding of a series of events that occurred on the industrial control network. The pcap file obtained by wireshark packet capture is a common network packet file storage format. According to a certain time interval, the total number of data packets in the time period is counted, and these data are stored in time order to obtain a statistical curve, namely the baseline, which can represent the network traffic situation in the production line under normal circumstances. The security baseline can be regarded as a basic line in data measurement, calculation or positioning. For information security, the security baseline can be considered as the minimum set that enables the system to operate. It is the baseline generated under normal conditions based on the minimum performance and functional requirements. When the subsequent baseline to be tested is not within the control range of the previous baseline, it is very likely that an abnormal situation has occurred.
[0160] The method of the embodiment of the present application relies on the DPDK technology and the XDP technology to collect the flow data of the power industrial control network control system in real time and expand the perception range, which can adapt to the deployment requirements of different scales; in addition, the application layer data outside the ordinary detection range is analyzed by using the deep packet inspection technology (DPI), which can deeply analyze the conventional network protocol, so it can provide more diverse behavior patterns and adapt to more complex environments. Finally, through the detection model, it aims to obtain a better normal baseline, thereby identifying malicious signaling and reducing false positives and false negatives.
[0161] The present invention not only improves the monitoring and analysis capabilities of the power industrial control network, but also provides an efficient, accurate and scalable solution for the network security field, which helps to ensure the safe and stable operation of key infrastructure. The method provided in the embodiment of the present application can be used to identify abnormal traffic and can adapt to scenarios such as traffic thresholds.
[0162] It is understandable that before using the technical solutions of each embodiment of the present disclosure, the type, scope of use, usage scenarios, etc. of the personal information involved will be informed to the user in an appropriate manner, and the user's authorization will be obtained.
[0163] For example, in response to receiving an active request from a user, a prompt message is sent to the user to clearly remind the user that the operation requested to be performed will require obtaining and using the user's personal information. Thus, the user can independently choose whether to provide personal information to software or hardware such as an electronic device, application, server, or storage medium that performs the operation of the technical solution of the present disclosure according to the prompt message.
[0164] As an optional but non-limiting implementation, in response to receiving the user's active request, the prompt information may be sent to the user in the form of a pop-up window, in which the prompt information may be presented in text form. In addition, the pop-up window may also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.
[0165] It is understandable that the above notification and the process of obtaining user authorization are merely illustrative and do not constitute a limitation on the implementation of the present disclosure. Other methods that meet relevant laws and regulations may also be applied to the implementation of the present disclosure.
[0166] It should be noted that the method of the embodiment of the present application can be performed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present application, and the multiple devices will interact with each other to complete the described method.
[0167] It should be noted that the above describes some embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0168] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides an abnormal flow detection device.
[0169] refer to Figure 8 The abnormal flow detection device 800 is applied to an industrial control system, and the device comprises:
[0170] The acquisition module 810 is used to acquire system flow data in the industrial control system in real time; the system flow data includes flow data of the kernel layer and flow data of the user layer;
[0171] A parsing module 820, configured to identify and parse the system traffic data based on a deep packet inspection technology to obtain protocol data in the system traffic data;
[0172] The processing module 830 is used to convert the protocol data and obtain a sequence to be tested based on the converted protocol data;
[0173] An input module 840 is used to input the sequence to be tested into a pre-trained abnormal traffic recognition model;
[0174] The judgment module 850 is used to determine whether the system flow data is abnormal according to the output result of the abnormal flow identification model.
[0175] In some embodiments, the abnormal traffic identification model is an abnormal traffic detection model; and determining whether the system traffic data is abnormal according to the output result of the model includes:
[0176] Get the predicted value of the sequence to be tested;
[0177] Whether the system flow data is abnormal is determined based on the detection value output by the abnormal flow detection model and the predicted value.
[0178] In some embodiments, the detection value includes the number of abnormal flow data, and determining whether the system flow data is abnormal based on the detection value and the predicted value includes:
[0179] In response to determining that the number of abnormal flow data in the detection value is greater than the number of abnormal flow in the predicted value, it is determined that the system flow data is abnormal.
[0180] In some embodiments, the real-time acquisition of system flow data in the industrial control system includes:
[0181] Obtain kernel layer traffic data in industrial control systems in real time through the data plane development kit;
[0182] The user layer traffic data in the industrial control system is obtained in real time through XDP based on XML data encapsulation.
[0183] In some embodiments, the converting the protocol data and obtaining the sequence to be tested based on the converted protocol data includes:
[0184] Encode all sub-data fields in the protocol data to obtain status data;
[0185] The state data within the preset time length is sorted according to time to obtain the sequence to be tested.
[0186] In some embodiments, encoding all sub-data fields in the protocol data includes:
[0187] Determining the number of codes of the sub-data field according to the number of changed values of the sub-data field in the protocol data;
[0188] All sub-data fields in the protocol data are encoded according to the number of the codes of the sub-data fields.
[0189] In some embodiments, the protocol data includes application information; the identification and parsing of the system traffic data based on the deep packet inspection technology to obtain the protocol data in the system traffic data includes:
[0190] Obtaining a corresponding protocol data packet in the system flow data by identifying a port number and a data packet structure in the system flow data;
[0191] Extract application information from the corresponding protocol data packet.
[0192] In some of the embodiments, the protocol data packet includes Modbus protocol data; the Modbus protocol data packet includes one of additional data and additional parameters, a device address, a function code, a data address, and a data quantity.
[0193] In some embodiments, the abnormal traffic identification model is a traffic threshold prediction model; determining whether the system traffic data is abnormal based on the output result of the abnormal traffic identification model includes: in response to determining that the traffic threshold output by the traffic threshold prediction model exceeds the threshold of a normal baseline, the system traffic data is abnormal.
[0194] For the convenience of description, the above device is described in terms of functions divided into various modules. Of course, when implementing the present application, the functions of each module can be implemented in the same or multiple software and / or hardware.
[0195] The device of the above embodiment is used to implement the corresponding abnormal flow detection method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.
[0196] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the abnormal traffic detection method described in any of the above embodiments is implemented.
[0197] Fig. 9 A more specific schematic diagram of the hardware structure of an electronic device provided in this embodiment is shown, and the device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are connected to each other through the bus 1050 in the device.
[0198] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.
[0199] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.
[0200] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.
[0201] The communication interface 1040 is used to connect a communication module (not shown) to realize communication interaction between the device and other devices. The communication module can realize communication through a wired mode (such as USB, network cable, etc.) or a wireless mode (such as mobile network, WIFI, Bluetooth, etc.).
[0202] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).
[0203] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.
[0204] The electronic device of the above embodiment is used to implement the corresponding abnormal flow detection method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.
[0205] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the abnormal traffic detection method described in any of the above embodiments.
[0206] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.
[0207] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute the abnormal traffic detection method described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.
[0208] Based on the same inventive concept, corresponding to the abnormal traffic detection method described in any of the above embodiments, the present disclosure also provides a computer program product, which includes computer program instructions. In some embodiments, the computer program instructions can be executed by one or more processors of a computer so that the computer and / or the processor execute the abnormal traffic detection method. Corresponding to the execution subject corresponding to each step in each embodiment of the abnormal traffic detection method, the processor that executes the corresponding step can belong to the corresponding execution subject.
[0209] The computer program product of the above embodiment is used to enable the computer and / or the processor to execute the abnormal traffic detection method as described in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.
[0210] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present application (including the claims) is limited to these examples. In line with the concept of the present application, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present application as described above, which are not provided in detail for the sake of simplicity.
[0211] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present disclosure (including the claims) is limited to these examples. Based on the concept of the present disclosure, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present disclosure as described above, which are not provided in detail for the sake of simplicity.
[0212] While the disclosure has been described in conjunction with specific embodiments thereof, many alternatives, modifications and variations of these embodiments will be apparent to those skilled in the art in light of the foregoing description.
[0213] The embodiments of the present disclosure are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present disclosure should be included in the scope of protection of the present disclosure.
Claims
1. A method for detecting abnormal traffic, characterized in that: Applied to industrial control systems, the abnormal flow detection method includes: Acquire system flow data in an industrial control system in real time; the system flow data includes flow data of a kernel layer and flow data of a user layer; Identify and parse the system traffic data based on deep data packet inspection technology to obtain protocol data in the system traffic data; Converting the protocol data and obtaining a sequence to be tested based on the converted protocol data; Inputting the sequence to be tested into a pre-trained abnormal traffic recognition model; According to the output result of the abnormal flow identification model, it is determined whether the system flow data is abnormal.
2. The abnormal flow detection method according to claim 1, characterized in that: The abnormal traffic identification model is an abnormal traffic detection model; Determining whether the system flow data is abnormal according to the output result of the abnormal flow identification model includes: Get the predicted value of the sequence to be tested; Whether the system flow data is abnormal is determined based on the detection value output by the abnormal flow detection model and the predicted value.
3. The abnormal flow detection method according to claim 2, characterized in that: The detection value includes the number of abnormal flow data, and the determination of whether the system flow data is abnormal based on the detection value output by the abnormal flow detection model and the predicted value includes: In response to determining that the number of abnormal flow data in the detection value is greater than the number of abnormal flow data in the predicted value, it is determined that the system flow data is abnormal.
4. The abnormal flow detection method according to claim 1, characterized in that: The real-time acquisition of system flow data in the industrial control system includes: Obtain kernel layer traffic data in industrial control systems in real time through the data plane development kit; The user layer traffic data in the industrial control system is obtained in real time through XDP based on XML data encapsulation.
5. The abnormal flow detection method according to claim 1, characterized in that: Converting the protocol data and obtaining a sequence to be tested based on the converted protocol data includes: Encode all sub-data fields in the protocol data to obtain status data; The state data within the preset time length is sorted according to time to obtain the sequence to be tested.
6. The abnormal flow detection method according to claim 5, characterized in that: The encoding of all sub-data fields in the protocol data comprises: Determining the number of codes of the sub-data field according to the number of changed values of the sub-data field in the protocol data; All sub-data fields in the protocol data are encoded according to the number of the codes of the sub-data fields.
7. The abnormal flow detection method according to claim 1, characterized in that: The protocol data includes application information; the identification and parsing of the system traffic data based on the deep data packet inspection technology to obtain the protocol data in the system traffic data includes: Obtaining a corresponding protocol data packet in the system flow data by identifying a port number and a data packet structure in the system flow data; Extract application information from the corresponding protocol data packet.
8. The abnormal flow detection method according to claim 7, characterized in that: The protocol data packet includes a Modbus protocol data packet; the Modbus protocol data packet includes one of additional data and additional parameters, a device address, a function code, a data address and a data quantity.
9. The abnormal flow detection method according to claim 1, characterized in that: The abnormal traffic identification model is a traffic threshold prediction model; Determining whether the system flow data is abnormal based on the output result of the abnormal flow identification model includes: determining that the system flow data is abnormal in response to determining that the flow threshold output by the flow threshold prediction model exceeds the threshold of a normal baseline.
10. An abnormal flow detection device, characterized in that: Applied to industrial control systems, the device comprises: An acquisition module is used to acquire system flow data in the industrial control system in real time; the system flow data includes flow data of the kernel layer and flow data of the user layer; A parsing module, used to identify and parse the system flow data based on deep data packet inspection technology to obtain protocol data in the system flow data; A processing module, used for converting the protocol data and obtaining a sequence to be tested based on the converted protocol data; An input module, used for inputting the sequence to be tested into a pre-trained abnormal traffic recognition model; The judgment module is used to determine whether the system flow data is abnormal according to the output result of the abnormal flow recognition model.
11. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method according to any one of claims 1 to 9 when executing the program.
12. A non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute the method according to any one of claims 1 to 9.
13. A computer program product, comprising computer program instructions, which, when executed on a computer, cause the computer to execute the method according to any one of claims 1 to 9.
Citation Information
Patent Citations
Equipment anomaly detection method and system based on deep learning and computer medium
CN112629905A
Method and device for detecting abnormal flow of intelligent manufacturing system
CN113285916A
Abnormality detection method and detection device thereof
CN114650186A
Anomaly detection method and cloud network platform
CN115514620A
Abnormal virtual dialing equipment detection method, device and equipment
CN116193021A
Cited By
Virus monitoring and early warning method based on deep analysis of network traffic
CN119583215A
Virus monitoring and early warning method based on deep analysis of network traffic
CN119583215B
Automatic defect analysis and protocol statistics method for ICS traffic database
CN122137762A
