Method and system for realizing secure access based on AI large model

Abstract

The invention particularly relates to a method and a system for realizing secure access based on an AI large model. The method for realizing security access based on the AI large model comprises the following steps: constructing an AI security large model, and creating a cloud host, a virtual router and a cloud firewall; detecting network flow data and task logs in real time, analyzing security risk data, obtaining service data, and detecting service security demand changes in real time; the AI security big model analyzes and reasones security risks and business security level requirements in real time, calls a cloud platform interface to dynamically adjust resources required by a cloud firewall, and calls a security control interface to intelligently adjust a security strategy; and training and optimizing the AI security big model in real time according to the detected new security risk and business data. According to the method and the system for realizing security access based on the AI large model, the security protection capability is improved, the security protection bottleneck is avoided, the reliability is improved, the complex and changeable service security requirements can be met, and the automatic security protection capability is realized.

Description

Technical Field

[0001] The present invention relates to the field of cloud native security technology, and in particular to a method and system for implementing secure access based on an AI big model. Background Art

[0002] Cloud firewall is a network security device provided in cloud computing environment. It protects data and applications in cloud environment through virtualized firewall. With the continuous development of cloud computing technology, the current cloud firewall can no longer meet the complex and changing business security needs of users.

[0003] Traditional cloud firewall solutions lack intelligence and flexibility. They are unable to dynamically expand and adjust resources based on complex and changing business needs, nor can they continuously assess security risks and intelligently configure security policies.

[0004] In order to solve the above problems, the present invention proposes a method and system for achieving secure access based on an AI big model. Summary of the invention

[0005] In order to make up for the deficiencies of the prior art, the present invention provides a simple and efficient method and system for achieving secure access based on an AI big model.

[0006] The present invention is achieved through the following technical solutions:

[0007] A method for implementing secure access based on an AI big model, characterized in that: a cloud native security firewall function is extended based on a cloud computing architecture, comprising the following steps:

[0008] Step S1: Build an AI security big model based on the pre-trained big model and industry security data;

[0009] Step S2: Based on the cloud computing framework, a cloud host, a virtual router and a cloud firewall are created through a cloud platform controller;

[0010] Step S3: Real-time detection of network traffic data and task logs of the cloud firewall, analysis of security risk data, and acquisition of business data from the business system to detect changes in business security requirements in real time;

[0011] Step S4: Based on security risk data and business security requirements, the AI ​​security big model analyzes and infers security risks and business security level requirements in real time;

[0012] Step S5: Based on security risk data and business security requirements, call the cloud platform interface to dynamically adjust the resources required by the cloud firewall itself, call the security control interface to intelligently adjust the security policy, and intelligently improve the security protection capability;

[0013] Step S6: Train and optimize the AI ​​security model in real time based on the new security risks and business data detected.

[0014] In step S3, the method for real-time detection of business security requirements adopts any one of real-time analysis of data packets through message parsing, real-time analysis of business log data through a large model, or real-time analysis of business process changes by acquiring business data.

[0015] In step S4, the AI ​​security big model uses knowledge enhancement generation technology to generate security rules and resource requirements, and uses N big model retrieval methods, N≥1, each method obtains K optimal results, K≥1, and after merging N×K results, the AI ​​security big model extracts the optimal answer from them.

[0016] In step S6, new security risks include new attack methods and new viruses, and new business data include changes in production processes and production procedures.

[0017] A system for achieving secure access based on an AI big model, using a cloud computing architecture, including a cloud host, a virtual router, and a cloud firewall; in addition, it also includes a real-time detection module, a security big model reasoning module, and a continuous security learning module;

[0018] The real-time detection module is responsible for real-time detection of network traffic data and task logs of the cloud firewall, analysis of security risk data, and acquisition of business data from the business system to detect changes in business security requirements in real time;

[0019] The security big model inference module is responsible for building an AI security big model based on the pre-trained big model and industry security data. Based on security risk data and business security requirements, the AI ​​security big model analyzes and infers security risks and business security level requirements in real time, calls the cloud platform interface to dynamically adjust the resources required by the cloud firewall itself, and calls the security control interface to intelligently adjust the security policy, thereby intelligently improving security protection capabilities.

[0020] The continuous security learning module is responsible for continuously training and optimizing the AI ​​security model based on new security risks and business security requirements, and automatically optimizing model responses based on feedback.

[0021] The method for the real-time detection module to detect business security requirements in real time adopts any one of the following methods: real-time analysis of data packets through message parsing, real-time analysis of business log data through a large model, or real-time analysis of business process changes by acquiring business data.

[0022] In the security big model reasoning module, the AI ​​security big model uses knowledge enhancement generation technology to generate security rules and resource requirements, and uses N big model retrieval methods, N≥1, each method obtains K optimal results, K≥1, and after merging N×K results, the AI ​​security big model extracts the optimal answer from them.

[0023] In the continuous security learning module, new security risks include new attack methods and new viruses, and new business data include changes in production processes and production procedures.

[0024] A computing device for realizing secure access based on an AI big model, characterized in that it includes:

[0025] One or more processors, one or more memories, and one or more programs, wherein the one or more programs are stored in the one or more memories and are configured to be executed by the one or more processors, and the one or more programs include instructions for executing any of the above methods.

[0026] A computer-readable storage medium storing one or more programs, characterized in that: the one or more programs include instructions, and when the instructions are executed by a computing device that implements secure access based on an AI big model, the computing device that implements secure access based on an AI big model executes any of the above methods.

[0027] The beneficial effects of the present invention are as follows: the method and system for realizing secure access based on the AI ​​big model can not only continuously detect security risks, adjust security rules in real time, and improve security protection capabilities; but also cooperate with actual business to dynamically adjust the resources required for the security firewall, avoid security protection bottlenecks caused by limited resources, and improve reliability; at the same time, by continuously learning new security knowledge, it can keep pace with the latest attack technologies, meet complex and changeable business security needs, and realize automated security protection capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0029] Attached Figure 1 This is a schematic diagram of a method for implementing secure access based on an AI big model according to the present invention.

[0030] Attached Figure 2 This is a schematic diagram of a system for implementing secure access based on an AI big model in the present invention. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0032] The method for implementing secure access based on the AI ​​big model extends the cloud native security firewall function based on the cloud computing architecture, and includes the following steps:

[0033] Step S1: Build an AI security big model based on the pre-trained big model and industry security data;

[0034] Step S2: Based on the cloud computing framework, a cloud host, a virtual router and a cloud firewall are created through a cloud platform controller;

[0035] Step S3: Real-time detection of network traffic data and task logs of the cloud firewall, analysis of security risk data, and acquisition of business data from the business system to detect changes in business security requirements in real time;

[0036] Step S4: Based on security risk data and business security requirements, the AI ​​security big model analyzes and infers security risks and business security level requirements in real time;

[0037] Step S5: Based on security risk data and business security requirements, call the cloud platform interface to dynamically adjust the resources required by the cloud firewall itself, call the security control interface to intelligently adjust the security policy, and intelligently improve the security protection capability;

[0038] Step S6: Train and optimize the AI ​​security model in real time based on the new security risks and business data detected.

[0039] In step S3, the method for real-time detection of business security requirements adopts any one of real-time analysis of data packets through message parsing, real-time analysis of business log data through a large model, or real-time analysis of business process changes by acquiring business data.

[0040] In step S4, the AI ​​security big model uses knowledge enhancement generation technology to generate security rules and resource requirements, and uses N big model retrieval methods, N≥1, each method obtains K optimal results, K≥1, and after merging N×K results, the AI ​​security big model extracts the optimal answer from them.

[0041] In step S6, new security risks include new attack methods and new viruses, and new business data include changes in production processes and production procedures.

[0042] The system for achieving secure access based on AI big models adopts cloud computing architecture, including cloud hosts, virtual routers and cloud firewalls; in addition, it also includes real-time detection modules, security big model reasoning modules and continuous security learning modules;

[0043] The real-time detection module is responsible for real-time detection of network traffic data and task logs of the cloud firewall, analysis of security risk data, and acquisition of business data from the business system to detect changes in business security requirements in real time;

[0044] The security big model inference module is responsible for building an AI security big model based on the pre-trained big model and industry security data. Based on security risk data and business security requirements, the AI ​​security big model analyzes and infers security risks and business security level requirements in real time, calls the cloud platform interface to dynamically adjust the resources required by the cloud firewall itself, and calls the security control interface to intelligently adjust the security policy, thereby intelligently improving security protection capabilities.

[0045] The continuous security learning module is responsible for continuously training and optimizing the AI ​​security model based on new security risks and business security requirements, and automatically optimizing model responses based on feedback.

[0046] The method for the real-time detection module to detect business security requirements in real time adopts any one of the following methods: real-time analysis of data packets through message parsing, real-time analysis of business log data through a large model, or real-time analysis of business process changes by acquiring business data.

[0047] In the security big model reasoning module, the AI ​​security big model uses knowledge enhancement generation technology to generate security rules and resource requirements, and uses N big model retrieval methods, N≥1, each method obtains K optimal results, K≥1, and after merging N×K results, the AI ​​security big model extracts the optimal answer from them.

[0048] In the continuous security learning module, new security risks include new attack methods and new viruses, and new business data include changes in production processes and production procedures.

[0049] The computing device that implements secure access based on the AI ​​big model includes:

[0050] One or more processors, one or more memories, and one or more programs, wherein the one or more programs are stored in the one or more memories and are configured to be executed by the one or more processors, and the one or more programs include instructions for executing any of the above methods.

[0051] The computer-readable storage medium stores one or more programs, wherein the one or more programs include instructions, which, when executed by a computing device that implements secure access based on an AI big model, enable the computing device that implements secure access based on an AI big model to execute any of the above methods.

[0052] The method and system for realizing secure access based on the AI ​​big model can not only continuously detect security risks, adjust security rules in real time, and improve security protection capabilities; it can also cooperate with actual business to dynamically adjust the resources required for the security firewall, avoid security protection bottlenecks caused by limited resources, and improve reliability; at the same time, by continuously learning new security knowledge, it can keep pace with the latest attack technologies, meet complex and changing business security needs, and realize automated security protection capabilities. The above-described embodiment is only one of the specific implementation methods of the present invention. The usual changes and substitutions made by technicians in this field within the scope of the technical solution of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for achieving secure access based on an AI big model, characterized in that: Expanding the cloud native security firewall function based on cloud computing architecture includes the following steps: Step S1: Build an AI security big model based on the pre-trained big model and industry security data; Step S2: Based on the cloud computing framework, a cloud host, a virtual router and a cloud firewall are created through a cloud platform controller; Step S3: Real-time detection of network traffic data and task logs of the cloud firewall, analysis of security risk data, and acquisition of business data from the business system to detect changes in business security requirements in real time; Step S4: Based on security risk data and business security requirements, the AI ​​security big model analyzes and infers security risks and business security level requirements in real time; Step S5: Based on security risk data and business security requirements, call the cloud platform interface to dynamically adjust the resources required by the cloud firewall itself, call the security control interface to intelligently adjust the security policy, and intelligently improve the security protection capability; Step S6: Train and optimize the AI ​​security model in real time based on the new security risks and business data detected.

2. The method for realizing secure access based on AI big model according to claim 1 is characterized in that: In step S3, the method for real-time detection of business security requirements adopts any one of real-time analysis of data packets through message parsing, real-time analysis of business log data through a large model, or real-time analysis of business process changes by acquiring business data.

3. The method for realizing secure access based on AI big model according to claim 1 is characterized in that: In step S4, the AI ​​security big model uses knowledge enhancement generation technology to generate security rules and resource requirements, and uses N big model retrieval methods, N≥1, each method obtains K optimal results, K≥1, and after merging N×K results, the AI ​​security big model extracts the optimal answer from them.

4. The method for realizing secure access based on AI big model according to claim 2 is characterized in that: In step S6, new security risks include new attack methods and new viruses, and new business data include changes in production processes and production procedures.

5. A system for achieving secure access based on an AI big model, characterized in that: It adopts cloud computing architecture, including cloud hosts, virtual routers and cloud firewalls; in addition, it also includes real-time detection modules, large security model reasoning modules and continuous security learning modules; The real-time detection module is responsible for real-time detection of network traffic data and task logs of the cloud firewall, analysis of security risk data, and acquisition of business data from the business system to detect changes in business security requirements in real time; The security big model inference module is responsible for building an AI security big model based on the pre-trained big model and industry security data. Based on security risk data and business security requirements, the AI ​​security big model analyzes and infers security risks and business security level requirements in real time, calls the cloud platform interface to dynamically adjust the resources required by the cloud firewall itself, and calls the security control interface to intelligently adjust the security policy, thereby intelligently improving security protection capabilities. The continuous security learning module is responsible for continuously training and optimizing the AI ​​security model based on new security risks and business security requirements, and automatically optimizing model responses based on feedback.

6. The system for implementing secure access based on AI big model according to claim 5 is characterized in that: The method for the real-time detection module to detect business security requirements in real time adopts any one of the following methods: real-time analysis of data packets through message parsing, real-time analysis of business log data through a large model, or real-time analysis of business process changes by acquiring business data.

7. The system for implementing secure access based on AI big model according to claim 5 is characterized in that: In the security big model reasoning module, the AI ​​security big model uses knowledge enhancement generation technology to generate security rules and resource requirements, and uses N big model retrieval methods, N≥1, each method obtains K optimal results, K≥1, and after merging N×K results, the AI ​​security big model extracts the optimal answer from them.

8. The system for implementing secure access based on AI big model according to claim 5 is characterized in that: In the continuous security learning module, new security risks include new attack methods and new viruses, and new business data include changes in production processes and production procedures.

9. A computing device for achieving secure access based on an AI big model, characterized in that: include: One or more processors, one or more memories, and one or more programs, wherein the one or more programs are stored in the one or more memories and are configured to be executed by the one or more processors, and the one or more programs include instructions for executing any of the methods according to claims 1 to 4.

10. A computer-readable storage medium storing one or more programs, characterized in that: The one or more programs include instructions, which, when executed by a computing device that implements secure access based on an AI big model, cause the computing device that implements secure access based on an AI big model to execute any one of the methods according to claims 1 to 4.

Images