Network attack processing method and device and electronic equipment
Feedback to the attacker through generative model forgery and simulation echo content, solving the problem of continuous evolution of attack methods in traditional defense methods, achieving misleading attackers without much loss, reducing the improvement and evolution of attack methods, and having certain Defensive effect.
Patent Information
- Application Number
- CN202411336839.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-24
- Publication Date
- 2024-12-06
AI Technical Summary
Traditional anti-cyber attack methods are blocked immediately after discovering attack behavior, resulting in continuous evolution of attack methods and difficulty in effective defense.
The content is falsified and echoed by generative models and feedback to the attacker, misleading the attacker to complete the attack task, achieving the effect of defending against cyber attacks.
Without major losses, mislead the attacker to complete the attack task, prevent the attacker from discovering defects in the attack method, reduce the evolution of the attack method, and have a certain degree of defense against cyber attacks.
Smart Images

Figure CN119094231A_ABST
Abstract
Description
Technical Field
[0001] This specification document belongs to the field of computer technology, and in particular, relates to a method, device and electronic device for processing network attacks. Background Art
[0002] A cyber attack is any type of offensive action against a computer network device. For computers and computer networks, destroying, exposing, modifying, disabling services, and accessing data without authorization are all considered cyber attacks.
[0003] The traditional method of countering network attacks is to block the attack immediately after it is discovered. In this confrontation mode, the attack methods are constantly evolving and becoming more and more diverse. Therefore, a more effective method of defending against network attacks is needed. Summary of the invention
[0004] The embodiments of this specification provide a method, device and electronic device for processing network attacks, which can forge simulated echo content based on a generative model and feed it back to the attacker, thereby misleading the attacker to complete the attack task without causing significant losses, so as to achieve the effect of defending against network attacks.
[0005] To solve the above technical problems, the embodiments of this specification are implemented as follows: First, a method for handling network attacks is proposed, including: Perform attack behavior detection on target users’ access requests; In the case where it is detected that the access request contains an attack behavior, generating simulation echo content matching the access request based on the generative model; The simulated echo content is fed back to the target user.
[0006] In the second aspect, a network attack processing device is proposed, comprising: Attack detection module, which detects attack behaviors on target users’ access requests; A content generation module, if it is detected that the access request contains attack behavior, generates simulation echo content matching the access request based on a generative model; The content feedback module feeds back the simulated echo content to the target user.
[0007] In a third aspect, an electronic device is provided, comprising: a processor; and a memory arranged to store computer executable instructions, wherein when the executable instructions are executed, the processor performs the following operations: Perform attack behavior detection on target users’ access requests; In the case where it is detected that the access request contains an attack behavior, generating simulation echo content matching the access request based on the generative model; The simulated echo content is fed back to the target user.
[0008] In a fourth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores one or more programs, and when the one or more programs are executed by an electronic device including a plurality of application programs, the electronic device performs the following operations: Attack detection module, which detects attack behaviors on target users’ access requests; A content generation module, if it is detected that the access request contains attack behavior, generates simulation echo content matching the access request based on a generative model; The content feedback module feeds back the simulated echo content to the target user.
[0009] In the embodiment of this specification, after the target user initiates an access request, the access request is firstly detected for attack behavior. If an attack behavior is detected, a generative model is used to forge simulated echo content to feed back to the target user. Since the simulated echo content is not data obtained by actual access, the target user cannot effectively use the simulated echo content to achieve a substantial attack effect after obtaining it, and may also mistakenly believe that the attack task has been successfully completed and stop evolving, thereby having a certain degree of ability to defend against network attacks. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] The drawings described herein are used to provide a further understanding of this specification and constitute a part of this specification. The illustrative embodiments and descriptions of this specification are used to explain this specification and do not constitute an improper limitation on this specification. In the drawings: Figure 1 The figure is a flowchart of a method for processing a network attack according to an embodiment of the present specification.
[0011] Figure 2 A schematic diagram of a method for processing network attacks according to an embodiment of the present specification when applied to a honeypot system.
[0012] Figure 3 Schematic diagram of the structure of the network attack processing device of the embodiment of this specification.
[0013] Figure 4 Schematic diagram of the structure of an electronic device according to an embodiment of this specification. DETAILED DESCRIPTION
[0014] In order to make the purpose, technical solutions and advantages of this document clearer, the technical solutions of this specification will be clearly and completely described in combination with the specific embodiments of this specification and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of this document, not all of them. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this document.
[0015] As mentioned above, the traditional method of countering network attacks is to block the attack immediately after it is discovered. In this confrontation mode, it promotes the continuous evolution of attack methods and does not necessarily have a good defensive effect.
[0016] To this end, this specification aims to propose a solution for handling network attacks, which can forge simulated echo content based on a generative model and feed it back to the attacker, thereby misleading the attacker to complete the attack task without major losses, so as to achieve a defensive effect. The solution for handling network attacks in this specification can be applied to the honeypot system. The honeypot system essentially arranges some hosts, network services or information as bait to induce the attacker to attack them, so that the attack behavior can be captured and analyzed, the tools and methods used by the attacker can be understood, and the attack intentions and motives can be inferred, so that the defender can clearly understand the security threats faced, and enhance the security protection capabilities of the actual system through technical and management means.
[0017] Figure 1 is a flowchart of a method for processing a network attack according to an embodiment of the present specification. Figure 1 The method shown can be performed by the corresponding device below, including the following steps: S102: Perform attack behavior detection on the target user's access request.
[0018] In the embodiments of this specification, the target user's access request may be, but is not limited to, a system file access request, a command line operation access request, a file upload / download access request, an API call access request, and a network detection / scanning access request, etc. It should be understood that the type of access request depends on the access object and is not specifically limited herein.
[0019] In one implementation, network traffic can be monitored and analyzed to find out access requests of target users suspected of attacking. For example, risk detection is performed on the overall network traffic data to determine the target network traffic data that is risky; then, the access request that generates the target network traffic data and its corresponding target user are determined.
[0020] It should be noted that risk detection of network traffic data belongs to the existing technology, which can be performed through deep learning models based on some characteristics of network traffic data, such as fluctuation patterns, time periods, data sensitivity, user screens, etc.; risk detection can also be performed based on IP addresses recorded in blacklists / graylists. This specification does not specifically limit the risk detection method.
[0021] S104: If it is detected that the access request contains an attack behavior, a simulation echo content matching the access request is generated based on the generative model.
[0022] The simulated echo content of the embodiment of this specification is not real content, and only serves to mislead the target user to complete the attack task, so the more realistic the simulated echo content is, the better. The following introduces how the generative model generates simulated echo content in combination with two implementation methods.
[0023] Implementation method 1 The generative model of the embodiment of this specification generates the simulated echo content in a predictive manner, that is, the generative model predicts the matching simulated echo content according to the access request of the target user.
[0024] The principle of training the prediction ability of the generative model is as follows: prepare attack behavior samples and the system response content corresponding to when the real system receives attack behaviors that match the attack behavior samples; then, use the system response content as the attack behavior sample label, and input the attack behavior sample data into the generative model to try to predict the corresponding simulation echo content; then, calculate the loss between the predicted simulation echo content and the system response content (label) given by the real system, and adjust the parameters of the generative model with the purpose of reducing the loss.
[0025] Through the above training method, the simulation echo content predicted by the generative model can be very close to the system response content of the real system. For the target users, it is difficult to distinguish the authenticity of the simulation echo content.
[0026] Implementation method 2 The generative model of the embodiment of this specification may be a large language model, so as to generate simulated echo content in a natural language understanding manner.
[0027] It should be understood that the simulation echo content generated in this way does not rely on the sample training in the previous implementation. Therefore, it is necessary to ensure that the simulation echo content is at least logical in terms of data format and data content.
[0028] Therefore, after detecting that the target user's access request contains attack behavior, it is necessary to first determine the access type and access intent of the access request; the access type can be, but is not limited to, system file access, web application interactive access, command line operation access, file upload access, file download access, API call access, network detection access, and network scanning access. The access intent corresponds to the access type. For example, if the access type is system file access, the access intent includes which system file is being accessed; for another example, if the access type is command line operation access, the access intent includes the specific command line operation and the purpose of executing the command line operation.
[0029] Afterwards, the access type and access intention corresponding to the access request are input into the generative model as prompt words, so that the generative model can analyze the data format of the access type and the data content of the access intention based on the natural language understanding ability, and generate simulation echo content that is consistent with the data format and data content through the content generation ability.
[0030] Obviously, when the data format and data content are logical, it is difficult for the target user, who is an attacker, to distinguish the authenticity of the simulated echo content.
[0031] In addition, the simulation echo content generated by natural language understanding has a certain degree of divergence, so as to avoid the logical problem that similar access requests are roughly the same but the simulation echo content is too different. To this end, on the basis of determining the access type and access intention of the target user's access request, the historical access records of the target user can be further refined to obtain continuous access information related to the current access intention. For example, the historical access type, historical access intention and historical simulation echo content corresponding to the historical access requests of the target user.
[0032] Then, the continuity access information, access type and access intention are input into the generative model as prompt words to create simulation echo content whose data format conforms to the access type, data content conforms to the access intention and data logic conforms to the continuity access information.
[0033] The data logic of the simulation echo content conforms to the continuous access information, which means that the first correspondence between the simulation echo content and the access type and access intention matches the first correspondence between the historical simulation echo content and the historical access type and historical access intention.
[0034] For example, the access type of the first access request initiated by the target user is API call, and the access intent is to query the information of account A. The generated simulated echo content is forged account A information. Afterwards, the access type and access intent of the second access request initiated by the target user are the same as the first access request. In this way, the account A information of the generated simulated echo content must also be consistent with that of the first access request to prevent the target user from discovering that the simulated echo content is forged.
[0035] That is, after adding continuity access information as a prompt word to the generative model, this embodiment can use the correspondence between the historical access type, historical access intention and historical simulation echo content of the target user's historical access requests as a regularization constraint to avoid the generative model from subsequently generating simulation echo content that is too different for access requests with similar access types and access intentions.
[0036] S106: Feedback the simulated echo content to the target user.
[0037] In summary, based on the method of this embodiment, after the target user initiates an access request, the access request is firstly subjected to attack behavior detection. If an attack behavior is detected, a generative model is used to forge simulated echo content to feed back to the target user. Since the simulated echo content is not data obtained by actual access, the target user cannot effectively use the simulated echo content to achieve a substantial attack effect after obtaining it, and may also mistakenly believe that the attack task has been successfully completed and stop evolving, thereby having a certain degree of ability to defend against network attacks.
[0038] The method of this embodiment is introduced below by taking the scenario of a honeypot system applied in an enterprise as an example.
[0039] In this application scenario, the honeypot system is configured with traffic monitoring and analysis components, generative artificial intelligence components, and response components. The functions of each component are as follows: 1. Traffic Monitoring and Analysis Components The traffic monitoring and analysis component is used to monitor the network traffic data accessed by the system in real time and perform risk detection. When risky target network traffic data is detected, the access request and target user corresponding to the target network traffic data are determined.
[0040] At the same time, the traffic monitoring and analysis component can analyze the access type and access intention based on the target user's access request.
[0041] 2. Generative AI Components The generative artificial intelligence component mainly involves the application of generative models. Taking the large language model as an example, the generative artificial intelligence component inputs the access type and access intent analyzed by the traffic monitoring and analysis component as prompt words into the generative model, and the generative model creates a simulated echo content whose data format conforms to the access type, data content conforms to the access intent, and data logic conforms to the continuous access information.
[0042] In addition, the honeypot system also stores the historical access records of each user. The historical access records include the historical access type, historical access intent, and historical simulation echo content corresponding to the historical access requests. After the traffic monitoring and analysis component analyzes the access type and access intent of this access request, the generative artificial intelligence component queries the target user's historical access records to see whether the target user has previously initiated a historical access request for the same access type and access intent. If the query is successful, the triple correspondence of the hit historical access type-historical access intent-historical simulation echo content is input into the generative model as a supplementary prompt word, so that the simulation echo content generated by the generative model for this access request is consistent with the data logic of the simulation echo content of the historical access request.
[0043] 3. Response Components The response component is used to send the simulated echo content generated by the generative model for the target user's access request to the target user's client device, so as to mislead the target user into successfully launching a network attack. For the target user, a successful network attack means that the task is completed, but it does not achieve a substantial attack effect. In addition, after the task is completed, the target user does not know what defects there are in his attack method, and cannot evolve the nature of the attack method to improve it.
[0044] It should be noted that the honeypot system in this application scenario can be applied to attacks of different access types, such as: 1) System file access: The simulated echo content corresponding to the system file access may include forged system files; for example, when an attacker attempts to access a system file, the honeypot system may feedback a false but correctly formatted file containing forged user information.
[0045] 2) Web application interactive access: The simulated echo content corresponding to the interactive access of the web application may include forged interactive pages of the web application, etc. For example, when an attacker attempts to log in through the web interface, the honeypot system can feedback a fake login success page or error message; in addition, when an attacker launches an SQL injection attempt against the website, the honeypot system can also feedback a forged page containing database query results.
[0046] 3) Command line operation: The simulated echo content corresponding to the command line operation access may include forged command line operation result information. For example, when an attacker attempts to execute a command through remote terminal access, the honeypot system can feedback the false output of the corresponding command, such as forged system status information or error messages.
[0047] 4) File upload The simulated echo content of file upload access includes forged file upload result information. For example, when an attacker attempts to upload a malicious file, the honeypot system can pretend to receive the file and feedback a successful upload message.
[0048] 5) File download: The simulated echo content of file download access includes forged download files. For example, when an attacker tries to download a target file, the honeypot system can feedback a fake or corrupted target file.
[0049] 6) API call: The simulated echo content of API call access includes forged API call result information. For example, when an attacker attempts to access or operate sensitive data through the API interface, the honeypot system can feedback an API response containing forged sensitive data.
[0050] 7) Network detection; The simulated echo content of network detection access includes forged service information. For example, when an attacker attempts to detect the number of servers on the network, the honeypot system can feedback the forged server number and server information.
[0051] 8) Network Scanning: The simulated echo content of network scanning access includes forged port scanning information. For example, when an attacker attempts to scan the port information that can be invaded, the honeypot system can feedback bait port information or forged port information.
[0052] In summary, when the method for handling network attacks of this embodiment is applied to a honeypot system, the honeypot system can adapt to the threat environment, analyze and identify access requests of various attack behaviors in real time, and forge corresponding simulation echo contents through a generative model to simulate the real system for feedback. The whole process is more about forging feedback results rather than relying on a forged operating environment, so the required resource overhead is relatively small. For attackers, there is no significant difference in perception from the real system, so it is difficult to discover, and there is no way to summarize the evolutionary direction of improving attack methods, thereby curbing the development of network attacks to a certain extent.
[0053] In addition, corresponding to Figure 1 In addition to the method shown, another embodiment of the present specification also provides a network attack processing device. Figure 3A schematic diagram of the structure of a processing device 300 provided in one embodiment of this specification includes: The attack detection module 310 performs attack behavior detection on the access request of the target user.
[0054] The content generation module 320 generates simulation echo content matching the access request based on a generative model if it is detected that the access request contains an attack behavior.
[0055] The content feedback module 330 feeds back the simulated echo content to the target user.
[0056] Based on the device of the embodiment of this specification, after the target user initiates an access request, the access request is firstly detected for attack behavior. If an attack behavior is detected, a generative model is used to forge simulated echo content to feed back to the target user. Since the simulated echo content is not data obtained by actual access, the target user cannot effectively use the simulated echo content to achieve a substantial attack effect after obtaining it, and may also mistakenly believe that the attack task has been successfully completed and stop evolving, thereby having a certain degree of ability to defend against network attacks.
[0057] Optionally, the generative model is trained based on attack behavior samples and corresponding labels, and the labels are system response content when a real system receives an attack behavior matching the attack behavior sample.
[0058] Optionally, the generative model is a large language model; the content generation module 320 generates simulated echo content that matches the access request based on the generative model, including: determining the access type and access intent of the access request; inputting the access type and the access intent as prompt words into the generative model to create simulated echo content whose data format conforms to the access type and whose data content conforms to the access intent.
[0059] Among them, the content generation module 320 inputs the access type and the access intention as prompt words into the generative model to create a simulation echo content whose data format conforms to the access type and whose data content conforms to the access intention, including: refining the historical access records of the target user to obtain continuity access information related to the access intention; inputting the continuity access information, the access type and the access intention as prompt words into the generative model to create a simulation echo content whose data format conforms to the access type, whose data content conforms to the access intention and whose data logic conforms to the continuity access information.
[0060] Optionally, the continuity access information includes the historical access type, historical access intention and historical simulation echo content corresponding to the historical access request of the target user; the data logic of the simulation echo content conforms to the continuity access information, which means that the first correspondence relationship matches the second correspondence relationship, and the first correspondence relationship refers to the correspondence between the simulation echo content and the access type and the access intention, and the second correspondence relationship refers to the correspondence between the historical simulation echo content and the historical access type and the historical access intention.
[0061] Optionally, the access type includes: at least one of: system file access, web application interactive access, command line operation access, file upload access, file download access, API call access, network detection access and network scanning access.
[0062] Optionally, the simulation echo content corresponding to the system file access includes: forged system files; the simulation echo content corresponding to the web application interactive access includes: forged interactive pages of web applications; the simulation echo content corresponding to the command line operation access includes: forged command line operation result information; the simulation echo content of the file upload access includes: forged file upload result information; the simulation echo content of the file download access includes: forged downloaded files; the simulation echo content of the API call access includes: forged API call result information; The simulated echo content of the network detection access includes: forged service information; the simulated echo content of the network scanning access includes: forged port scanning information.
[0063] Optionally, before performing attack behavior detection on the target user's access request, the attack detection module 310 also performs risk detection on the network traffic data, and when risky target network traffic data is detected, determines the access request of the target user that generates the target network traffic data.
[0064] Obviously, the processing device of the embodiment of this specification can achieve Figure 1 The steps and functions in the illustrated embodiment will not be described in detail here.
[0065] Figure 4 This is a schematic diagram of the structure of an electronic device provided by an embodiment of this specification. Figure 4At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.
[0066] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.
[0067] The memory is used to store the program. Specifically, the program may include a program code, and the program code includes a computer operation instruction. The memory may include a memory and a non-volatile memory, and provides instructions and data to the processor.
[0068] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming the above-mentioned network attack processing device at the logical level. The processor executes the program stored in the memory and is specifically used to perform the following operations: Perform attack behavior detection on target users' access requests. If it is detected that the access request contains an attack behavior, a simulation echo content matching the access request is generated based on the generative model.
[0069] The simulated echo content is fed back to the target user.
[0070] Based on the electronic device of the embodiment of this specification, after the target user initiates an access request, the access request is firstly detected for attack behavior. If an attack behavior is detected, a generative model is used to forge simulated echo content to feed back to the target user. Since the simulated echo content is not data obtained by actual access, the target user cannot effectively use the simulated echo content to achieve a substantial attack effect after obtaining it, and may also mistakenly believe that the attack task has been successfully completed and stop evolving, thereby having a certain degree of ability to defend against network attacks.
[0071] The above is as in this manual Figure 1 The method disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in one or more embodiments of this specification can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in one or more embodiments of this specification can be directly embodied as a hardware decoding processor for execution, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field such as random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory, register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.
[0072] Of course, in addition to software implementation, the electronic device of this specification does not exclude other implementation methods, such as logic devices or a combination of software and hardware, etc., that is to say, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.
[0073] The embodiments of this specification also provide a computer-readable storage medium, which stores one or more programs.
[0074] The one or more programs include instructions that, when executed by a portable electronic device including a plurality of application programs, enable the portable electronic device to execute Figure 1 The method of the embodiment shown is specifically used to perform the following operations: Perform attack behavior detection on target users' access requests. If it is detected that the access request contains an attack behavior, a simulation echo content matching the access request is generated based on the generative model.
[0075] The simulated echo content is fed back to the target user.
[0076] In short, the above description is only a preferred embodiment of this specification and is not intended to limit the protection scope of this specification. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included in the protection scope of one or more embodiments of this specification.
[0077] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
[0078] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules 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, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape 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. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0079] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0080] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
Claims
1. A method for processing network attacks, applied to honeypot systems, include: Perform attack behavior detection on target users’ access requests; If it is detected that the access request contains an attack behavior, generating a simulation echo content matching the access request based on the generative model; The simulated echo content is fed back to the target user.
2. The method according to claim 1, The generative model is trained based on attack behavior samples and corresponding labels, and the labels are system response content when a real system receives an attack behavior matching the attack behavior sample.
3. The method according to claim 1, The generative model is a large language model; based on the generative model, a simulation echo content matching the access request is generated, include: Determining the access type and access intent of the access request; The access type and the access intention are input into the generative model as prompt words to create simulation echo content whose data format conforms to the access type and whose data content conforms to the access intention.
4. The method according to claim 3, The access type and the access intention are input into the generative model as prompt words to create a simulation echo content whose data format conforms to the access type and whose data content conforms to the access intention, include: Refining the historical access records of the target user to obtain continuous access information related to the access intention; The continuity access information, the access type and the access intention are input into the generative model as prompt words to create simulation echo content whose data format conforms to the access type, data content conforms to the access intention and data logic conforms to the continuity access information.
5. The method according to claim 4, The continuity access information includes the historical access type, historical access intention and historical simulation echo content corresponding to the historical access request of the target user; the data logic of the simulation echo content conforms to the continuity access information, which means that the first correspondence relationship matches the second correspondence relationship, and the first correspondence relationship refers to the correspondence between the simulation echo content and the access type and the access intention, and the second correspondence relationship refers to the correspondence between the historical simulation echo content and the historical access type and the historical access intention.
6. The method according to claim 4, The access type include: At least one of system file access, web application interactive access, command line operation access, file upload access, file download access, API call access, network detection access and network scanning access.
7. The method according to claim 6, The system file accesses the corresponding simulation echo content include: Falsification of system files; The simulated echo content corresponding to the web application interactive access includes: a web application forged interactive page; The simulation echo content corresponding to the command line operation access includes: forged command line operation result information; The simulated echo content of the file upload access includes: forged file upload result information; The simulated echo content of the file download access includes: forged download file; The simulated echo content of the API call access includes: forged API call result information; The simulated echo content of the network detection access includes: forged service information; The simulated echo content of the network scanning access includes: forged port scanning information.
8. The method according to any one of claims 1 to 7, Before performing attack behavior detection on the target user's access request, the method also include: Conduct risk detection on network traffic data; In the case where risky target network traffic data is detected, an access request of the target user generating the target network traffic data is determined.
9. A network attack processing device, include: Attack detection module, which detects attack behaviors on target users’ access requests; A content generation module, if it is detected that the access request contains attack behavior, generates simulation echo content matching the access request based on a generative model; The content feedback module feeds back the simulated echo content to the target user.
10. An electronic device, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the following operations when executed by the processor: Perform attack behavior detection on target users’ access requests; If it is detected that the access request contains an attack behavior, generating a simulation echo content matching the access request based on the generative model; The simulated echo content is fed back to the target user.
11. A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the following operations: Perform attack behavior detection on target users’ access requests; If it is detected that the access request contains an attack behavior, generating a simulation echo content matching the access request based on the generative model; The simulated echo content is fed back to the target user.
Citation Information
Cited By
Intention recognition method, large language model, electronic equipment and medium
CN120353891A
Full-automatic intelligent simulation method and system for power system driven by large language model
CN120430024A
Large language model driven power system full-automatic intelligent simulation method and system
CN120430024B
