Router secure internet surfing method and system

Abstract

The invention relates to the field of network communication, aims to provide an efficient and accurate intelligent interception method, and provides a router safe internet surfing method and system, a multi-layer detection mechanism is realized through a local grading risk library, an AI risk prediction model and a cloud comparison library which are designed in sequence, and various risk scenes can be covered more comprehensively, the local grading risk library can quickly identify high-frequency risks, the AI prediction model can process complex mode identification, and the cloud comparison library can provide more comprehensive data support. Through a layered detection mechanism and local and cloud combined detection, the detection efficiency can be improved, the network bandwidth consumption can be reduced, and the detection accuracy can also be improved. Through multi-band parallel processing and a network slicing technology, compatibility of security detection and high-speed transmission is realized, and the delay is reduced to be less than 10ms.

Description

Technical Field

[0001] This invention relates to the field of network communication, specifically to a method and system for secure internet access via router. Background Technology

[0002] With the widespread adoption of Wi-Fi 7 technology, home networks have entered a new phase of multi-band concurrent and low-latency transmission. However, the threat posed by illegal websites (such as scams, vulgar content, and phishing sites) to users remains unresolved. Existing technologies have the following shortcomings: Traditional routers' security detection and data transmission are processed serially, which can easily lead to increased network latency and is incompatible with the high-speed characteristics of Wi-Fi 7. The blocking mode is limited to "Allow / Block" only, which cannot meet the diverse needs of children's devices, adult devices, and IoT devices in the home. Relying on static databases for regular updates results in a delayed response to new types of illegal websites, and the user feedback mechanism is inadequate, leading to a high rate of false / false blocking. Insufficient privacy protection, with some solutions transmitting raw access data to the cloud, poses a risk of user information leakage. Summary of the Invention

[0003] In order to provide an efficient and accurate intelligent blocking method, this application provides a method and system for secure internet access via router.

[0004] The technical solution adopted by the present invention to solve the above problems is: Methods for secure internet access via router include: When a user initiates a network access request through the router, the access information in the access request is extracted in parallel. Access information is compared with a local risk classification database. If the risk is not classified as low-risk, an interception operation is performed based on the corresponding risk level. If the risk is classified as low-risk, access is allowed. If the risk is not identified, an AI risk prediction model is used to determine the risk level. If the risk is not classified as low-risk or no-risk, an interception operation is performed based on the corresponding risk level. If the risk is classified as low-risk or no-risk, access is allowed. If the risk is not identified, a comparison is performed with a cloud-based database. If a match is found, access is blocked; otherwise, access is allowed.

[0005] Furthermore, the access information includes the target URL, device identifier, and access time.

[0006] Furthermore, risks other than low risk and no risk include medium risk and high risk; If the risk level is high, it will be directly blocked; if the risk level is medium, it will be determined based on the device identifier and the preset device customization rules, either by mandatory blocking or by user decision on whether to grant access.

[0007] Furthermore, it also includes collecting user feedback on the interception results and optimizing the local risk classification database, AI risk prediction model, and / or cloud comparison database based on the feedback results.

[0008] Furthermore, when comparing data with the cloud-based comparison database, the URL hash value is used for comparison.

[0009] Furthermore, the URL hash value is sent to the cloud using the 6GHz frequency band.

[0010] Furthermore, the local tiered risk database caches risks within the past H time period.

[0011] Furthermore, if the risk is other than no risk, it also includes pushing risk warning information to the user.

[0012] A secure internet access system for routers includes: Local risk grading database: used for local caching of information on risky websites and their corresponding risk levels; AI risk prediction model: used to predict the risk level of a website; Cloud-based comparison database: Used to store information about risky websites and their corresponding risk levels in the cloud; Parsing Unit: When a user initiates a network access request through the router, it extracts the access information from the access request in parallel. Intelligent comparison unit: The access information is compared with the local risk classification database. If the risk is other than low risk, the corresponding blocking operation is performed; if it is low risk, access is allowed; if not identified, an AI risk prediction model is used to determine the risk level. If the risk is other than low risk or no risk, the corresponding blocking operation is performed; if it is low risk or no risk, access is allowed; if not identified, the comparison is performed with the cloud comparison database. If a match is found, access is blocked; otherwise, access is allowed.

[0013] Furthermore, it also includes a device customization rule unit, which is used to preset the access rules for the device.

[0014] The advantages of this invention compared to existing technologies are as follows: By sequentially designing a local tiered risk database, an AI risk prediction model, and a cloud-based comparison database, a multi-layered detection mechanism is achieved, which can more comprehensively cover various risk scenarios. Specifically, the local tiered risk database can quickly identify high-frequency risks, the AI ​​prediction model can handle complex pattern recognition, and the cloud-based comparison database provides more comprehensive data support. Through this layered detection mechanism and the combination of local and cloud-based detection, detection efficiency is improved, network bandwidth consumption is reduced, and detection accuracy is enhanced. By employing multi-band parallel processing and network slicing technology, security detection is compatible with high-speed transmission, reducing latency to less than 10ms. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0016] Methods for secure internet access via router include: When a user initiates a network access request through the router, the access information in the access request is extracted in parallel, including the target URL, IP address, device identifier (MAC address, operating system characteristics) and access time; Access information is compared with a local risk classification database. If the risk is not classified as low-risk, an interception operation is performed based on the corresponding risk level. If the risk is classified as low-risk, access is allowed. If the risk is not identified, an AI risk prediction model is used to determine the risk level. If the risk is not classified as low-risk or no-risk, an interception operation is performed based on the corresponding risk level. If the risk is classified as low-risk or no-risk, access is allowed. If the risk is not identified, the data is compared with a cloud-based comparison database. If a match is found, access is blocked; otherwise, access is allowed and logged.

[0017] In this embodiment, the local tiered risk database is used to cache recent (e.g., two days) high-frequency risks, allowing for rapid local comparison, significantly reducing the number of requests to the cloud, lowering network latency, and improving system response speed; it is suitable for handling common, known risks. The AI ​​prediction model can handle complex pattern recognition. In this embodiment, the AI ​​prediction model is trained based on URL keywords, domain structure, and historical access records, making it suitable for identifying risks that are difficult to detect through simple comparisons. The cloud-based comparison database provides more comprehensive risk comparisons and can be updated in real time, ensuring that the system can detect newly emerging risks promptly; it is suitable for handling rare, newly emerging risks. The tiered detection mechanism not only improves detection efficiency but also detection accuracy.

[0018] In this embodiment, risk levels include no risk, low risk (such as controversial information), medium risk (such as vulgar content), and high risk (such as phishing websites), with different handling methods corresponding to different risk levels. No risk and low risk access are allowed, high risk access is directly blocked, and for medium risk access, a preset device-customized rule determines whether to forcibly block access or allow user confirmation. Device-customized rules are used to set access rules for the device. For example, under medium risk, if the accessing device is a child's device, it will be forcibly blocked; if the accessing device is an adult's device, a pop-up window will prompt "Access?". Device-customized rules meet the differentiated access needs of devices, improving the user experience. Furthermore, risk warning information can be pushed to users for low, medium, and high risk levels. This risk warning information can be prioritized for transmission through Wi-Fi 7's QoS mechanism, ensuring no delivery delay.

[0019] Furthermore, it also includes collecting user feedback on the interception results and optimizing the local risk classification database, AI risk prediction model, and / or cloud comparison database based on the feedback results.

[0020] To enhance network transmission security, URL hash values ​​are used for comparison when comparing data with the cloud-based database. To improve transmission efficiency, the 6GHz frequency band is used to send URL hash values ​​to the cloud, ensuring that data transmission on the main frequency bands (2.4GHz / 5GHz) is not affected.

[0021] Taking a child's device accessing a vulgar website as an example: The child tablet initiates an access request via the 5GHz band of the Wi-Fi 7 router; the device type is identified as a "children's device," and the target URL is extracted; a local rapid comparison reveals that the URL is in the "children's device blocking list" and marked as "medium risk"; access is forcibly blocked, and a notification is pushed to the tablet via the QoS priority channel: "This website contains vulgar content and has been blocked for children's devices"; parents confirm the blocking is effective through the router's app, and the feedback information is updated to the AI ​​prediction model to strengthen the identification of similar websites.

[0022] Correspondingly, this embodiment also provides a router secure internet access system, including: Local risk grading database: used for local caching of information on risky websites and their corresponding risk levels; AI risk prediction model: used to predict the risk level of a website; Cloud-based comparison database: Used to store information about risky websites and their corresponding risk levels in the cloud; Parsing Unit: When a user initiates a network access request through the router, it extracts the access information from the access request in parallel. Intelligent comparison unit: The access information is compared with the local risk classification database. If the risk is other than low risk, the corresponding blocking operation is performed. If it is low risk, access is allowed. If not identified, the AI ​​risk prediction model is used to determine the risk level. If the risk is other than low risk or no risk, the corresponding blocking operation is performed. If it is low risk or no risk, access is allowed. If not identified, the comparison is performed with the cloud comparison database. If a match is found, access is blocked; otherwise, access is allowed and logged.

[0023] Furthermore, it also includes a device customization rule unit, which is used to preset the access rules for the device.

Claims

1. A method for secure internet access via router, characterized in that: include: When a user initiates a network access request through the router, the access information in the access request is extracted in parallel. The access information is compared with the local risk classification database. If it is a risk other than low risk, the blocking operation is performed according to the corresponding risk. If the risk level is low, access is allowed; if the risk level is not identified, an AI risk prediction model is used to determine the risk level. If the risk level is other than low risk or no risk, an interception operation is performed according to the corresponding risk level. Access is allowed if the risk level is low or no; if the risk level is not identified, the system compares the data with the cloud database. If a match is found, access is blocked; otherwise, access is allowed.

2. The method for secure internet access via router according to claim 1, characterized in that, Access information includes the target URL, device identifier, and access time.

3. The method for secure internet access via router according to claim 2, characterized in that, Risks other than low risk and no risk include medium risk and high risk; If it is high-risk, it will be intercepted directly; If the risk level is medium, the system will determine whether to forcibly block access based on the device identifier and the preset device customization rules, or allow the user to decide whether to grant access.

4. The method for secure internet access via router according to claim 1, characterized in that, It also includes collecting user feedback on the interception results and optimizing the local risk classification database, AI risk prediction model, and / or cloud comparison database based on the feedback.

5. The method for secure internet access via router according to claim 1, characterized in that, When comparing data with the cloud-based comparison database, the URL hash value is used for comparison.

6. The method for secure internet access via router according to claim 5, characterized in that, The URL hash value is sent to the cloud using the 6GHz frequency band.

7. The method for secure internet access via router according to claim 1, characterized in that, The local risk classification database caches risks within the last H time period.

8. The method for secure internet access via router according to claim 1, characterized in that, If the risk is other than no risk, it also includes pushing risk warning information to the user.

9. A router-based secure internet access system, characterized in that: include: Local risk grading database: used for local caching of information on risky websites and their corresponding risk levels; AI risk prediction model: used to predict the risk level of a website; Cloud-based comparison database: Used to store information about risky websites and their corresponding risk levels in the cloud; Parsing Unit: When a user initiates a network access request through the router, it extracts the access information from the access request in parallel. Intelligent comparison unit: compares the access information with the local risk classification database. If the risk is not low-risk, it will perform an interception operation according to the corresponding risk. If the risk level is low, access is allowed; if it is not identified, an AI risk prediction model is used to determine the risk level. If the risk level is other than low risk or no risk, an interception operation is performed according to the corresponding risk level. Access is allowed if the risk level is low or no; if the risk level is not identified, the system compares the data with the cloud database. If a match is found, access is blocked; otherwise, access is allowed.

10. The router secure internet access system according to claim 9, characterized in that, It also includes a device customization rule unit, which is used to preset the access rules for the device.