A multi-agent privacy-preserving control system against DOS attacks
Through the multi-agent privacy protection control system, combined with machine learning and rule engine technology, real-time monitoring and analysis of network traffic is achieved, and the problem of insufficient flexibility and privacy protection of existing systems in the face of DOS attacks is solved, and the problem of existing systems is improved. The ability to identify and defend new attack patterns and ensure the continuity and privacy of network services.
Patent Information
- Application Number
- CN202411136039.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-19
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-08-19
AI Technical Summary
Existing network security systems lack flexibility and adaptability in the face of DOS attacks, making it difficult to identify and respond to new attack modes, privacy protection measures are not comprehensive enough, and task scheduling and resource allocation are not intelligent enough, making it difficult to achieve effective in large-scale network environments. Protect.
A multi-agent privacy protection control system was designed, including multiple agent units, each unit consisting of data acquisition module, communication interface module, security policy decision unit and privacy protection unit, using machine learning, rule engine and deep learning technology , through the integration of data acquisition, communication interface, security policy decision-making and privacy protection modules, comprehensive monitoring and analysis of network traffic is achieved, and the continuity of network services is ensured through system-level and device-level privacy protection policies.
Improve the system's response speed and adaptability, enhance the ability to identify and defend new attack patterns, achieve comprehensive and systematic privacy protection, be able to dynamically adjust resource allocation and task execution, optimize network resource utilization efficiency, and ensure network Security and privacy protection.
Smart Images

Figure CN118677699B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of network security and privacy protection, and more specifically, to a multi-agent privacy protection control system for counteracting DOS attacks. Background Art
[0002] In the field of network security, distributed denial of service (DDoS) attacks are a common means of network attack. They use a large number of requests or malicious traffic to exhaust the resources of the target network, making it impossible for normal users to access services. Existing network security systems often have some limitations and shortcomings when facing such attacks. On the one hand, traditional security systems may lack sufficient flexibility and adaptability, making it difficult to cope with rapidly changing network environments and attack methods; on the other hand, privacy protection measures are often not comprehensive enough, making it difficult to effectively resist external attacks while protecting user privacy.
[0003] In the process of implementing the embodiments of the present invention, the inventors found that there are at least the following problems or defects in the prior art: First, existing systems often rely on fixed rules and thresholds when processing network traffic data, which limits their ability to identify and respond to new attack patterns; second, privacy protection measures are usually relatively isolated and lack system-level integration and coordination, which makes it difficult to achieve effective privacy protection in large-scale network environments; finally, existing systems may not be smart enough in task scheduling and resource allocation, and cannot be dynamically adjusted according to real-time network status and security requirements. Summary of the invention
[0004] The present invention provides a multi-agent privacy protection control system for resisting DOS attacks, comprising:
[0005] N1 intelligent units, each of which includes a data acquisition module, a communication interface module, a security policy decision unit, a privacy protection unit, and a task priority scheduling module;
[0006] Said N1≥2;
[0007] The data acquisition module is used to collect network traffic data, and after preprocessing, transmit the preprocessing results to the security policy decision unit;
[0008] The communication interface module is used to receive operation control instructions from other intelligent units, and transmit the privacy protection instructions therein to the privacy protection unit, and transmit other operation control instructions to the task priority scheduling module;
[0009] The security policy decision unit is used to compare the preprocessing result with the preset threshold value to determine whether the network operation state is safe. If it is not safe, the security policy is triggered; otherwise, it is not triggered, and the threshold comparison result is transmitted to the privacy protection unit in real time; at the same time, according to the above-mentioned privacy protection instruction, the privacy protection operation is performed;
[0010] The privacy protection unit is used to issue a system-level privacy protection instruction and a device control-level instruction according to the threshold comparison result and the other operation control instructions, and transmit the system-level privacy protection instruction and the device control-level instruction to the task priority scheduling module;
[0011] The task priority scheduling module is used to judge the priority level of the system-level privacy protection instructions and the device control-level instructions, and send a scheduling signal to schedule and control the network tasks according to the judgment result.
[0012] Further, the data acquisition module includes a flow analyzer, a data cleaner, and a feature extraction module connected in sequence;
[0013] The traffic analyzer is used to collect network traffic data, analyze the network traffic data, and transmit the obtained analysis results to the data cleaner;
[0014] The data cleaner is used to clean the analysis results, remove invalid or redundant data, and transmit the cleaned data to the feature extraction module;
[0015] The feature extraction module is used to extract features from the cleaned data and transmit the obtained features as preprocessing results to the security policy decision unit.
[0016] Further, the security policy decision unit includes N2 policy units, each of which includes an input module, a communication module, a logic judgment module and an output module;
[0017] The input module is used to receive the preprocessing result output by the data acquisition module and transmit it to the logic judgment module;
[0018] The communication module is used for data transmission between different policy units and between the security policy decision unit and the privacy protection unit;
[0019] The logic judgment module is used to compare the preprocessing result output by the data acquisition module with the preset threshold value, judge whether the network operation status is safe according to the obtained threshold comparison result, and transmit the threshold comparison result to the privacy protection unit through the communication module, and transmit the judgment result to the output module;
[0020] The output module is used to trigger the security policy and issue a corresponding security warning instruction when the judgment result is unsafe.
[0021] Further, the privacy protection unit includes N3 protection units, each protection unit includes 1 system-level privacy protection group and N4 device-level privacy protection groups; the system-level privacy protection group and the device-level privacy protection group include an input module, a communication module, a logic module and an output module respectively;
[0022] The system-level privacy protection group is used to receive the privacy protection instruction from the third-party system through the input module, issue the system-level privacy protection instruction, and receive the threshold comparison result output by the security policy decision unit in the intelligent unit where it is located, and the threshold comparison result output by the security policy decision unit in other intelligent units through the corresponding communication module, and comprehensively analyze the threshold comparison results of the above-mentioned intelligent units through the logic module, and issue the system-level privacy protection instruction according to the analysis result, and transmit the system-level privacy protection instruction to the task priority scheduling module through the output module;
[0023] The device-level privacy protection group is used to collect privacy protection requirements and status feedback instructions from a third-party system through an input module, receive user operation control instructions issued by a communication interface module through a corresponding communication module, convert the user operation control instructions, privacy protection requirements and status feedback instructions into device control-level instructions through a logic module, and then transmit the device control-level instructions to a task priority scheduling module through a communication module.
[0024] Furthermore, the task priority scheduling module includes a priority evaluation module, a communication module and a task scheduling control module;
[0025] The priority evaluation module is used to judge the priority of the system-level privacy protection instructions, device control-level instructions, diverse privacy protection instructions from third-party systems or hard-operated switch devices, and test and maintenance instructions according to preset rules, and issue control instructions to the task scheduling control module according to the judgment results;
[0026] The communication module is used to receive device control level instructions, diverse privacy protection instructions from third-party systems or hard-operated switch devices, and test and maintenance instructions, and transmit them to the priority assessment module;
[0027] The task scheduling control module is used to control the network task to be scheduled accordingly according to the judgment result.
[0028] Furthermore, it also includes a security testing unit; the security testing unit is connected to the security policy decision unit and the privacy protection unit;
[0029] The security testing unit is used to periodically issue test requests to the security policy decision unit or the privacy protection unit, and control the isolation of the security policy decision unit or the privacy protection unit from the security policy decision unit or the privacy protection unit in other intelligent units based on the response to the periodic test request returned by the security policy decision unit or the privacy protection unit, and further trigger the judgment logic of the security policy decision unit and the system-level privacy protection logic of the privacy protection unit, and judge whether the operating status of the security policy decision unit or the privacy protection unit is normal by reading back the output signal of the security policy decision unit or the privacy protection unit.
[0030] Furthermore, N1=4, the composition and structure of each intelligent unit are the same, wherein two intelligent units adopt intelligent analysis technology based on machine learning, and the other two intelligent units adopt rapid response technology based on rule engine;
[0031] Said N2≥1, N3≥1, N4≥1, said comprehensive analysis adopts at least one of weighted voting, fuzzy logic or neural network.
[0032] Furthermore, in the data acquisition module, the traffic analyzer adopts a high-precision traffic analysis algorithm, the data cleaner adopts an adaptive data cleaning technology, and the feature extraction module adopts a deep learning feature extraction technology.
[0033] Furthermore, in the security policy decision unit, the input module adopts a high-reliability input interface, the communication module adopts a secure encrypted communication protocol, the logic judgment module adopts an advanced logic judgment algorithm, and the output module adopts a real-time response output technology.
[0034] Further, the processors of the system-level privacy protection group adopt a parallel redundant configuration, and the processors of the device-level privacy protection group adopt a hot standby redundant configuration;
[0035] The transmission interface of the task priority scheduling module includes a hard-wired interface and a secure encrypted communication interface.
[0036] According to the above embodiments of the present invention, at least the following beneficial effects are achieved: the system of the present invention can realize comprehensive monitoring and analysis of network traffic by integrating multiple modules such as data acquisition, communication interface, security policy decision, privacy protection and task priority scheduling. This integrated design can not only improve the response speed of the system, but also enhance the adaptability to complex network environments. Secondly, the advanced technologies such as machine learning, rule engine and deep learning adopted by the system make the formulation and execution of security policies more intelligent and automated, which can effectively improve the recognition and defense capabilities of new attack patterns. In addition, the present invention also emphasizes the importance of privacy protection. By setting a privacy protection unit in each intelligent unit and adopting a privacy protection strategy combining system level and device level, the system can ensure the continuity and stability of network services while protecting user privacy. The comprehensiveness and systematicness of this privacy protection measure are lacking in the prior art. Finally, the modular design of the system and the task priority scheduling mechanism make the scheduling of network tasks more flexible and efficient, and can dynamically adjust resource allocation and task execution according to real-time network status and security requirements, thereby optimizing the utilization efficiency of network resources while ensuring network security. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] The above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the detailed description below with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, in which:
[0038] Figure 1 A schematic diagram of the structure of a multi-agent privacy protection control system for countering DOS attacks provided by one embodiment of the present invention. DETAILED DESCRIPTION
[0039] The principles and spirit of the present invention will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and implement the present invention, and are not intended to limit the scope of the present invention in any way. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to fully convey the scope of the present invention to those skilled in the art.
[0040] Those skilled in the art know that the embodiments of the present invention can be implemented as a system, device, apparatus, method or computer program product. Therefore, the present invention can be specifically implemented in the following forms, namely: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.
[0041] It should be noted that any number of elements in the drawings is for illustration rather than limitation, and any naming is only for distinction and does not have any limiting meaning.
[0042] Reference below Figure 1 , Figure 1 This is a schematic diagram of the structure of a multi-agent privacy protection control system for resisting DOS attacks provided by an embodiment of the present invention. Figure 1 As shown, a multi-agent privacy protection control system 100 for resisting DOS attacks includes:
[0043] N1 intelligent units, each intelligent unit includes a data acquisition module, a communication interface module, a security policy decision unit, a privacy protection unit, and a task priority scheduling module; N1≥2;
[0044] The data collection module 201 is used to collect network traffic data, and after preprocessing, transmit the preprocessing results to the security policy decision unit;
[0045] The communication interface module 202 is used to receive operation control instructions from other intelligent units, and transmit the privacy protection instructions therein to the privacy protection unit, and transmit other operation control instructions to the task priority scheduling module;
[0046] The security policy decision unit 203 is used to compare the preprocessing result with the preset threshold value to determine whether the network operation state is safe. If it is not safe, the security policy is triggered; otherwise, it is not triggered, and the threshold comparison result is transmitted to the privacy protection unit in real time; at the same time, according to the above-mentioned privacy protection instruction, the privacy protection operation is performed;
[0047] The privacy protection unit 204 is used to issue a system-level privacy protection instruction and a device control-level instruction according to the threshold comparison result and the other operation control instructions, and transmit the system-level privacy protection instruction and the device control-level instruction to the task priority scheduling module;
[0048] The task priority scheduling module 205 is used to determine the priority of the system-level privacy protection instructions and the device control-level instructions, and send a scheduling signal to schedule and control the network tasks according to the determination result.
[0049] It should be noted that the multi-agent privacy protection control system is designed with N1 agent units, where N1 is an integer greater than or equal to 2, which means that the system has multiple agent units to achieve distributed processing and control. Each agent unit is independent and can autonomously perform data collection, communication, security policy decision-making, privacy protection, and task scheduling. This design increases the flexibility and scalability of the system, allowing the system to adapt to network environments of different scales and requirements.
[0050] Specifically, the modules contained in each intelligent unit have clear functional divisions. The data acquisition module is responsible for collecting network traffic data and performing necessary preprocessing; the communication interface module is responsible for exchanging information with other intelligent units or external systems; the security policy decision unit determines the network status based on the preprocessed data and the preset security threshold, and makes corresponding security policy decisions; the privacy protection unit performs privacy protection operations according to security policies and privacy requirements; the task priority scheduling module prioritizes and schedules network tasks according to the current security status and task requirements. The collaborative work of these modules ensures the efficient operation and privacy protection of the system.
[0051] Preferably, the data acquisition module can adopt a high-precision traffic analysis algorithm, such as a traffic identification technology based on deep learning, to improve the accuracy and efficiency of data collection. The communication interface module can adopt a secure encrypted communication protocol, such as TLS or SSL, to ensure the security of data transmission. The logic judgment module of the security policy decision unit can adopt advanced algorithms, such as an artificial intelligence-based risk assessment algorithm, to achieve more accurate security judgments. The privacy protection unit can customize the system-level and device-level privacy protection strategies according to the specific needs of different intelligent units. The task priority scheduling module can adopt a dynamic adjustment mechanism to flexibly adjust the priority of the task according to the real-time network status and security requirements to optimize the use of network resources.
[0052] In some embodiments, the data acquisition module includes a flow analyzer, a data cleaner, and a feature extraction module connected in sequence;
[0053] A traffic analyzer, used to collect network traffic data, analyze the network traffic data, and transmit the obtained analysis results to a data cleaner;
[0054] A data cleaner is used to clean the analysis results, remove invalid or redundant data, and transmit the cleaned data to a feature extraction module;
[0055] The feature extraction module is used to extract features from the cleaned data and transmit the obtained features as preprocessing results to the security policy decision unit.
[0056] It should be noted that the data acquisition module is composed of a traffic analyzer, a data cleaner, and a feature extraction module connected in sequence. This structured design allows the system to achieve more sophisticated and efficient processing during the data acquisition process. The traffic analyzer is responsible for preliminary analysis of network traffic, the data cleaner removes invalid or redundant data, and the feature extraction module further extracts key information to provide necessary input for the security policy decision unit.
[0057] Specifically, traffic analyzers can use statistical learning or machine learning-based methods to identify traffic patterns and abnormal behaviors. Data cleaners may use techniques such as rule matching, outlier removal, or data smoothing to purify data. Feature extraction modules may use deep learning techniques such as convolutional neural networks (CNNs) or recurrent neural networks (RNNs) to extract features from cleaned data that are helpful for security policy decisions.
[0058] Preferably, the traffic analyzer can be configured with specific algorithm parameters, such as time window size, traffic threshold, etc., to adapt to different network environments and traffic characteristics. The data cleaner can set adaptive parameters to dynamically adjust the cleaning strategy according to the changes in real-time traffic. The feature extraction module can use a pre-trained deep learning model to improve the accuracy and efficiency of feature recognition. In addition, the feature extraction module can be designed to support multiple feature extraction techniques to adapt to different types of network traffic data and provide a more flexible feature extraction solution.
[0059] In some embodiments, the security policy decision unit includes N2 policy units, each of which includes an input module, a communication module, a logic judgment module and an output module;
[0060] The input module is used to receive the preprocessing result output by the data acquisition module and transmit it to the logic judgment module;
[0061] The communication module is used for data transmission between different policy units and between the security policy decision unit and the privacy protection unit;
[0062] The logic judgment module is used to compare the preprocessing result output by the data acquisition module with the preset threshold value, judge whether the network operation status is safe according to the obtained threshold comparison result, and transmit the threshold comparison result to the privacy protection unit through the communication module, and transmit the judgment result to the output module;
[0063] The output module is used to trigger the security policy and issue a corresponding security warning instruction when the judgment result is unsafe.
[0064] It should be noted that the security policy decision unit is composed of N2 policy units, each of which contains an input module, a communication module, a logic judgment module, and an output module. This design allows the system to flexibly adjust and apply different security policies when facing different security scenarios. The setting of N2 ensures that the system has enough policy units to cope with diverse security needs.
[0065] Specifically, the input module is responsible for receiving the preprocessing results from the data acquisition module and transmitting these data to the logic judgment module for further analysis. The communication module is responsible for transmitting data between different policy units and between the privacy protection unit to ensure real-time sharing and synchronization of information. The logic judgment module compares the preprocessing results with the preset threshold to determine whether the network status is safe and makes decisions accordingly. When the output module detects an unsafe state, it triggers the corresponding security policy and issues a security warning instruction.
[0066] Preferably, the input module can adopt efficient data receiving technology to ensure fast and accurate data transmission. The communication module can adopt advanced encryption technology, such as end-to-end encryption, to ensure the security of data transmission. The logic judgment module can adopt complex algorithms, such as support vector machine (SVM) or deep neural network (DNN), to improve the accuracy of judgment. The output module can be designed to have a fast response mechanism so that it can take immediate action when a threat is detected. In addition, the policy unit can be expanded or reduced according to actual needs to adapt to network security needs of different scales.
[0067] Preferably, the threshold comparison is as shown in calculation formula (1);
[0068] (1)
[0069] In the formula, is the trigger value, used to determine whether to trigger the security policy; A quantitative indicator of the preprocessing result, such as a traffic anomaly score; is the preset safety threshold.
[0070] Formula (1) is used to calculate quantitative indicators With safety threshold If Greater than , If it is a positive value, the security policy is triggered.
[0071] In some embodiments, the privacy protection unit includes N3 protection units, each protection unit includes 1 system-level privacy protection group and N4 device-level privacy protection groups; the system-level privacy protection group and the device-level privacy protection group include an input module, a communication module, a logic module and an output module respectively;
[0072] The system-level privacy protection group is used to receive the privacy protection instruction from the third-party system through the input module, issue the system-level privacy protection instruction, and receive the threshold comparison result output by the security policy decision unit in the intelligent unit where it is located, and the threshold comparison result output by the security policy decision unit in other intelligent units through the corresponding communication module, and comprehensively analyze the threshold comparison results of the above-mentioned intelligent units through the logic module, and issue the system-level privacy protection instruction according to the analysis result, and transmit the system-level privacy protection instruction to the task priority scheduling module through the output module;
[0073] The device-level privacy protection group is used to collect privacy protection requirements and status feedback instructions from a third-party system through an input module, receive user operation control instructions issued by a communication interface module through a corresponding communication module, convert the user operation control instructions, privacy protection requirements and status feedback instructions into device control-level instructions through a logic module, and then transmit the device control-level instructions to a task priority scheduling module through a communication module.
[0074] It should be noted that the privacy protection unit consists of N3 protection units, each of which is further subdivided into system-level privacy protection groups and N4 device-level privacy protection groups. This layered design allows the system to perform privacy protection measures at different levels, ensuring the comprehensiveness and pertinence of privacy protection. The settings of N3 and N4 provide flexibility to adapt to privacy protection tasks of different scales and requirements.
[0075] Specifically, the system-level privacy protection group receives privacy protection instructions from third-party systems through the input module and issues corresponding system-level privacy protection instructions. It receives the threshold comparison results from the security policy decision unit in the intelligent unit through the communication module and conducts a comprehensive analysis of these results through the logic module. The device-level privacy protection group collects privacy protection requirements and status feedback instructions through the input module, and converts user operation control instructions into device control-level instructions through the logic module. Each group includes an input module, a communication module, a logic module, and an output module to realize its functions.
[0076] Preferably, the input module of the system-level privacy protection group can be designed to support multiple data formats and protocols to be compatible with different third-party systems. The communication module can adopt efficient data transmission technology, such as message queues or real-time data streams, to ensure the rapid transmission of information. The logic module can adopt complex data analysis algorithms, such as fuzzy logic or neural networks, to improve the efficiency and accuracy of the generation of privacy protection instructions. The output module can be designed to be highly customizable to adapt to different privacy protection strategies and requirements. The device-level privacy protection group can adopt similar technologies, but focus more on device-specific privacy protection requirements and controls. In addition, the system-level and device-level privacy protection groups can be designed to support modular expansion so that they can be quickly updated and upgraded according to new privacy protection requirements.
[0077] In some embodiments, it also includes calculating privacy protection weights. The privacy protection unit consists of N3 protection units, each of which is further subdivided into system-level privacy protection groups and N4 device-level privacy protection groups. These groups are responsible for receiving, processing, and responding to privacy protection requirements and instructions from different sources. In order to effectively manage and execute these privacy protection tasks, a quantitative method is needed to evaluate and balance the importance and urgency of different privacy protection requirements.
[0078] By assigning weights and importance scores to each privacy factor, a comprehensive privacy protection weight can be calculated. This weight reflects the importance of different privacy factors relative to the overall privacy protection requirements in the current situation, as shown in formula (2);
[0079] (2)
[0080] In the formula, is the comprehensive privacy protection weight; is the weight of the i-th privacy factor;
[0081] Score the importance of the i-th privacy factor; n is the total number of privacy factors.
[0082] Comprehensive privacy protection weight It can guide the privacy protection unit on how to allocate resources and priorities. For example, if the importance score of a privacy factor is Higher and given greater weight , then in the comprehensive weight It will account for a larger proportion and will be given priority in the formulation and implementation of privacy protection measures.
[0083] In some embodiments, the task priority scheduling module includes a priority evaluation module, a communication module, and a task scheduling control module;
[0084] The priority evaluation module is used to judge the priority of the system-level privacy protection instructions, device control-level instructions, diverse privacy protection instructions from third-party systems or hard-operated switch devices, and test and maintenance instructions according to preset rules, and issue control instructions to the task scheduling control module according to the judgment results;
[0085] The communication module is used to receive device control level instructions, diverse privacy protection instructions from third-party systems or hard-operated switch devices, and test and maintenance instructions, and transmit them to the priority assessment module;
[0086] The task scheduling control module is used to control the network task to be scheduled accordingly according to the judgment result.
[0087] It should be noted that the task priority scheduling module consists of a priority evaluation module, a communication module, and a task scheduling control module. This modular design makes the task scheduling process more orderly and efficient, and can reasonably allocate and control network tasks according to different priorities. In this way, the system can ensure that critical tasks are executed first while taking into account the needs of privacy protection and other operations.
[0088] Specifically, the priority assessment module is responsible for prioritizing the received instructions according to preset rules. These instructions may include system-level privacy protection instructions, device control-level instructions, third-party system instructions, and test and maintenance instructions. The communication module, as a bridge for information transmission, is responsible for receiving and forwarding these instructions to the priority assessment module. The task scheduling control module schedules network tasks based on the evaluation results to ensure that tasks are executed in the established priority order.
[0089] Preferably, the priority evaluation module can be designed to support dynamic adjustment rules so that the priority can be flexibly adjusted according to the real-time network status and security requirements. The communication module can adopt a variety of communication protocols to adapt to different network environments and communication requirements. The task scheduling control module can adopt advanced scheduling algorithms, such as shortest job first (SJF) or round-robin scheduling (RR), to optimize the efficiency and fairness of task execution. In addition, the task priority scheduling module can be designed to have the ability of self-learning and optimization, and continuously improve the task scheduling strategy through machine learning technology to adapt to the changing network conditions.
[0090] In some embodiments, a security testing unit is further included; the security testing unit is connected to the security policy decision unit and the privacy protection unit;
[0091] The security testing unit is used to periodically issue test requests to the security policy decision unit or the privacy protection unit, and control the isolation of the security policy decision unit or the privacy protection unit from the security policy decision unit or the privacy protection unit in other intelligent units based on the response to the periodic test request returned by the security policy decision unit or the privacy protection unit, and further trigger the judgment logic of the security policy decision unit and the system-level privacy protection logic of the privacy protection unit, and judge whether the operating status of the security policy decision unit or the privacy protection unit is normal by reading back the output signal of the security policy decision unit or the privacy protection unit.
[0092] It should be noted that the security test unit is connected to the security policy decision unit and the privacy protection unit, and regular security tests are performed to verify and ensure the normal operation of other units. The introduction of the security test unit is an important supplement to the existing multi-agent privacy protection control system, which improves the system's self-detection and self-repair capabilities through an active testing mechanism.
[0093] Specifically, the security test unit is responsible for issuing test requests, which can be periodic or executed according to specific trigger conditions. The test requests will be sent to the security policy decision unit or the privacy protection unit to verify whether they can correctly respond and execute the expected security policy or privacy protection measures. The test results will be used to evaluate the operating status of these units. If an abnormality is found, the security test unit can take measures to isolate the faulty unit and trigger the corresponding judgment logic and protection logic to ensure the stable operation of the system.
[0094] Preferably, the security testing unit can be designed to have the ability to generate and execute automated test plans to reduce manual intervention and improve testing efficiency. It can set a variety of test scenarios and parameters to cover different security threats and privacy leakage risks. In addition, the security testing unit can use artificial intelligence technologies, such as natural language processing (NLP) or pattern recognition, to analyze test results and provide suggestions for improvement. More specifically, the security testing unit can be integrated with other parts of the system to achieve real-time monitoring and instant feedback so that potential problems can be responded to quickly when they are discovered. Alternatives may include using third-party security testing services or developing dedicated hardware testing equipment to provide more diverse testing methods and higher testing accuracy.
[0095] In some embodiments, N1=4, the composition and structure of each intelligent unit are the same, wherein two intelligent units adopt intelligent analysis technology based on machine learning, and the other two intelligent units adopt rapid response technology based on rule engine;
[0096] Said N2≥1, N3≥1, N4≥1, said comprehensive analysis adopts at least one of weighted voting, fuzzy logic or neural network.
[0097] It should be noted that N1 is equal to 4, and the composition and structure of each intelligent unit are the same, among which half of the intelligent units use intelligent analysis technology based on machine learning, and the other half use rapid response technology based on rule engine. This design aims to combine the advantages of the two technologies to achieve more comprehensive security and privacy protection.
[0098] Specifically, intelligent analysis technologies based on machine learning can include supervised learning, unsupervised learning, or reinforcement learning methods, which enable intelligent units to learn from historical data and identify potential security threats. Rapid response technologies based on rule engines rely on predefined rule sets and can take quick action when events that meet the rules are detected. The parameters of N2, N3, and N4 are set to at least 1, which means that each policy unit, protection unit, and device-level privacy protection group contains at least one instance to ensure the basic functions of the system.
[0099] Preferably, the machine learning models in the intelligent unit can be updated regularly to adapt to new security threats and changes. The rule engine can be designed to be configurable, allowing security experts to update the rule set based on the latest security intelligence. The comprehensive analysis can adopt a weighted voting mechanism, in which the output of each intelligent unit is given different weights according to its importance or accuracy. More specifically, an adaptive mechanism can be designed to dynamically adjust the weights based on real-time feedback to optimize the decision-making process. Alternatives may include using fuzzy logic or neural networks to enhance the ability of comprehensive analysis to handle security scenarios with higher uncertainty and complexity.
[0100] In some embodiments, in the data acquisition module, the traffic analyzer adopts a high-precision traffic analysis algorithm, the data cleaner adopts an adaptive data cleaning technology, and the feature extraction module adopts a deep learning feature extraction technology.
[0101] It should be noted that the traffic analyzer, data cleaner and feature extraction module are key components of the data acquisition module, which use high-precision traffic analysis algorithms, adaptive data cleaning technology and deep learning feature extraction technology respectively. These technologies are selected to improve the accuracy and efficiency of data collection, thereby providing high-quality input data for subsequent security policy decisions.
[0102] Specifically, traffic analyzers may use statistical analysis or machine learning-based methods to identify and quantify network traffic patterns. Data cleaners may use automated rules or algorithms to identify and remove invalid or redundant data, such as rules based on data consistency and integrity. Feature extraction modules may use deep learning techniques, such as convolutional neural networks (CNNs) or recurrent neural networks (RNNs), to extract key features from cleaned data. These features are then used to formulate security policies and implement privacy protection.
[0103] Preferably, the traffic analyzer can be configured with specific parameters, such as sampling rate, analysis window size, etc., to adapt to different network traffic characteristics. The data cleaner can be designed to have self-learning capabilities and be able to adjust its cleaning strategy based on real-time feedback from the data. The feature extraction module can adopt a pre-trained deep learning model to improve the speed and accuracy of feature extraction. More specifically, a modular architecture can be designed for the feature extraction module, allowing flexible selection or replacement of models based on different data types and security requirements. Alternatives may include using other types of machine learning algorithms, such as support vector machines (SVM) or random forests, to extract features to adapt to different application scenarios and performance requirements.
[0104] In some embodiments, in the security policy decision unit, the input module adopts a high-reliability input interface, the communication module adopts a secure encrypted communication protocol, the logic judgment module adopts an advanced logic judgment algorithm, and the output module adopts real-time response output technology.
[0105] It should be noted that the modules in the security policy decision unit include input module, communication module, logic judgment module and output module, which work together to ensure the efficient and accurate execution of security policies. By adopting high reliability, security and real-time technology, the unit can quickly and accurately analyze and respond to network traffic data.
[0106] Specifically, the input module may use high-reliability interface technologies, such as fiber-optic communication or wireless communication, to ensure stable data transmission. The communication module may use secure encrypted communication protocols, such as TLS or IPSec, to protect the security of data during transmission. The logic judgment module may use advanced algorithms, such as decision trees or Bayesian networks, to analyze the comparison results between preprocessing results and preset thresholds. The output module may use real-time response technologies, such as fast hardware interrupts or event-driven programming, to ensure that security policies can be quickly triggered when security threats are detected.
[0107] Preferably, the input module can be designed to have error detection and correction functions to improve the reliability of data transmission. The communication module can integrate multiple encryption technologies to adapt to different security needs and compliance requirements. The logic judgment module can adopt an adaptive algorithm to dynamically adjust its judgment logic based on historical data and real-time feedback. The output module can be designed to have a priority scheduling function to ensure that when multiple security warnings occur at the same time, the corresponding security policies can be executed in a predetermined priority order. More specifically, a machine learning framework can be implemented for the logic judgment module so that it can continuously learn and optimize its judgment logic to cope with ever-changing network threats. Alternatives may include using different encryption protocols or algorithms, or using hardware accelerators to increase the processing speed of the logic judgment and output modules.
[0108] In some embodiments, the processors of the system-level privacy protection group adopt a parallel redundant configuration, and the processors of the device-level privacy protection group adopt a hot standby redundant configuration;
[0109] The transmission interface of the task priority scheduling module includes a hard-wired interface and a secure encrypted communication interface.
[0110] It should be noted that the processors of the system-level privacy protection group adopt parallel redundant configuration, which means that there are multiple processors working at the same time to improve processing power and fault tolerance. The processors of the device-level privacy protection group adopt hot standby redundant configuration, that is, when the main processor fails, the backup processor can immediately take over the work to ensure the continuity of service. The transmission interface of the task priority scheduling module includes a hard-wired interface and a secure encrypted communication interface. The hard-wired interface provides a stable and reliable physical connection, while the secure encrypted communication interface ensures the security of data in wireless or network transmission.
[0111] Preferably, the processors of the system-level and device-level privacy protection groups can adopt load balancing technology to optimize resource usage and reduce the risk of single-point overload. The redundant configuration of the processor can be designed as an automatic fault detection and recovery mechanism to achieve seamless switching. The hard-wired interface of the task priority scheduling module can adopt a high-fidelity, low-latency design to adapt to application scenarios with high real-time requirements. The secure encrypted communication interface can integrate the latest encryption standards, such as quantum encryption, to combat potential future computing attacks. More specifically, an intelligent scheduling algorithm can be implemented for the task priority scheduling module, which can dynamically adjust the transmission strategy according to the network conditions and task characteristics to optimize the overall network performance. Alternatives may include using different redundancy strategies, such as dynamic redundancy or multi-level redundancy, and adopting multiple communication technologies, such as satellite communications or 5G networks, to provide a wider range of connection options.
[0112] The above-mentioned embodiments of the present invention have the following beneficial effects: the system of the present invention can realize real-time monitoring and analysis of network traffic and rapid response to security threats through modular and intelligent design. This design can not only improve the adaptability and flexibility of the system, but also improve the recognition and defense capabilities of new attack patterns by integrating multiple advanced technologies such as machine learning, rule engines and deep learning. Secondly, the privacy protection measures of the system are comprehensive and systematic, which can ensure the continuity and stability of network services while protecting user privacy. In addition, the multi-agent privacy protection control system of the present invention can enhance the self-detection and self-repair capabilities of the system by introducing a security test unit, thereby further improving the reliability and security of the system. The redundant configuration of the system-level and device-level privacy protection groups, as well as the diversified transmission interfaces of the task priority scheduling module, can ensure that the system can continue to operate stably in the face of high load or single point failure. These measures work together to provide users with a safe and reliable network environment, effectively resist network threats such as distributed denial of service (DOS) attacks, and protect user data privacy and network security.
[0113] The above descriptions are only some preferred embodiments of the present invention and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present invention is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the embodiments of the present invention.
Claims
1. A multi-agent privacy protection control system against DOS attacks, It is characterized in that It includes N1 intelligent units, each of which includes a data acquisition module, a communication interface module, a security policy decision unit, a privacy protection unit, and a task priority scheduling module; N1≥2; The data acquisition module is used to collect network traffic data, and after preprocessing, transmit the preprocessing results to the security policy decision unit; The communication interface module is used to receive operation control instructions from other intelligent units, and transmit the privacy protection instructions therein to the privacy protection unit, and transmit other operation control instructions to the task priority scheduling module; The security policy decision unit is used to compare the preprocessing result with the preset threshold value to determine whether the network operation state is safe. If it is not safe, the security policy is triggered; otherwise, it is not triggered, and the threshold comparison result is transmitted to the privacy protection unit in real time; at the same time, according to the above-mentioned privacy protection instruction, the privacy protection operation is performed; The privacy protection unit is used to issue a system-level privacy protection instruction and a device control-level instruction according to the threshold comparison result and the other operation control instructions, and transmit the system-level privacy protection instruction and the device control-level instruction to the task priority scheduling module; The task priority scheduling module is used to judge the priority of the system-level privacy protection instructions and the device control-level instructions, and send a scheduling signal to schedule and control the network task according to the judgment result. The privacy protection unit includes N3 protection units, each of which includes 1 system-level privacy protection group and N4 device-level privacy protection groups; the system-level privacy protection group and the device-level privacy protection group respectively include an input module, a communication module, a logic module and an output module; The system-level privacy protection group is used to receive the privacy protection instruction from the third-party system through the input module, issue the system-level privacy protection instruction, and receive the threshold comparison result output by the security policy decision unit in the intelligent unit where it is located, and the threshold comparison result output by the security policy decision unit in other intelligent units through the corresponding communication module, and comprehensively analyze the threshold comparison results of the above-mentioned intelligent units through the logic module, and issue the system-level privacy protection instruction according to the analysis result, and transmit the system-level privacy protection instruction to the task priority scheduling module through the output module; The device-level privacy protection group is used to collect privacy protection requirements and status feedback instructions from a third-party system through an input module, receive user operation control instructions issued by a communication interface module through a corresponding communication module, convert the user operation control instructions, privacy protection requirements and status feedback instructions into device control-level instructions through a logic module, and then transmit the device control-level instructions to a task priority scheduling module through a communication module.
2. The multi-agent privacy protection control system according to claim 1, It is characterized in that The data acquisition module includes a flow analyzer, a data cleaner, and a feature extraction module connected in sequence; The traffic analyzer is used to collect network traffic data, analyze the network traffic data, and transmit the obtained analysis results to the data cleaner; The data cleaner is used to clean the analysis results, remove invalid or redundant data, and transmit the cleaned data to the feature extraction module; The feature extraction module is used to extract features from the cleaned data and transmit the obtained features as preprocessing results to the security policy decision unit.
3. The multi-agent privacy protection control system according to claim 2, It is characterized in that The security policy decision unit includes N2 policy units, each of which includes an input module, a communication module, a logic judgment module and an output module; The input module is used to receive the preprocessing result output by the data acquisition module and transmit it to the logic judgment module; The communication module is used for data transmission between different policy units and between the security policy decision unit and the privacy protection unit; The logic judgment module is used to compare the preprocessing result output by the data acquisition module with the preset threshold value, judge whether the network operation status is safe according to the obtained threshold comparison result, and transmit the threshold comparison result to the privacy protection unit through the communication module, and transmit the judgment result to the output module; The output module is used to trigger the security policy and issue a corresponding security warning instruction when the judgment result is unsafe.
4. The multi-agent privacy protection control system according to claim 3, It is characterized in that The task priority scheduling module includes a priority evaluation module, a communication module and a task scheduling control module; The priority evaluation module is used to judge the priority of the system-level privacy protection instructions, device control-level instructions, diverse privacy protection instructions from third-party systems or hard-operated switch devices, and test and maintenance instructions according to preset rules, and issue control instructions to the task scheduling control module according to the judgment results; The communication module is used to receive device control level instructions, diverse privacy protection instructions from third-party systems or hard-operated switch devices, and test and maintenance instructions, and transmit them to the priority assessment module; The task scheduling control module is used to control the network task to be scheduled accordingly according to the judgment result.
5. The multi-agent privacy protection control system according to any one of claims 1 to 4, It is characterized in that It also includes a security testing unit; the security testing unit is connected to the security policy decision unit and the privacy protection unit; The security testing unit is used to periodically issue test requests to the security policy decision unit or the privacy protection unit, and control the isolation of the security policy decision unit or the privacy protection unit from the security policy decision unit or the privacy protection unit in other intelligent units based on the response to the periodic test request returned by the security policy decision unit or the privacy protection unit, and further trigger the judgment logic of the security policy decision unit and the system-level privacy protection logic of the privacy protection unit, and judge whether the operating status of the security policy decision unit or the privacy protection unit is normal by reading back the output signal of the security policy decision unit or the privacy protection unit.
6. The multi-agent privacy protection control system according to claim 3, It is characterized in that The N1=4, each intelligent unit has the same composition and structure, wherein two intelligent units adopt intelligent analysis technology based on machine learning, and the other two intelligent units adopt rapid response technology based on rule engine; Said N2≥1, N3≥1, N4≥1, said comprehensive analysis adopts at least one of weighted voting, fuzzy logic or neural network.
7. The multi-agent privacy protection control system according to claim 2, It is characterized in that In the data acquisition module, the traffic analyzer adopts a high-precision traffic analysis algorithm, the data cleaner adopts an adaptive data cleaning technology, and the feature extraction module adopts a deep learning feature extraction technology.
8. The multi-agent privacy protection control system according to claim 3, It is characterized in that In the security policy decision unit, the input module adopts a high-reliability input interface, the communication module adopts a secure encrypted communication protocol, the logic judgment module adopts an advanced logic judgment algorithm, and the output module adopts a real-time response output technology.
9. The multi-agent privacy protection control system according to claim 1 or 4, It is characterized in that The processors of the system-level privacy protection group adopt a parallel redundant configuration, and the processors of the device-level privacy protection group adopt a hot standby redundant configuration; The transmission interface of the task priority scheduling module includes a hard-wired interface and a secure encrypted communication interface.
